From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.8) Gecko/20020204 Description of problem: I just worked at configuring an IBM LCD monitor whose name included a double quote (They were being cute to say it was a '17"' monitor. The code in XConfigurator that fetches the vendor strings has to get a bit more clever to filter out that double quote, because the resulting XF86Config files would say "IBM 6251 LCD 17"" And that chokes XFree86. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.Run Xconfigurator on an IBM NetVista X series system. 2.View the error messages from the X server when it chokes on the syntax error. Additional info:
I can't find that string located anywhere in any of the hardware databases. Where exactly is this string coming from?
pts/28 mharris@devel:~/src/redhat/hwdata$ dir total 1216 drwxrwx--- 3 mharris 4096 Mar 28 05:44 ./ drwxrwx--- 31 mharris 4096 Mar 18 07:56 ../ -rw-rw---- 1 mharris 17992 Jan 30 19:02 COPYING drwxrwx--- 2 mharris 4096 Mar 28 05:44 CVS/ -rw-rw---- 1 mharris 2747 Mar 20 2001 CardMonitorCombos -rw-rw---- 1 mharris 69727 Mar 23 04:35 Cards -rw-rw---- 1 mharris 66964 Feb 1 01:52 Cards.ORIGINAL-before-commenting-out-UNSUPPORTED -rw-rw---- 1 mharris 23077 Feb 8 18:49 Cards.diff -rw-rw---- 1 mharris 151 Jan 30 19:02 LICENSE -rw-rw---- 1 mharris 1154 Jan 30 19:02 Makefile -rw-rw---- 1 mharris 202369 Mar 7 11:34 MonitorsDB -rw-rw---- 1 mharris 202430 Feb 3 17:37 MonitorsDB.Iiyama -rw-rw---- 1 mharris 150 Mar 17 12:20 README -rw-rw---- 1 mharris 1264 Mar 22 17:04 hwdata.spec -rwxrwx--- 1 mharris 6114 Feb 27 20:11 inf2mondb.py* -rwxrwx--- 1 mharris 3048 Aug 10 2001 makeids* -rwxrwx--- 1 mharris 171 Dec 2 1998 mconvert* -rw-rw---- 1 mharris 160200 Mar 22 17:04 pci.ids -rw-rw---- 1 mharris 158202 Mar 28 05:44 pcitable -rw-rw---- 1 mharris 166724 Dec 29 1999 printertable -rw-rw---- 1 mharris 2540 Mar 13 19:09 translations -rw-rw---- 1 mharris 374 Mar 14 07:14 upgradelist -rw-rw---- 1 mharris 64140 Mar 14 07:14 usb.ids pts/28 mharris@devel:~/src/redhat/hwdata$ grep 17\" * grep: CVS: Is a directory pcitable:0x1039 0x5517 "unknown" "Silicon Integrated Systems [SiS]|5517" pcitable:0x104e 0x0017 "unknown" "Oak Technology, Inc|OTI-64017" pcitable:0x104e 0x0217 "unknown" "Oak Technology, Inc|OTI-64217" pcitable:0x104e 0x0317 "unknown" "Oak Technology, Inc|OTI-64317" pcitable:0x108a 0x0001 "unknown" "Bit3 Computer Corp.|VME Bridge Model 617" pcitable:0x1119 0x0006 "gdth" "ICP Vortex|GDT 6x17" pcitable:0x11d5 0x0117 "unknown" "Ikon Corporation|10117" I did not see that this bug was filed against RHL 7.1 until just now. Since the data provided doesn't seem to exist in any current databases, I am assuming that the problem no longer exists in Rawhide using the latest versions of Xconfigurator and hwdata (new package).
I have not studied this very deeply, so I'll ask: When one does a VESA query of a video monitor to have it describe itself, isn't there a string returned as the name of the hardware? It was my guess that the string returned by the monitor contained a double quote, and that that string was being passed to soemthing that was crafting the Monitor section of XF86Config. Am I being silly here? -wdc
The X server does a DDC probe of the monitor if the video card supports doing such. If the monitor supports DDC, then the monitor's EDID info is returned and the EDID value is looked up in the MonitorsDB database. The EDID value is typically a string of 7-8 characters generally in the form xxxnnnn. The field you are refering to is only a textual string for humans to read, and is not involved in the autodetection process.
This is very perplexing. Well the system I was playing with was a loaner. We sent it back. If the problem happens with something I care more about, I'll do additional sleuthing. Thanks very much for the help and the clarification.