From Peter.Pregler@risc.uni-linz.ac.at Tue, 02 Nov 1999 09:34:04 +0100 (CET) Date: Tue, 02 Nov 1999 09:34:04 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Further Progress... On 21-Oct-99 Peter Pregler wrote: > > I guess I can help you with that. ;) I have w3cam (a v4l-webcam) running > with my driver. The w3cam also includes a utility called vidcat and I > can do a 'vidcat -f jpeg -s 352x240|xv -' without a problem. Guess that > is what you need. Have a look at http://lugburz.risc.uni-linz.ac.at/ for > my test-setup. I have no idea if it is stable. But on the localhost it > is working well. I will let it run and if things look stable enough I > will release the first v4l-driver before weekend. I spent the holiday yesterday doing some cleanups of the code. The webcam at http://lugburz.risc.uni-linz.ac.at/ is online again. The first error seems to be exposure control in none-streaming mode. It is a bit dark. :) If you are bored give it a try. If I find some time in the evening I will put out a first version. It runs now with most of my test application (w3cam, xawtv 2/3, patched gqcam, but not kwintv-0.7.4 [kde software is junk as always ;]). So at least using it for a webcam should be possible. -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From allan@ostenfeld.dk Tue, 02 Nov 1999 19:23:05 +0100 Date: Tue, 02 Nov 1999 19:23:05 +0100 From: Allan Frank allan@ostenfeld.dk Subject: Webcam II resolutions Hi, I have a Creative Video Blaster Webcam II, and in the manual it states that it can capture still pictures at a resolution of up to 704x576. Therefore I installed the cam at a windows 98 computer and sure enough, the software can capture pictures in that resolution. Then I looked in developer.pdf under the SetFormat function. In this function VideoSize can be set to QCIF(176x144) or CIF(352x288). How does this add up to 704x576? Is it a software trick? Thanks for your help and have a great day! -- /Allan ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From madsdyd@challenge.dk Tue, 2 Nov 1999 20:35:53 +0100 (CET) Date: Tue, 2 Nov 1999 20:35:53 +0100 (CET) From: Mads Bondo Dydensborg madsdyd@challenge.dk Subject: Webcam II resolutions On Tue, 2 Nov 1999, Allan Frank wrote: > Hi, > > I have a Creative Video Blaster Webcam II, and in the manual it states > that it can capture still pictures at a resolution of up to 704x576. > Therefore I installed the cam at a windows 98 computer and sure enough, > the software can capture pictures in that resolution. > Then I looked in developer.pdf under the SetFormat function. In this > function VideoSize can be set to QCIF(176x144) or CIF(352x288). How does > this add up to 704x576? Is it a software trick? AFAICT the 704x576 are software interpolation done by the windows driver. Mads -- Mads Bondo Dydensborg. madsdyd@challenge.dk My shell can beat your shell. So there! ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Wed, 03 Nov 1999 00:10:57 +0100 (CET) Date: Wed, 03 Nov 1999 00:10:57 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: first v4l release Hi all, I will send Jarl my current v4l-release in a few minutes. Note that this is still a alpha-release. So better read the README.v4l. I am pretty sure that it will not crash most of the boxes out there immediatly. But for other things I cannot give any guarantee. ;) Compared to the ppcpia-module you should get a better frame rate and a nicer interface to write programs. The down side is right now that the used kernel-thread is not totally nice to user-programs, i.e. your mouse might jump from time to time. Any comments on how to improve this (see function fetch_frame_timer in v4l.c source) are _highly_ welcome. Additionally, not all features of the camera are supported by the driver. This is due to the limited scope of the v4l-API. If you find any backdoors in the API you can implement any additional proprietary functions (control/set auto exposure, gain, ...) in no time. For anything else please really read the README.v4l in the source. Enjoy, Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Wed, 03 Nov 1999 10:58:42 +0100 Date: Wed, 03 Nov 1999 10:58:42 +0100 From: Bas Rijniersce bas@brijn.nu Subject: V4L: results Hi, Well, I have good news and bad news :) The goods news is that the driver/module loads OK: Nov 3 09:59:46 brainbox kernel: Linux video capture interface: v1.00 Nov 3 09:59:47 brainbox kernel: parport0: PC-style at 0x378, dma 3 [SPP,ECP,ECPEPP,ECPPS2] Nov 3 09:59:47 brainbox kernel: parport0: detected irq 7; use procfs to enable interrupt-driven operation. Nov 3 10:01:40 brainbox kernel: V4L Parport CPiA Webcam driver (C) 1999 P.Pregler v0.3 Nov 3 10:01:40 brainbox kernel: using DMA mode (irq 7, DMA 3) Nov 3 10:01:40 brainbox kernel: CPIA Version: 1.20 (2.0) Nov 3 10:01:40 brainbox kernel: CPIA PnP-ID: 0553:0002:0100 Nov 3 10:01:40 brainbox kernel: VP-Version: 1.0 0100 The bad news is that gqcam and vidcat crash my machine after they have fetched one image. How it looks: gqcam start OK. I see the control buttons etc, there is a excellent looking picture from the cam. But that's it, the machine is locked solid. No ping, no Control-Alt-Backspace, no VT switching. I downloaded w3cam en tried vidcat. I just entered ./vidcat at the prompt to get some help. But it produced a screen of garbage (i guess that was the jpg :) and it also locks solid. Then I loaded evyrhing except the ppv4l module and loaded the old module, clients/main still works OK (nothing changed there?). This machine: Abit BP6 Dual Celeron 333 128 Mb PC100 mem Linux 2.2.13/RH 6.0 + updates Creative Webblaster II I think we are lucky that it gets at least one image, I guess that it makes tracking down the problem easier :) Do you have any suggestions on what to try. I can dump printk's in the module to get some output on where it crashes? Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Majordomo@errors.no Wed, 3 Nov 1999 18:25:55 +0100 Date: Wed, 3 Nov 1999 18:25:55 +0100 From: Majordomo@errors.no Majordomo@errors.no Subject: SUBSCRIBE vision-webcam jose.j.serrano@uv.es -- jose.j.serrano@uv.es has been added to vision-webcam. No action is required on your part. From jose.j.serrano@uv.es Wed, 03 Nov 1999 18:22:15 +0100 Date: Wed, 03 Nov 1999 18:22:15 +0100 From: Jose Serrano Rodenas jose.j.serrano@uv.es Subject: ERRORS I have RedHat  6.1 linux machine. I hava been trying to install the Parport driver v0.3. but I can't do it successfully. I am going to describe the steps that I have followed: Command prompt | Kernel messages 1>modprobe videodev | Linux video capture interface:v1.00 2>modprobe partport | 3>modprobe partport_pc | partport0: PC-style at 0x378 [SPP,PS2] 4>insmod ppv4lcpia | V4L Parport CPiA Webcam driver (C) 1999 P.Pregler v0.3 | parport0: read2 timeout | parport0: Multimedia device, VLSI Vision Ltd PPC2 Camera | ppcpia:interface_pp.c:ppcpia_register(802):port is not ECP capable | ppcpia:interface_pp.c:if_Init(967):No cameras found I have got the parallell webcamII well connected into the parallell port. My question is: have I doing anything wrong? I think so, then can anybody help me? Thanks in advanced. Jose -- Jose J. Serrano Rodenas, Ingeniero Informatico Becario del Servicio de Informática Universidad de Valencia Dr. Moliner, 50 46100 Burjassot (Valencia) Spain Tel: +34-963864310 Fax: +34-963864200 e-mail: Jose.J.Serrano@uv.es ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Wed, 03 Nov 1999 19:44:56 +0100 Date: Wed, 03 Nov 1999 19:44:56 +0100 From: Bas Rijniersce bas@brijn.nu Subject: ERRORS Hi, Jose Serrano Rodenas wrote: > I have RedHat  6.1 linux machine. > I hava been trying to install the Parport driver v0.3. but I can't do > it successfully. | > ppcpia:interface_pp.c:ppcpia_register(802):port is not ECP capable > ppcpia:interface_pp.c:if_Init(967):No cameras found My guess is that the first message is relevant. Go into your bios and look for the parallel port settings. You can often changes this from EPP to ECP+EPP and some other options that I don't recall right now. I think you currently have something like EPP, change it (often with PageUp/PageDown) to the option that has at least ECP (I think ECP+EPP is the option I have used). > Thanks in advanced. > Jose Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From zioivy@libero.it Wed, 03 Nov 1999 23:51:21 +0100 Date: Wed, 03 Nov 1999 23:51:21 +0100 From: ZioIvy zioivy@libero.it Subject: Info about drivers Hi, my name is Ivan Ferrario, and I'm an italian (so sorry for my terrible english...) aerospace engineering student. I decided to use a WebCam-II on a martian rover (my thesys) for objects recognition. First of all, I would like to use the existing drivers on my rover (obviously iI will specify who is the real author); it is possible? I have to use an 8 bit laptop display and I'd like to know if the drivers support this colour depth. If not, can you help me in modifing the original drivers? Thank you! ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 04 Nov 1999 08:22:30 +0100 (CET) Date: Thu, 04 Nov 1999 08:22:30 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Info about drivers On 03-Nov-99 ZioIvy wrote: > Hi, my name is Ivan Ferrario, and I'm an italian (so sorry for my > terrible english...) aerospace engineering student. I decided to use a > WebCam-II on a martian rover (my thesys) for objects recognition. > First of all, I would like to use the existing drivers on my rover > (obviously iI will specify who is the real author); it is possible? The software is published under the GPL. Just have a look at the COPYING file and than consult your lawyer. :) > I have to use an 8 bit laptop display and I'd like to know if the > drivers support this colour depth. > If not, can you help me in modifing the original drivers? There are good an bad news. First of all display depth is not really a matter of the hardware-driver. The camera itself supports yuvu and uyvy formats. The v4l-layer I have written does conversions of this to rgb15/rgb16/rgb24/rg32 for the sole reason that any v4l-application expects that and it is easy and fast to implement. You could also implement to just return the camera formats. Tell me if you need that. The more basic cpia-module does no conversion at all, i.e. it returns the basic yuvu/uyvy-stream. As for rgb8 it really does only make sense if you have an accompanying color-map. So this is best done in user space, e.g. by means of existing library such as gtk. Note that the v4l-test client will do just that, i.e. it should (untested, but strong believe ;) run on rgb8-displays. It gathers rgb24 from the camera. Greetings, Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 04 Nov 1999 08:31:24 +0100 (CET) Date: Thu, 04 Nov 1999 08:31:24 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: ERRORS On 03-Nov-99 Bas Rijniersce wrote: > Hi, > > Jose Serrano Rodenas wrote: >> I have RedHat  6.1 linux machine. >> I hava been trying to install the Parport driver v0.3. but I can't do >> it successfully. > | >> ppcpia:interface_pp.c:ppcpia_register(802):port is not ECP capable >> > ppcpia:interface_pp.c:if_Init(967):No cameras found > > My guess is that the first message is relevant. Go into your bios and > look for the parallel port settings. You can often changes this from EPP > to ECP+EPP and some other options that I don't recall right now. I think > you currently have something like EPP, change it (often with > PageUp/PageDown) to the option that has at least ECP (I think ECP+EPP is > the option I have used). After insertion of the parallel port module you must get a console message similar to this: parport0: PC-style at 0x378, dma 3 [SPP,ECP,ECPEPP,ECPPS2] The protocol ECP _must_ be in the list above. If it is not there stop fiddling around with the camera-driver. It is not the cause and the camera will _not_ work as long as you do not solve that problem first. I can think of several possible causes: the bios/jumper-settings for the parport are wrong as suggested by Bas; the linux parport driver might not support ECP-mode on your parallel card chip (guess only possible for rather exotic/old PCs); your parallel-port does not support ECP-mode (also only for very old PCs). -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From madsdyd@challenge.dk Thu, 4 Nov 1999 10:04:58 +0100 (CET) Date: Thu, 4 Nov 1999 10:04:58 +0100 (CET) From: Mads Bondo Dydensborg madsdyd@challenge.dk Subject: ERRORS On Thu, 4 Nov 1999, Peter Pregler wrote: > wrong as suggested by Bas; the linux parport driver might not support > ECP-mode on your parallel card chip (guess only possible for rather > exotic/old PCs); I have an IBM pc 330 /486 sx and a toshiba laptop, where I believe the Linux kernel does not support ECP. It may be more common then normally assumed? Mads -- Mads Bondo Dydensborg. madsdyd@challenge.dk Faced with the prospect of rereading this book, I would rather have my brains ripped out by a plastic fork - Charles Cooper, ZD net, in review of B@TSOT by Bill Gates. ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 04 Nov 1999 10:12:50 +0100 (CET) Date: Thu, 04 Nov 1999 10:12:50 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: ERRORS On 04-Nov-99 Mads Bondo Dydensborg wrote: > On Thu, 4 Nov 1999, Peter Pregler wrote: > >> wrong as suggested by Bas; the linux parport driver might not support >> ECP-mode on your parallel card chip (guess only possible for rather >> exotic/old PCs); > > > I have an IBM pc 330 /486 sx and a toshiba laptop, where I believe the > Linux kernel does not support ECP. > > It may be more common then normally assumed? Maybe ask the guys that wrote the parport-code or directly on the linux-kernel list. The cpia-driver just uses that code. As for cpia: you could implement nibble-mode which also works with SPP. Performance will be worse and DMA won't be possible or hard. -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From madsdyd@challenge.dk Thu, 4 Nov 1999 10:34:59 +0100 (CET) Date: Thu, 4 Nov 1999 10:34:59 +0100 (CET) From: Mads Bondo Dydensborg madsdyd@challenge.dk Subject: ERRORS On Thu, 4 Nov 1999, Peter Pregler wrote: > > > It may be more common then normally assumed? > > Maybe ask the guys that wrote the parport-code or directly on the > linux-kernel list. The cpia-driver just uses that code. As for cpia: you > could implement nibble-mode which also works with SPP. Performance will be > worse and DMA won't be possible or hard. There already is a nible based implementaion, based on (probably your work) the driver; http://www.panteltje.demon.nl/mcam/ However, I just wanted to mention it - I just got some other hardware to run it on. (Still using it as survalience - works pretty good). Mads -- Mads Bondo Dydensborg. madsdyd@challenge.dk Faced with the prospect of rereading this book, I would rather have my brains ripped out by a plastic fork - Charles Cooper, ZD net, in review of B@TSOT by Bill Gates. ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 04 Nov 1999 11:24:37 +0100 (CET) Date: Thu, 04 Nov 1999 11:24:37 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: ERRORS On 04-Nov-99 Mads Bondo Dydensborg wrote: > > There already is a nible based implementaion, based on (probably your > work) the driver; > > http://www.panteltje.demon.nl/mcam/ It looks like a direct port of the VLSI-sample code. All stuff runs in user space and therefore needs root to access the parport directly. -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Fri, 05 Nov 1999 16:41:46 +0100 (CET) Date: Fri, 05 Nov 1999 16:41:46 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: V4L: results On 03-Nov-99 Bas Rijniersce wrote: > This machine: > Abit BP6 > Dual Celeron 333 > 128 Mb PC100 mem > Linux 2.2.13/RH 6.0 + updates > Creative Webblaster II > > I think we are lucky that it gets at least one image, I guess that it > makes tracking down the problem easier :) Do you have any suggestions on > what to try. I can dump printk's in the module to get some output on > where it crashes? I think it is some kind of smp-locking problem (related to the thread that queries the camera - fetch_frame_timer). The printk-stuff will not help much. I review the locking stuff. As for now consider the driver smp-unsafe. :( -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Fri, 05 Nov 1999 17:39:29 +0100 Date: Fri, 05 Nov 1999 17:39:29 +0100 From: Bas Rijniersce bas@brijn.nu Subject: V4L: results Peter Pregler wrote: > I think it is some kind of smp-locking problem (related to the thread that > queries the camera - fetch_frame_timer). The printk-stuff will not I guess so, I don't know to much about C, but looking thru v4l.c there wasn't to much that looked very SMP unsafe :) Just for fun I tried to replace the smart loop that idles until an image is ready with a plain while loop. Didn't help to much :) it hadn't looked further in the file :) My first post that said that I got a first image... that't isn'r really true. It sometimes crashes before the complete GTK window is drawn, sometime I see one "speed stats" line :(( I've been searching the web for +linux +SMP +scheduling but didn't find anything :( > help much. I review the locking stuff. As for now consider the driver > smp-unsafe. :( Any SMP guru's lurking on this list??? Please ... ;-)) I guess a lot of people read this list, more test reports would be nice. See if there are SMP configs that work for example.. > -Peter Thanx for the good work! Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Jarl.Totland@bdc.no Wed, 10 Nov 1999 10:33:30 +0100 Date: Wed, 10 Nov 1999 10:33:30 +0100 From: Jarl Totland Jarl.Totland@bdc.no Subject: New driver available (cpia-0.3.1.tgz) New driver available at the webpage. Also, the list archive should be up and running now. Vyrdsamt, Jarl Totland http://home.eunet.no/~jtotland ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Wed, 10 Nov 1999 13:25:36 +0100 Date: Wed, 10 Nov 1999 13:25:36 +0100 From: Bas Rijniersce bas@brijn.nu Subject: cpia-0.3.1 SMP results Hi, The last version doesn't lock my machine anymore, thanx Peter :) But it doesn't work correctly yet: >From dmesg --------- Linux video capture interface: v1.00 parport0: PC-style at 0x378, dma 3 [SPP,ECP,ECPEPP,ECPPS2] parport0: detected irq 7; use procfs to enable interrupt-driven operation. V4L Parport CPiA Webcam driver (C) 1999 P.Pregler v0.3 using DMA mode (irq 7, DMA 3) ppcpia(38576):v4l.c:v4l_initialize(1398):v4l_initialize ppcpia(38577):v4l.c:goto_low_power(759):camera now in LOW power state ppcpia(38585):v4l.c:goto_high_power(748):camera now in HIGH power state CPIA Version: 1.20 (2.0) CPIA PnP-ID: 0553:0002:0100 VP-Version: 1.0 0100 ppcpia(38597):v4l.c:save_camera_state(797):32/6/92 ppcpia(38598):v4l.c:save_camera_state(815):0/0/185/0/220/214/214/230 ppcpia(38599):v4l.c:goto_low_power(759):camera now in LOW power state ppcpia(46742):v4l.c:v4l_open(895):v4l_open ppcpia(46750):v4l.c:goto_high_power(748):camera now in HIGH power state ppcpia(46763):v4l.c:dispatch_commands(512):0 1 0 0 1 1 1 0 ppcpia(46781):v4l.c:v4l_ioctl(1020):VIDIOCGCAP ppcpia(46781):v4l.c:v4l_ioctl(1109):VIDIOCGWIN ppcpia(46781):v4l.c:v4l_ioctl(1074):VIDIOCGPICT ppcpia(46781):v4l.c:v4l_ioctl(1084):VIDIOCSPICT ppcpia(46781):v4l.c:v4l_ioctl(1090):palette: 4 ppcpia(46781):v4l.c:v4l_ioctl(1093):depth: 24 ppcpia(46781):v4l.c:v4l_ioctl(1102):VIDIOCSPICT: 24 / 4 // 32768 / 32768 / 32768 / 32768 ppcpia(46781):v4l.c:v4l_ioctl(1119):VIDIOCSWIN ppcpia(46781):v4l.c:v4l_ioctl(1199):1/352:288 --------- The last lines (i don't know from which one) is when I start ./gqcam. GQcam prints one line: flags: 49 Start drawing the GTK window, but never finishes. Then gqcam is frozen, no way to kill it (kill -9 doesn't work!). I have left the program for 15-20 minutes, no change. The rest of the machines is working OK. The only way to kill gqcam is to reboot. Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Wed, 10 Nov 1999 14:02:07 +0100 (CET) Date: Wed, 10 Nov 1999 14:02:07 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: cpia-0.3.1 SMP results On 10-Nov-99 Bas Rijniersce wrote: > Hi, > > The last version doesn't lock my machine anymore, thanx Peter :) That is simply because it does not use polled streaming mode anymore for a simple read. It will still crash if you use an application that does use mmap. I have now a smp-machine here at work that I can use to fix the smp-stuff (until the box is used for production). So I guess I will fix the smp-stuff sooner or later. There is one thing about the new driver that might be of interest for the non-v4l driver users. The new v4l_read uses frame capture instead of stream capture mode (see the source and the developers guide for more info). With this method which seems to be more suitable to polling I got frame rates of about 30 for CIF and up to 60 for QCIF (both with auto compression). For application writers using the non-v4l module this means that the streaming-based implementation of the test-client (main.cc) is sub-optimal. I guess streaming does only make sense for a true interrupt driven capture though this is not my primary target anymore. -Peter > But it doesn't work correctly yet: >>From dmesg > --------- > Linux video capture interface: v1.00 > parport0: PC-style at 0x378, dma 3 [SPP,ECP,ECPEPP,ECPPS2] > parport0: detected irq 7; use procfs to enable interrupt-driven > operation. > V4L Parport CPiA Webcam driver (C) 1999 P.Pregler v0.3 > using DMA mode (irq 7, DMA 3) > ppcpia(38576):v4l.c:v4l_initialize(1398):v4l_initialize > ppcpia(38577):v4l.c:goto_low_power(759):camera now in LOW power state > ppcpia(38585):v4l.c:goto_high_power(748):camera now in HIGH power state > CPIA Version: 1.20 (2.0) > CPIA PnP-ID: 0553:0002:0100 > VP-Version: 1.0 0100 > ppcpia(38597):v4l.c:save_camera_state(797):32/6/92 > ppcpia(38598):v4l.c:save_camera_state(815):0/0/185/0/220/214/214/230 > ppcpia(38599):v4l.c:goto_low_power(759):camera now in LOW power state > ppcpia(46742):v4l.c:v4l_open(895):v4l_open > ppcpia(46750):v4l.c:goto_high_power(748):camera now in HIGH power state > ppcpia(46763):v4l.c:dispatch_commands(512):0 1 0 0 1 1 1 0 > ppcpia(46781):v4l.c:v4l_ioctl(1020):VIDIOCGCAP > ppcpia(46781):v4l.c:v4l_ioctl(1109):VIDIOCGWIN > ppcpia(46781):v4l.c:v4l_ioctl(1074):VIDIOCGPICT > ppcpia(46781):v4l.c:v4l_ioctl(1084):VIDIOCSPICT > ppcpia(46781):v4l.c:v4l_ioctl(1090):palette: 4 > ppcpia(46781):v4l.c:v4l_ioctl(1093):depth: 24 > ppcpia(46781):v4l.c:v4l_ioctl(1102):VIDIOCSPICT: 24 / 4 // 32768 / 32768 > / 32768 / 32768 > ppcpia(46781):v4l.c:v4l_ioctl(1119):VIDIOCSWIN > ppcpia(46781):v4l.c:v4l_ioctl(1199):1/352:288 > --------- > > The last lines (i don't know from which one) is when I start ./gqcam. > GQcam prints one line: > flags: 49 > Start drawing the GTK window, but never finishes. Then gqcam is frozen, > no way to kill it (kill -9 doesn't work!). I have left the program for > 15-20 minutes, no change. The rest of the machines is working OK. The > only way to kill gqcam is to reboot. > > Bas > ---- > Bas Rijniersce Phone +31 341 550545 > Oude Telgterweg 81 Fax +31 341 562940 > 3851 EA Ermelo http://www.brijn.nu > The Netherlands bas@brijn.nu > ------------------------------------------------------------------------- > ---- > To unsubscribe from this mailinglist, send the line "unsubscribe > vision-webcam" in the > body of a message to "majordomo@errors.no". > ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Jarl.Totland@bdc.no Thu, 11 Nov 1999 08:21:39 +0100 Date: Thu, 11 Nov 1999 08:21:39 +0100 From: Jarl Totland Jarl.Totland@bdc.no Subject: CPiA driver v0.4.0 available A major rewrite of the driver has taken place, courtesy of Scott Bertin. These are the highlights: - Completely restructured module to be more modular and integrate with the kernel better. - Reformated module code to follow kernel coding style. - kernel patch - new make targets: + kernel will patch your kernel. WARNING this will overwrite the cpia.h file needed by the current USB driver in the 2.3 kernels. + unkernel will undo the changes made by the kernel target + install will install the modules in /lib/modules/`uname -r`/misc - /proc/cpia/camX interface - removed old module and client] - new color conversion code - default sensorFPS to 15 instead of 30. This seems to work better with indoor lighting. - removed all mention of overlay support, it is not appropriate for this camera. Video capture boards that capture directly to the framebuffer or do overlaying themselves use this. - completely nonfunctional USB driver :) Very nice work, Scott. You all know where to find it. Vyrdsamt, Jarl Totland http://home.eunet.no/~jtotland ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Thu, 11 Nov 1999 10:19:27 +0100 Date: Thu, 11 Nov 1999 10:19:27 +0100 From: Bas Rijniersce bas@brijn.nu Subject: Test results cpia-0.4 Hi, The modules load OK :) parport0: PC-style at 0x378, dma 3 [SPP,ECP,ECPEPP,ECPPS2] parport0: detected irq 7; use procfs to enable interrupt-driven operation. Linux video capture interface: v1.00 V4L-Driver for Vision CPiA based cameras v0.4.0 cpia_pp.c:cpia_pp_init(1341):Parallel port driver for Vision CPiA based cameras v0.4 <6> using DMA mode (irq 7, DMA 3) 1 cameras found But when I try to use vidcat ore gqcam I get the message: Can't open device /dev/video I created the /dev/video as shown in the gqcam docs. Scott mentioned that /dev/cpia now exists, should I create an entry for cpia[0-x] in dev (how???) and link video to it? Or is this my problem: cpia.c(117081):cpia_open(2299):cpia_open cpia_pp.c(117081):cpia_pp_open(1174):Found camera[0] cpia.c(117081):printbytes(1230):04 00 00 00 00 00 10 00 Trying to vfree() nonexistent vm area (c8908000) It's the same for gqcam and vidcat. TIA, Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Thu, 11 Nov 1999 08:11:51 -0500 (EST) Date: Thu, 11 Nov 1999 08:11:51 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Test results cpia-0.4 On 11 Nov, Bas Rijniersce wrote: > But when I try to use vidcat ore gqcam I get the message: > Can't open device /dev/video > I created the /dev/video as shown in the gqcam docs. Scott mentioned > that /dev/cpia now exists, should I create an entry for cpia[0-x] in dev > (how???) and link video to it? /dev/cpia[0-x] are no longer supported by this verion of the driver. I think you confused this with /proc/cpia. > Or is this my problem: > cpia.c(117081):cpia_open(2299):cpia_open > cpia_pp.c(117081):cpia_pp_open(1174):Found camera[0] > cpia.c(117081):printbytes(1230):04 00 00 00 00 00 10 00 printbytes shown the camera status after a failed goto_low_power or goto_high_power. In this case it shows that the camera is in state 4. I've never seen that state before, and don't know what it means. reset_camera now calls goto_low_power first, and this is failing or you would have seen a message that the camera is now in LOW power state. I'm late for work, so I'll have another look at this tonight and see if I can figure out what might be happening. > Trying to vfree() nonexistent vm area (c8908000) This is a bug, but unrelated to your problem. I'll try to track it down tonight. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jm.merliot@wanadoo.fr Fri, 12 Nov 1999 08:03:03 +0100 (CET) Date: Fri, 12 Nov 1999 08:03:03 +0100 (CET) From: Jean-Michel Merliot jm.merliot@wanadoo.fr Subject: Driver V4.0 Hi, I tried the driver v4.0. I just added the stuff to make it on my SPM system. Compiles nicely, installs nicely, but I got an I/O error when I try to open /dev/video with the built-in gqcam ;-(. In fact, the same problem than Bas has. Driver v3.1 worked well afer reading the mailing list and remembering my PC was dual CPU ;-). If any test needed, I can help... Is it me or does images tend to turn yellow when lighting is low ? In any case, it's a real nice job you've all done !!! E-Mail: Jean-Michel Merliot Linux and Blue Oyster Cult Fan Deadly Opponent To microsoft Misconception Of Computing Art... ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Fri, 12 Nov 1999 09:53:02 +0100 Date: Fri, 12 Nov 1999 09:53:02 +0100 From: Bas Rijniersce bas@brijn.nu Subject: Driver V4.0 Jean-Michel Merliot wrote: > Driver v3.1 worked well afer reading the mailing list and remembering my PC was > dual CPU ;-). Did you change anything in the source. I coudn't get 0.3.1 to work on my SMP machine: Abit BP 6 dual celeron 333 Webcam II > E-Mail: Jean-Michel Merliot Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Fri, 12 Nov 1999 09:57:07 +0100 (CET) Date: Fri, 12 Nov 1999 09:57:07 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Driver V4.0 On 12-Nov-99 Jean-Michel Merliot wrote: > Hi, > > I tried the driver v4.0. > > I just added the stuff to make it on my SPM system. > > Compiles nicely, installs nicely, but I got an I/O error when I > try to open > /dev/video with the built-in gqcam ;-(. > > In fact, the same problem than Bas has. Please have a look /proc/cpia/cam0 and mail us the content. IIRC it is populated right after the first open. Also any kernel-syslog messages are a must for any help. One thing, did you use the ./loadpp script to load the modules of kernel-patch? If not (i.e. you do insmod by hand), keep in mind that you have to load cpia.o _and_ cpia_pp.o (and in that order). > Driver v3.1 worked well afer reading the mailing list and > remembering my PC was dual CPU ;-). > > If any test needed, I can help... That is nice to hear. It means that the basic v4l with the read-mehtod seems to work with SMP. BUT :) take care with any v4l-application that does use the mmap to get the data from the driver. It will definitly kill your machine right now. I will try to fix mmap-stuff over the weekend, i.e. consider mmap as broken for now. -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Fri, 12 Nov 1999 04:14:11 -0500 (EST) Date: Fri, 12 Nov 1999 04:14:11 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Driver V4.0 On 12 Nov, Jean-Michel Merliot wrote: > Hi, > > I tried the driver v4.0. > > I just added the stuff to make it on my SPM system. > > Compiles nicely, installs nicely, but I got an I/O error when I try to open > /dev/video with the built-in gqcam ;-(. > > In fact, the same problem than Bas has. Is anybody having this problem on a uni-processor system, or does this occur only on SMP machines? > Is it me or does images tend to turn yellow when lighting is low ? Try decreasing the sensor FPS. 15 works well for me. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jm.merliot@wanadoo.fr Fri, 12 Nov 1999 20:29:44 +0100 (CET) Date: Fri, 12 Nov 1999 20:29:44 +0100 (CET) From: Jean-Michel Merliot jm.merliot@wanadoo.fr Subject: Driver V4.0 This message is in MIME format --_=XFMail.1.3.p0.Linux:991112202944:6216=_ Content-Type: text/plain; charset=iso-8859-1 On 12-Nov-99 Bas Rijniersce wrote: > Jean-Michel Merliot wrote: > > >> Driver v3.1 worked well afer reading the mailing list and remembering my PC >> was >> dual CPU ;-). > > Did you change anything in the source. I coudn't get 0.3.1 to work on my > SMP machine: > Abit BP 6 dual celeron 333 > Webcam II > Yep, I modified the Makefile in the module directory. I attach it here ( it's not that big ;-) ) My config : Intel DK440LX Dual PII 333 Webcam II >> E-Mail: Jean-Michel Merliot > > Bas > > ---- > Bas Rijniersce Phone +31 341 550545 > Oude Telgterweg 81 Fax +31 341 562940 > 3851 EA Ermelo http://www.brijn.nu > The Netherlands bas@brijn.nu > ----------------------------------------------------------------------------- > To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the > body of a message to "majordomo@errors.no". --_=XFMail.1.3.p0.Linux:991112202944:6216=_ Content-Disposition: attachment; filename="Makefile" Content-Transfer-Encoding: base64 Content-Description: Makefile Content-Type: application/octet-stream; name=Makefile; SizeOnDisk=1539 IwojIFBvc3NpYmxlIHNldHRpbmdzOgojIF9ERUJVR18gd2lsbCBtYWtlIHRoZSBkcml2ZXIgX3Zl cmJvc2VfICh2aWEgc3lzbG9nIEtFUk5fREVCVUcpICwgcGxlYXNlIGxldAojICAgdGhpcyBzd2l0 Y2hlZCBvbiB1bnRpbCB3ZSBoYXZlIHZlcnNpb24gMS4wLCB3ZSBuZWVkIHRoZSBvdXRwdXQgaW4g Y2FzZQojICAgb2YgdHJvdWJsZXMKIyBfVVNFX0RNQV8gd2lsbCBlbmFibGUgZG1hIGNvZGUsIHNl ZSBSRUFETUUuRE1BIGZvciBmdXJ0aGVyIGluZm8KIyBfX1NNUF9fIGlmIHlvdSBoYXZlIGNvbXBp bGVkIHlvdXIga2VybmVsIHdpdGggU01QIHN1cHBvcnQgKG1vcmUgdGhhbgojICAgb25lIENQVSkK IyAKCkNDICAgICAgID0gZ2NjCkRFQlVHICAgID0gLURfREVCVUdfCldBUk4gICAgID0gLVdhbGwg LVdzdHJpY3QtcHJvdG90eXBlcwpPUFRJTUlaRSA9IC1PMiAtcGlwZQpDT0RFU1RZTCA9IC1mb21p dC1mcmFtZS1wb2ludGVyIC1mbm8tc3RyZW5ndGgtcmVkdWNlCklOQ0xVREVTID0gLUkuLi9pbmNs dWRlIC1JL3Vzci9zcmMvbGludXgvaW5jbHVkZSAjLUk8cGF0aC10by15b3VyLWtlcm5lbC1oZWFk ZXI+CkRFRklORVMgID0gLURNT0RWRVJTSU9OUyAtaW5jbHVkZSAvdXNyL3NyYy9saW51eC9pbmNs dWRlL2xpbnV4L21vZHZlcnNpb25zLmggLURfX0tFUk5FTF9fIC1ETU9EVUxFIC1EX1VTRV9ETUFf IC1EX19TTVBfXwpDRkxBR1MgICA9ICQoREVCVUcpICQoV0FSTikgJChPUFRJTUlaRSkgJChDT0RF U1RZTCkgJChJTkNMVURFUykgJChERUZJTkVTKQpMSUJTICAgICA9CkxGTEFHUyAgID0KUFBDVVMg ICAgPSBtb2R1bGUubyBpbnRlcmZhY2VfcHAubyBpbnRlcmZhY2VfcHBfbG93Lm8gYnVmZmVyLm8K UFBWNEwgICAgPSB2NGwubyBpbnRlcmZhY2VfcHAubyBpbnRlcmZhY2VfcHBfbG93Lm8gY29udmVy dC5vIGJ1ZmZlci5vClBQVjRMMiAgID0gdjRsMi5vIGludGVyZmFjZV9wcC5vIGludGVyZmFjZV9w cF9sb3cubyAKCmFsbDogLmRlcGVuZCBwcGN1c2NwaWEubyBwcHY0bGNwaWEubwojcHB2NGwyY3Bp YS5vCgouYy5vOgoJJChDQykgLWMgJChDRkxBR1MpICQ8CgpwcGN1c2NwaWEubzogJChQUENVUykK CSQoTEQpIC1yIC1vICRAICReIC4uL3NoYXJlZC9wcm9jcy5vCgpwcHY0bGNwaWEubzoJJChQUFY0 TCkKCSQoTEQpIC1yIC1vICRAICReIC4uL3NoYXJlZC9wcm9jcy5vCgpwcHY0bDJjcGlhLm86ICQo UFBWNEwyKQoJJChMRCkgLXIgLW8gJEAgJF4gLi4vc2hhcmVkL3Byb2NzLm8KCmNsZWFuOgoJcm0g LWYgKi5vICp+IGNvcmUgLmRlcGVuZCBERUFESk9FCgpjbGVhbmJpbjoKCXJtIC1mIG1haW4gY3Bp YS5vCgpkZXA6CglybSAtZiAuZGVwZW5kCgltYWtlIC5kZXBlbmQKCi5kZXBlbmQ6CgllY2hvICcj IFByb2dyYW0gZGVwZW5kZW5jaWVzJyA+LmRlcGVuZAoJZ2NjIC1NTSAkKENGTEFHUykgKi5jID4+ LmRlcGVuZAoKCiMKIyBpbmNsdWRlIGEgZGVwZW5kZW5jeSBmaWxlIGlmIG9uZSBleGlzdHMKIwpp ZmVxICguZGVwZW5kLCQod2lsZGNhcmQgLmRlcGVuZCkpCmluY2x1ZGUgLmRlcGVuZAplbmRpZgoK --_=XFMail.1.3.p0.Linux:991112202944:6216=_-- End of MIME message ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jm.merliot@wanadoo.fr Fri, 12 Nov 1999 20:38:40 +0100 (CET) Date: Fri, 12 Nov 1999 20:38:40 +0100 (CET) From: Jean-Michel Merliot jm.merliot@wanadoo.fr Subject: Driver V4.0 On 12-Nov-99 Peter Pregler wrote: > > On 12-Nov-99 Jean-Michel Merliot wrote: >> Hi, >> >> I tried the driver v4.0. >> >> I just added the stuff to make it on my SPM system. >> >> Compiles nicely, installs nicely, but I got an I/O error when I >> try to open >> /dev/video with the built-in gqcam ;-(. >> >> In fact, the same problem than Bas has. > > Please have a look /proc/cpia/cam0 and mail us the content. IIRC it is > populated right after the first open. Also any kernel-syslog messages are > a must for any help. One thing, did you use the ./loadpp script to load > the modules of kernel-patch? If not (i.e. you do insmod by hand), keep in > mind that you have to load cpia.o _and_ cpia_pp.o (and in that order). > I loaded them by hand, cpia then cpia_pp read-only ----------------------- CPIA Version: 1.2 (1.0) CPIA PnP-ID: 0553:0001:0100 VP-Version: 0.0 0000 system_state: 0x2 grab_state: 0x0 stream_state: 0x0 fatal_error: 0x20 cmd_error: 0x60 debug_flags: 0x0 vp_status: 0x10 error_code: 0x62 video_size: CIF sub_sample: 422 yuv_order: YUYV roi: (0, 0) to (352, 288) actual_fps: 0 read-write ----------------------- colour_params: 50 50 50 exposure: 0x2 0x2 0x1 0x1 0x0 0x0 0xb9 0x0 0xdc 0xd6 0xd6 0xe6 color_balance: 0x2 0x20 0x6 0x5c sensor_fps: 15.0 apcor: 0x1c 0x1a 0x2d 0x2a flicker_control: 0x0 0x0 0x0 vl_offset: 0x18 0x1c 0x1e 0x1e stream_start_line: 240 compression_mode: none compression_target: quality 7 10 yuv_threshold: 15 15 compression_params: 0x3 0xb 0x1 0x3 0x2 0x5 0x3 0x2 ecp_timing: normal >> Driver v3.1 worked well afer reading the mailing list and >> remembering my PC was dual CPU ;-). >> >> If any test needed, I can help... > > That is nice to hear. It means that the basic v4l with the read-mehtod > seems to work with SMP. BUT :) take care with any v4l-application that does > use the mmap to get the data from the driver. It will definitly kill your > machine right now. I will try to fix mmap-stuff over the weekend, i.e. > consider mmap as broken for now. > > -Peter > > > ------------------------------- > Email: Peter.Pregler@risc.uni-linz.ac.at > WWW: http://www.risc.uni-linz.ac.at/people/ppregler E-Mail: Jean-Michel Merliot Linux and Blue Oyster Cult Fan Deadly Opponent To microsoft Misconception Of Computing Art... ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Nick.Holloway@alfie.demon.co.uk 15 Nov 1999 11:31:34 -0000 Date: 15 Nov 1999 11:31:34 -0000 From: Nick Holloway Nick.Holloway@alfie.demon.co.uk Subject: cpia 0.4.0 --==UUD_=_942665469 Content-Type: text/plain Content-Transfer-Encoding: 8bit This version has bought the progress on by leaps and bounds. It has the two important ingredients: video4linux support, and merging of PP and USB camera support. Well done. A minor point about the picture produced. The image appears to be darker than that produced by the old cpia client. I know this can be adjusted, but should the default be lighter? The biggest stumbling block in getting to see this was compiling the client. I eventually discovered I had some old gtk headers in /usr/local. Leaving that aside, I did have to make some changes to compile with gtk 1.1 (from Debian 2.1). I've attached the changes in case they are useful to anyone else. -- `O O' | Nick.Holloway@alfie.demon.co.uk // ^ \\ | http://www.alfie.demon.co.uk/ --==UUD_=_942665469 Content-Type: Application/Octet-Stream Content-Transfer-Encoding: Base64 Content-Disposition: attachment; filename="gqcam-gtk11.diff.gz" H4sICETrLzgCA2dxY2FtLWd0azExLmRpZmYA7Zhtb9pIEMdfh08xQurJlCcbMAiqVHGQQVyJnYDR XaVK1mIv4CvZpfYCd1f1u3fWS1pImzsBLa1EecHgh9n/+v+bwdKE0WQCxfEyhlJ5EnMmKAtLAZRK 5WARkaJeqpX0cjCPKBNb1zPFYnEn4eKGM3D4CgwTDKNlVFsV/NFsNjP5fP7/Vrv4g4YqWYdKpWUa LV1XyVdXUKw1moUG5FW4usoAwFS89ZNoysjcDzhjNBCgdb1Xvnv9u932QFtHLOTrXAGyIZ1TQX26 QsVsIQMXF5B+5N3DXtex+n5n5LRB274RE51Rv597IbUyRaWHQoJEjMb+mMchhnUUipnSbbuOZ/Uc e7AlXcH0/OPUhIo90kHql59DO6ZEUBB8MeZ/w/OyPL05uEwFVvjTZ3QNWsfqD+0C6JgtvTMbhvRO BeVdTMKIj5dCIDCVnZ7x1Sm5Cm5NzPw5GdM5aNII9LHd62SVHzJDyi1IgBAEiTfeX7t/ghYhjBiv 4gNs6RTAG4zsh29drRNNQFtFIYqxkvLi8hKqZiUnDVcygk+nc/qwM+kd6qERqZ7ndrt9278eeZ7r gLYll1NKyv+nViKBiFZ7LPXVqvP5+K8vim83PbuIaZLQMFt4XHMabiOJ/qV+EE12Kk7qoCNT+bwz vt5dUdWkYttUbJuHsVXLTmO+XOA2705E2GjUf27C/+3896qBd4+LQCFuminiNBzTvowz+h3pTuZk mvzWvu1ZfqdvdfEP7eZ2YA+HPdfxHdexN8S/JfOfu68Dfp9ejyQS9H6v/k7R1/WqRK/C0d1NloL/ IP7WyHPPmr/0/gD+RvrmVuFo/veELcn8B1XAjeWMrP5Z14Dy/4AqqFbSKkjDMS+AO5SPxD8nrADP GnRt7w7J97zX5wUfDZVY3ynPD6Be01PqaTi69zsxuacxGnxy+J2BdWMPLM8+S/yTB98PKACznhZA Go5p+6oOk0VyQvCd26Ff1c+LNzosn3l/yvVaSjkNR7e5YZ4Y9Vn3uGRumAcwb6gXeqPyLZg3Sr+g nxh6w9+Terg1YE3Iij41XFXXNoNVdbAzVNVbNf3TXPTJoeom8dFAVW9Vqp8HqnVTvVw+VSDWIBXL mKUP9OFh6Jlg1QUz0OQTMi7omPO3frCMY9TBGpvS1ETH9exr132lhUSQXC4H7yXGr+ZKj3byVQpm oGJAEgqLxX1LbUjW4BekNZnl400l1SExWcujXPGlKo7cewCVDtgWccQErsKXYhLNKYK9rb9hz0J4 FsrwhiHgnenU58MZjaYzUYCKaSLCjzei+9kiFwAA --==UUD_=_942665469-- ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Jarl.Totland@bdc.no Mon, 15 Nov 1999 18:36:46 +0100 Date: Mon, 15 Nov 1999 18:36:46 +0100 From: Jarl Totland Jarl.Totland@bdc.no Subject: Michal Kral's V4L parport CPiA driver I have recieved an alternative V4L CPiA parport driver from Michal Kral (Michal.Kral@csob.iqnet.cz). It shows how to take advantage of extended parport support in recent kernels. Might be some code worth looting here and there? Here's what Michal wrote: > I've started developing this driver some time ago while your driver was > not really that useful. However, your implementation got much better > than mine since then. > > So this driver I'm sending is rather an example of implementation with the > IEEE1284 support in the new kernel versions. Get the source at: http://home.eunet.no/~jtotland/vision/driver/kralcpia-0.1.tgz Vyrdsamt, Jarl Totland http://home.eunet.no/~jtotland ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Chris@informinteractive.com Tue, 16 Nov 1999 16:48:07 -0500 Date: Tue, 16 Nov 1999 16:48:07 -0500 From: Chris Whiteford Chris@informinteractive.com Subject: Driver V4.0 This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF307C.44DD09F0 Content-Type: text/plain; charset="windows-1252" -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I just finally got around to compileing the module and client and I am also getting the error about opening the /dev/video device... I have yet to reboot since I recompiled the kernel with v4l support enabled. I will do that later tonight when I get home... (Remote boots with a cable modem... mmmmm fun...DHCP... more fun!) I will post my results when I have them... P.S. I am Running a single Pii 266 (I wish it was a SMP System :)) chris. - -----Original Message----- From: sbertin@mindspring.com [mailto:sbertin@mindspring.com] Sent: Friday, November 12, 1999 4:14 AM To: vision-webcam@errors.no Subject: Re: Driver V4.0 On 12 Nov, Jean-Michel Merliot wrote: > Hi, > > I tried the driver v4.0. > > I just added the stuff to make it on my SPM system. > > Compiles nicely, installs nicely, but I got an I/O error when I try > to open /dev/video with the built-in gqcam ;-(. > > In fact, the same problem than Bas has. Is anybody having this problem on a uni-processor system, or does this occur only on SMP machines? > Is it me or does images tend to turn yellow when lighting is low ? Try decreasing the sensor FPS. 15 works well for me. Scott J. Bertin sbertin@mindspring.com - ---------------------------------------------------------------------- - ------ To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.5.1 for non-commercial use iQA/AwUBODHRnnCyV2ve66GWEQL8PgCdH4tOMkc+reRFlilD+yE0Jr3m+zgAoIRV HvuVGqKo8Wzf0UFFn8bELhoZ =98Ft -----END PGP SIGNATURE----- ------_=_NextPart_001_01BF307C.44DD09F0 Content-Type: text/html; charset="windows-1252" Content-Transfer-Encoding: quoted-printable RE: Driver V4.0

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I just finally got around to compileing the module = and client and I
am also getting the error about opening the = /dev/video device...  I
have yet to reboot since I recompiled the kernel = with v4l support
enabled.  I will do that later tonight when I = get home... (Remote
boots with a cable modem...  mmmmm = fun...DHCP...  more fun!)  I will
post my results when I have them...

P.S. I am Running a single Pii 266 (I wish it was a = SMP System :))

chris.

- -----Original Message-----
From: sbertin@mindspring.com [mailto:sbertin@mindspring.com= ]
Sent: Friday, November 12, 1999 4:14 AM
To: vision-webcam@errors.no
Subject: Re: Driver V4.0


On 12 Nov, Jean-Michel Merliot wrote:
> Hi,
>
> I tried the driver v4.0.
>
> I just added the stuff to make it on my SPM = system.
>
> Compiles nicely, installs nicely, but I got an = I/O error when I try
> to open /dev/video with the built-in gqcam = ;-(. 
>
> In fact, the same problem than Bas has.

Is anybody having this problem on a uni-processor = system, or does
this
occur only on SMP machines?

> Is it me or does images tend to turn yellow when = lighting is low ?

Try decreasing the sensor FPS.  15 works well = for me.

Scott J. Bertin
sbertin@mindspring.com


- = ----------------------------------------------------------------------
- ------
To unsubscribe from this mailinglist, send the line = "unsubscribe
vision-webcam" in the
body of a message to = "majordomo@errors.no".

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 for non-commercial use = <http://www.pgp.com>

iQA/AwUBODHRnnCyV2ve66GWEQL8PgCdH4tOMkc+reRFlilD+yE0Jr3m+zgAoIR= V
HvuVGqKo8Wzf0UFFn8bELhoZ
=3D98Ft
-----END PGP SIGNATURE-----

------_=_NextPart_001_01BF307C.44DD09F0-- ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From chris@tennille.com Wed, 17 Nov 1999 00:43:04 -0500 Date: Wed, 17 Nov 1999 00:43:04 -0500 From: Chris Whiteford chris@tennille.com Subject: Driver V4.0 - My Findings Well I managed to get it up an running once and I even captured a picture. It was running great. I was getting really good frame rates. somewhere about 13fps fullsize on my pii 266 and I was very impressed... But (there is always a but) then I closed the test client and tried to open it again and I ran into problems... The application could not open the /dev/video device again.. But this time there was a diffrent error... So on to see if I could reproduce it.... here goes... the following is from just after using the loadpp script /dev/log/messages *************************************************************************************** Nov 16 23:43:50 tennille kernel: Linux video capture interface: v1.00 Nov 16 23:44:06 tennille kernel: V4L-Driver for Vision CPiA based cameras v0.4.0 Nov 16 23:44:06 tennille modprobe: can't locate module cpia_pp Nov 16 23:44:06 tennille modprobe: can't locate module cpia_usb Nov 16 23:44:06 tennille kernel: cpia_pp.c:cpia_pp_init(1341):Parallel port driver for Vision CPiA based cameras v0.4 <6> using PIO mode Nov 16 23:44:06 tennille kernel: 1 cameras found lsmod *************************************************************************************** Module Size Used by cpia_pp 18496 0 (unused) cpia 36164 0 [cpia_pp] videodev 2400 1 [cpia] vmnet 11200 3 vmppuser 5216 0 (unused) parport_pc 5620 1 [cpia_pp vmppuser] parport 7092 1 [cpia_pp vmppuser parport_pc] vmmon 15072 0 (unused) Now time to run the test program.... hmmmm.............. No luck [root@tennille /home/chris/cpia/client]# ./gqcam /dev/video: No such device When infact the /dev/video device does actually exist. hmmmm and low and behold in the system log we get Nov 16 23:52:51 tennille kernel: Trying to vfree() nonexistent vm area (c991a000) which was what I was getting before when the applicaiton would not load I had set up the devices using the documentation provided in the gqcam install text file. So another dead end here. Just as a side note. I have also rebooted between the first succesfull capture and the third try. (It worked, Did not work, reboot, did not work) I have also tried reseting the camera. I pulled the keyboard power and the put it back. Still no luck. hmmmm. So close yet still so far away. Now I have managed to trace the vfree problem to the camera reset code in cpia_open. I added some debug lines to my code, but it is around 2339 of cpia.c All this debuging seamed to tell me that It was never making it to the end of the cpia_open function. I then tried to reset the camera the hard way (i.e. Power Cycle - Keyboard unplug - plug) and I try again and I get the following in my system log. Nov 17 00:25:38 tennille kernel: Unable to handle kernel paging request at virtual address c990e000 Nov 17 00:25:38 tennille kernel: current->tss.cr3 = 03b0e000, %cr3 = 03b0e000 Nov 17 00:25:38 tennille kernel: *pde = 08fec063 Nov 17 00:25:38 tennille kernel: *pte = 00000000 Nov 17 00:25:38 tennille kernel: Oops: 0002 Nov 17 00:25:38 tennille kernel: CPU: 0 Nov 17 00:25:38 tennille kernel: EIP: 0010:[] Nov 17 00:25:38 tennille kernel: EFLAGS: 00010256 Nov 17 00:25:38 tennille kernel: eax: 00000019 ebx: c81e7e60 ecx: 00000019 edx: 00000019 Nov 17 00:25:38 tennille kernel: esi: c990e000 edi: 00031ca0 ebp: 00000000 esp: c202fee0 Nov 17 00:25:38 tennille kernel: ds: 0018 es: 0018 ss: 0018 Nov 17 00:25:38 tennille kernel: Process cat (pid: 1428, process nr: 78, stackpage=c202f000) Nov 17 00:25:38 tennille kernel: Stack: 00031ca0 c990e000 c81e7be0 c81e7e60 c990e000 c770b318 00000019 c81e7be0 Nov 17 00:25:38 tennille kernel: 00000000 c9842518 c81e7e60 c990e000 00031ca0 c81e7e60 00000000 c990e000 Nov 17 00:25:38 tennille kernel: c770b318 c9842fe2 c81e7e60 c990e000 00031ca0 c770b200 c202e000 c770b34c Nov 17 00:25:38 tennille kernel: Call Trace: [] [] [] [] [] [] [] Nov 17 00:25:38 tennille kernel: [] [] [] [] [sys_read+174/196] [system_call+52/56] Nov 17 00:25:38 tennille kernel: Code: 88 16 3c ff 75 07 bd 01 00 00 00 eb 02 31 ed 46 c7 44 24 28 The program that was using the video device segfulted, but I did notice that the camera made it to the end of the cpia_open command (ha ha my debuging lines helped woo hoo) I also got this camera init stuf in the syslog Nov 17 00:25:38 tennille kernel: CPIA Version: 1.2 (1.0) Nov 17 00:25:38 tennille kernel: CPIA PnP-ID: 0553:0001:0100 Nov 17 00:25:38 tennille kernel: VP-Version: 1.0 0100 and then boom... see the kernel syslog output above Also now any further tries to work with the video device result in a /dev/video0: Device or resource busy Also any try to remove the module cpia_pp results in a rmmod: cpia_pp: Device or resource busy  Well that is about as far as I have come with my debuging... I hope all this info is of help and I hope that it is just not wasting bandwith... :-) P.S. To all the active developers - Thanks... This is Wicked. And the progress has been great. There have been leeps and bounds made in the last little bit and the quailty of application that is surfacing is great. Keep up the good work. Also if there is anything I can do to help in the development just let me know. I am a compientant C and C++ developer (although I do not really have any experence in kernel programing :-)) thanks. chris. ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jm.merliot@wanadoo.fr Wed, 17 Nov 1999 06:48:37 +0100 (CET) Date: Wed, 17 Nov 1999 06:48:37 +0100 (CET) From: Jean-Michel Merliot jm.merliot@wanadoo.fr Subject: Driver V4.0 Hi, I just found the v4.1 driver on the ftp site, tried it and got the same result./dev/video, input output error. I don't have that much interesting messages in the logs, so I'm not sure what's really going on... On 16-Nov-99 Chris Whiteford wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > I just finally got around to compileing the module and client and I > am also getting the error about opening the /dev/video device... I > have yet to reboot since I recompiled the kernel with v4l support > enabled. I will do that later tonight when I get home... (Remote > boots with a cable modem... mmmmm fun...DHCP... more fun!) I will > post my results when I have them... > > P.S. I am Running a single Pii 266 (I wish it was a SMP System :)) > Ouch, seems the problem isn't SMP related... > chris. > > - -----Original Message----- > From: sbertin@mindspring.com [mailto:sbertin@mindspring.com] > Sent: Friday, November 12, 1999 4:14 AM > To: vision-webcam@errors.no > Subject: Re: Driver V4.0 > > > On 12 Nov, Jean-Michel Merliot wrote: >> Hi, >> >> I tried the driver v4.0. >> >> I just added the stuff to make it on my SPM system. >> >> Compiles nicely, installs nicely, but I got an I/O error when I try >> to open /dev/video with the built-in gqcam ;-(. >> >> In fact, the same problem than Bas has. > > Is anybody having this problem on a uni-processor system, or does > this > occur only on SMP machines? > >> Is it me or does images tend to turn yellow when lighting is low ? > > Try decreasing the sensor FPS. 15 works well for me. > > Scott J. Bertin > sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Wed, 17 Nov 1999 11:12:07 +0100 (CET) Date: Wed, 17 Nov 1999 11:12:07 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: ERRORS On 04-Nov-99 Jose Serrano Rodenas wrote: > After that I closedthe gqcam aplication, and I run a v4l aware > aplication (vic v2.8ucl4) but when I begin to transmit with the cpcia > driver the image appear in green ... Okay, I found some time to look at vic and the problem is that vic only works with two video-palettes (VIDEO_PALETTE_YUV422 and VIDEO_PALETTE_YUV422P) which are both not supported by the driver right now. It is not so hard to add a converter for those palettes if we can find a reasonable description. -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Wed, 17 Nov 1999 18:10:43 +0100 (CET) Date: Wed, 17 Nov 1999 18:10:43 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Driver V4.0 - My Findings On 17-Nov-99 Chris Whiteford wrote: > > [root@tennille /home/chris/cpia/client]# ./gqcam > /dev/video: No such device > > When infact the /dev/video device does actually exist. hmmmm The message has nothing to do with the existence of the inode. It does mean that no kernel-routine feels responsible to handle the open, i.e. if the module seems load it means the kernel driver 'crashed'. As for the errors I try to reproduce them. They seem to be just by some unclean close/release of the module. I use some new allocation stuff that I will put together for an update. That might solve your problems too. Some more comments: - I was able to try out isabel - a video conferencing tool - (Austria to Spain, TEN-155 as the basic network provider). It was quite nice. :) - The kernel did not crash on my SMP-test machine during operation with the gqcam-application. Interrupts were taken even by both processors. Actually I think that that part of the driver _is_ SMP-safe. However, I will some more nasty tests to find possible race conditions etc. Has anyone a good (graphics) tool where one can see the cpu-usage of each of the processors? - Regarding SMP mmap is still a killer. Do not try it out. I will fix this until next week. Greetings, Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Nick.Holloway@alfie.demon.co.uk 17 Nov 1999 08:09:27 -0000 Date: 17 Nov 1999 08:09:27 -0000 From: Nick Holloway Nick.Holloway@alfie.demon.co.uk Subject: Driver V4.0 - My Findings chris@tennille.com (Chris Whiteford) writes: > [How cpia-0.4.0 doesn't work for him] > Nov 17 00:25:38 tennille kernel: CPIA Version: 1.2 (1.0) > Nov 17 00:25:38 tennille kernel: CPIA PnP-ID: 0553:0001:0100 > Nov 17 00:25:38 tennille kernel: VP-Version: 1.0 0100 I wonder if the problems are due to different versions of the camera. I have it working fine on UP Pentium 133. However, I notice that the camera version is later. Nov 17 07:44:56 alfie kernel: CPIA Version: 1.20 (2.0) Nov 17 07:44:56 alfie kernel: CPIA PnP-ID: 0553:0002:0100 Nov 17 07:44:56 alfie kernel: VP-Version: 1.0 0100 I've also tried vidcat from w3cam-0.6.2, and that has also managed to grab an image. -- `O O' | Nick.Holloway@alfie.demon.co.uk // ^ \\ | http://www.alfie.demon.co.uk/ ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Wed, 17 Nov 1999 19:06:12 +0100 (CET) Date: Wed, 17 Nov 1999 19:06:12 +0100 (CET) From: Bas Rijniersce bas@brijn.nu Subject: Driver V4.0 - My Findings Hi, > etc. Has anyone a good (graphics) tool where one can see the cpu-usage of > each of the processors? There are two that I know of, I can't get to my machine at the moment. But I'm pretty sure one is called ascpu (an Afterstep/Windowmaker applet) that you can start for separate cpu's. It also shows kernel, user and idle time seperately. But it's all in a small window, so it's maybe to small for what you need. The other that I know of, but never tested it, is the tool that comes with KDE. They have a Windows rip-off CPU monitor. From the screenshots I remember it's very much like the cpu monitor from WinNT And thanx for working on SMP :-) > Greetings, Peter Bas ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Wed, 17 Nov 1999 17:30:17 -0500 (EST) Date: Wed, 17 Nov 1999 17:30:17 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Driver V4.0 - My Findings On 17 Nov, Nick Holloway wrote: > chris@tennille.com (Chris Whiteford) writes: >> [How cpia-0.4.0 doesn't work for him] >> Nov 17 00:25:38 tennille kernel: CPIA Version: 1.2 (1.0) >> Nov 17 00:25:38 tennille kernel: CPIA PnP-ID: 0553:0001:0100 >> Nov 17 00:25:38 tennille kernel: VP-Version: 1.0 0100 > > I wonder if the problems are due to different versions of the camera. > I have it working fine on UP Pentium 133. However, I notice that the > camera version is later. > > Nov 17 07:44:56 alfie kernel: CPIA Version: 1.20 (2.0) > Nov 17 07:44:56 alfie kernel: CPIA PnP-ID: 0553:0002:0100 > Nov 17 07:44:56 alfie kernel: VP-Version: 1.0 0100 > > I've also tried vidcat from w3cam-0.6.2, and that has also managed to > grab an image. > It's good to see this. There are some known problems with the 1.2 version that aren't in the 1.20 version. I thought I was handling this properly, but maybe I missed something. I'll look at it again this weekend. chris@tennille.com (Chris Whiteford) writes: > and then boom... see the kernel syslog output above > > Also now any further tries to work with the video device result in a > /dev/video0: Device or resource busy > > Also any try to remove the module cpia_pp results in a > rmmod: cpia_pp: Device or resource busy Unfortunately, any crash in the kernel will leave the module in use until a reboot. This is probably related to the vfree problem. The vfree problem is fixed in the 0.4.1 driver (Yes, it's hiding. Look in the software archive for the patch.) Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Chris@informinteractive.com Mon, 22 Nov 1999 15:22:49 -0500 Date: Mon, 22 Nov 1999 15:22:49 -0500 From: Chris Whiteford Chris@informinteractive.com Subject: More findings and a wonderful wonderful little picture. This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF3527.5DBEAD60 Content-Type: text/plain; charset="iso-8859-1" -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Well I have been using the unofficial (i think) 0.4.1 version, the vfree problems have gone away, I still encounter a problem with the reset camera code, it seams to have to do with the fatalerror stuff that checks for a bug in the older firmware, if you just ignore the error everything works fine, so it looks like the logic that decideds to ignore or actuall fail is a little weird, I am looking into that. After I got over that hurddle, everything just worked wounderfully. I have a working web cam...... WOOOOOO HOOOOOOOO. I am currently developing and running it of my Libretto 70CT which is a P120mmx and I get around 12fps using the PIO Mode. I have not yet ventured and tried DMA mode, that is next on my list. For all to see. http://209.47.106.152 The camera will be up during bussiness hours (when I am at work); Please be gental with my laptop, it is just a baby and our line here sucks crap. It is only 128k ISDN chris. -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.5.1 for non-commercial use iQA/AwUBODmmrHCyV2ve66GWEQLefgCdG49XQVnsd38qB6Hg4G+f6frgmaIAn2yq +njLwkjGSrNJBmDonmxBTgNK =iqSE -----END PGP SIGNATURE----- ------_=_NextPart_001_01BF3527.5DBEAD60 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable More findings and a wonderful wonderful little picture.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Well I have been using the unofficial (i think) 0.4.1 = version, the vfree problems have gone away, I still encounter a problem = with the reset camera code, it seams to have to do with the fatalerror = stuff that checks for a bug in the older firmware, if you just ignore = the error everything works fine, so it looks like the logic that = decideds to ignore or actuall fail is a little weird, I am looking into = that.  After I got over that hurddle, everything just worked = wounderfully.  I have a working web cam......    = WOOOOOO HOOOOOOOO.  I am currently developing and running it of my = Libretto 70CT which is a P120mmx and I get around 12fps using the PIO = Mode.  I have not yet ventured and tried DMA mode, that is next on = my list. 

For all to see. 
http://209.47.106.152

The camera will be up during bussiness hours (when I = am at work);
Please be gental with my laptop, it is just a baby = and our line here sucks crap.  It is only 128k ISDN

chris.

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 for non-commercial use = <http://www.pgp.com>

iQA/AwUBODmmrHCyV2ve66GWEQLefgCdG49XQVnsd38qB6Hg4G+f6frgmaIAn2y= q
+njLwkjGSrNJBmDonmxBTgNK
=3DiqSE
-----END PGP SIGNATURE-----

------_=_NextPart_001_01BF3527.5DBEAD60-- ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From webmaster@softgallery.com Tue, 23 Nov 1999 14:08:50 +0100 Date: Tue, 23 Nov 1999 14:08:50 +0100 From: J-F Mammet webmaster@softgallery.com Subject: mmap smp problem ? HI everybody. I'm new on this list but I have read most of the recent archive. I saw there is a mmap problem. Is it only with SMP ? I just crashed my machine at home (damn !) trying to use laucch camserv on my dual celeron. I did not see the mmap calls in a file. Is a fix in the way. I just saw 0.4.2 but I did not see any mmap related change in the patch. If I recompile my kernels/modules with SMP disabled, will it work ok ? Thanks a lot for the great work ! J-F Mammet webmaster@softgallery.fr ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From u950057@giki.edu.pk Tue, 23 Nov 1999 19:00:38 +0500 Date: Tue, 23 Nov 1999 19:00:38 +0500 From: Tasnim Ahmed u950057@giki.edu.pk Subject: Compilation for kernel 2.3.24 Hello: I am using kernel 2.3.24 and I tried to compile cpia-0.4.1 but the make kernel has some problems and I cant compile the module without doing make kernel. Secondly, the README file is too obselete which does not mentio any thing about the make kernel stuff. Please someone help me. Thanks, Tasnim debugger@bigfoot.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Tue, 23 Nov 1999 17:53:37 +0100 (CET) Date: Tue, 23 Nov 1999 17:53:37 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: mmap smp problem ? On 23-Nov-99 J-F Mammet wrote: > HI everybody. > > I'm new on this list but I have read most of the recent archive. I saw > there is a mmap problem. Is it only with SMP ? I just crashed my machine > at > home (damn !) trying to use laucch camserv on my dual celeron. I did not > see the mmap calls in a file. There is some warning in the README that mmap on SMP is a killer. Normal read operation, e.g. the gqcam-application, should work fine. Can you check that? > Is a fix in the way. I just saw 0.4.2 but I did not see any mmap related > change in the patch. Correct, there is no change. > If I recompile my kernels/modules with SMP disabled, will it work ok ? Yes and no. The mmap stuff was never working too well. However, I will release a 'working' version as soon as possible. The mmap-stuff was implemented as a asynchronous 'pseudo kernel-thread'. To make that stuff really working stable it has to be implemented as a real kernel thread so one can also schedule it reasonably in case of polling. Also to make it smp-save one has to serialize the stuff with wait-queues and the like. All in all that is a major rewrite that needs quite some thinking first. So I made the decision to implement mmap synchronous just as the normal read (I do have that at home, it just needs some cleanups). This will go to the first real beta-release that we should try to add to the 2.2.X kernel series sooner later. The drawbacks of this synchronous mmap-stuff: - no interrupt driven operation possible - therefore USB pointless - no gain whatsoever on SMP-hosts - maybe slower frame rate due to unneccessary waits Still I think this will be usefull enough for most of the normal users. After that I will rewrite the mmap-stuff to get some real practice with parallel kernel threads and SMP-machines. :) -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From gleicon@engebras.com Tue, 23 Nov 1999 15:33:42 -0200 Date: Tue, 23 Nov 1999 15:33:42 -0200 From: Gleicon S. Moraes gleicon@engebras.com Subject: mmap smp problem ? > > There is some warning in the README that mmap on SMP is a killer. Normal > read operation, e.g. the gqcam-application, should work fine. Can you > check that? is there a link pointing to a simple implementation of the read operation using cpia driver ? gqcam gave an ideia, but i want to make a layer to put in some v4l programs. The point is: i use bttv, and most of these programs use mmap to get the picture. Can i just set the image size and depth , and after thist, read the entire picture ? I just want to make those programs run with cpia driver. Thanks in advance, regards []'s Gleicon S. Moraes ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From ancar@tin.it Tue, 23 Nov 1999 20:45:56 +0100 Date: Tue, 23 Nov 1999 20:45:56 +0100 From: Antonio Carofiglio ancar@tin.it Subject: Device busy When I load (with modprobe -a ) the bw-qcam module, linux report me: "/lib/modules/2.2.12-20/bw-qcam.o: init_module: device or resource busy parport: device or resource busy". Do you have any idea about? All the test (modprobe, lsmod, etc.) in /client/README it's ok, but... (prg mcam works fine). thanks in advance Antonio ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jm.merliot@wanadoo.fr Tue, 23 Nov 1999 21:31:07 +0100 (CET) Date: Tue, 23 Nov 1999 21:31:07 +0100 (CET) From: Jean-Michel Merliot jm.merliot@wanadoo.fr Subject: CPIA-0.4.2 & SMP Hi everybody, I tried driver version 0.4.2 on my SMP box. It worked much much better than earlier version ( in fact, v3.1 worked pretty nice ). Gqcam worked perfectly. A few annoying things : - I got a message in syslog ( repeated many times ) : cpia_pp.c(7921713):Negotiate2SetupPhase(326):nFault should be low at beginning That just makes the hard drive noisy ( linearly with frame rate ;-) ) More disturbing things : The driver worked the first time in PIO mode, as I had forgotten to activate the IRQ in my parport settings. After quitting gqcam, I've been unable to start it again, IRQ activated or not. I unloaded the modules ( all of them), reloaded them, nothing made them work again ;-). That's what I got after activating the IRQ : Nov 23 21:07:51 linux kernel: cpia.c(7964331):cpia_open(2545):cpia_open Nov 23 21:07:51 linux kernel: cpia_pp.c(7964332):cpia_pp_open(1175):Found camera [0] Nov 23 21:07:51 linux kernel: cpia_pp.c(7964332):cpia_pp_irq_handler(974):got IRQ(7) when in CPIA_PHASE_idle Nov 23 21:07:51 linux kernel: cpia_pp.c(7964332):cpia_pp_irq_handler(979):image skipped? Nov 23 21:07:51 linux kernel: cpia.c(7964332):goto_low_power(2392):camera now in LOW power state When I tried to stgart gqcam again, I got an input/output error... Hope this may help... But I know this will be solved soon... E-Mail: Jean-Michel Merliot Linux and Blue Oyster Cult Fan Deadly Opponent To microsoft Misconception Of Computing Art... ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Tue, 23 Nov 1999 21:44:15 +0100 (CET) Date: Tue, 23 Nov 1999 21:44:15 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: mmap smp problem ? On 23-Nov-99 Gleicon S. Moraes wrote: >> >> There is some warning in the README that mmap on SMP is a killer. Normal >> read operation, e.g. the gqcam-application, should work fine. Can you >> check that? > > is there a link pointing to a simple implementation of the read operation > using cpia > driver ? gqcam gave an ideia, but i want to make a layer to put in some v4l > programs. The > point is: i use bttv, and most of these programs use mmap to get the picture. > Can i just set the image size and depth , and after thist, read the entire > picture ? > I just want to make those programs run with cpia driver. That is exactly what the 'synchronous' mmap-implementation I was talking about does; but at driver level and not at applocation level. It maps the mmap-calls to the same routines used for read calls (ioctl-CMCAPTURE will actually read the frame, ioctl-CSYNC will just mark the buffer free). So old v4l-applications will work (e.g. xawtv, webcam, w3cam do work on my test box). For new applications use the read-interface as implemented in the gqcam. Memory mapping does not make that much sense (compared to video capture boards as the bttv) for a parallel port webcam unless we get interrupt driven operation to work.z Just stay tuned for a few more days. There are some small problems with exposure control and stability during crash tests (e.g. unplugging the camera during read should not crash the box). I will get rid of those over the weekend and release the stuff afterwards. -Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Tue, 23 Nov 1999 16:37:06 -0500 (EST) Date: Tue, 23 Nov 1999 16:37:06 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Compilation for kernel 2.3.24 On 23 Nov, Tasnim Ahmed wrote: > Hello: > > I am using kernel 2.3.24 and I tried to compile cpia-0.4.1 but the make kernel > has some problems and I cant compile the module without doing make kernel. The driver does not work with 2.3 kernels at this time, but I'll be working on it this weekend. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From webmaster@softgallery.com Wed, 24 Nov 1999 10:03:45 +0100 Date: Wed, 24 Nov 1999 10:03:45 +0100 From: J-F Mammet webmaster@softgallery.com Subject: mmap smp problem ? Ok I looked all the Emails about the mmap problem. I just recompiled my kernel without SMP and mmap works fine, so it's definitely a SMP problem, like everybody said, but I like to check by myself. I'm not competent enough to write some threaded mmap functions for the driver so I'll have to wait for the corrected one 8). Do you know some image grabber (I want to do a webcam) that uses a simple read for a v4l device instead of mmap like gqcam does ? Having a dual celeron and using a non SMP kernel is a little bit annoying ;-) Best Regards. J-F Mammet webmaster@softgallery.fr ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Wed, 24 Nov 1999 10:32:26 +0100 (CET) Date: Wed, 24 Nov 1999 10:32:26 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: mmap smp problem ? On 24-Nov-99 J-F Mammet wrote: > > Do you know some image grabber (I want to do a webcam) that uses a > simple read for a v4l device instead of mmap like gqcam does ? Having a > dual celeron and using a non SMP kernel is a little bit annoying ;-) Nope, I will send you a patch in the evening. You have the honour to beta-test. :) -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From webmaster@softgallery.fr Wed, 24 Nov 1999 10:40:38 +0100 Date: Wed, 24 Nov 1999 10:40:38 +0100 From: J-F Mammet webmaster@softgallery.fr Subject: mmap smp problem ? >Nope, I will send you a patch in the evening. You have the honour to >beta-test. :) Thanks It will be a pleasure to destroy an ext3 partition for you. Ext3 is great for testing some kernel killing applications because there is no fsck to do after rebooting. Great stuff ! J-F Mammet webmaster@softgallery.fr ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Wed, 24 Nov 1999 11:20:02 +0100 Date: Wed, 24 Nov 1999 11:20:02 +0100 From: Bas Rijniersce bas@brijn.nu Subject: mmap smp problem ? Hi, >> Do you know some image grabber (I want to do a webcam) that uses a >> simple read for a v4l device instead of mmap like gqcam does ? Having a >> dual celeron and using a non SMP kernel is a little bit annoying ;-) >Nope, I will send you a patch in the evening. You have the honour to >beta-test. :) I would be happy to give it a try as well :-) >-Peter ---- Bas Rijniersce bas@brijn.nu Tel. +31 (0)341-550545 http://www.brijn.nu Fax. +31 (0)341-562940 ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 25 Nov 1999 15:29:36 +0100 (CET) Date: Thu, 25 Nov 1999 15:29:36 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: backlight and filters - algorithms needed Hi all, I recently compared the output of our driver with the m$-stuff and found the following obvious difference. There is a 'backlight' option and some 'filters (50/60Hz)' that one can switch on/off in the m$-video panel. The first one is usefull in dark scenes. In case of compressed mode the image looks less grainy. The second one eliminates horizontal stripes that seem to occur during exposure changes. Has anyone an idea what kind of image processing filters those are? -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From madsdyd@challenge.dk Thu, 25 Nov 1999 16:43:15 +0100 (CET) Date: Thu, 25 Nov 1999 16:43:15 +0100 (CET) From: Mads Bondo Dydensborg madsdyd@challenge.dk Subject: backlight and filters - algorithms needed On Thu, 25 Nov 1999, Peter Pregler wrote: > Hi all, > > I recently compared the output of our driver with the m$-stuff and found > the following obvious difference. There is a 'backlight' option and some > 'filters (50/60Hz)' that one can switch on/off in the m$-video panel. The > first one is usefull in dark scenes. In case of compressed mode the image > looks less grainy. Compensation gains? > The second one eliminates horizontal stripes that seem > to occur during exposure changes. I think this would be the flicker control. It is documented in the "Notes on use of the CPiA command set", p. 3-34, section 3.12.1 ff. > Has anyone an idea what kind of image > processing filters those are? My best guess is above. Mads -- Mads Bondo Dydensborg. madsdyd@challenge.dk 7) Sacrifice a small animal for Good Luck (step 7 of 8 in configuring the Open Source Global File System for Linux) ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Henry.BRUCE@st.com Thu, 25 Nov 1999 15:52:42 GMT Date: Thu, 25 Nov 1999 15:52:42 GMT From: Henry Bruce Henry.BRUCE@st.com Subject: backlight and filters - algorithms needed > I recently compared the output of our driver with the m$-stuff and found > the following obvious difference. There is a 'backlight' option and some > 'filters (50/60Hz)' that one can switch on/off in the m$-video panel. The > first one is useful in dark scenes. In case of compressed mode the image > looks less grainy. The second one eliminates horizontal stripes that seem > to occur during exposure changes. Has anyone an idea what kind of image > processing filters those are? These are no filters as such, just variations in how the automatic exposure controller (AEC) can be configured from the driver. Note that the AEC runs in firmware, so you have limited control on the way it operates The backlight option (as in other cameras) doesn't use the full image to calculate the exposure value - just the bottom part of it. As you would expect, this option should be used when the subject is lit from behind. The horizontal stripes are caused by mains lighting and are avoided, rather than cancelled. To do this, the driver must know if it is working in a 50 or 60Hz environment. Details of how to set this up are in the developer's guide, but it is a tricky business. Hope this helps. Henry Bruce ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Thu, 25 Nov 1999 17:34:03 +0100 Date: Thu, 25 Nov 1999 17:34:03 +0100 From: Bas Rijniersce bas@brijn.nu Subject: Peters patch Hi, Maybe I misunderstood, but the patch was against 0.4.1? I did get some reject, but it compiled anyway :-) And it works great! Vidcat from w3cam no longer crashes by machine! I saw there is another 0.4.2 version on the webpage with some patches from Scott, that 0.4.2 doesn't have your patch Peter? If you want me to try some more programs... just ask :) Bas ---- Bas Rijniersce Phone +31 341 550545 Oude Telgterweg 81 Fax +31 341 562940 3851 EA Ermelo http://www.brijn.nu The Netherlands bas@brijn.nu ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 25 Nov 1999 17:44:37 +0100 (CET) Date: Thu, 25 Nov 1999 17:44:37 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: backlight and filters - algorithms needed On 25-Nov-99 Mads Bondo Dydensborg wrote: > On Thu, 25 Nov 1999, Peter Pregler wrote: > >> The second one eliminates horizontal stripes that seem >> to occur during exposure changes. > > I think this would be the flicker control. It is documented in the > "Notes on use of the CPiA command set", p. 3-34, section 3.12.1 ff. No, flicker control is to prevent rapid exposure changes between frames not 'stripes' within a frame. At least that is what I thought it is. :) -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Thu, 25 Nov 1999 12:08:04 -0500 (EST) Date: Thu, 25 Nov 1999 12:08:04 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: camera control program Starting with version 0.4.2, the /proc interface should be stable. The only changes I can see coming are bug fixes and better values for the limits of some of the parameters. If someone would like to write a video-control window for the camera to change the settings via the /proc interface we can drop the ioctl-kludge that is currently used. Scott J. Bertin sbertin@mindspring.com PS: If I have some of the limits wrong, please let me know what the correct ones are so I can fix it. ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Thu, 25 Nov 1999 12:08:41 -0500 (EST) Date: Thu, 25 Nov 1999 12:08:41 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: backlight and filters - algorithms needed On 25 Nov, Henry Bruce wrote: > The horizontal stripes are caused by mains lighting and are avoided, rather > than cancelled. To do this, the driver must know if it is working in a 50 or > 60Hz environment. Details of how to set this up are in the developer's guide, > but it is a tricky business. 0.4.2 contains code to turn the flicker control on/off via the /proc interface. I haven't tested it (I have incandescent lights) and I'm sure it need more work for 1-02 firmware cameras, but it is there for people to try. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 25 Nov 1999 21:39:37 +0100 (CET) Date: Thu, 25 Nov 1999 21:39:37 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Peters patch On 25-Nov-99 Bas Rijniersce wrote: > Hi, > > Maybe I misunderstood, but the patch was against 0.4.1? I did get some > reject, but it compiled anyway :-) And it works great! Vidcat from w3cam > no longer crashes by machine! Yep, it was against 0.4.1 by accident. I missed Scotts 0.4.2 patch by accident. > I saw there is another 0.4.2 version on the webpage with some patches > from Scott, that 0.4.2 doesn't have your patch Peter? Yep, I will send a revised patch vs. that 0.4.2 to Jarl to put it on the web as version 0.4.3. This is now close to all I want to do for the driver for now. Please give it extensive testing. My plans for the near future: Things that should work now: - mmap interface now implemented as synchronous read - DMA read has now timeout to prevent lockup if transfer fails - dirty reset_camera patch to handle systemState 0x4 - check for skipsize in compressed image data since we cannot trust what we get from the camera - SMP-safe - works with w3cam/xawtv/gqcam/webcam/isabel-beta (partially) and maybe more, if you know of any other v4l-application that makes sense with a webcam let me know Things that might be added by me before 1.0: - VIDEO_PALETTE_YUVU422, ... needed by vic - gqcam patch for gtk-1.1 backwards compatibility - remove ioctl-kludge used by gqcam, this is better done with the proc-interface - kwintv does not work, but I guess that's the application not the driver Things to do for a later release: - SMP-safe streaming with kernel thread - thruput optimizations - overlay support - v4l2 ;) I guess Scott has some more things in his mind. And it would be really nice if someone wrote a camera control application. That's just reading and writing strings from/to a file and a fency GUI. Not that hard. :) Greetings, Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Fri, 26 Nov 1999 10:32:11 -0500 (EST) Date: Fri, 26 Nov 1999 10:32:11 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Peters patch On 25 Nov, Peter Pregler wrote: > I guess Scott has some more things in his mind. Here are some more things I'd like to see added to the driver, nothing that is critical for 1.0 however. - USB support - nibble mode support - handle decimation in the conversion code - handle YUV420 data from the camera - add VIDEO_TYPE_SUBCAPTURE support If anybody wants to do some of these, please jump in. You don't need to be an expert kernel programmer to do this stuff. I'm not. USB support definately needs someone to work on it. I don't have a USB camera, so can't do it myself. The driver currently won't compile for 2.3 kernels, but I'm working on that right now. If you want to do USB support before I release these changes, please contact me and I'll give you what I've gort so far. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From debugger@bigfoot.com Fri, 26 Nov 1999 21:49:42 +0500 Date: Fri, 26 Nov 1999 21:49:42 +0500 From: Tasnim Ahmed debugger@bigfoot.com Subject: the beginning... Hi: I finally got a 2.2 kernel and compiled the module. BUT, I got 4-5 fps in CIF mode with auto compression. then I tried to put my camera in DMA mode which I finally did but performance was not increased, however bands are introduced into the picture. The best I got on my RH 6.0 2.2 Celeron 333a 64 MB system is 10 fps in QCIF and thats when nothing is moving in front of the camera. Now I think that this is the beginning of the driver so we can expect some real improvement in the future. I am waiting for the 2.3 support because my other hardware performs better under 2.3 so may be this follows the trend. anyways, I appreciate all the great work done by you guys out there. --- Tasnim Ahmed debugger@bigfoot.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Thu, 25 Nov 1999 21:47:51 +0100 (CET) Date: Thu, 25 Nov 1999 21:47:51 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Peters patch On 25-Nov-99 Peter Pregler wrote: > > Things that might be added by me before 1.0: > ... > - gqcam patch for gtk-1.1 backwards compatibility Ooops, that one was already added. Never trust a cut-and-paste. :( -Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Sat, 27 Nov 1999 13:15:54 +0100 (CET) Date: Sat, 27 Nov 1999 13:15:54 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: the beginning... On 26-Nov-99 Tasnim Ahmed wrote: > I got 4-5 fps in CIF mode with auto compression. then I tried to put my > camera > in DMA mode which I finally did but performance was not increased, however > bands are introduced into the picture. > > The best I got on my RH 6.0 2.2 Celeron 333a 64 MB system is 10 fps in QCIF > and > thats when nothing is moving in front of the camera. This is really _not_ normal and should have nothing to do with 2.2/2.3. Please send me the related part of your syslog, the /proc/cpia, /proc/parport, /proc/interrupts, /proc/dma and all other information regarding your parport/webcam hardware and the test you did (program your run, load on the machine etc.). You should really get about 30pfs in QCIF autocompression mode depending on the camera settings (targetting, sensor frame rate, ...). At least I do get it. :) -Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Nick.Holloway@alfie.demon.co.uk 27 Nov 1999 17:06:14 -0000 Date: 27 Nov 1999 17:06:14 -0000 From: Nick Holloway Nick.Holloway@alfie.demon.co.uk Subject: Peters patch sbertin@mindspring.com writes: > Here are some more things I'd like to see added to the driver, nothing > that is critical for 1.0 however. A minor nit is the way that the cpia core requests the cpia_usb and cpia_pp modules. This isn't required. If you have the modules in /lib/modules, and your dependencies are up to date, then "modprobe cpia_pp" will also load "videodev" and "cpia". If the webcam is your only v4l device, then this in /etc/modules.conf will cause it to be autoloaded: alias char-major-81 cpia_pp When trying the driver out, I run "insmod cpia" by hand from the compilation directory. As it currently stands, this then leads to an attempt to load the lowlevel modules -- which just causes errors to be syslogged. For the non-modular case, instead of cpia_init being added to videodev.c, and that being used to initialise the lowlevel driver, there should be calls to cpia_pp_init and cpia_usb_init as appropriate. -- `O O' | Nick.Holloway@alfie.demon.co.uk // ^ \\ | http://www.alfie.demon.co.uk/ ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From lists@sprawl.net Sat, 27 Nov 1999 15:26:30 -0500 Date: Sat, 27 Nov 1999 15:26:30 -0500 From: Adam Lynch lists@sprawl.net Subject: Lots of fun with w3cam and friends Hi all. Just found out about the CPiA driver yesterday, and I'm pretty psyched. The old Creative webcam lives! Anyway, I snagged the current achive (0.4.3), and went through installing the driver. Went fine, modules loaded fine with no errors or warnings. As did the gqcam client supplied with it. So I fired up gqcam, and had a jolly 'ole time. Now, my problem lies with other V4L packages. w3cam especially. w3cam seems to be able do what I want to be able to do; that is, provide a simple online webcam of my office. So I go out and grab the current distribution of w3cam (0.6.2). This compiles and installes fine. However, when w3cam's CGI goes to grab an image, it tells me it can't open /dev/video. As does vidcat after grabbing 1 image. And following this, gqcam gets this error: /dev/video: Input/output error The only way to get the whole setup to work again is to power off, then back on. Doesn't lock up my machine, just the V4L interface is wacky. Any ideas? The system is a PPro 200, 64mb of RAM, kernel 2.2.5... Thanks! -- --- AdamL. alynch@sprawl.net http://sprawl.net Never attribute to malice that which can be adequately explained by stupidity. -Robert Heinlein, 'Logic of Empire' Any view of the universe that is not strange is false. -Neil Gaiman ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From jfm@softgallery.com Sat, 27 Nov 1999 21:58:39 +0100 Date: Sat, 27 Nov 1999 21:58:39 +0100 From: J-F Mammet jfm@softgallery.com Subject: Lots of fun with w3cam and friends Hi ! w3cam runs as nobody, so do a chmod 777 /dev/video and will be ok. However I only get a blank image with it. Any idea ?? J-F Mammet webmater@softgallery.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From sbertin@mindspring.com Sat, 27 Nov 1999 19:49:58 -0500 (EST) Date: Sat, 27 Nov 1999 19:49:58 -0500 (EST) From: sbertin@mindspring.com sbertin@mindspring.com Subject: Peters patch On 27 Nov, Nick Holloway wrote: > A minor nit is the way that the cpia core requests the cpia_usb and > cpia_pp modules. This isn't required. You're right, it isn't required. It is a convenience, as is all module auto loading. > If you have the modules in /lib/modules, and your dependencies are up > to date, then "modprobe cpia_pp" will also load "videodev" and "cpia". Yes, but if someone has both parallel and USB cameras, then they must load 2 modules. The current scheme allows loading just one module and getting support for all cpia based cameras on the system. > For the non-modular case, instead of cpia_init being added to videodev.c, > and that being used to initialise the lowlevel driver, there should be > calls to cpia_pp_init and cpia_usb_init as appropriate. I disagree. The way the driver is structured, cpia_pp and cpia_usb are simply transports for use by cpia. Only cpia needs to know about videodev, and videodev only needs to know about cpia. Videodev calling cpia_pp_init and cpia_usb_init would not be appropriate. Scott J. Bertin sbertin@mindspring.com ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Nick.Holloway@alfie.demon.co.uk 28 Nov 1999 09:55:19 -0000 Date: 28 Nov 1999 09:55:19 -0000 From: Nick Holloway Nick.Holloway@alfie.demon.co.uk Subject: Divide error in fetch_frame when enabling compression (0.4.3) To calculate the transfer rate, the following is used within fetch_frame: rate = image_size * HZ / 1024 / (jiffies-oldjif); When compression is enabled, it would appear that you can still be on the same jiffy, and so you get a divide error (which leads to a SEGV being sent to the application, and a reboot before further use of the device is allowed). As I didn't care about the accuracy, I just added a "+1" to the jiffy difference. This confirmed that this was the problem. Alternatively, treat a zero difference observed as 1/2 jiffy: oldjif = jiffies-oldjif; rate = image_size * HZ / 512 / (oldjif ? 2*oldjif : 1 ); -- `O O' | Nick.Holloway@alfie.demon.co.uk // ^ \\ | http://www.alfie.demon.co.uk/ ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Nick.Holloway@alfie.demon.co.uk 28 Nov 1999 12:13:11 -0000 Date: 28 Nov 1999 12:13:11 -0000 From: Nick Holloway Nick.Holloway@alfie.demon.co.uk Subject: Lots of fun with w3cam and friends jfm@softgallery.com (J-F Mammet) writes: > w3cam runs as nobody, so do a chmod 777 /dev/video and will be ok. > However I only get a blank image with it. Any idea ?? I think this is due to the auto-exposure getting reset (at least that is what it looks like). When you restart gqcam, with the camera pointing outside, you can see the auto-exposure. The picture starts off white, and adjusts so you see a reasonable picture. With 0.4.1, you could see that every other time, the picture would reset to white, and recalibrate. This meant that you could get vidcat to grab a usable picture once after gqcam had run. The next time it would come out white. With 0.4.3, the reset appears to happen every time. So, when used indoors, you get a picture (from the default exposure), but outside you get just white. With single frame captures, I don't think just using the last exposure will be valid. I think waiting until the auto-exposure has settled down before grabbing the frame would be valid. I don't know if you can always tell the difference between a single frame grab, and a stream grab. It appears that "vidcat" uses capture_frame, and "gqcam" doesn't, but I don't know if this is universally true. Secondly, my attempts to look at the values from GetExposure in capture_frame didn't reveal the values changing with time. -- `O O' | Nick.Holloway@alfie.demon.co.uk // ^ \\ | http://www.alfie.demon.co.uk/ ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Sun, 28 Nov 1999 13:24:46 +0100 (CET) Date: Sun, 28 Nov 1999 13:24:46 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Lots of fun with w3cam and friends On 27-Nov-99 Adam Lynch wrote: > w3cam seems to be able do what I want to be able to do; that is, provide a > simple online webcam of my office. So I go out and grab the current > distribution > of w3cam (0.6.2). > > This compiles and installes fine. However, when w3cam's CGI goes to grab an > image, it tells me it can't open /dev/video. As does vidcat after grabbing 1 > image. > > And following this, gqcam gets this error: > > /dev/video: Input/output error Hmmm, I just grabbed the 0.6.2 version (used 0.6.1 for my testing) and had no problem after the upgrade. It still works Also my test-webcam at http://lugburz.risc.uni-linz.ac.at is running w3cam 0.6.1 rather stable now since the release of the 0.4.3 driver. Can you send the relevant syslog output, the proc-stuff and everything else that seems relevant :) directly to me. I'll try to fix things. -Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Sun, 28 Nov 1999 18:54:09 +0100 (CET) Date: Sun, 28 Nov 1999 18:54:09 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: Lots of fun with w3cam and friends On 28-Nov-99 Nick Holloway wrote: > > With 0.4.3, the reset appears to happen every time. So, when used > indoors, you get a picture (from the default exposure), but outside you > get just white. This reset is a bug I am aware of. The driver is meant to save the exposure and color-balance state. It does save the state, but there is something wrong with the reinitiasation of the exposure. > With single frame captures, I don't think just using the last exposure > will be valid. I think waiting until the auto-exposure has settled down > before grabbing the frame would be valid. I think this would really kill any performance and there is no API to control this. My solution to this problem was to save the current camera state and use that as the starting point for the new capture. I have to test it but I think if implemented correctly this should result in auto-exposure taking place even in single frame capture mode, just a bit slower. Of cause any other reasonable ideas are highly welcome. > I don't know if you can always tell the difference between a single frame > grab, and a stream grab. It appears that "vidcat" uses capture_frame, > and "gqcam" doesn't, but I don't know if this is universally true. Nope, right now the camera is used in just one way, single frame grab mode. -Peter ------------------------------- Even the thought that a ship was waiting to take him back to Earth did not wipe out the sense of loss he felt at that moment. ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at Peter Pregler / RISC, University of Linz, Austria ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Peter.Pregler@risc.uni-linz.ac.at Mon, 29 Nov 1999 12:04:19 +0100 (CET) Date: Mon, 29 Nov 1999 12:04:19 +0100 (CET) From: Peter Pregler Peter.Pregler@risc.uni-linz.ac.at Subject: VIC - testers needed This message is in MIME format --_=XFMail.1.4.2.Linux:19991129120419:25817=_ Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable On 04-Nov-1999 Jose Serrano Rodenas wrote: >=20 > After that I closedthe gqcam aplication, and I run a v4l aware > aplication (vic v2.8ucl4) > but when I begin to transmit with the cpcia driver the image appear in > green and > in the kernel appear the following message: >=20 > ppcpia:streamtoyuv.c:stream2rgb24c(121):line length was not 1 but -2 > after 19887/203649 > bytes > Do you know the cause of this behaviouf? Please apply the attached patch to the latest driver (0.4.3) and try it out. I have vic installed as a plain debian-potato package (2.8ucl4-1). It is running with that setup. You might be able toconnect to lugburz.risc.unilinz.ac.at/8000. There should be a 128KB stream. Unfortunatly I cannot connect myself to it. Has anyone an idea how to do that? Or is multicasting needed? -Peter ------------------------------- Email: Peter.Pregler@risc.uni-linz.ac.at WWW: http://www.risc.uni-linz.ac.at/people/ppregler --_=XFMail.1.4.2.Linux:19991129120419:25817=_ Content-Disposition: attachment; filename="vic-patch" Content-Description: vic-patch Content-Type: application/octet-stream; name=vic-patch; SizeOnDisk=1187 Content-Transfer-Encoding: base64 KioqIC4uLy4uL2NwaWEtMC40LjMvbW9kdWxlL2NwaWEuYwlUaHUgTm92IDI1IDE0OjAyOjIyIDE5 OTkKLS0tIGNwaWEuYwlNb24gTm92IDI5IDEwOjQzOjA3IDE5OTkKKioqKioqKioqKioqKioqCioq KiAxNDUxLDE0NTYgKioqKgotLS0gMTQ1MSwxNDU3IC0tLS0KICAJICAgICAgIChwYWxldHRlID09 IFZJREVPX1BBTEVUVEVfUkdCNTY1ICYmIGRlcHRoID09IDE2KSB8fAogIAkgICAgICAgKHBhbGV0 dGUgPT0gVklERU9fUEFMRVRURV9SR0IyNCAmJiBkZXB0aCA9PSAyNCkgfHwKICAJICAgICAgIChw YWxldHRlID09IFZJREVPX1BBTEVUVEVfUkdCMzIgJiYgZGVwdGggPT0gMzIpIHx8CisgCSAgICAg ICAocGFsZXR0ZSA9PSBWSURFT19QQUxFVFRFX1lVVjQyMiAmJiBkZXB0aCA9PSAxNikgfHwKICAJ ICAgICAgIChwYWxldHRlID09IFZJREVPX1BBTEVUVEVfWVVZViAmJiBkZXB0aCA9PSAxNikgfHwK ICAJICAgICAgIChwYWxldHRlID09IFZJREVPX1BBTEVUVEVfVVlWWSAmJiBkZXB0aCA9PSAxNik7 CiAgfQoqKioqKioqKioqKioqKioKKioqIDE4NjQsMTg2OSAqKioqCi0tLSAxODY1LDE4NzEgLS0t LQogIAkJKnJnYisrID0geTsKICAJCSpyZ2IgPSB5MTsKICAJCXJldHVybiAyOworIAljYXNlIFZJ REVPX1BBTEVUVEVfWVVWNDIyOgogIAljYXNlIFZJREVPX1BBTEVUVEVfWVVZVjoKICAJCSpyZ2Ir KyA9IHk7CiAgCQkqcmdiKysgPSB1OwoqKioqKioqKioqKioqKioKKioqIDE4ODcsMTg5MiAqKioq Ci0tLSAxODg5LDE4OTUgLS0tLQogIAkJcmV0dXJuIGNvdW50OwogIAljYXNlIFZJREVPX1BBTEVU VEVfUkdCNTU1OgogIAljYXNlIFZJREVPX1BBTEVUVEVfUkdCNTY1OgorIAljYXNlIFZJREVPX1BB TEVUVEVfWVVWNDIyOgogIAljYXNlIFZJREVPX1BBTEVUVEVfWVVZVjoKICAJY2FzZSBWSURFT19Q QUxFVFRFX1VZVlk6CiAgCQlyZXR1cm4gMipjb3VudDsKKioqKioqKioqKioqKioqCioqKiAyOTYw LDI5NjUgKioqKgotLS0gMjk2MywyOTY5IC0tLS0KICAJCQlicmVhazsKICAJCWNhc2UgVklERU9f UEFMRVRURV9SR0I1NTU6CiAgCQljYXNlIFZJREVPX1BBTEVUVEVfUkdCNTY1OgorIAkJY2FzZSBW SURFT19QQUxFVFRFX1lVVjQyMjoKICAJCWNhc2UgVklERU9fUEFMRVRURV9ZVVlWOgogIAkJY2Fz ZSBWSURFT19QQUxFVFRFX1VZVlk6CiAgCQkJY2FtLT52cC5kZXB0aCA9IDE2Owo= --_=XFMail.1.4.2.Linux:19991129120419:25817=_-- End of MIME message ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Chris@informinteractive.com Mon, 29 Nov 1999 11:40:55 -0500 Date: Mon, 29 Nov 1999 11:40:55 -0500 From: Chris Whiteford Chris@informinteractive.com Subject: /dev/video: Input/output error This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF3A88.8447D720 Content-Type: text/plain; charset="iso-8859-1" -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 There still seams to be some problems with the 1.02 FW version and the 2 latest driver versions. The 0.4.2 seams a bit better with the 1.02 FW but I can still cause a /dev/video: Input/output error when trying to load the gqcam program. I basically open and close the program a few times and then the error pops. I give the program about 10s of time to capture and display then I close it, then I try and repeat. With the 0.4.3 version it seams to happen after 1 cycle of opening, closing, repeat. With 0.4.2 it takes longer, and it does not seam to be a fixed number. As always I will be digging into the driver later tonight and I should have some line numbers and better debuging output then. chris. -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.5.1 for non-commercial use iQA/AwUBOEKtN3CyV2ve66GWEQJDeACgxV0iN/Mww/rqgD0DJ8thH60ZwnoAn3Wb ODlWqT3Sz0Q/1fRLlM4Yfq0q =GIo6 -----END PGP SIGNATURE----- ------_=_NextPart_001_01BF3A88.8447D720 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable /dev/video: Input/output error

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

There still seams to be some problems with the 1.02 = FW version and the 2 latest driver versions.  The 0.4.2 seams a = bit better with the 1.02 FW but I can still cause a /dev/video: = Input/output error when trying to load the gqcam program.  I = basically open and close the program a few times and then the error = pops.  I give the program about 10s of time to capture and display = then I close it, then I try and repeat.  With the 0.4.3 version it = seams to happen after 1 cycle of opening, closing, repeat.  With = 0.4.2 it takes longer, and it does not seam to be a fixed number.  =

As always I will be digging into the driver later = tonight and I should have some line numbers and better debuging output = then.

chris.

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.1 for non-commercial use = <http://www.pgp.com>

iQA/AwUBOEKtN3CyV2ve66GWEQJDeACgxV0iN/Mww/rqgD0DJ8thH60ZwnoAn3W= b
ODlWqT3Sz0Q/1fRLlM4Yfq0q
=3DGIo6
-----END PGP SIGNATURE-----

------_=_NextPart_001_01BF3A88.8447D720-- ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Tue, 30 Nov 1999 11:55:27 +0100 Date: Tue, 30 Nov 1999 11:55:27 +0100 From: Bas Rijniersce bas@brijn.nu Subject: Off-topic: How to persuade a company to release specs Hi, It's a bit off-topic, but anyway. For learning C, the cpia code is a bit to complex for me to follow :-) So I started writing a driver for a simple digital camera (Jenoptik JD11e, 640x480, serial port). From captured communications under Windows I can talk to the camera and get the same answers the Windows driver gets. But most of them don't mean anything to me. I wrote a short message to Jenoptik with the question if they would be willing to release protocol specs. They didn't. Now I want to send a bit longer message explaining the benefits of open protocol descriptions etc. Jarl, since you had succes with VLSI, what arguments did you use? Are there any good examples on the net? Are there other people who have experience in this. TIA, Bas ---- Bas Rijniersce bas@brijn.nu Tel. +31 (0)341-550545 http://www.brijn.nu Fax. +31 (0)341-562940 ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From bas@brijn.nu Tue, 30 Nov 1999 11:57:25 +0100 Date: Tue, 30 Nov 1999 11:57:25 +0100 From: Bas Rijniersce bas@brijn.nu Subject: Off-topic: How to persuade a company to release specs Hi, It's a bit off-topic, but anyway. For learning C, the cpia code is a bit to complex for me to follow :-) So I started writing a driver for a simple digital camera (Jenoptik JD11e, 640x480, serial port). From captured communications under Windows I can talk to the camera and get the same answers the Windows driver gets. But most of them don't mean anything to me. I wrote a short message to Jenoptik with the question if they would be willing to release protocol specs. They didn't. Now I want to send a bit longer message explaining the benefits of open protocol descriptions etc. Jarl, since you had succes with VLSI, what arguments did you use? Are there any good examples on the net? Are there other people who have experience in this. TIA, Bas ---- Bas Rijniersce bas@brijn.nu Tel. +31 (0)341-550545 http://www.brijn.nu Fax. +31 (0)341-562940 ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no". From Jarl.Totland@bdc.no Tue, 30 Nov 1999 12:22:33 +0100 Date: Tue, 30 Nov 1999 12:22:33 +0100 From: Jarl Totland Jarl.Totland@bdc.no Subject: Off-topic: How to persuade a company to release specs At 30.11.1999 11:55 +0100, Bas wrote: >Jarl, since you had succes with VLSI, what arguments did you use? Are there >any good examples on the net? Are there other people who have experience in >this. You might want to check out the Commercial Port Advocacy HOWTO at http://www.seul.org/pub/howto/cpah.html. That one is about software, should nevertheless be interesting... It also links to more resources for Linux Advocacy. Also, if you go to my page, the very first post in the news archive has a link to a page with all my early correspondence with Vision. Obviously, other people were bugging them to, still I'd like to think my letters mattered ;-) Good Luck! Vyrdsamt, Jarl Totland http://home.eunet.no/~jtotland ----------------------------------------------------------------------------- To unsubscribe from this mailinglist, send the line "unsubscribe vision-webcam" in the body of a message to "majordomo@errors.no".