From guichaz at yahoo.fr Tue Apr 6 17:50:49 2004 From: guichaz at yahoo.fr (Guillaume Chazarain) Date: Tue, 6 Apr 2004 19:50:49 +0200 Subject: How to launch a second X server Message-ID: <20040406195049.705052e1.guichaz@yahoo.fr> Hi, I have a Voodoo Banshee (16 Mo), so if I want to have OpenGL acceleration, I have to use a depth of 16 bits, but if I want to see videos, 24 bits are preferred. The only way I found was to launch a secondary X server on :1 but then the troubles begin. If I launch this X server using "xinit -- :1" in a terminal from a gnome environment brought by gdm, the new display shows up in a black screen where I can move the mouse. But nothing can access this server, even xinit gets: AUDIT: Wed Mar 31 21:37:11 2004: 11509 X: client 1 rejected from local host Xlib: connection to ":1.0" refused by server Xlib: No protocol specified waiting for X server to begin accepting connections . Xlib: connection to ":1.0" refused by server Xlib: No protocol specified From guichaz at yahoo.fr Tue Apr 6 17:59:32 2004 From: guichaz at yahoo.fr (Guillaume Chazarain) Date: Tue, 6 Apr 2004 19:59:32 +0200 Subject: How to launch a second X server In-Reply-To: <20040406195049.705052e1.guichaz@yahoo.fr> References: <20040406195049.705052e1.guichaz@yahoo.fr> Message-ID: <20040406195932.0f34c1cf.guichaz@yahoo.fr> [Sorry to reply to myself, but the '..' in the message caused its end] Hi, I have a Voodoo Banshee (16 Mo), so if I want to have OpenGL acceleration, I have to use a depth of 16 bits, but if I want to see videos, 24 bits are preferred. The only way I found was to launch a secondary X server on :1 but then the troubles begin. If I launch this X server using "xinit -- :1" in a terminal from a gnome environment brought by gdm, the new display shows up in a black screen where I can move the mouse. But nothing can access this server, even xinit gets: AUDIT: Wed Mar 31 21:37:11 2004: 11509 X: client 1 rejected from local host Xlib: connection to ":1.0" refused by server Xlib: No protocol specified waiting for X server to begin accepting connections . Xlib: connection to ":1.0" refused by server Xlib: No protocol specified [ Here there is ".." ] Xlib: connection to ":1.0" refused by server Xlib: No protocol specified with these 3 last lines continuously. If I try "xinit -- :1" on the console, the display appears and I can use it but the initial xterm is on display :0 and "echo $DISPLAY" in it gives :0. I can then "DISPLAY=:1 xterm" to use this new display, but it's quite inconvenient. This is on FC2 test2, but it happened before, so it should not be a problem with SELinux or Xorg. Thanks for any help. Guillaume. From ted at cypress.com Tue Apr 6 20:37:01 2004 From: ted at cypress.com (Thomas Dodd) Date: Tue, 06 Apr 2004 15:37:01 -0500 Subject: How to launch a second X server In-Reply-To: <20040406195932.0f34c1cf.guichaz@yahoo.fr> References: <20040406195049.705052e1.guichaz@yahoo.fr> <20040406195932.0f34c1cf.guichaz@yahoo.fr> Message-ID: <407314ED.70109@cypress.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Guillaume Chazarain wrote: | If I launch this X server using "xinit -- :1" in a terminal from | a gnome environment brought by gdm, the new display shows up in a | black screen where I can move the mouse. But nothing can access this | server, even xinit gets: | | AUDIT: Wed Mar 31 21:37:11 2004: 11509 X: client 1 rejected from local host | Xlib: connection to ":1.0" refused by server | Xlib: No protocol specified Since XSESSION and DISPLAY are set, I'm not surprized you have trouble starting a second Xserver for GNOME. Running two GNOME sessions at once is problematic. You might want to use KDE, WindowMaker, or somother windowmanager on the second display instead. Search the archives for GNOME and VNC problems for more info. | If I try "xinit -- :1" on the console, the display appears and I can | use it but the initial xterm is on display :0 and "echo $DISPLAY" in | it gives :0. I can then "DISPLAY=:1 xterm" to use this new display, | but it's quite inconvenient. That seams odd. xinit should set DISPLAY correctly. You might try a custom xinitrc file (~/.xinitrc) that checks for 1st or 2nd xserver and changes behavior. Or just a script that starts the server, sets DISPLAY, starts a termional, and a window manager. -Thomas -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (SunOS) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFAcxTs+mwggWqA1pQRAtRhAJ4xV+cxcR7Su2ahv+QP0QRsT0QxhgCgguJp broUyl6QZIIFgzR8+eHeIxQ= =oR07 -----END PGP SIGNATURE----- From michal at harddata.com Tue Apr 6 22:30:20 2004 From: michal at harddata.com (Michal Jaegermann) Date: Tue, 6 Apr 2004 16:30:20 -0600 Subject: How to launch a second X server In-Reply-To: <407314ED.70109@cypress.com>; from ted@cypress.com on Tue, Apr 06, 2004 at 03:37:01PM -0500 References: <20040406195049.705052e1.guichaz@yahoo.fr> <20040406195932.0f34c1cf.guichaz@yahoo.fr> <407314ED.70109@cypress.com> Message-ID: <20040406163020.A21026@mail.harddata.com> On Tue, Apr 06, 2004 at 03:37:01PM -0500, Thomas Dodd wrote: > > Guillaume Chazarain wrote: > | If I launch this X server using "xinit -- :1" in a terminal from > | a gnome environment brought by gdm, > > Since XSESSION and DISPLAY are set, I'm not surprized you have trouble > starting a second Xserver for GNOME. > Running two GNOME sessions at once is problematic. Hm, interesting, I am doing that all the time for years now. :-) Running from a console, not from a terminal emulator in already running X session, 'startx -- :1 &' never gave me trouble. Even simpler on a long run is to edit '[servers]' section in /etc/X11/gdm/gdm.conf, with details you will find in comments, and to add extra servers there. Yes, you can run from gdm, say, a gnome session on vt7 and a kde one on vt8 or whatever else you desire. Two different gnome sessions are also not a problem. Michal From guichaz at yahoo.fr Thu Apr 8 13:35:13 2004 From: guichaz at yahoo.fr (Guillaume Chazarain) Date: Thu, 8 Apr 2004 15:35:13 +0200 Subject: How to launch a second X server In-Reply-To: <407314ED.70109@cypress.com> References: <20040406195049.705052e1.guichaz@yahoo.fr> <20040406195932.0f34c1cf.guichaz@yahoo.fr> <407314ED.70109@cypress.com> Message-ID: <20040408153513.6d4aae50.guichaz@yahoo.fr> > | If I try "xinit -- :1" on the console, the display appears and I can > | use it but the initial xterm is on display :0 and "echo $DISPLAY" in > | it gives :0. I can then "DISPLAY=:1 xterm" to use this new display, > | but it's quite inconvenient. > > That seams odd. xinit should set DISPLAY correctly. You might try a > custom xinitrc file (~/.xinitrc) that checks for 1st or 2nd xserver and > changes behavior. Or just a script that starts the server, sets DISPLAY, > starts a termional, and a window manager. Upgrading xinitrc and xorg from rawhide, replacing 'xinit -- :1' by 'startx -- :1' with a 'echo xterm > ~/.xinitrc' worked well, even from another X session. I discovered that startx wanted the display number as its first argument : startx -- :1 -depth 24 => good startx -- -depth 24 :1 => bad Thanks Michal and Thomas. Guillaume. From bhughes at elevating.com Tue Apr 13 21:43:56 2004 From: bhughes at elevating.com (Bret Hughes) Date: 13 Apr 2004 16:43:56 -0500 Subject: via cle266 and xv Message-ID: <1081892644.10467.19.camel@bretsony> I am trying to configure a fedora 1 machine to, among other things, use hardware decoding for mpegs. We use mplayer on ati and proSavage chipsets with no problem using -v0 xv flags to tell mplayer to use xvideo extensions. I have upgraded all packages to the latest released updates including : XFree86-4.3.0-55 kernel-2.4.22-1.2174.nptl current device section in XF86Config after copious surfing: Section "Device" Identifier "Videocard0" Driver "via" VendorName "VIA" BoardName "via driver supporting cle266" VIdeoRam 32786 Option "ActiveDevice" "CRT" Option "PciRetry" "true" EndSection xvinfo outputs a bunch of stuff so I assume that it is available I see the folowing lines in XFree86.0.log:Symbol REGION_EQUAL from module /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved! Symbol xf86XVCopyYUV12ToPacked from module /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved! Could this be part of the problem? Does not sound like it but hey, what do I know? ( not enough - by definition ) mplayer plays using -vo x11 but is slow mplayer -vo xv crashes X altogether Any tips? I found a thread about rawhide having updated XFree rpms but it did not sound like this was fixing anything like this so I have not tries it TIA Bret From agd5f at yahoo.com Tue Apr 13 22:20:10 2004 From: agd5f at yahoo.com (Alex Deucher) Date: Tue, 13 Apr 2004 15:20:10 -0700 (PDT) Subject: via cle266 and xv In-Reply-To: <1081892644.10467.19.camel@bretsony> Message-ID: <20040413222010.87679.qmail@web11308.mail.yahoo.com> --- Bret Hughes wrote: > I am trying to configure a fedora 1 machine to, among other things, > use > hardware decoding for mpegs. We use mplayer on ati and proSavage > chipsets with no problem using -v0 xv flags to tell mplayer to use > xvideo extensions. > > I have upgraded all packages to the latest released updates including > : > > XFree86-4.3.0-55 > kernel-2.4.22-1.2174.nptl > > current device section in XF86Config after copious surfing: > > Section "Device" > Identifier "Videocard0" > Driver "via" > VendorName "VIA" > BoardName "via driver supporting cle266" > VIdeoRam 32786 > Option "ActiveDevice" "CRT" > Option "PciRetry" "true" > > EndSection > > > xvinfo outputs a bunch of stuff so I assume that it is available > > I see the folowing lines in XFree86.0.log:Symbol REGION_EQUAL from > module /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved! > Symbol xf86XVCopyYUV12ToPacked from module > /usr/X11R6/lib/modules/drivers/via_drv.o is unresolved! > > > Could this be part of the problem? Does not sound like it but hey, > what > do I know? ( not enough - by definition ) > > mplayer plays using -vo x11 but is slow > > mplayer -vo xv crashes X altogether > > Any tips? I found a thread about rawhide having updated XFree rpms > but > it did not sound like this was fixing anything like this so I have > not > tries it I'm not sure what version of the via driver was included in fedora, but the initial drop from via was pretty buggy. the version in cvs and the one here are substantially better. also, if you are using a KM/N chip Xv support was not added until recently. See this page for more info and newer versions of the driver: http://unichrome.sourceforge.net/ Alex > > TIA > > Bret > __________________________________ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html From bhughes at elevating.com Tue Apr 13 23:57:08 2004 From: bhughes at elevating.com (Bret Hughes) Date: 13 Apr 2004 18:57:08 -0500 Subject: via cle266 and xv In-Reply-To: <20040413222010.87679.qmail@web11308.mail.yahoo.com> References: <20040413222010.87679.qmail@web11308.mail.yahoo.com> Message-ID: <1081900629.10467.33.camel@bretsony> On Tue, 2004-04-13 at 17:20, Alex Deucher wrote: > > --- Bret Hughes wrote: > > I am trying to configure a fedora 1 machine to, among other things, > > use > > hardware decoding for mpegs. We use mplayer on ati and proSavage > > chipsets with no problem using -v0 xv flags to tell mplayer to use > > xvideo extensions. > > Any tips? I found a thread about rawhide having updated XFree rpms > > but > > it did not sound like this was fixing anything like this so I have > > not > > tries it > > I'm not sure what version of the via driver was included in fedora, but > the initial drop from via was pretty buggy. the version in cvs and the > one here are substantially better. also, if you are using a KM/N chip > Xv support was not added until recently. > > See this page for more info and newer versions of the driver: > http://unichrome.sourceforge.net/ Thanks Alex what are the odds that a mandrake src.rpm will build on a fedora box? I will try it later tonight. Right now I am in the middle of blowing a rh9 install on the box over the fedora install (sigh). Bret From agd5f at yahoo.com Wed Apr 14 00:07:23 2004 From: agd5f at yahoo.com (Alex Deucher) Date: Tue, 13 Apr 2004 17:07:23 -0700 (PDT) Subject: via cle266 and xv In-Reply-To: <1081900629.10467.33.camel@bretsony> Message-ID: <20040414000723.46750.qmail@web11309.mail.yahoo.com> --- Bret Hughes wrote: > On Tue, 2004-04-13 at 17:20, Alex Deucher wrote: > > > > --- Bret Hughes wrote: > > > I am trying to configure a fedora 1 machine to, among other > things, > > > use > > > hardware decoding for mpegs. We use mplayer on ati and proSavage > > > chipsets with no problem using -v0 xv flags to tell mplayer to > use > > > xvideo extensions. > > > > Any tips? I found a thread about rawhide having updated XFree > rpms > > > but > > > it did not sound like this was fixing anything like this so I > have > > > not > > > tries it > > > > I'm not sure what version of the via driver was included in fedora, > but > > the initial drop from via was pretty buggy. the version in cvs and > the > > one here are substantially better. also, if you are using a KM/N > chip > > Xv support was not added until recently. > > > > See this page for more info and newer versions of the driver: > > http://unichrome.sourceforge.net/ > > Thanks Alex what are the odds that a mandrake src.rpm will build on a > fedora box? I will try it later tonight. Right now I am in the > middle > of blowing a rh9 install on the box over the fedora install (sigh). no idea. never tried it. It may be easier to try and patch an X tree and rebuild. Alex > > > Bret > > > __________________________________ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html From A.Verheij at econ.usyd.edu.au Wed Apr 14 05:34:05 2004 From: A.Verheij at econ.usyd.edu.au (Andre Verheij) Date: Wed, 14 Apr 2004 15:34:05 +1000 Subject: RedHat 9 and Xorg and Dual Head Message-ID: <78292DA462BCDF41B4A7DF6EAF1AABA9012CABF0@evs2.econ.usyd.edu.au> All, Do you reckon it should be doable to install the latest Xorg onto RH9 and have a working dual head on a HP XW4100 system with a NVIDIA Quadro NV280 Card? I use Ximian Evolution with the Exchange Connector and RH9 is the only OS I have that supports the connector. RHEL WS or Fedora are not supported yet :-( So I am stuck with RH9...... Kind regards, Andr? Verheij Web Support Officer / Blackboard Administrator Faculty of Economics & Business (Building H69) University of Sydney a.verheij at econ.usyd.edu.au Work +61 2 9351 3346 Mobile +61 416 065 628 From ddembrow at nlxcorp.com Wed Apr 14 16:22:59 2004 From: ddembrow at nlxcorp.com (ddembrow at nlxcorp.com) Date: Wed, 14 Apr 2004 12:22:59 -0400 Subject: Touch Screen Driver Message-ID: I have a multi-headed application which uses four touch screens (3m Micro Touch). The linux touch screen driver provided by 3m only supports a single touch screen. Can someone point me to an XFree86 resource for developing a touch screen driver that will support multi-headed configurations, or better yet an existing driver that supports the EXII, SMT3, MT3000, MT410, and MT510 controllers? Thanks, ---d.dembrow From agd5f at yahoo.com Wed Apr 14 16:39:16 2004 From: agd5f at yahoo.com (Alex Deucher) Date: Wed, 14 Apr 2004 09:39:16 -0700 (PDT) Subject: Touch Screen Driver In-Reply-To: Message-ID: <20040414163916.34132.qmail@web11308.mail.yahoo.com> --- ddembrow at nlxcorp.com wrote: > I have a multi-headed application which uses four touch screens (3m > Micro > Touch). > The linux touch screen driver provided by 3m only supports a single > touch > screen. > > Can someone point me to an XFree86 resource for developing a touch > screen > driver > that will support multi-headed configurations, or better yet an > existing > driver that supports > the EXII, SMT3, MT3000, MT410, and MT510 controllers? > your best bet is to try one of the xfree86 or XORG mailing lists. http://www.xfree86.org/lists.html http://www.freedesktop.org/Software/xorg Alex > Thanks, > ---d.dembrow > > __________________________________ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html From bhughes at elevating.com Wed Apr 14 18:02:57 2004 From: bhughes at elevating.com (Bret Hughes) Date: 14 Apr 2004 13:02:57 -0500 Subject: via cle266 and xv In-Reply-To: <20040414000723.46750.qmail@web11309.mail.yahoo.com> References: <20040414000723.46750.qmail@web11309.mail.yahoo.com> Message-ID: <1081965778.10467.85.camel@bretsony> On Tue, 2004-04-13 at 19:07, Alex Deucher wrote: > > --- Bret Hughes wrote: > > On Tue, 2004-04-13 at 17:20, Alex Deucher wrote: > > > > > > --- Bret Hughes wrote: > > > > I am trying to configure a fedora 1 machine to, among other > > things, > > > > use > > > > hardware decoding for mpegs. We use mplayer on ati and proSavage > > > > chipsets with no problem using -v0 xv flags to tell mplayer to > > use > > > > xvideo extensions. > > > > > > Any tips? I found a thread about rawhide having updated XFree > > rpms > > > > but > > > > it did not sound like this was fixing anything like this so I > > have > > > > not > > > > tries it > > > > > > I'm not sure what version of the via driver was included in fedora, > > but > > > the initial drop from via was pretty buggy. the version in cvs and > > the > > > one here are substantially better. also, if you are using a KM/N > > chip > > > Xv support was not added until recently. > > > > > > See this page for more info and newer versions of the driver: > > > http://unichrome.sourceforge.net/ > > > > Thanks Alex what are the odds that a mandrake src.rpm will build on a > > fedora box? I will try it later tonight. Right now I am in the > > middle > > of blowing a rh9 install on the box over the fedora install (sigh). > > no idea. never tried it. It may be easier to try and patch an X tree > and rebuild. > A download and build of the current 4.4 source works great! I have not even got to the unichrome source yet. scaling of a small mpeg to 1024x768 with 11% processor usage. Way cool. An interesting anomoly is that unclutter flickers the cursor if the window changes. I did not see this in 4.3. running sawfish instead of no wm fixes that issue. Now for the real question. How big a deal is it to build a rpm package for the 4.4 code base? I have never packaged something as large as XFree86 and really need this to be part of a kickstart installation. I am downloading the 4.3 srpm now to take a look at it. Tips appreciated and thanks for the pointer to 4.4 Bret From bhughes at elevating.com Wed Apr 14 18:58:00 2004 From: bhughes at elevating.com (Bret Hughes) Date: 14 Apr 2004 13:58:00 -0500 Subject: via cle266 and xv In-Reply-To: <1081965778.10467.85.camel@bretsony> References: <20040414000723.46750.qmail@web11309.mail.yahoo.com> <1081965778.10467.85.camel@bretsony> Message-ID: <1081969080.10512.106.camel@bretsony> On Wed, 2004-04-14 at 13:02, Bret Hughes wrote: > On Tue, 2004-04-13 at 19:07, Alex Deucher wrote: > > > > no idea. never tried it. It may be easier to try and patch an X tree > > and rebuild. > > > > A download and build of the current 4.4 source works great! I have not > even got to the unichrome source yet. scaling of a small mpeg to > 1024x768 with 11% processor usage. Way cool. > > An interesting anomoly is that unclutter flickers the cursor if the > window changes. I did not see this in 4.3. running sawfish instead of > no wm fixes that issue. > > Now for the real question. How big a deal is it to build a rpm package > for the 4.4 code base? I have never packaged something as large as > XFree86 and really need this to be part of a kickstart installation. > > I am downloading the 4.3 srpm now to take a look at it. > > Tips appreciated and thanks for the pointer to 4.4 > Man that is one scary specfile Mike, It looks like you earn your $. is there an easy way to build a binary tarball from the 4.4 tree I built? As an interim kickstart step I could untar the new binaries over the 4.3 install. Is there a better place to ask? Bret From agd5f at yahoo.com Wed Apr 14 19:40:28 2004 From: agd5f at yahoo.com (Alex Deucher) Date: Wed, 14 Apr 2004 12:40:28 -0700 (PDT) Subject: via cle266 and xv In-Reply-To: <1081969080.10512.106.camel@bretsony> Message-ID: <20040414194028.73799.qmail@web11304.mail.yahoo.com> --- Bret Hughes wrote: > On Wed, 2004-04-14 at 13:02, Bret Hughes wrote: > > On Tue, 2004-04-13 at 19:07, Alex Deucher wrote: > > > > > > > no idea. never tried it. It may be easier to try and patch an X > tree > > > and rebuild. > > > > > > > A download and build of the current 4.4 source works great! I have > not > > even got to the unichrome source yet. scaling of a small mpeg to > > 1024x768 with 11% processor usage. Way cool. > > > > An interesting anomoly is that unclutter flickers the cursor if the > > window changes. I did not see this in 4.3. running sawfish > instead of > > no wm fixes that issue. > > > > Now for the real question. How big a deal is it to build a rpm > package > > for the 4.4 code base? I have never packaged something as large as > > XFree86 and really need this to be part of a kickstart > installation. > > > > I am downloading the 4.3 srpm now to take a look at it. > > > > Tips appreciated and thanks for the pointer to 4.4 > > > > Man that is one scary specfile Mike, It looks like you earn your $. > > is there an easy way to build a binary tarball from the 4.4 tree I > built? As an interim kickstart step I could untar the new binaries > over > the 4.3 install. take a look at the xc/config/cf/host.def there's an option you can use to specify a non-default installation directory. #define ProjectRoot /path/to/Your/XFree86/installation Alex > > Is there a better place to ask? > > Bret > > __________________________________ Do you Yahoo!? Yahoo! Tax Center - File online by April 15th http://taxes.yahoo.com/filing.html From bhughes at elevating.com Wed Apr 14 21:32:14 2004 From: bhughes at elevating.com (Bret Hughes) Date: 14 Apr 2004 16:32:14 -0500 Subject: via cle266 and xv In-Reply-To: <20040414194028.73799.qmail@web11304.mail.yahoo.com> References: <20040414194028.73799.qmail@web11304.mail.yahoo.com> Message-ID: <1081978335.10467.113.camel@bretsony> On Wed, 2004-04-14 at 14:40, Alex Deucher wrote: > > --- Bret Hughes wrote: > > On Wed, 2004-04-14 at 13:02, Bret Hughes wrote: > > > On Tue, 2004-04-13 at 19:07, Alex Deucher wrote: > > > > > > > > > > no idea. never tried it. It may be easier to try and patch an X > > tree > > > > and rebuild. > > > > > > > > > > A download and build of the current 4.4 source works great! I have > > not > > > even got to the unichrome source yet. scaling of a small mpeg to > > > 1024x768 with 11% processor usage. Way cool. > > > > > > An interesting anomoly is that unclutter flickers the cursor if the > > > window changes. I did not see this in 4.3. running sawfish > > instead of > > > no wm fixes that issue. > > > > > > Now for the real question. How big a deal is it to build a rpm > > package > > > for the 4.4 code base? I have never packaged something as large as > > > XFree86 and really need this to be part of a kickstart > > installation. > > > > > > I am downloading the 4.3 srpm now to take a look at it. > > > > > > Tips appreciated and thanks for the pointer to 4.4 > > > > > > > Man that is one scary specfile Mike, It looks like you earn your $. > > > > is there an easy way to build a binary tarball from the 4.4 tree I > > built? As an interim kickstart step I could untar the new binaries > > over > > the 4.3 install. > > take a look at the xc/config/cf/host.def > there's an option you can use to specify a non-default installation > directory. > #define ProjectRoot /path/to/Your/XFree86/installation > Thanks. THis pointed me in the right direction. I looks like from the README in config/cf that ProjectRoot will place the files differently on the target machine and since some of these paths are hardcoded into the binary that it would not do what I wanted. I believe I resolved it using make DESTDIR=dir_for_binary_tree install rather the original make install I now have a handy dandy tree with all the installed files in it ready to tar up. Thank you. THank you Thank you. Now if I can just find time to get my head around Mikes spec file I could do this right. Bret From laurent.jean-rigaud at regis-dgac.net Fri Apr 23 16:06:08 2004 From: laurent.jean-rigaud at regis-dgac.net (laurent.jean-rigaud at regis-dgac.net) Date: Fri, 23 Apr 2004 18:06:08 +0200 Subject: Problem with locale and XFree86 under RedHat Message-ID: Hi, We currently porting some applications from HP to LINUX RedHat and we meet a problem with locale. The problem appears with RH7.2+XFREE430, RH9 et RHEL3.0. The problem is if we set a locale to en_GB, our test app seems to crash in XIM code. If the locale is set to en_GB.utf8 , the problem disapears... So it is mandatory to set the locale to something UTF8 compatible ? Here is the code for testing : it compiles with "gcc -DLINUX -g -static -o pb_label_accent pb_label_accent.c -I/usr/X11R6/include -L/usr/X11R6/lib -lMrm -lXm -lXp -lXext -lXt -lSM -lICE -lXmu -lX11 -lm" #---------------- #include #include #include #include #include #include #include #include #include #include #include #include #include static Widget g_toplevel = NULL; static Widget g_wts_configu = NULL; static XmString valider_simple; static XtAppContext g_app_context; #ifdef hp700 static const char *const NOM_LOCALE = "english.iso88591"; #endif /* hp700 */ /************************************************************* * Sous Robin, l'utilisation de la locale qui suit, marche. * * Sous samekh, le meme positionnement entraine un coredump * *************************************************************/ /* #ifdef LINUX static const char *const NOM_LOCALE = "en_GB"; #endif */ /************************************************************* * Sous samekh, l'utilisation de la locale qui suit empeche * * la plante mais elle entraine un probl?me d'affichage des * * accents dans les label lorsque l'on adresse un display * * autre qu'un display LINUX * *************************************************************/ #ifdef LINUX static const char *const NOM_LOCALE = "en_GB"; #endif int main ( int argc, char *argv[] ) { Widget wf_pere, wl_texte; char buffer[255]; if (! setlocale (LC_ALL, NOM_LOCALE)) XtWarning ("locale not supported by C library, locale unchanged"); g_toplevel = XtVaAppInitialize (&g_app_context, "Test_samekh", NULL, 0, &argc, argv, NULL, NULL); g_wts_configu = XtVaCreatePopupShell ("CONF", transientShellWidgetClass, g_toplevel, XtNtitle, "Test Label Accent", XmNwidth, 200, XmNheight, 200, XmNx, 100, XmNy, 100, NULL); wf_pere = XtVaCreateManagedWidget ("PERE", xmFormWidgetClass, g_wts_configu, XmNborderWidth, 0, XmNshadowThickness, 0, XmNresizePolicy, XmRESIZE_GROW, NULL); sprintf (buffer, "%s", "Pr?c?dent"); valider_simple = XmStringGenerate (buffer, XmSTRING_DEFAULT_CHARSET, XmCHARSET_TEXT, NULL); wl_texte = XtVaCreateManagedWidget ("Valider", xmLabelWidgetClass, wf_pere, XmNx, 0, XmNy, 100, XmNlabelString, valider_simple, NULL); XtPopup(g_wts_configu, XtGrabNone); XmStringFree(valider_simple); XtAppMainLoop(g_app_context); return 0; } #---------------- Thanks for you help. PS: maybe the problem come from GLIBC ? From Patil.Vijay at honeywell.com Wed May 12 06:21:58 2004 From: Patil.Vijay at honeywell.com (Patil, Vijay N (IE03x) (SW)) Date: Wed, 12 May 2004 11:51:58 +0530 Subject: X server problem Message-ID: <5CA9809542E9D4118A54000629A87ECB039CC35F@IE03M01> Hi All, I have one application written using Motif and X11, but when i run it im getting following errors X11TransSocketOpen: socket() failed for tcp _X11TransSocketOpenCOTSClient: Unable to open socket for tcp _X11TransOpen: transport open failed for tcp/145.49.0.179:0 Error: Can't open display: 145.49.0.179:0.0 and some times X Error of failed request: BadGC (invalid GC parameter) Major opcode of failed request: 60 (X_FreeGC) Resource id in failed request: 0x1000007 Serial number of failed request: 108 Current serial number in output stream: 116 ./fmlTest -x -i 1 Im running it on Redhat Linux using X emulator( Exceed ) I don't whether this is the right place to post this problem. any king of help is highly appreciated regards, Vijay From bhughes at elevating.com Thu May 13 01:43:11 2004 From: bhughes at elevating.com (Bret Hughes) Date: 12 May 2004 20:43:11 -0500 Subject: X server problem In-Reply-To: <5CA9809542E9D4118A54000629A87ECB039CC35F@IE03M01> References: <5CA9809542E9D4118A54000629A87ECB039CC35F@IE03M01> Message-ID: <1084412592.3110.3.camel@bretsony> On Wed, 2004-05-12 at 01:21, Patil, Vijay N (IE03x) (SW) wrote: > Hi All, > I have one application written using Motif and X11, but when i run > it im getting > following errors > > X11TransSocketOpen: socket() failed for tcp > _X11TransSocketOpenCOTSClient: Unable to open socket for tcp > _X11TransOpen: transport open failed for tcp/145.49.0.179:0 > Error: Can't open display: 145.49.0.179:0.0 > > and some times > > X Error of failed request: BadGC (invalid GC parameter) > Major opcode of failed request: 60 (X_FreeGC) > Resource id in failed request: 0x1000007 > Serial number of failed request: 108 > Current serial number in output stream: 116 > ./fmlTest -x -i 1 > > Im running it on Redhat Linux using X emulator( Exceed ) > I don't whether this is the right place to post this problem. > any king of help is highly appreciated > Can you run other X apps, say xclock? If so I can't help at all. If you can't even run xclock then it looks like exceed is either not listening on the windows (I am assuming windows here.) or the client does not have perms to attach but I think that would give you a different error. Bret From wenbian at hotmail.com Mon Jun 21 21:04:48 2004 From: wenbian at hotmail.com (Wen Bian) Date: Mon, 21 Jun 2004 16:04:48 -0500 Subject: setting up PseudoColor visual Message-ID: Hi there, Could anyone please give me some advice on setting the display to 8-bit PseudoColor (I'm trying to run a legacy program)? I'm running Red Hat Enterprise Linux WS 3.0 on a DELL Precision 450n workstation with DELL P1230 monitor and video card of 128MB NVIDIA Quadro FX 1000. Simply changing the "XF86Config" or using redhat-config-xfree86 doesn't work.. and if you know of any free software that provides PseudoColor overlay mode over TrueColor etc, would you please let me know also.. Thank you very much! Sincerely, Wen Bian Center of Structural Biology Vanderbilt University From r_sat at hotmail.com Fri Jun 25 16:02:20 2004 From: r_sat at hotmail.com (SATISH RAMANATHAN) Date: Fri, 25 Jun 2004 09:02:20 -0700 Subject: xvfb Message-ID: Hi, I have a quick, may be silly, question. My Linux (that runs on RHEL, RH8) application requires Xvfb be installed and running on the user machine. And we are coming across cases where not all the user machines have Xvfb installed by default. I believe Xvfb is not part of the standard package,though it comes with the CD. I would like to know if it is okay to package and ship the xvfb binary along with my application. Are there any issues(technical, legal,...) that I need to consider before doing this? Thanks in advance, -Satish _________________________________________________________________ FREE pop-up blocking with the new MSN Toolbar ? get it now! http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/ From mharris at www.linux.org.uk Sat Jun 26 11:56:23 2004 From: mharris at www.linux.org.uk (Mike A. Harris) Date: Sat, 26 Jun 2004 07:56:23 -0400 (EDT) Subject: xvfb In-Reply-To: References: Message-ID: On Fri, 25 Jun 2004, SATISH RAMANATHAN wrote: >I have a quick, may be silly, question. >My Linux (that runs on RHEL, RH8) application requires Xvfb be installed and >running on the user machine. And we are coming across cases where not all >the user machines have Xvfb installed by default. I believe Xvfb is not >part of the standard package,though it comes with the CD. I would like to >know if it is okay to package and ship the xvfb binary along with my >application. Are there any issues(technical, legal,...) that I need to >consider before doing this? Xvfb is a standard part of the XFree86 source code, and is always built and included with all Red Hat OSs. If you use Red Hat Kickstart to perform installations, you can have Kickstart ensure that Xvfb is always installed. You could also include Xvfb with your application, as it is open source software, however I'd recommend against that for various reasons: - Using the Red Hat supplied Xvfb, when you apply security updates, if there are any updates which affect Xvfb, you'll automatically get them. Otherwise you'd have to monitor X security and update it manually in your application bundle. - Red Hat supports the Red Hat supplied Xvfb, provided the Red Hat supplied rpm packages have been installed. You may need to use mailing lists for Xvfb support issues if you supply it yourself. - By using Red Hat packages, you ensure you don't have files being installed on the filesystem that conflict with potentially rpm managed files. This can cause problems during upgrades, unless of course the file paths are unique. The XFree86 sources included with Red Hat Enterprise Linux, and Red Hat Linux 8.0 are under the various licenses included in the XFree86 documentation. You'd have to check the Xvfb license for the specifics, but I believe it is MIT/X11 like most of the rest of the source tree. DISCLAIMER: This is just friendly information, not legal advice. Please consult the license files or an attourney for legally binding licencing information. ;o) From wanqing.wu at openwave.com Tue Jun 29 05:12:01 2004 From: wanqing.wu at openwave.com (Wanqing Wu) Date: Mon, 28 Jun 2004 22:12:01 -0700 Subject: ATI Mach64 Rage II video card driver in Red Hat 9 (Fedora Core 2) Message-ID: <010201c45d97$9c862b60$6500a8c0@myopwv.com> Hi, I installed the latest red hat Linux, Fedora Core 2, in my machine. But the X server cannot be started. I got the following error message: Module: init10 X11R6.7.0 Driver: atimisc_drv.o Error: ATI cannot read V_BIOS My machine has the video card, ATI Mach64 3D Rage IIC. The driver installed is atimisc_drv.o of X11R6. Could you please let me know what is wrong ? Do I have the correct driver ? If I don't have the correct driver, where can I get the one for my video card ? Thanks very much, -Wanqing -------------- next part -------------- An HTML attachment was scrubbed... URL: From mharris at www.linux.org.uk Tue Jun 29 19:40:33 2004 From: mharris at www.linux.org.uk (Mike A. Harris) Date: Tue, 29 Jun 2004 15:40:33 -0400 (EDT) Subject: ATI Mach64 Rage II video card driver in Red Hat 9 (Fedora Core 2) In-Reply-To: <010201c45d97$9c862b60$6500a8c0@myopwv.com> References: <010201c45d97$9c862b60$6500a8c0@myopwv.com> Message-ID: On Mon, 28 Jun 2004, Wanqing Wu wrote: >I installed the latest red hat Linux, Fedora Core 2, in my machine. But the X server cannot be >started. I got the following error message: > >Module: init10 X11R6.7.0 >Driver: atimisc_drv.o >Error: ATI cannot read V_BIOS ^^^^^^^^^^^^^^^^^^ The driver can't read the video BIOS. >My machine has the video card, ATI Mach64 3D Rage IIC. The driver installed >is atimisc_drv.o of X11R6. > >Could you please let me know what is wrong ? Do I have the correct driver ? >If I don't have the correct driver, where can I get the one for my video card ? You have the correct (and only) driver. For one reason or another it can't read the video BIOS, and will not function. You need to determine the cause of that first, and then try to fix it. Make sure that your CMOS is configured correctly, and that you're using Red Hat supplied XFree86 rpms and kernel. By the way, in the Subject line of your email you state: "Re: ATI Mach64 Rage II video card driver in Red Hat 9 (Fedora Core 2)" I presume you mean "Red Hat Linux 9" there. "Red Hat Linux 9" and "Fedora Core 2" are two totally different operating system releases which have nothing in common with each other. From your email though, it appears that you are using Fedora Core 2, not Red Hat Linux 9. Post your X server log file and config file somewhere and pass on the URL, and we'll have a look.