Re: transcode fails
Nicolas FRANCOIS wrote: > Le Thu, 17 Apr 2008 10:10:16 +0600 "Alexander E. Patrakov" > <[EMAIL PROTECTED]> a écrit : > >> Nicolas FRANCOIS wrote: >> >>> I encountered this error while compiling transcode : >> Could you please explain why you use transcode instead of calling ffmpeg >> directly? Shouldn't transcode be removed from the book, because it stayed >> broken >> for so long and nobody noticed it before you? > > Don't know. It is a dependency of K3b, so I built it. Thanks for your reply. This package is indeed useful (and used bu K3B) for ripping DVDs (although I prefer to use "mplayer -dumpstream dvd://" for this), and should not be removed. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: transcode fails
Nicolas FRANCOIS wrote: > I encountered this error while compiling transcode : Could you please explain why you use transcode instead of calling ffmpeg directly? Shouldn't transcode be removed from the book, because it stayed broken for so long and nobody noticed it before you? -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Running firefox as root once no longer necessary
Wilco Beekhuizen wrote: > I don't know why the BLFS book still mentions to run firefox once as > the root user, but I don't think this is necessary. I checked with > firefox 2 and firefox 3 but both run fine without running firefox as > root once. > > It's this line: > "You should run /usr/bin/firefox once as the root user (or any user > with write privileges) to create some necessary additional files in > the /usr hierarchy. " > > So I guess this can be removed or am I missing something? If you attempt to start Seamonkey (not tested with Firefix) from Xfce, it will work only from the second try, unless you registered the shrome by starting Seamonkey once as root, or by running this as root: MFH=/usr/lib/seamonkey MOZILLA_FIVE_HOME=$MFH LD_LIBRARY_PATH=$MFH $MFH/regxpcom MOZILLA_FIVE_HOME=$MFH LD_LIBRARY_PATH=$MFH $MFH/regchrome So the book should be updated. The fact that one can't automate Firefox build using only the book's instructions is a bug. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xfdesktop from CBLFS
Richard Melville wrote: > Any idea why XFCE is no longer supported in BLFS? Basically, it is my own decision. I made it because of the following reasons: 1) No editors except myself use XFCE. Nobody cared about updating the ancient 4.2.x version when 4.4.0 came out. So even if I put the new version into the book, the text will suffer from "bit rot". 2) It has a lot of external optional dependencies not in the book, such as PolicyKit and libnotify that do provide the functionality that developers would expect to be available. 3) I disagree with these developers, because of critical (for me) bugs that come with the code that uses them. Such bugs are known for a very long time (see, e.g., http://bugzilla.xfce.org/show_bug.cgi?id=2891), but developers wait for a miracle (and a miracle from a wrong party - in this example, PolicyKit has nothing relevant to "default mount options" in its stated goals) and do nothing. 4) Newbies are likely to compile with all optional dependencies even if warned that they are harmful. I don't want to see the resulting bug reports on the blfs-support list. In short, the package is too buggy, requires workarounds, and cannot be maintained in a good shape in the book. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xfdesktop from CBLFS
Arnie Stender wrote: > Hi Folks, > I am sending this post from my new albeit minimal for now > workstation (CLFS/CBLFS multi-lib) running on AMD64 Athlon. I'm running > xorg-7.3 and XFCE window manager. XFCE is no longer supported in BLFS (and you didn't even specify its version!). There is some activity on the XFCE mailing list, but it is mostly about the upcoming 4.6 release. As for the segfault, strace logs are always useless in this situation, please provide a gdb backtrace with full debug info (i.e., compile with -g3 in CFLAGS). OTOH, XFCE seems to be functional on the official LFS LiveCD, even for non-multilib x86-64. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Stable LFS LiveCD 6.3-r2160 released
The LFS LiveCD Team is proud to announce the release of the 6.3-r2160 stable version of the LFS LiveCD for x86 and x86_64 architectures. The CD contains the LFS book (version 6.3 for x86 CD, and its unofficial x86_64 equivalent for x86_64), and a lot of software that you need to build your own LFS system, ask for support and perform various recovery tasks. For each architecture, the CD is available in -min, -nosrc and full variants. The -min CD contains only a text-mode environment, -nosrc adds the X window system and the Xfce desktop, while the full CD also contains all LFS source tarballs. All of the above variants are suitable as hosts for building x86 and x86_64 LFS and CLFS systems. There are the following user-visible changes as compared to the 6.3-r2145 release: * Speechd-up has been patched in order to be able to recognize one-character messages from git snapshots of speakup properly. This fixes the inability of visually-impaired users to examine punctuation characters (which are important, e.g., in command lines) and to distinguish between upper- and lowercase letters. * The Xfmedia player has been replaced with gxine due to old video-related bugs and upstream inactivity. If these changes don't affect your setup, there is no reason to upgrade. Note that the speechd-up and speech-dispatcher maintainer has not responded timely to my bug report about the "speech-dispatcher listens on 0.0.0.0 without access controls" security issue and has not responded at all to the "single character messages are not recognized as such" bug report (http://lists.freebsoft.org/pipermail/speechd/2007q4/000977.html). The last commit in speechd-up repository happened 9 months ago. There are severe problems (kernel crashes!) with copying and pasting text in speakup. This, essentially, means that speakup is dead and will not be present on the next development version of the LFS LiveCD. Feedback is requested from its users about possible migration paths: 1) replace speakup with a purely userspace solution for a console screen reader (yasr, http://yasr.sourceforge.net/) 2) maybe stop providing accessibility on the console (and note that Fedora 8 doesn't provide console screen readers at all), default to starting X, and use one of GNOME-based screen readers. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Problem with iconv ?
Ken Moffat wrote: > on all my other builds I use UTF-8 for manpages. How? -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg-7.2 and NLS
Andrey wrote: > Here is the lists of fonts. May be you see something... Sure - there are no fonts with Cyrillic characters available via Xft! I.e., your action about placing Windows fonts into /usr/share/fouts had no effect, because your fontconfig doesn't look there. Please reinstall it, verify that /usr/share/fonts is mentioned somewhere under /etc/fonts, and read the "Setting up Fonts" section on the "X Window System Components" page. BTW, posting 7z attachments to linux lists is considered extremely rude. Plain text attachments should be sent uncompressed. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg-7.2 and NLS
Andrey wrote: > Yes, I use xterm. But I tried kdm too, with full KDE. No difference. Hm. Strange. Are you saying that e.g. Konqueror cannot display Cyrillic characters, too? What about gtk2 apps such as Firefox? > I cannot show full list of fonts, that shown by fc-list or xlsfonts. As I > said earlier, I use vmware, and copy-paste of course don't work. Typing by > hand is very difficult, list is too long. Not a good excuse. You can redirect the output of those commands to files, and transfer them to the outside world. You can also build mutt inside VMware and use it to mail these files to the list. > I added some windows core fonts to /usr/share/fonts , run 'mkfontdir' and > 'mkfontscale', and 'fc-cache'. It doesn't help too. Good step, bad result. If you want help, you either have to post the list of Xft fonts (and the output of "fc-match Monospace" command) to the list, or try to debug the issue with me privately (Jabber: [EMAIL PROTECTED], VoIP: [EMAIL PROTECTED], use the free 3CX phone from http://www.3cx.com/VOIP/voip-phone.html if you have only Windows). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg-7.2 and NLS
Andrey wrote: > 1) Summary: all fields "ru_RU.KOI8-R" > 2) just type "startx" in console OK > 3) any application. For example I run midnight commander. All menus just a > squares. > 4) Only basic fonts that installed after building xorg according to BLFS > book. > A year ago I read a tip "add /usr/lib/X11/fonts/cyrillic to xorg.conf in > section files". I did it, but same result. This only affects core-protocol fonts, while you are probably interested in Xft fonts. And you, again, provided not all required information. 3) It is important to know the name of the terminal emulator. xterm? 4) "Only basic fonts" is not the answer. You can obtain the list of core X fonts with the following command: xlsfonts You can obtain the list of Xft fonts with the following command: fc-list The BLFS book configures all terminal emulators (possibly except recently-added rxvt-unicode) to use Xft fonts. Unfortunately, scalable fonts that come with Xorg have no Cyrillic glyphs. For this reason, the book relies on third-party fonts to cover national alphabets. Thus, you should install DejaVu fonts for Cyrillic glyphs in Xft-based applications. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg-7.2 and NLS
Andrey wrote: > Yesterday I did all compiling for my new BLFS system. After that I tried to > run it in VMWARE. It is good that I'm not use it in my real computer. All > NLS in X window system disappeared. > I use KOI8-R as my system locale. In X I see only squares, but only for my > native language. All English strings displayed correctly. > May be someone know, how to beat that? All system is almost ready. I don't > want to rebuild all. You provided no required information. 1) the output of the "locale" command 2) the way you start X 3) the list of applications that display squares 4) the list of installed fonts. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
LFS LiveCD 6.3-r2145 released
A new version of LFS LiveCD (6.3-r2145) is available for download: http://kerrek.linuxfromscratch.org/pub/lfs-livecd/ As usual, the CD is available for x86 and x86_64 architectures. The CD contains the LFS book (version 6.3 for x86 CD, and its unofficial x86_64 equivalent for x86_64), and a lot of software that you need to build your own LFS system, ask for support and perform various recovery tasks. The -min CD contains only text-mode environment, -nosrc adds X window system and the Xfce desktop, while the full CD also contains all source tarballs. If the old 6.3-r2130 CD works for you, there is no need to upgrade, as the changes are minimal: * jhalfs (a tool for automating the LFS build) was patched so that it now checks host requirements correctly (with r2130, it was unusable in French locales, because it extracted findutils version incorrectly and thought that it was too old) * some dead symlinks were removed * module-init-tools manual pages are now installed in the correct location * documentation has been updated with instructions how to change the default boot arguments, and a recommendation has been added for visually-impaired users to do this instead of typing long boot command lines blindly in (unfamiliar to some of them) US English keyboard layout. -- Alexander E. Patrakov P.S. I am really disappointed that I still have to patch jhalfs manually instead of just updating it, even though the issue has been reported (with patch!) 3 weeks ago: http://linuxfromscratch.org/pipermail/alfs-discuss/2007-November/009609.html. Is the ALFS project dead? The last commit was 2 months ago. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Pseudo TTy's deprecated
TheOldFellow wrote: > Having just built a new LFS/Xorg using mostly book methods, I came to > configure the kernel (2.6.23.8) and running oldconfig suggested that > the BSD-style Ptys were deprecated - I seem to remember this has been > the case for some while. However, I don't know of an X-terminal > emulator that doesn't need ptys, and they are used for other purposes. BSD-style ptys are things like /dev/ptyp8. They are indeed deprecated, and /dev/pts/* should be used instead (and the openpty() function from glibc, that xterm calls, does the right thing with /dev/pts). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: wvstreams configure fails with "openssl>=0.9.7 required"
Randy McMurchy wrote: > I'm not sure that anyone can help you here. I can't imagine > anyone that contributes to BLFS-Support having dial-up capability > any longer. I know I don't even have an analog line capable of > doing dial-up, much less a modem. GPRS is very much like dialup, and I still use it as a backup Internet connection when ADSL fails. However, I don't use wv* programs. Please see http://wiki.linuxfromscratch.org/blfs/wiki/PPP or inspect the (very similar) LiveCD setup. Maybe it is time to integrate this information (and the updated pppoe service) into BLFS, if the "removal" proposal doesn't pass? Note, however, that the whole "Dealing with Passwords" section in the Wiki has been created only to deal with Randy's objection to storing password in cleartext even in a file with 0600 permissions. There was at least one report that the "pcall" method doesn't work. I store my ADSL password in /etc/ppp/pap-secets and don't use pcall. > I don't know what to do here, so I suppose this is a call for > advice. If there is nobody who uses, or can support dial-up, > dial-up support probably should be removed from the book. Well, I am always for removing unsupportable parts of the book. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: X over SSH not working?
Alan Lord wrote: > Hi all, > > I need some help to point me in the right direction here: > > I have built a home server based on LFS6.3 and some BLFS including > xorg-7.2. All works fine and dandy on the local console. > > My main desktop machine is Ubuntu 7.04. I can connect to "other" servers > using ssh (such as the Debian Etch managed virtual machine that hosts > all our web sites etc) and successfully run remote X applications on my > Ubuntu desktop. > > However, when I ssh into my new LFS/BLFS server I keep getting the > dreaded "Error: Can't open display:". sshd hard-codes the path to the "xauth" command, and defaults to /usr/X11R6/bin/xauth. Rebuilding it after Xorg should help. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: howto create LFS livecd
Lauri Kasanen wrote: > can someone tell me how is the device mapper used in the lfs livecd? device-mapper is used for pretending that the CD is fully read-write. This is done with the help of two loop devices. /dev/loop0 is set on top of the root.ext2 file, /dev/loop1 is set on top of a huge (1.5 GB) but sparse (i.e., zeroed-out regions don't take space) file in /dev/shm. Then a "snapshot" device-mapper target is used to combine the two loops into a /dev/mapper/lfs-cd device that has initially the same content as root.ext2, but can be written to (changed sectors go to /dev/loop1, i.e. to tmpfs) > I'm also interested in why was it chosen. Another alternative is to use unionfs in order to pretend that the CD is fully RW. The 6.1.1 CD did this, but it was easy to cause a kernel oops or a full system hang. unionfs has been removed exactly because it was too buggy and unstable. > At least on my older > laptops it tends to cause some "block reading errors" that fill the > screen, and then the system hangs. On the other hand, cloop-based > livecd's work fine on these lappies too.. Cloop was never used on LFS LiveCDs. Older CDs used squashfs for compression, while modern CDs use zisofs. The change was done because, with device-mapper based RW mechanism, one has to compress a single file (root.ext2), not a full filesystem tree, and both zisofs and squashfs are equally efficient for that - but zisofs is simpler to set up. BTW, this question is more suitable for the [EMAIL PROTECTED] list. And please post the errors from the latest LFS LiveCD (r2052 or self-built from SVN trunk) - they are bugs. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: howto create LFS livecd
[EMAIL PROTECTED] wrote: > I compiled my LFS system now. and I think to create a LiveCD . > Howto create it? There is no fully-working and documented procedure to do this. I.e.: you can't. > I readed lfscd-remastering-howto ,but i don't know howto create the > root.ext2 file. This is a wrong HOWTO. It is about making changes to the already-existing official LiveCD, not about creating your own. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Older system, older network cards, which one is better
Justin The Cynical wrote: > I have an older system running LFS 6.1 (currently using kernel 2.6.16.32 > P4), and I am finding the on-board SiS900 network chip to be lacking. > > I have sitting around in my parts collection an HP J3171-60011 PCI LAN > 10/100TX card (Intel 82557 based) and a 3Com 3C905-TX. > > I've looked around, and I can't seem to find any performance information > on these two cards. Both are well supported, but which one would be the > better choice to use? Tests at the Urals State University show that on "Pentium MMX 166 MHz with 128 MB of RAM" machines, "3COM 3C905B" and "Intel Pro 100" cards behave approximately identically, and are both slower than a cheap Realtek 8139 card. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote: > I rebuilt xorg-server, the freetype module appears to be there, but my > fonts still look like junk. I'm going to try and compile Qt and KDE > and see if compiling without Xft works, as I recall this solved this > problem in the past. Please don't do this - it is a workaround. Xft fonts are beautiful by default, but something went wrong on your system. You may also want to start from scratch, using LFS-6.3 and BLFS SVN, because they contain a lot of fixes as compared to the 6.2 release. > I really just don't know why these fonts are > being so stubborn, but like I said, I'll probably kick myself when I > figure it out. You may also want to download LFS LiveCD and examine and/or copy the setup from it. It relies solely on Xft, and has no bitmap fonts (even though this requires a one-line xorg-7.2 patch - but xorg-7.3 has a configure option for this). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Missing memory?
Simon Geard wrote: > However, booted into Linux, more than half of it isn't being seen. 'top' > reports 906740k, which is also the figure listed in /proc/meminfo. You have to configure the kernel with CONFIG_HIGHMEM4G=y -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote: > Noticed something interesting in my xorg logs: > (EE) Failed to load module "freetype" (module does not exist, 0) This means that old applications that use the X core font protocol (instead of xft) will not be able to use TrueType fonts). This should not affect Qt or GTK, however. BTW, could you please post the output of "ldd /usr/lib/libgtk-x11-2.0.so" and the same for the qt library? > (EE) Failed to load module "xft" (module does not exist, 0) This module is not supposed to exist. Xft works on top of the RENDER extension, or as a completely client-side thingy. > I tried to 'find /usr -name "*freetype*"' it didn't find anything in > /usr/lib/X11/modules/font! What gives? let me try reinstalling > freetype *again* and see if I can't figure this out. Do I need to > recompile some part of Xorg to get this working now? Yes, freetype installs only /usr/lib/libfreetype.so (so you don't need to recompile freetype), and the needed module comes from Xorg. > I'm having > problems with this as there's a perl module dependancy broken somehow > and I can't seem to compile anything needing perl anymore! Ack, my > lfs is falling apart (or I'm just really good at breaking it hehe)! Starting from scratch again is always an option. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote: > Hmm... fontconfig didn't seem to install anything resembling > "conf.avail" when I ran "make install", nor does it seem to be > mentioned in the book. What did I miss? Which version of fontconfig? And please tar up /etc/fonts again. Also check if merely removing /etc/fonts/conf.d improves the situation. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote: > Hmm... fontconfig didn't seem to install anything resembling > "conf.avail" when I ran "make install", nor does it seem to be > mentioned in the book. What did I miss? Which version of fontconfig? And please tar up /etc/fonts again. Also check if merely removing /etc/fonts/conf.d improves the situation. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote: > The font in the Xterm seems to look just fine (am I wrong?), but the > fonts in the browser are just heart-breaking. You are comparing apples and oranges. xterm in your configuration seems to be using a bitmap font via X core font protocol. This is a valid configuration, although the book configures it to use a TrueType font via XFT with antialiasing. Konqueror prefers XFT, but seems to fall back to non-antialiased fonts. > I must be missing > something obvious, and I'm sure I'll have the Dept. of Redundancy > Dept. bonk me several times repeatedly when some kind soul points it > out to me :). Impossible to say without actually looking at your setup. Please provide the following information: 1) The same screenshot with any GTK2 application added (to distinguish between "miscompiled qt" and "misconfigured fontconfig"). 2) The output of "fc-list". 3) Tar up the contents of /etc/fonts and attach. 4) Attach xorg.conf. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Martin McCourt wrote:
> I followed the advisories of Alexander and Ken
I did my suggestion without actually looking at your screenshots. The
problem is that your fontconfig uses bitmap fonts (or toolkits use core
X fonts instead of those provided by fontconfig) instead of nice
TrueType outlines. This is exactly why BLFS puts X fonts into a
directory where fontconfig doesn't look.
In summary:
0) clean up the mess in /usr/share/fonts by deleting this directory
completely, and make sure that /usr/lib/X11/fonts is not mentioned
anywhere under /etc/fonts
1) install X fonts into /usr/lib/X11/fonts, don't mention this directory
in fontconfig configuration
2) Make (if desired) symlinks /usr/share/fonts/X11-{TTF,OTF} ->
/usr/lib/X11/fonts/{TTF,OTF}
3) Install DejaVu fonts into /usr/share/fonts/dejavu
--
A. E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: Fonts look bad in X.org
Ken Moffat wrote: > On Wed, Sep 12, 2007 at 01:10:44PM -0400, Martin McCourt wrote: >> Hi all, >> >> I have this problem where all my fonts look really small and actually >> are pretty hard to look at. See this: >> http://img528.imageshack.us/my.php?image=badaavg8.jpg > > Possibly a red herring (and nothing to do with aliasing), but on > exactly one of my machines it somehow thinks that 75dpi is a > sensible size for a 20" 1600x1200 display (all the others find, > and use, 100dpi). For my own desktop, adding the following to > ~/.Xresources fixes e.g. the toolbar in icewm, and font size in > urxvt. > > Xft.dpi: 100 > > Note that doesn't cover all the cases - I use gdm, so that still > presents the login dialog with a tiny font, but I can live with that > as long as there is only me using it. A better solution is to specify the display size in Xorg.conf. This way, it will apply to all programs, not only Xft. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
The first CD with LFS-6.3
The LiveCD team is proud to announce the release of the first LiveCD with LFS-6.3 book for x86-based computers. The CD is available at: http://kerrek.linuxfromscratch.org/pub/lfs-livecd/lfslivecd-x86-6.3-min-r2052.iso (214 MB) - the minimal CD without sources and without X window system. http://kerrek.linuxfromscratch.org/pub/lfs-livecd/lfslivecd-x86-6.3-r2052-nosrc.iso (424 MB) - this CD is identical to the full CD in all aspects, except that it doesn't contain tarballs with LFS sources. http://kerrek.linuxfromscratch.org/pub/lfs-livecd/lfslivecd-x86-6.3-r2052.iso (622 MB) - the full CD. There are some known bugs and missing features that were present in the 6.2 series of CDs (e.g., no loop-aes support), thus the CDs are not called "stable", but for most people, they should be good enough to build LFS-6.3 and ask for support online. The full CD contains X window system, a visually pleasing desktop environment (Xfce), the Seamonkey web browser, the LFS book, the multiprotocol chat client (GAIM), and many console programs. It also includes jhalfs-2.3.1 - a tool that automates building of LFS, HLFS and CLFS systems. You can add software of your own choice by compiling and installing it directly on the CD, or by remastering the CD image. New features as compared to 6.2 CDs: * Improved support for wireless network cards * Added wpa_supplicant * Added support for IBM JFS filesystem * Added proprietary video drivers * Made it possible to forcibly load, blacklist and specify arbitrary options for kernel modules from the boot prompt * Added experimental libata-based drivers for IDE devices (add "pata" to the boot prompt to activate them, old drivers are used without this option) * Added the ability to load the entire CD contents (except sources and proprietary drivers) to RAM. This requires at least 512 MB of RAM, allows the use of the CD drive for reading other CDs, and makes the programs on the LiveCD operate much faster. * Added the ability to boot the CD image on systems without the CD-ROM drive (slightly broken on FAT and NTFS filesystems due to this bug: http://wiki.linuxfromscratch.org/livecd/ticket/1749) More information: The README file: http://wiki.linuxfromscratch.org/livecd/browser/trunk/doc/README?rev=2052 Boot loader help screens: http://wiki.linuxfromscratch.org/livecd/browser/trunk/isolinux/?rev=2052 The remastering HOWTO: http://wiki.linuxfromscratch.org/livecd/browser/trunk/doc/lfscd-remastering-howto.txt?rev=2052 In order for the stable release to be possible, the following bug has to be either confirmed and fixed, or proven to be non-existing on the CD: http://wiki.linuxfromscratch.org/livecd/ticket/1736 (see also questions at http://linuxfromscratch.org/pipermail/livecd/2007-August/004680.html). The best thing you can do is to reproduce the "no available native/legacy port" error message with your own self-compiled kernel (so that we know that the kernel problem is reproducible at all with your hardware), boot the CD without the "pata" argument and see whether this error message appears. Also, we need to know whether your self-compiled kernel can be configured to assign the same device names to IDE and SATA disks as the LiveCD, because of the complicated initramfs setup on the CD that plays dirty tricks with the module loading order. Please send your feedback to the livecd list. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
New LiveCD for testing: r2014
Hello, I have put 4 new CD images for testing on ums.usu.ru: http://ums.usu.ru/~patrakov/test/lfslivecd-x86-6.3-r2014.iso http://ums.usu.ru/~patrakov/test/lfslivecd-x86-6.3-r2014-nosrc.iso http://ums.usu.ru/~patrakov/test/lfslivecd-x86_64-6.3-r2014.iso http://ums.usu.ru/~patrakov/test/lfslivecd-x86_64-6.3-r2014-nosrc.iso The x86 versions, as before, contain both 32-bit (linux) and 64-bit (linux64) kernels and 32-bit userspace. The x86_64 version contains 64-bit kernel and 64-bit userspace, and should be used with CLFS or this book: http://www.linuxfromscratch.org/~jhuntwork/lfs-x86_64/ Since LFS is close to its 6.3 release, these CD versions should give you a close approximation of what will become the 6.3 LiveCD release, if any. I say "if any" because of bugs and untested updates that I don't want to go into a "stable" release. The worst possibly-existing non-documentation bug is http://wiki.linuxfromscratch.org/livecd/ticket/1736, but if you have information about other bugs, please share it too. Owners of motherboards with Intel IDE and/or SATA controllers can help by downloading and testing the CD, and answering the following questions: 1) How (i.e., with IDE or SATA) are your CD-ROMs and hard disks connected? 2) Does your motherboard use Intel or third-party (e.g., Marvel) chipset for the CD-ROM and hard disk connection? Please provide the output of "lspci" if in doubt. If there is something like "Native/AHCI" chioce in the BIOS, please test both options. 3) Does the LFS LiveCD detect the CD drive and the hard disk without special boot arguments? What are the device nodes created for them? What are the device nodes created for them in other distributions and/or your self-compiled kernels (please name the distribution and its version)? 4) Do "IDE port busy" or "failed to request/iomap BARs" errors like those in http://lkml.org/lkml/2007/7/20/171 appear in the "dmesg" command output? Does "hdparm -tT" display reasonable speed for your hard drives? 5) Does the CD boot if you type "linux pata" at the boot loader screen instead of just pressing Enter? Any errors in dmesg? What are the device nodes? If at least one of "Intel IDE only", "mixed Intel IDE/SATA", "Intel SATA plus third party IDE" and "Intel SATA only" case doesn't get covered by your replies, there will be no LFS-6.3 LiveCD. To get this task easier for you, here are my answers: 1) Both the CD-ROM and the hard disk are connected to SATA ports. 2) SATA controller is by Intel (both "Legacy", "Native" and "AHCI" options work fine), IDE is by Marvel (but there is nothing connected to the IDE port, so can't test). 3) Yes, /dev/sda is the hard drive, /dev/scd0 is the CD-ROM in both cases, the hard disk speed is around 80 MB/s, the same applies to Debian Lenny. 4) No such messages, probably because they are supposed to appear only in the "mixed Intel IDE/SATA" case. 5) Yes, it works exactly the same as without the "pata" option. Another task for testers: open http://wiki.linuxfromscratch.org/livecd/browser/trunk/packages, and test each package if you can. Send reports what works (even if you don't find any bugs - simply for me to see what users test and what they don't) and what doesn't. Untested packages are likely to be removed in the future. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Good video editor for linux?
Lauri Kasanen wrote: > Hi > > It seems there are three good video editors for Linux: Kino, Lives and > Avidemux. and there is one more: Cinelerra. It does work for joining several videos into one with transition effects (just drop the needed pieces from the viewer window onto the timeline, and then drop the needed effects from the resources window onto junction points). It also does non-linear tricks, e.g., with color keys, projections and transparency. http://cvs.cinelerra.org/ Unfortunately, Cinelerra crashes often, so don't forget to save your project after each change. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Trouble with the bootscript called console (/etc/init.d/console)
Guillermo Ibañez wrote: > the prompt flash pointer jumps a few lines before the right position, I mean > the cursor places where it has not have to place. I think it is related with > my locale configuration and with the UTF on the console output. Please try the following in /etc/sysconfig/console: FONT="lat1-16 -m 8859-1" KEYMAP="es" UTF-8 on console should not be used with non-UTF-8 locale (in other words, the console bootscript informs the kernel about locale settings, and "I use UTF-8" would be a misinformation). > using Framebuffer Which driver? Could you please try without framebuffer and report whether the console works correctly? Many framebuffer drivers are just broken beyond repair when a font change occurs. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: links graphics browers fonts look horrible
Alan wrote: > what's the deal with default settings on links and the font it chooses? > Mine looks TERRIBLE, the menus are non legible. Google can be made out, > but looks like an inkjet print-out with missing nozzles. links contains its own bitmap fonts that it tries to downscale to the required resolution. IOW: this is not configurable. > Also it does not render pages correctly, slashdot runs down the page > instead of being in it's frames. Links the wrong choice for browsing now? Slashdot uses CSS for layout, and links doesn't support it. Please install Firefox or Konqueror. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: BRLTTY on LFS LiveCD
I wrote: > If I do not get a reply within 10 days, brltty (and, thus, support for > blind users) will be removed from the CD. r1908 contains a workaround, but better solutions (and testers) are still welcome. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
BRLTTY on LFS LiveCD
Hello, booting the latest development LFS LiveCD (plus brltty-3.8) as "linux brltty=,usb:," (as one wants for autodetection of a USB braille device) results in the following bug: 1) brltty does not find usbfs anywhere and mounts it on /etc/brltty/usbfs (what a non-standard location!) without also registering it in /etc/mtab. Adding usbfs to /etc/fstab does not solve this issue. 2) when shutting the system down, the bootscripts do not unmount this filesystem, because it is not in /etc/mtab. 3) this prevents proper unmounting of the root filesystem (a LiveCD-specific step, see scripts/shutdown-helper or, equivalently, /usr/bin/shutdown-helper) and makes ejecting the CD impossible. It even prints "please report this as a bug" - but nobody did this, presumably because one must be sighted in order to see this request. You can try this boot line even without a braille device in order to reproduce the bug. If I do not get a reply within 10 days, brltty (and, thus, support for blind users) will be removed from the CD. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Flash-9 with GLIB-2 Problem
[EMAIL PROTECTED] wrote: > all "locale" parameters: "en_US.iso88591". > > I upgraded FlashPlayer from 7 to 9 (mostly to keep up with the Joneses and to > avoid various site warnings about my Flash being behind the times.) > Ever since, when I exit XFree to come back to the command line I am greeted > by a ton of messages, > > "Flash Player: Warning: environment variable G_FILENAME_ENCODING is set and > is not UTF-8" > > It seems the new Flash (or Adobe?) is more picky about the 'i18n.sh' line > recommended in BLFS GLIB-2.12.x instructions: > [EMAIL PROTECTED] > > Not surprisingly, if I set G_FILENAME_ENCODING to UTF-8 everything quiets > down. but if you save a file from any GTK2 app and then try to "ls" it in the terminal, the file name will be shown incorrectly if it contains any non-ASCII character. So, please ignore this stupid and pointless warning. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Kernel 2.6.21 BUG?
[EMAIL PROTECTED] wrote: > Hi Tijnema, > > On 4/29/07, Tijnema ! <[EMAIL PROTECTED]> wrote: > >> It might be useful to post the whole log, and not only a few lines. > > If by "log" you mean the whole screen, I tried to avoid it for two reasons > 1. It's hard to write down and then retype all those 24 lines (I have no idea > how to do otherwise (snapshot, etc.) with a frozen machine.) Next time you hit this need, please use a digital camera, and upload the resulting jpeg file. No need to retype anything. And no need to do this now, because the issue is probably already fixed. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Kernel 2.6.21 BUG?
[EMAIL PROTECTED] wrote: > 5. Call Trace: > [] sis900_interrupt+0x625/0x6a0 [sis900] Should be fixed by this patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=dc5a144991ba803bc8afded105c9db1dea0e57ab -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Kernel 2.6.21 BUG?
[EMAIL PROTECTED] wrote: > - Alex: > > BUG: at arch/i386/kernel/smp.c:546 smp_call_function() > [] smp_call_function+0x125/0x130 > [<>] ... (Twenty lines of this nature) > [] unknown_bootoption+0x0/0x240 > > > - Alexander E. Patrakov: > 1) Tell us the kernel line in your /boot/grub/menu.lst file > 2) Complain to LKML <[EMAIL PROTECTED]>, >including 20 lines that you snipped > - 1. I haven't touched that line since day one. > I don't see its relevance, anyway. It is relevant because "unknown_bootoption" is in the oops. > Even the 2.6.21 kernel works fine until you try to network. Try 2.6.21.1, it includes a network-related fix. If it still doesn't work, complain to LKML. > 2. I'm a little shell-shocked since I last got entangled into kernel > mailing lists. To avoid re-traumatization, you may want to submit a draft of your bug report to me before actually sending it to LKML. Just be sure to include all lines that are present in your oops (and a bit of information above and below that), and the kernel line in menu.lst. Then you will be probably asked to rebuild your kernel several times with exactly the same config, but with some patches applied. > It still crashes on the slightest attempt at networking > but at least with a different screen - to keep it interesting, I suppose. > FWIW (maybe an astute and generous soul might detect a clue): > 1. Process swapper ... > 2. Stack ... > ... > 5. Call Trace: > [] sis900_interrupt+0x625/0x6a0 [sis900] > ... > 19. === > 20. Code: 00 ... > 24 ... > c0 ... > 23. EIP [c024fde4] skb_over_panic+0x54/0x60 SS:ESP 0068:c033debc > 24. Kernel panic - not syncing: Fatal exception in interrupt The usual procedure to deal with this is to either use a serial console, or take a photo of the screen, upload somewhere, and post a link to it. But I see that the sis problem is already reported by Brett Ryland, so you may just sit and watch. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Kernel 2.6.21 BUG?
[EMAIL PROTECTED] wrote: > BUG: at arch/i386/kernel/smp.c:546 smp_call_function() > [] smp_call_function+0x125/0x130 > [<>] ... (Twenty lines of this nature) > [] unknown_bootoption+0x0/0x240 1) Tell us the kernel line in your /boot/grub/menu.lst file 2) Complain to LKML <[EMAIL PROTECTED]>, including 20 lines that you snipped -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xfce 4.4
Scott Czepiel wrote: > I highly recommend Xfce 4.4 to anyone looking for a fast, > asethetically-pleasing, and bloat-free alternative to Gnome and Kde. > 4.4 has many new features over version 4.2 which is currently featured > in the BLFS book, and it is well worth upgrading. If you already enjoy > 4.2, you will love 4.4. > > I had a trouble-free installation loosely following the existing > directions in the book, and adapting them according to Wilco's excellent > suggestions here: > > http://linuxfromscratch.org/pipermail/blfs-support/2007-March/062206.html > > Three things I had to do that weren't readily obvious in the documentation: > > 1) vte 0.12.2 complained loudly (aka. make error) about not having > python until explicitly passing --disable-python to configure. > > 2) vte also complained about not being able to find a freetype header > file. The code was looking in /usr/include/freetype, but my freetype > installation has put its header files into > /usr/include/freetype2/freetype. I assume other applications will run > into the same expectation so my solution was to simply make a symlink > that everybody will be happy with: > > cd /usr/include && > ln -sv freetype2/freetype freetype > > 3) D-BUS. If you haven't done this already, you need to do the > following so that dbus will run correctly. I found this out after > suffering repainting issues that made the desktop unusable. Running > this fixed the problems: > > dbus-uuidgen --ensure > > Read 'man dbus-uuidgen' for details. Thanks for your findings. Unfortunately, I was very busy with my scientific work until today in order to update xfce in the book. Maybe I will do the update on this week. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xvnc and GLX
Dan Nicholson wrote: > CVS sucks. Try again, but don't use -z3 for the checkout. I've solved > many "hangs" by just not using the compression. I suspect that this might be related to our patch that makes CVS use unmodified external zlib instead of its own _modified_ copy. Nobody else does this (even Debian, using a private copy of any library without very significant reason is a release-critical bug). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPP and module autoloading
MickB wrote: > Ken Moffat a écrit : >> If the modules aren't loaded by the time you can log in (use lsmod >> to check), and they *are* installed in /lib/modules, add them to >> /etc/sysconfig/modules. >> >> ĸen >> > > Thanks for your answer. Indeed, I'll always have this solution. But > that's not elegant :))) Please read http://www.linuxfromscratch.org/lfs/view/6.2/chapter07/udev.html and note that "Udev is not intended to load “wrapper” drivers such as snd-pcm-oss and non-hardware drivers such as loop at all". So /etc/sysconfig/modules is the correct solution for a non-hardware "pppoe" module. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Setup of serial port
Koenraad Lelong wrote: > Hi, > I think this is for this list, if not please redirect me to the proper list. > I have 4 serial ports in my embedded PC. /dev/ttyS0 and /dev/ttyS1 work > fine, but /dev/ttyS2 and /dev/ttyS3 don't work. In the BIOS I found that > those ports have "non-standard" interrupts : 11 and 10. Since I need > ttyS3 for the touchscreen this is rather important. > How can I set those ? I think this belongs to the udev rules but I don't > know how to do this. Any pointers ? udev is completely irrelevant here. Look for the "setserial" program. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
LFS LiveCD 6.2-5 released
The new stable version of LFS LiveCD, 6.2-5, is out. The only changes to the 6.2-4 CD are: * Updated the kernel to 2.6.16.38 * Reflected this in the book * Fixed inability of Midnight Commander to go into directories containing the underscore character * Fixed GRUB non-installability to filesystems different from reiser4 (reported by Piet Delaney) Temporary location: http://ums.usu.ru/~patrakov/x86/lfslivecd-x86-6.2-5.iso MD5: e7a075618b257c91eae196eb05b12263 SHA1: b3e3947bf2c3616fa45541c0643a2adfa0618207 When the mirroring process finishes, the CD will be available from the usual places mentioned at http://www.linuxfromscratch.org/livecd/download.html Please mail any comments to the LFS LiveCD mailing list. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Bootable CD with X-Window
Dan Nicholson wrote:
> VESA is pretty solid for getting X to "just work" with minimal
> configuration and some additional power over VGA. The only real draw
> back is no acceleration of any kind and no DRI.
No. Please find someone with a NEC 20WGX2 monitor connected via DVI, and try
VESA on it.
1) VESA doesn't have the mode that matches the native 1680x1050 resolution.
Thus, the monitor will display the annoying "please change resolution" box
every minute, and there is no way to disable this annoyance.
2) Even if one ignores the box, the image still has the wrong aspect ratio.
So, because of such beasts, the "least common denominator" approach (even
with VESA) is busted. The only way forward is to detect the video card by
its PCI ID, and the official LFS LiveCD does this.
Just to be fair, I couldn't configure Windows Vista RC1 to use the native
resolution either (bug reported to Microsoft).
> Regardless of which implementation you want to support, the key is
> that you need support in the kernel and in X. So, you basically want
> to compile all the graphics and all the input drivers in the kernel
> (preferably as modules). Then, you want to build all the drivers for
> X. Those are the xf86-{input,video}-* packages. Some of them you could
> probably ignore, but it doesn't hurt to have them installed.
It does hurt. Sun-specific modules such as xf86-video-sunleo make "X
-configure" always fail.
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: XFCE in non-english language (french for example) [solved]
Thierry Nuttens wrote: > It took me quite a bit of time to find where to put this. > Even I google it there always unsolved solutions. That's why, I bring here > the solution which did work for me at least. > > export LC_LANG=fr_CH > export LC_LANG=fr_CH > ... > In XFCE 4, it is necessairy to have a xinitrc file in > ~/.config/xfce4/xinitrc which is going to hold this > variable definition. > Hurra, all the applications which support the native language of xfce are > switching to the right language except > OpenOffice Rrrr! Because you didn't read the book correctlly. The correct variable is LANG, not LC_LANG, and it has to be in either /etc/profile or /etc/profile.d/i18n.sh. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xorg7.1 graphics chip errors
Pedro Rodrigues de Figueiredo wrote: > - (WW) Open ACPI failed (/var/run/acpid.socket) (No such file or > directory) [log, line 35] > ACPI is enabled in the kernel... Is this something I should look into? This is only a warning. It is safe to ignore, although installing acpid will remove it (warning: I don't know if acpid and HAL tolerate the presence of each other). > - (EE) I810(0): [dri] DRIScreenInit failed. Disabling DRI. [line 2158] > - (EE) AIGLX: Screen 0 is not DRI capable [line 2212] > I manually added DRI to xorg.conf, since it was autodetected on several > other linux systems that I had running on this hardware. Apparently it's > not working here... If I comment out the DRI section, these go away. I > don't even know what DRI is, but since the other systems had it on... DRI is needed for accelerated 3d graphics, so that you can, e.g., play Quake. > - [drm] failed to load kernel module "i915" [line 2156] > This is the one that's bothering me most, as it appears even if I switch > off DRI. I don't see any i915 stuff on the kernel, and, as far as i > know, the i810 code should be loaded instead. no, your chip is new and has to use the i915 kernel module. Look at character devices -> DRM. If i915 is not there, you have to build a new kernel (e.g. 2.6.19.2). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xkeyboard-config
Ken Moffat wrote: > One for the archives (I would have put it straight into the wiki, > but I wasn't sure where: I couldn't find xkeyboard-config among the > apps, libraries, or data for the xorg releases). > > If you are using xkeyboard-config-0.9 you may hit the following > problem when you come to build xorg-server: > > make[2]: Entering directory `/usr/src/xorg-server-1.1.1/xkb' > make[2]: Nothing to be done for `install-exec-am'. > test -z "/usr/lib/X11/xkb/compiled" || mkdir -p -- > "/usr/lib/X11/xkb/compiled" > mkdir: cannot create directory `/usr/lib/X11/xkb/compiled': File > exists > make[2]: *** [install-dist_xkbcompiledDATA] Error 1 > make[2]: Leaving directory `/usr/src/xorg-server-1.1.1/xkb' > make[1]: *** [install-am] Error 2 > make[1]: Leaving directory `/usr/src/xorg-server-1.1.1/xkb' > make: *** [install-recursive] Error 1 > > This is because the 'compiled' directory is now a symlink (with > xkbdata it didn't exist). It is created by the install-compiled > target of xkeyboard-config at the end of the install. In CVS, > it has been removed from Makefile.am so I think it is safe to delete > it with > > sed -i '/^install-data-hook/s/install-compiled//' Makefile.in > > before running configure. Even better fix is to add --with-xkb-path=/usr/share/X11/xkb --with-xkb-output=/var/lib/xkb to xorg-server. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Can't get kbd to behave the right way !!
Kevin Williams wrote: > Thanks very much for the explanation. I guess that explains why I can't > Ctrl-C > and Ctrl-Z working in single user mode. I just thought that, it should work > in all runlevels as this is the normal behaviour you'll see in all distros. > Maybe, they patch it to enable job control, I guess !! They won't. The problem is that many daemon processes use /dev/console as a last-resort output device. They really don't want to acquire a controlling tty. If you really want a tty, try typing this as your first command in single-user mode: exec /bin/bash /dev/tty1 2>&1 -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Can't get kbd to behave the right way !!
Kevin Williams wrote: > 1) Ctrl+C and Ctrl+Z do not work while in single user mode. But these key > > combinations do work in other runlevels This is simply not supposed to work. In single user mode, the sulogin program starts a shell directly on /dev/console. Job control (with keys such as Ctrl+C and Ctrl+Z) works only when a shell has a controlling tty, but /dev/console in the kernel is deliberately made so that it can never be a controlling tty for anything. > 2) Ctrl+Alt+Fn combination doesn't work when the kdm login screen is > displayed !! The only way I can use Ctrl+Alt+Fn is either to login into > kde or select "Console Login" from the drop down menu and thus terminate > kdm session. While in other runlevels, these key combinations work > without a hitch ! Sorry, I don't use KDM and thus can't help. There was a bug in xorg-server-1.1.0 (fixed in 1.1.1) that sometimes prevented Ctrl+Alt+Fn from working, but what you describe doesn't look like this bug. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: help with firmware loading problems with DBVT device
lux-integ wrote: > However the boot is ugly. Updating LFS bootscripts may help. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: help with firmware loading problems with DBVT device
lux-integ wrote: >> [EMAIL PROTECTED]:# ls -l /lib/udev >> ls: /lib/udev: No such file or directory Here is your problem. Reinstall udev and its rules as the book says (i.e.: update), and the problem will go away. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
LFS LiveCDs 6.2-4 and 6.3-pre1 released
The LFS LiveCD Team is proud to announce the release of x86-6.2-4 and x86-6.3-pre1 versions of the LFS LiveCD. Both versions should be treated as betas, the stable release is still x86-6.2-3. If you are unable to start Xorg on the 6.2-3 CD because your monitor reports its maximum dot clock as 0 MHz, as a workaround, add the "NoDDC" option to the "Device" xorg.conf section that corresponds to your video card. The bug (caused by a LiveCD-specific Xorg patch) is fixed properly in the new CDs. The following features are (temporarily?) dropped: * PATA patches from Alan Cox - not available officially for linux-2.6.18, and nobody is interested in continued testing of the outdated 2.6.16 patch. * Proprietary video drivers: NVIDIA driver crashes even on 2D graphics, ATI driver is too fat. * PACO patch for jhalfs - not available for jhalfs-2.1 The CD images should soon become available from the usual places mentioned at http://www.linuxfromscratch.org/livecd/download.html . For now, download from http://ums.usu.ru/~patrakov/x86/ , but the files will be removed from this temporary location once all the mirrors become up to date. The version 6.2-4 of the CD is built using LFS 6.2 and many BLFS packages from the SVN branch. Source packages for LFS 6.2, and the LFS book itself, are included on the LiveCD. The CD is also suitable as a host for building x86 and x86_64 CLFS systems. The main new feature of this CD is major updates of BLFS packages as compared to the previous (6.2-3) release. Simultaneously, the x86-6.3-pre1 CD is released. Its only difference from the 6.2-4 version is that it is based on LFS SVN (plus upstream gcc fixes and installation of asm/unaligned.h). The BLFS packages are _exactly_ the same as in 6.2-4. It is recommended that you download both versions and compare them. We are especially interested in bugs that can be reproduced only on one of them, because this would indicate either a bug in LFS SVN, or something for the BLFS team to consider (because they still target LFS-6.2). Please mail any comments to the LFS LiveCD mailing list. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Ctrl+Alt+Fn in X
Alberto Hernando wrote: > Hi. > > I've lost the ability to change to console or other user in X with > Ctrl+Alt+Fn. I can do it from console to X, and I could when I had x.org-6.9, > but with x.org-7.1 this seems to have gone. I can use the "switch user" > option in the K menu (I'm using kde-3.5.5), but for some reason, I can't use > the keyboard. I don't know if this is kde related, but already in the kdm > screen I can't change either. I guess that some defalt behaviour has changed, > or there is a bug in x.org-7.1. > > Any tip? Known bug. Upgrade xorg-server. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[ANNOUNCE] LFS LiveCD x86-6.2-3
The LFS LiveCD Team is proud to announce the release of the x86-6.2-3 version of the LFS LiveCD. This version is built using LFS 6.2 and many BLFS packages from the SVN branch. Source packages for LFS 6.2, and the LFS book itself, are included on the LiveCD. The CD is also suitable as a host for building x86 and x86_64 CLFS systems. Mirrors: http://www.linuxfromscratch.org/livecd/download.html In this release, some bugs have been fixed: * Added Italian locales on the language configuration screen * Now it is possible to boot even if a different CD is inserted in another drive * The CD no longer prints a warning about the network configuration when hibernating * Xfsprogs now work * The cpqarray driver is now functional (the official fix is scheduled for linux-2.6.16.29) New features: * Reiser4 filesystem support in GRUB * The loop-AES driver is included in the kernel If you have not hit any of the bugs mentioned above, and don't need new features, there is no reason to upgrade. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[ANNOUNCE] LFS LiveCD x86-6.2-2
The LFS LiveCD Team is proud to announce the release of the x86-6.2-2 version of the LFS LiveCD. This version is built using LFS 6.2 and many BLFS packages from the SVN branch. Source packages for LFS 6.2, and the LFS book itself, are included on the LiveCD. Mirrors: http://www.linuxfromscratch.org/livecd/download.html The main change is that the CD now includes a 64-bit kernel for x86_64 (type "linux64" at the boot prompt). This makes it possible to use the "chroot" scenario from the CLFS book when building a CLFS x86_64 system (either pure 64-bit, or multilib) from this CD. Userspace on the CD is still 32-bit, and the old 32-bit kernel is still available for those people who have 32-bit PCs. Other changes: * Irssi now in understands the "/charset" command as an alias of "/recode add". So, only GAIM doesn't understand this command now on the CD. * Lilo has been upgraded to a version that has the device mapper patch integrated. * Keymap configuration: fixed English (GB), added Dutch (Belgium). If the old 6.2-1 CD works for you, there is no reason to upgrade. Thanks to defumar from #lfs-support and to Joe Ciccone for pre-release testing of the 64-bit kernel. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [ANNOUNCE] LFS LiveCD x86-6.2-1 - bug
I wrote: The LFS LiveCD Team is proud to announce the release of the x86-6.2-1 version of the LFS LiveCD. Sorry, there is a bug with British keymap: "uk" and "gb" should be swapped on the second keymap related screen. This will be fixed in the next release, coming in a week together with fixes for other reported bugs, if any. For now, British users should just edit the "Console keymap" and "X keymap" fields on the console configuration screen. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[ANNOUNCE] LFS LiveCD x86-6.2-1
The LFS LiveCD Team is proud to announce the release of the x86-6.2-1 version of the LFS LiveCD. This version is built using LFS 6.2 and many BLFS packages from the SVN branch. Source packages for LFS 6.2, and the LFS book itself, are included on the LiveCD. Other features: * The CD supports hibernation, AKA software suspend, or suspend-to-disk. This allows to save the system state to a swap partition and turn the computer off. The saved state can be completely restored later. * The CD filesystem can be written to. The changes actually go to RAM and don't survive a reboot (but do survive hibernation). * The CD contains a visually pleasing and easy to use window manager, Xfce. * jhalfs-1.0, a system for autobuilding LFS according to the book, is included. Please mail any comments to the LFS LiveCD mailing list. To download the new LiveCD, visit the LFS LiveCD download page for a list of mirrors: http://www.linuxfromscratch.org/livecd/download.html (it still lists the 6.1.1-4 as a stable version, which is wrong, but I can't fix the page) -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg 7.1 ugly fonts
Wilco Beekhuizen wrote: The fonts in xorg 7.1 look bad compared to xorg 6.9. I have this problem since upgrading from 6.9 to 7.1. I followed the instructions for Fontconfig, freetype and xorg exactly as stated in the book. These problems happen with the default xorg.conf (from Xorg -configure), and even with no xorg.conf. I checked the font paths and they exist and contain almost the same fonts as in xorg 6.9. The links in /usr/share exists like the book says. Any ideas? "Ugly" is a very subjective word. In the past, such problems were caused by the reader not installing additional TTF fonts that don't come with Xorg. If installing DejaVu fonts according to the instructions in the book doesn't fix your problem, please post a screenshot of a small part of the screen (no more than 10 KB in PNG or JPG with 80% quality). Be sure to include the application name. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: missing euro-sign in KDE
Uwe Kramer wrote: When I do this on my system I see many fonts which show me the euro-sign on the right side but don't show it with altgr-e. I never before had this trouble with kde, maybe I should try reinstalling the whole thing. You should have said from the very beginning that you have a problem with _typing_, not _displaying_ the Euro character. For debugging, please post the "InputDevice" section from your xorg.conf that corresponds to your keyboard. Note that layout names did change since the 6.8.2 version, so you may want to replace the "XkbLayout". -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Getting in a mess with pats - X11 and fonts
Andrey Voropaev wrote: Well, I can't say why BLFS insists that /usr/X11R6/lib/X11/fonts is outside of fontconfig path. fontconfig-2.3.2 already includes that path in the /etc/fonts/fonts.conf, so one does not need to do anything at all. Because if one includes /usr/X11R6/lib/X11/fonts in the font path, bad things happen: 1) Go to http://www.linux.org.ru/ in Seamonkey or Konqueror, and watch how the bitmapped Helvetica font is picked up for Russian characters. This font is non-antialiased, too-small for big screens, and overall ugly. Also it doesn't print properly on paper. 2) Add a new user, log in, start Konsole. Russian shows as double-width characters because of the "Courier 10 pitch" font and bad substitution. Both problems can be avoided by not including /usr/X11R6/lib/X11/fonts in the font path for fontconfig. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: missing euro-sign in KDE
Uwe Kramer wrote: Hello, I have a blfs-book-cvs-2006-06-08 system with KDE-3.5.3 and I cannot get the euro-sign anywhere. It works on the text console and with twm, I can also start xterm in KDE and it works, also OpenOffice and firefox are no problem, but in all KDE apps - no euro. I tried several option in the control centre but nothing works. Does anybody know where to look for this? Yes. Please provide the output of: 1) echo $LANG 2) locale 3) fc-list You have a problem either with an incorrectly named locale (then fix it), or with the fact that your TrueType fonts don't contain that sign (then install more TrueType fonts). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Does establishing a PPPoE connection create an interface?
Angel Tsankov wrote: When I establish a PPPoE connection to my ISP they assign me an IP address. Is this address associated with an interface that I can use with the iptables options "--out-interface" and "--in-interface"? You use the "ppp0" interface there (assuming that you have no other ppp connections). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Try moving mtu and mru statements in /etc/ppp/peers/megalan53 to the very top. If this doesn't help, I give up. Unfortunately, this did not help. Could you explain in a few words why I get this message and where I could find further assistance? You get this harmless message because the default mtu for ppp connections is 1500 (big mtu = low overhead for packet headers = better speed), and this is not valid for pppoe (where the maximum allowed mtu/mru is 1492, because one IP packet plus PPP headers must fit in one 1500-byte Ethernet packet). The end result is correct, so you have nothing to worry about. If you really want to know everythig about packet headers, try sniffing the traffic on eth0 with WireShark (formerly known as Ethereal) and googling for unknown words you see on the screen, plus the word "RFC". -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Hmm, eth0 has mtu=1500, I guess this might be the problem: No, that's correct. eth0 must have mtu at least mtu_of_ppp0 + 8. I have also attached the /etc/ppp/options and /etc/ppp/peers/megalan53 files. Try moving mtu and mru statements in /etc/ppp/peers/megalan53 to the very top. If this doesn't help, I give up. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: pppd call pppoe dryrun eth0 This seems to produce some more info. Here it is (the peer file for this PPPoE connection is called megalan53): Plugin rp-pppoe.so loaded. RP-PPPoE plugin version 3.3 compiled against pppd 2.4.3 pppd options in effect: holdoff 15 # (from /etc/ppp/peers/megalan53) persist # (from /etc/ppp/peers/megalan53) dryrun # (from command line) plugin rp-pppoe.so # (from /etc/ppp/peers/megalan53) noauth # (from /etc/ppp/peers/megalan53) user sven # (from /etc/ppp/peers/megalan53) eth0 # (from command line) rp_pppoe_service megalan53 # (from /etc/ppp/peers/megalan53) eth0 # (from command line) rp_pppoe_service megalan53 # (from /etc/ppp/peers/megalan53) mru 1492 # (from /etc/ppp/peers/megalan53) mtu 1492 # (from /etc/ppp/peers/megalan53) lcp-echo-failure 3 # (from /etc/ppp/peers/megalan53) lcp-echo-interval 30 # (from /etc/ppp/peers/megalan53) defaultroute # (from /etc/ppp/peers/megalan53) usepeerdns # (from /etc/ppp/peers/megalan53) If you need further assistance in tackling this issue just let me know. Still doesn't help. Could you please, after connecting, issue this command: ip link list and paste the entire output. But if that doesn't show the problem, I give up. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: You tried the correct options, not sure why they don't work. The correct value for mru and mtu is 1492, BTW. Hmm, I put "mtu 1492" and "mru 1492" in the /etc/ppp/peers/pppoe file, but I still get the "Couldn't increase MTU to 1500" message: I need the output of "pppd call pppoe dryrun" in order to debug this. Plugin rp-pppoe.so loaded. RP-PPPoE plugin version 3.3 compiled against pppd 2.4.3 Hope this helps. No, that doesn't. pppd call pppoe dryrun eth0 -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: You tried the correct options, not sure why they don't work. The correct value for mru and mtu is 1492, BTW. Hmm, I put "mtu 1492" and "mru 1492" in the /etc/ppp/peers/pppoe file, but I still get the "Couldn't increase MTU to 1500" message: I need the output of "pppd call pppoe dryrun" in order to debug this. How can I specify which PPPoE server I want to connect to when more than one is available? There are the following options: rp_pppoe_service : Desired PPPoE service name rp_pppoe_ac : Desired PPPoE access concentrator name rp_pppoe_verbose : Be verbose about discovered access concentrators These options seem to work. Where are they documented? I got it directly from the source code :( but I guess that they are documented somewhere in the rp-pppoe package. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: idiots guide to cups and samba?
DJ Lucas wrote: Where is the XP Home guide? Why not do it _without_ samba? That's much easier, and already documented at: http://archives.linuxfromscratch.org/mail-archives/blfs-support/2005-September/056770.html -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Is the "Couldn't increase MTU to 1500" message harmless? Yes. If not, how can I avoid it (I already tried adding "mru 1280" and "mtu 1280" to the /etc/ppp/peers/pppoe file)? You tried the correct options, not sure why they don't work. The correct value for mru and mtu is 1492, BTW. How can I specify which PPPoE server I want to connect to when more than one is available? There are the following options: rp_pppoe_service : Desired PPPoE service name rp_pppoe_ac : Desired PPPoE access concentrator name rp_pppoe_verbose : Be verbose about discovered access concentrators It seems that /etc/ppp/ip-up does not get called; how do I inform bind where to forward DNS queries? You probably didn't chmod the file to 755. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Installing X in other than /usr/X11R6/
Tijnema ! wrote: well, only twm would be enough to run, it's just that i can acces a terminal without putty, and that i'm running a X server so i can run most apps on it, can u explain, or give me a URL which explains how i can install Xorg or Xfree? That's way too complex to explain here, and off-topic for this list (BLFS assumes that you have root access). Also, I already hinted how to install XOrg 7.1 into your home directory in my previous mail. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: error compiling xorg-server-1.1.0
Andrew Benton wrote: Uwe Kramer wrote: Hello, I am trying to build a new system with xorg-7.1 according to BLFS-book-cvs-html-2006-06-08. While compiling xorg-server I get following error: In file included from lnx_agp.c:24: /usr/include/linux/agpgart.h:55: error: syntax error before '__u16' /usr/include/linux/agpgart.h:60: error: field 'version' has incomplete type Can anybody help me with this? sed -i '/CONFIG_H/i #include ' \ hw/xfree86/os-support/linux/lnx_agp.c Or, better, since the problem is really in glibc and affects other apps (e.g. dosemu): cd /usr/include/sys wget -qO- http://www.linuxfromscratch.org/patches/lfs/development/glibc-2.3.6-linux_types-1.patch | patch -Np6 -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Installing X in other than /usr/X11R6/
Tijnema ! wrote: I'm doing this, becuase i have a server with SSH login, and so i wanted to install VNC server, as VNC server needs X server to run, i wanted to install X. is there any other way i can install VNC? One can probably build and install XOrg 7.1 libraries, apps and fonts into the home directory (i.e., do everything the SVN BLFS book says, but pass --prefix=$HOME, and add $HOME/lib/pkgconfig to $PGK_CONFIG_PATH), and then build the VNC server. However, I have not tested this. Also, what are you going to see in the VNC server? Without X applications, it is going to show you only a blank screen. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Installing X in other than /usr/X11R6/
Tijnema ! wrote: Hi, i want to install the X server into my home directory. is that possible? It's because i don't have acces to /usr The X server has to be installed setuid root (or started by root), in order to access the registers of graphics cards. Since you indicated that you have no access to /usr, I assume you are not root. Result: you cainstall the server into your home directory, but it won't work. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
I wrote:
killproc -p /var/run/ppp-${1}.pid
Oops, should be:
killproc -p /var/run/ppp-${1}.pid pppd
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
[OT] Re: latest wine 0.9.14 and BLFS-SVN-2006-05-28 [Solution]
Dan Nicholson wrote: And the proprietary ATI drivers won't even start up on Xorg-7.1 due to the ABI breakage. And I don't use them :) Open-source drivers work just fine with 3D. But for those who have Radeon X1000 or newer, the situation is worse. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote:
Now there is one more thing about the PPPoE connection I would like to
setup - I want to have it established at boot time. As far as I know I
need to have the following files:
1) /etc/sysconfig/network-devices/ifconfig.eth0/pppoe:
ONBOOT="yes"
SERVICE="pppoe"
2) /etc/sysconfig/network-devices/services/pppoe:
#!/bin/sh
. /etc/sysconfig/rc
. $rc_functions
case "$2" in
up)
boot_mesg "Bringing up the PPPoE interface..."
ip link set up
pppd call
evaluate_retval
;;
down)
boot_mesg "Bringing down the PPPoE interface..."
killall pppd
evaluate_retval
;;
*)
echo "Usage: $0 {up|down}"
exit 1
;;
esac
The question is how do I know the value of N in .
ethN is passed as the argument to the bootscript and, thus, is accessible as
${1}.
Probably (untested, I have a custom script not using the service model) you can
do the following:
1) Don't specify the network interface name in /etc/ppp/peers/pppoe
2) In the "up" part, run this:
ip link set ${1} up
loadproc -p /var/run/ppp-${1}.pid pppd call pppoe nic-${1} updetach linkname
${1}
No need to evaluate_retval
(the "nic-" prefix is needed so that it works even with network interfaces
renamed to "realtek" or something like that. "updetach" is needed so that you
can be 100% sure that pppd doesn't go into background until establishing the
connection, so that e.g. ntpd could start correctly. "linkname" is for a
predictable pid file name)
3) In the "down" part, it is better to do this:
killproc -p /var/run/ppp-${1}.pid
Again, no need to evaluate_retval
Please report if this works for you with or without modifications.
The other thing I want is to have the PPPoE connection reestablished if
it breaks. Is this automatically handled or do I need to setup smth?
You have to add the following options to /etc/ppp/peers/pppoe:
persist
holdoff 15
lcp-echo-interval 30
lcp-echo-failure 3
This way, pppd will retry in 15 seconds after it knows that the connection
broke. Also, it will send echo requests to the peer every 30 seconds, and
failure to answer 3 requests counts as link failure.
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: latest wine 0.9.14 and BLFS-SVN-2006-05-28 [Solution]
Tasos Drosopoulos wrote: Anyway, my box uses an nvidia card and for 3d graphics accel I use the nvidia driver (version 8762). Switching back to the slower nv that comes with Xorg7.1, lo and behold, the fonts were visible and wine seems to work fine. Indeed, there is an incompatibility between 8xxx nvidia drivers and Xorg 7.1. It can be (reportedly) worked around by adding to xorg.conf: Option "RenderAccel" "no" but I can't check this because I have an ATI video card. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: How can I be sure that I'm using kernel-mode PPPoE in this case? Very easy. 1) "pidof pppoe" should return nothing. 2) "pidof pppd" should return only one process ID, not two IDs. 3) "ls -l /proc/`pidof pppd`/fd" should not reference /dev/pts -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: I verified I have CONFIG_PACKET=y in my .config file. If this guess is wrong, you have two roads for debugging this: 1) Mail your kernel .config to me privately, so that I can look at it instead of guessing. Sure, I have attached the compressed kernel configuration file. After adding support for my hardware and making "oldconfig" with newer kernel, this config _worked_ for me. I can't run an old kernel due to new udev, sorry. So let's assume this is either a 2.6.11 specific problem, or not a kernel problem at all. Please install strace and post the output of: strace -e open,socket pppd call pppoe debug nodetach together with all files in /etc/ppp that are mentioned in that log, plus your firewall configuration, if any. 2) Download the known good config from http://wiki.linuxfromscratch.org/livecd/browser/trunk/packages/linux/config.x86?format=raw I will probably try these steps later on as the last way out of this problem. My argument in favor of this decision is that being able to make the PPPoE connection work (using RP-PPPoE along with WvDial, WvStreams and OpenSSL) is an indication that the problem has smth to do with configuration of existing software (ppp package and/or rp-pppoe.so module for example) rather that the software itself (kernel version for example)? Could this assumption be wrong? It could, because there are two ways to configure pppoe with rp-pppoe: kernel-mode pppoe and user-mode pppoe. If one works, there is no reason to assume that the other will work. With the PPP package alone, only kernel-mode pppoe is available, so your assumption breaks if you were using user-mode pppoe (the default with rp-pppoe). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: How do I build the PPPoE module into the kernel? CONFIG_EXPERIMENTAL=y (that's under Code maturity level options -> Prompt for development and/or incomplete code/drivers) CONFIG_PPPOE=y (that's under Device drivers -> Network device support -> PPP over Ethernet (EXPERIMENTAL)) Yes, I do have the above two lines in the .config file. This means that the PPPoE module is built into the kernel. and the suggestion about the PPPoE module not loaded should be wrong. What other reason could there be for the PPPoE connection not being established? Maybe you missed CONFIG_PACKET=y. If this guess is wrong, you have two roads for debugging this: 1) Mail your kernel .config to me privately, so that I can look at it instead of guessing. 2) Download the known good config from http://wiki.linuxfromscratch.org/livecd/browser/trunk/packages/linux/config.x86?format=raw Download the 2.6.16.20 kernel from kernel.org, save the config as linux-2.6.16.20/.config, run "make oldconfig" (because the LiveCD kernel is patched and thus the config above is not exactly for 2.6.16.20 kernel), then build the kernel as usual (you may want to add sound, put please don't remove anything). Old udev won't fully work with that kernel, but you still should be able to: ip link set eth0 up modprobe pppoe pppd call pppoe -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: How do I build the PPPoE module into the kernel? CONFIG_EXPERIMENTAL=y (that's under Code maturity level options -> Prompt for development and/or incomplete code/drivers) CONFIG_PPPOE=y (that's under Device drivers -> Network device support -> PPP over Ethernet (EXPERIMENTAL)) -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Now it does, but when I reboot the system after installing the kernel I get a lot of lines reading: ./hotplug.functions: line 113: /lib/modules/2.6.11.12/modules/isapnpmap: No such file or directory Known bug with /etc/hotplug/pnp.rc. Remove that file, because it only loads modules and you don't even have support for them. Do I need to enable "Loadable module support" in order to use the PPPoE module? No. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Hmm, I had disabled "Loadable module support" prior to rebuilding the kernel. This should mean that the PPPoE module is built into the kernel and is therefore always loaded, right? No, this means that the module is not built into the kernel and never loaded. You can build it into the kernel, though, and this would fix your problem. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: 1. created the /etc/ppp/ip-up and the /etc/ppp/peers/pppoe files; 2. populated the /etc/ppp/pap-secrets and the /etc/ppp/chap-secrets files; 3. issued "ip link set eth0 up" (this produced no output on the terminal); 4. issued "pppd call pppoe" and I got: Plugin rp-pppoe.so loaded. RP-PPPoE plugin version 3.3 compiled against pppd 2.4.3 Failed to create PPPoE socket: Address family not supported by protocol That's because you didn't load the "pppoe" kernel module before following the instructions. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: PPPoE connection using only the PPP package
Angel Tsankov wrote: Then I executed pppd eth0 which produced: Plugin rp-pppoe.so loaded. RP-PPPoE plugin version 3.3 compiled against pppd 2.4.3 Then I ping'ed an Internet host by IP address and got: ping: sendto: Network is unreachable Can someone help me setup a PPPoE connection using only the PPP package? Yes, sorry for inaccurate instructions in the Wiki. You need to run: ip link set eth0 up and only then "pppd eth0" (assuming that everything ele is correct) -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: /dev/bugreport (was: BLFS progress: udev/hotplug dbus/hal)
Warren Wilder wrote: I've got a /dev/bugreport for you. UDEV_LOG=3 ACTION=add DEVPATH=/devices/pci:00/:00:0d.0/fw-host0/00406354efd3 SUBSYSTEM=ieee1394 SEQNUM=469 PHYSDEVBUS=ieee1394 UDEVD_EVENT=1 _SEPARATOR=--- UDEV_LOG=3 ACTION=add DEVPATH=/class/ieee1394_node/00406354efd3 SUBSYSTEM=ieee1394_node SEQNUM=470 PHYSDEVPATH=/devices/pci:00/:00:0d.0/fw-host0/00406354efd3 PHYSDEVBUS=ieee1394 UDEVD_EVENT=1 _SEPARATOR=--- It appears that all leaks reported up to now are unfixable from udev side (they are due to drivers probing their hardware in a separate thread), so let's just ignore bugreports and remove the "bug" rule, the /lib/udev/bug program and the corresponding parts of "udev" and "udev_retry" bootscripts. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gcc-3.3.6 "tail" command fix.
Randy McMurchy wrote: Too bad. Sounds to me as though it's time to update your software inventory to modern standards. :-) This means junking too-new ATI PCI-E cards that work only with proprietary driver (it requires libstdc++.so.5). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Language setting - Probably missed something
Georgina Joyce wrote: Hi All Well, I've got gnome running as root however, as a general user x / gnome complains of my language settings. Which I should have addressed before now probably as perl complained when I had to install the XML::Parser. I've tried stating in my /etc/profile: export LANG=en_GB.ISO-8859-1 Try en_GB -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: BLFS progress: udev/hotplug dbus/hal
Dan Nicholson wrote: Unfortunately, there isn't a complete solution at this time that I know of. HAL provides the info and methods to do things like automounting removable drives and suspending the system. However, another application needs tell hal to do this. In gnome, this is gnome-volume-manager and gnome-power-manager. I'm not aware of any daemon on the linux console. ivman. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: detex
Uli Fahrenberg wrote: I'm trying to compile detex-2.7, I was able to compile it using the following commands: sed -f states.sed detex.l > xxx.l lex xxx.l rm -f xxx.l mv lex.yy.c detex.c sed -i -e '1i#include' -e '/\]\]/d' detex.c cc -O -DNO_MALLOC_DECL -c -o detex.o detex.c cc -O -o detex detex.o -ll The need for #include and for -DNO_MALLOC_DECL are detex bugs. The need to remove the ']]' line is definitely a bug in flex (reproducible on Debian too). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox cashes in file-open and file-save dialogs
Dan Nicholson wrote: On 5/12/06, Alexander E. Patrakov <[EMAIL PROTECTED]> wrote: Also I have to check that the problem is indeed still relevant for Seamonkey-1.0.1 (it was checked only for 1.0). I'd be surprised if it has changed. I've opened up simple bugs like this before, and they never get responded to. I'm thinking that maybe it hasn't been an issue in the past because pangoxft was pulled in by GTK+'s .pc file. Just a guess, but it seems unlikely that this has been a problem for a long time. There really should be a separate PKG_CHECK macro for pango. The problem indeed still exists with Seamonkey-1.0.1 (but my initial description about a delayed crash was wrong: it tells about the undefined symbol immediately). The patch helps. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox cashes in file-open and file-save dialogs
Dan Nicholson wrote: Shouldn't be any harm. Alexander, could you test out the attached patch without the LDFLAGS next build you do? AFAICT, this is the only directory that has sources that need to link to pangoxft. In fact, the program gfx/src/gtk/mozilla-decoder.cpp includes . Sorry, my attempt to build SeaMonkey today failed for a completely different reason. Could you please wait until tomorrow? Also I have to check that the problem is indeed still relevant for Seamonkey-1.0.1 (it was checked only for 1.0). -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox cashes in file-open and file-save dialogs
Randy McMurchy wrote: Alexander E. Patrakov wrote these words on 05/10/06 08:27 CST: If it doesn't: bug in Firefox (and in the book). If this is a bug in the book, why haven't you ever mentioned it? Because the book doesn't tell the user to add --enable-pango (i.e.: doesn't trigger the bug by default) If we were to put it in the book, what would be the reason why it is in the .mozconfig. Unlike your .mozconfig, the BLFS versions describe each entry in the file. The ones not documented are: ac_add_options --enable-pango Required for rendering complex scripts like Thai ac_add_options --disable-updater Disables checks for updates (they are pointless on a LiveCD) LDFLAGS="-Wl,--as-needed -lpangoxft-1.0" Fixes runtime chash because of undefined symbol while rendering Thai. IOW: seamonkey forgets to link one of its libraries with libpangoxft-1.0 but uses functions from it. This goes unnoticed until the function gets actually called. The LD_BIND_NOW=1 allowed me to move this crash to the startup phase and thus to debug it. The --as-needed was added because I don't want every library or program shipped with seamonkey to be needlessly linked with libpangoxft-1.0. There will be no harm if I add -Wl,--as-needed -lpangoxft-1.0 -Wl,--no-as-needed. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox cashes in file-open and file-save dialogs
Leo Peschier wrote:
Hello all,
After upgrading Firefox from BLFS 6.1 (version 1.0.6) to version
1.5.0.2, following the svn-book, it crashes in file-open and file-save
dialogs when:
- moving the scrollbar, or
- using the mouse wheel, or
- moving below the visual part with the down-arrow,
with: /usr/lib/firefox-1.5.0.2/run-mozills.sh: line 131:
segmentation fault "$Prog" ${1+"$@"}
Otherwise it is stable.
Could you please do this:
LD_BIND_NOW=1 firefox
Does firefox start at all?
If it does: sorry for the trouble, the problem is not what I expected.
If it doesn't: bug in Firefox (and in the book). Rebuild with the library
containing the needed symbol explicitly mentioned in LDFLAGS, as the LiveCD does
with SeaMonkey:
http://wiki.linuxfromscratch.org/livecd/browser/trunk/packages/seamonkey/.mozconfig
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: arts and glib2=>2.10
Wilco Beekhuizen wrote: Hi, After a glib update arts stopped working. I've read about this problem on several lists. The KDE bugtrackers say it's a glib error but the glib ML says it's an application error, which sound more logical to me. This is the error I get: ***MEMORY-ERROR***: []: GSlice: failed to allocate 248 bytes (alignment: 256): Invalid argument I know you can prevent this error by using G_SLICE=always-malloc but that way all glib programs use malloc and it's not a very elegant solution. Does anyone have more insight? Should I fix arts or really wait for an updated glib? You should fix arts. The problem is that developers on each side put the responsibility onto the other side, and there is no fix available. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Keyboard problem
Nicolas FRANCOIS wrote: I've got a problem with my new LFS (SVN from may 24). Under X, when I try to use the "Alt Gr" key, I get a full keypress under Emacs, so that for example I can't call the Auctex debugger with "AltGr `" to go to the next tex error. As you can see, I can use this key with other apps. The strange thing is that I copied the emacs config files from my old LFS 6.0, so this can't be a mistake in .emacs or something like that. I have a french keyboard. Which file should I look for investigation ? /etc/X11/xorg.conf You can download the 6.2-pre3 LFS LiveCD, boot it, select French, and, if it works, copy xorg.conf from there. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gphoto permissions
Alberto Hernando wrote:
I'm trying to connect to my digital camera using digikam&gphoto, and I have a
problem with permissions, because root can access. When I connect the camera,
I get a /dev/usbdevX device, owned by root.root and with 660. I've read all
my udev rules and there is nothing for this, so I don't know what to do. I
tried to install the latest udev (kernel i 2.6.16) but it conflicts with
hotplug and even if I remove the hotplug init script, the system doesn't work
well, or doesn't boot. So I'm bak to udev-056. What do I need to do to access
my camera?
The answer depends upon the kernel and udev/hotplug version.
Old-style setup, works with any kernel and udev version (in fact, it works even
without udev/hotplug):
Please be sure that there is a /proc/bus/usb line in /etc/fstab, and it mentions
the "14" gid and "660" permissions. This creates pseudofiles like
/proc/bus/usb/001/002 owned by the "usb" group. Add yourself to this group in
order to use your camera. There were racy and non-working attempts to change
permissions from hotplug scripts, they are not in the book.
New-style setup (linux >=2.6.14, udev >=081 and libusb >=0.1.12 are needed):
Please be sure that you have the long rule mentioned in
http://wiki.linuxfromscratch.org/blfs/ticket/1680. This will create symlinks
like /dev/bus/usb/1/2 -> /dev/usbdev1.2 to files owned by the "usb" group. Add
yourself to this group in order to be able to use the devices. There is also a
(currently debian-specific) script that generates udev rules for all known
cameras and specifies a different group. You can do it yourself for your camera,
matching on vendor and product IDs, like this:
# UMAX AstraPix 320s
BUS=="usb", SYSFS{idVendor}=="0d64", SYSFS{idProduct}=="1001", MODE="0660",
GROUP="camera"
This rule catches all PTP cameras:
# USB PTP Class Camera
BUS=="usb", SYSFS{bDeviceClass}=="06", MODE="0660", GROUP="camera"
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: Font changing by their own ???
Nicolas FRANCOIS wrote: Hi. I rebooted to recompile the kernel of my new LFS system. Now I'm back on the old one... I think I'm going slightly mad ! The fonts of my Gtk apps (Pan, Sylpheed-claws) are now EXTERMELY small, I can hardly read the menus or the messages ! How come that by doing NOTHING, configuration can change ? Hao can I get it back to the old setting ? Thanks for any help, this font story is driving me crazy. This probably has to do with bugs in either your monitor or a video driver that result in inability to read the monitor size via DDC. If this is indeed the case, putting a "DisplaySize x y" where x and y represent the physical size of your screen in millimeters, into the "Monitor" section of xorg.conf should work around the problem. If this helps, you need to buy a new monitor. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Dosemu not compiling, "pci_power_t" problem
Nicolas FRANCOIS wrote: While compiling dosemu-1.2.2 on a LFS 6.0 (dev, gcc 3.4.3, kernel 2.6.13), I get this error message : gcc -c -I/usr/X11R6/include -MP -MMD -I../src/include -I../src/plugin/include -Wall -Wstrict-prototypes -Wmissing-declarations -Wnested-externs -O2 -fomit-frame-pointer -fno-strict-aliasing -mtune=i686 -pipe -o emu.o emu.c In file included from /usr/include/sys/pci.h:23, from ../src/include/pci.h:10, from emu.c:119: /usr/include/linux/pci.h:454: erreur: erreur d'analyse syntaxique avant « pci_power_t » make[1]: *** [emu.o] Erreur 1 make[1]: Leaving directory `/sources/dosemu-1.2.2/src' make: *** [default] Erreur 2 I encounterd it twice on Google, but not on the dosemu bugtracking system. Can someone help me on this one ? It is probably caused by a bug in glibc. Apply this patch to the installed /usr/include/sys/kd.h file: http://sources.redhat.com/cgi-bin/cvsweb.cgi/libc/sysdeps/unix/sysv/linux/sys/kd.h.diff?cvsroot=glibc&r1=1.4&r2=1.5 -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xfce error
Faheem Mohammed wrote: Could not init font path element /usr/X11R6/lib/X11/fonts/local/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/CID/, removing from list! That's almost normal, these directories don't exist so don't mention them in xorg.conf to get rid of this. FreeFontPath: FPE "/usr/X11R6/lib/X11/fonts/misc/" refcount is 2, should be 1; fixing. Xorg bug (but a harmless one), shows up for everyone. Please ignore. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
