Well not so fast... I was using the "stable" 10.0.12 pwc driver for a several time, and I never experimented that issue. Today I've tested it with the (old) 10.0.12 again (nowadays I use the most recent snapshot), and the issue now appears!, with the actual aMSN. So I think it's not a pwc driver bug, because in the past (when I was using stable pwc), this bug was not present, but now, with the actual aMSN, it is (using the same stable pwc driver I've used in the past).
On Sunday 28 May 2006 13:30, Vivia Nikolaidou wrote: > Always works perfectly with ov511, tested thousands of times on my > side. Another person with spca5xx also confirmed that webcam is > released correctly. > > So let's just assume it's a pwc bug? :) > > On 5/28/06, Harry Vennik <[EMAIL PROTECTED]> wrote: > > About 'webcam not released': I can confirm this is true, and have already > > searched through the code for the reason why this happens. It *might* be > > related to the change from getimage() to nextframe(), but I am not at all > > sure. Because all necessary calls seem to be done correctly, which is > > approved by the fact that we have not had this complaint often yet (so > > apparently things work well with other webcam drivers), I think it is a > > pwc driver bug (not the first one eh...). > > > > I proofed that there is actually done a close() on the device: > > > > vid-open: trying: v4l2... > > v4l2: open > > v4l2: init > > v4l2: device info: > > pwc 10.0.11 / Logitech QuickCam Pro 4000 @ usb-0000:00:10.2-2 > > v4l2: close > > vid-open: ok: v4l2 > > vid-open: flags: 2 > > v4l2: open > > ng_dev_open: opened Logitech QuickCam Pro 4000 [refcount 1] > > v4l2: new capture params (320x240, YU12, 115200 byte) > > convert-in : 320x240 12 bit YUV 4:2:0 (planar) > > convert-out: 320x240 24 bit TrueColor (BE: rgb) > > v4l2: buf 0: video-cap 0x0+460800, used 115200 > > v4l2: buf 1: video-cap 0x71000+460800, used 115200 > > v4l2: start ts=1148814415628311000 > > v4l2: buf 0: video-cap 0x0+115200, used 115200 > > v4l2: buf 1: video-cap 0x71000+115200, used 115200 > > v4l2: close > > ng_dev_close: closed Logitech QuickCam Pro 4000 [refcount 0] > > v4l2: fini > > > > The line of code printing 'v4l2: close' is right before the actual > > close() in the code. > > > > Op zondag 28 mei 2006 11:54, schreef Sander Hoentjen: > > > corellian if I stop sending my webcam (using pwc driver), aMSN still > > > reads or uses /dev/v4l/video0 corellian until I reset aMSN > > > corellian that means, I only can send webcam one time, then I have to > > > reset aMSN to send to another person, or send again to the same person. > > > corellian that's the 'lsof' output, when I'm not sending webcam, but > > > I've used it: corellian wish 29227 corellian mem CHR > > > 81,0 > > > 44270654 /dev/v4l/video0 corellian wish 29227 corellian mem > > > REG 3,8 73472 845256 > > > /home/corellian/amsn/utils/linux/capture/libng/plugins/drv0-v4l2.so > > > corellian wish 29227 corellian mem REG 3,8 44960 > > > 845259 > > > /home/corellian/amsn/utils/linux/capture/libng/plugins/drv1-v4l.so > > > corellian wish 29227 corellian mem REG 3,8 85852 > > > 845236 /home/corellian/amsn/utils/webcamsn/webcamsn.so * march1980 has > > > quit (Read error: 110 (Connection timed out)) > > > corellian and, when I send and receive, sometimes says that's rejected, > > > but that's not rejected/cancelled! corellian well I hope any devel will > > > read this :S > > > > _______________________________________________ > > Amsn-devel mailing list > > Amsn-devel@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/amsn-devel -- GPG Key-ID: 0x564903FA - JID: [EMAIL PROTECTED] _______________________________________________ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
