Rox Filer
Rox Filer is a gtk2 file browser with few dependencies but it is not a straightforward CMMI to install it. I think it would be appreciated by the people who read the book so I've made the attached patch which will add it to the book. Please apply. Andy diff -Naur BLFS/general/genutils/genutils.xml BLFS-rox/general/genutils/genutils.xml --- BLFS/general/genutils/genutils.xml 2010-03-18 19:16:10.0 + +++ BLFS-rox/general/genutils/genutils.xml 2010-03-30 14:50:49.0 +0100 @@ -44,5 +44,6 @@ http://www.w3.org/2003/XInclude"; href="pinentry.xml"/> http://www.w3.org/2003/XInclude"; href="gperf.xml"/> http://www.w3.org/2003/XInclude"; href="rarian.xml"/> + http://www.w3.org/2003/XInclude"; href="rox-filer.xml"/> diff -Naur BLFS/general/genutils/rox-filer.xml BLFS-rox/general/genutils/rox-filer.xml --- BLFS/general/genutils/rox-filer.xml 1970-01-01 01:00:00.0 +0100 +++ BLFS-rox/general/genutils/rox-filer.xml 2010-03-30 18:02:36.0 +0100 @@ -0,0 +1,163 @@ + +http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd"; [ + + %general-entities; + + http://sourceforge.net/projects/rox/files/rox/&rox-filer-version;/rox-filer-&rox-filer-version;.tar.bz2/download";> + + + + + +]> + + + + + +$LastChangedBy: nobody $ + + + + rox-filer-&rox-filer-version; + + +rox-filer + + + +Introduction to rox-filer + +Rox-filer is a fast, lightweight, gtk2 file +manager. + +Package Information + + +Download (HTTP): + + +Download (FTP): + + +Download MD5 sum: &rox-filer-md5sum; + + +Download size: &rox-filer-size; + + +Estimated disk space required: &rox-filer-buildsize; + + +Estimated build time: &rox-filer-time; + + + +rox-filer Dependencies + +Required + + and + + + + + +Installation of rox-filer + +Compile rox-filer with the following +commands: + +cd ROX-Filer && +sed -i 's:g_strdup(getenv("APP_DIR")):"/usr/share/rox":' src/main.c && +mkdir build && +cd build && +../src/configure && +make + +Now install it as the root +user: + +mkdir /usr/share/rox && +mv ../{Help,images,Messages,Options.xml,ROX,style.css,.DirIcon} \ +/usr/share/rox && +install -m644 ../../rox.1 /usr/share/man/man1 && +mv ../ROX-Filer /usr/bin/rox && +chown -R root:root /usr/{bin,share}/rox && +cd /usr/share/rox/ROX/MIME && +ln -sv text-x-{diff,patch}.png && +ln -sv application-x-font-{afm,type1}.png && +ln -sv application-xml{,-dtd}.png && +ln -sv application-xml{,-external-parsed-entity}.png && +ln -sv application-{,rdf+}xml.png && +ln -sv application-x{ml,-xbel}.png && +ln -sv application-{x-shell,java}script.png && +ln -sv application-x-{bzip,xz}-compressed-tar.png && +ln -sv application-x-{bzip,lzma}-compressed-tar.png && +ln -sv application-x-{bzip-compressed-tar,lzo}.png && +ln -sv application-x-{bzip,xz}.png && +ln -sv application-x-{gzip,lzma}.png && +ln -sv application-{msword,rtf}.png + + + + +Command Explanations + + +sed -i 's:g_strdup(getenv("APP_DIR")):"/usr/share/rox":' src/main.c +: This command hard codes /usr/share/rox as the directory for its +private files. Without this sed rox needs the environment variable $APP_DIR +to be set. + +ln -sv application-...: These commands duplicate +the icons for some common mime types. Without these links +rox would just display the default "unkown +binary blob" icon. + + + + +Configuring Rox Filer + Rox itself doesn't need any configuration + but if you want it to be able to update its view of the contents of a + directory when changes are made to files by other applications you will + need to enable dnotify support (on the filesystems menu in make + menuconfig) in your kernel. + + + +Contents + + + Installed Programs + Installed Directories + + +rox +/usr/share/rox + + + + + Short Descriptions + + + + +rox + + is the rox file manager. + +rox-filer + + + + + + + + + + diff -Naur BLFS/general.ent BLFS-rox/general.ent --- BLFS/general.ent 2010-03-30 14:50:17.0 +0100 +++ BLFS-rox/general.ent 2010-03-30 14:50:49.0 +0100 @@ -235,6 +235,7 @@ + -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
ghostscript jasper patch
Hello, when installing ghostscript, the jasper patch can be replaced by a sed: sed -i '/e j/,+5d;/b(s/d' base/sjpx.c Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
faac-1.28
The faac-1.28 page says that it installs libmp4v2.so. I couldn't get it to install libmp4v2.so (I needed it for easytag - I had some m4a files with broken metadata) so I downloaded mp4v2 http://code.google.com/p/mp4v2/ and used that (easytag still needed some hacking but I got there in the end). Then I turned my attention to faac. To get it to use the external libmp4v2 I had to: sed -i '/xternal /aMY_DEFINE(HAVE_LIBMP4V2)\ LIBS="-lmp4v2 $LIBS"' configure.in && ./bootstrap && sed -i 's:mp4\.h:mp4v2/mp4v2.h:' {configure,frontend/main.c} Then it was just: ./configure --prefix=/usr; make; make install The glibc patch is only needed if you use the libmp4v2 included in the faac source. I find this sed is easier than a patch: sed -i 's/strcasestr/str_casestr/' common/mp4v2/mpeg4ip{,_win32}.h Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Xorg download a little easier
On Mon, 27 Sep 2010 17:52:42 -0400 linux fan wrote: > wgetter Launch 2 subprocesses when you could do it in bash? Instead of: here=$(basename $(pwd)) Try: here=${PWD##*/} Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: xorg luit?
On Wed, 29 Sep 2010 00:15:09 -0400 linux fan wrote: > Luit is in the X Window System table of contents, xterm is not. > Does anything really need luit? Can I build without it? > -- I don't build luit. xterm doesn't need it Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Variability in build times.
On Mon, 18 Oct 2010 18:38:48 -0400 linux fan wrote: > start=$(date +%s) > ... stuff happens ... > end=$(date +%s) > elaps=$(( end - start )) > That's an interesting suggestion. Thanks! Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Boost building
On Sat, 12 Mar 2011 07:43:48 +0600 Vladislav Kuzkokov wrote: > Hi everyone, > > In order to keep LFS/BLFS the major source of concise building instructions. > > Boost.IOStreams optionally requires zlib and libbz2. > Alternatively you can use > ./bjam -sNO_ZLIB=1 -sNO_BZIP2=1 > Otherwise compiling IOStreams fails. The instructions in BLFS assume that you're starting from an LFS system. They do not list things installed while completing the LFS build as dependencies. Otherwise every page would list GCC, Glibc, Binutils, Bash and so on as dependencies. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: blfs svn 20110620 ntp 4.2.4p7 -- failure to start, possible bootscript typo
On Sat, 2 Jul 2011 01:44:11 -0500 MBW wrote: > Hi, > > I just went through the installation procedure for NTP-4.2.4p7 as > outlined in BLFS svn 20110620 and all was well until I tried to start > /etc/rc.d/init.d/ntp. > > After issuing (as root): > /etc/rc.d/init.d/ntp start > > I was greeted with the error message: > Starting ntpd... > ntpd: Command line arguments not allowed > ntpd - NTP daemon program - Ver. 4.2.4p7 > USAGE: ntpd [ - [] | --[{=| }] ]... > > (snipped -- followed by a list of valid options) > > I took a look at the wiki/FAQ and did not find any mention of this > problem - so I decided to peruse the /etc/rc.d/init.d/ntp script > itself. As best I can tell, the error seems to occur on line 13 with > the following code block: > > start) > boot_mesg "Starting ntpd..." > loadproc /usr/sbin/ntpd gx > ;; > > I changed this to (based upon the USAGE argument above): > > start) > boot_mesg "Starting ntpd..." > loadproc /usr/sbin/ntpd -gx > ;; > > and everything seems to work fine. > > It appears the "gx" should be changed to "-gx" in the bootscript - but > I would like the opinion of someone a bit more versed in shell > scripting. > I think you're right. CCing blfs-dev Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: The State of BLFS
On Sat, 16 Jul 2011 12:11:10 -0500 Rob Landley wrote: > I'm building a dozen targets and only x86-64 tries to install stuff in > lib64. I create a temporary lib64->lib symlink during the gcc build and > then delete it again after gcc finishes crapping into it. gcc builds fine for me with no lib64->lib symlink. Before I compile it I run this sed inside the gcc source: sed -i 's:\.\./lib64::;s/lib64/lib/' gcc/config/i386/{linux64.h,t-linux64} I also s/lib64/lib/ during the gcc -dumpspecs. glibc needs a sed too: sed -i 's|/lib64|/lib|' sysdeps/unix/sysv/linux/configure Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: The State of BLFS
On Sat, 30 Jul 2011 18:22:52 -0500 Rob Landley wrote: > On 07/29/2011 07:29 PM, Randy McMurchy wrote: > > I am open to suggestions, let the community speak out! C'mon, we need > > to keep BLFS alive! > > > So I grabbed the SVN, which produced nothing interesting in the top > level directory, guessed I wanted "book", there was a readme in there, > told me I could run make, and I did. It sat there for three minutes > "validating the book" before I gave up and killed it. Looked at the > readme some more, it said I could do "make nochunks", which i did, and > it started validating xml again. Killed it, opened up the makefile, > searched for nochunks, it didn't have an obvious dependency on > validating (oh great, another crawling makefile horror) searched for the > "Validating the book" string, put an echo in front of its targets, that > did nothing, searched again and found a SECOND instance of this, put an > echo in front of THOSE targets, and then make nochunks died with: > > Generating profiled XML for XHTML... > warning: failed to load external entity "tmp/blfs-full.xml" > unable to parse tmp/blfs-full.xml > make: *** [tmp/blfs-html.xml] Error 6 > > That was about the limit of my interest for the moment. I have a guess > why you have no new developers. Getting the book to render is easy. The hard part of editing BLFS is dealing with the bile filled hateful emails you get from Randy McMurchy after you've committed a change. Life's too short to deal with that shit. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: ca-bundle and openssl
On Fri, 14 Oct 2011 14:21:37 -0500 Bruce Dubbs wrote: > This is mostly for DJ. > > I think we should add information to the BLFS openssl page on how to > create a ca bundle. I tried looking at some older messages, but > cvs.fedoraproject.org no longer exists. > > I can find a list from debian. > > RH wants to put the files in /etc/pki. > > openssl puts them in /etc/ssl/certs. > > My questions are: > > 1. What procedure is used to generate the BLFS-ca-bundles? The curl makefile has a make ca-bundle target. > 2. Should we continue to put them in /etc/ssl/certs or should we > consider another location? I like /etc/ssl/certs. I wrote about it here: http://linuxfromscratch.org/pipermail/blfs-support/2011-October/068687.html Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: ca-bundle and openssl
On Fri, 14 Oct 2011 21:45:00 -0400 Jonathan Oksman wrote: > /etc/ssl/certs makes sense to me. I'm an /etc minimalist though and > don't like having extra content without a really good reason. > > /etc/pki could make sense if you wanted your certificate information > available in a single location independent of ssl specifically, but I > can't think of any reason for BLFS since there are no other packages > that directly handle certificates in use, as far as I recall. You raise a good point; what program is going to use these certificates? It seems to me that different programs handle ssl certificates in different ways. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Python 3.2.2
On Fri, 28 Oct 2011 21:57:15 +1100 Wayne Blaszczyk wrote: > On 28/10/11 21:13, Wayne Blaszczyk wrote: > > Some observations about the latest Python build. > > Besides the > > http://www.linuxfromscratch.org/patches/blfs/svn/Python-3.2.2-bdb_4.8-1.patch > > missing, there is no actual instruction to apply the patch. > > The ftp://mirror.ovh.net/gentoo-distfiles/distfiles/Python-3.2.2.tar.bz2 > > url is also broken. > > If we are going to have two links to the same package, we should keep > > the file type the same otherwise the MD5 sum is only going to apply to > > the one url and not the other. > > Regards, > > Wayne. > > I would also like to add that libxcb will not build with Python 3.2. > I get the following error during make: > > Making all in src > make[1]: Entering directory `/sources/libxcb-1.7/src' > /usr/bin/python3 ./c_client.py -p /usr/lib/python3.2/site-packages > /usr/share/xcb/xproto.xml > File "./c_client.py", line 1013 > except getopt.GetoptError, err: > ^ > SyntaxError: invalid syntax > make[1]: *** [xproto.c] Error 1 > make[1]: Leaving directory `/sources/libxcb-1.7/src' > make: *** [all-recursive] Error 1 > > When I tried to build libxml2, it did not detect Python. > libxslt also did not detect Python. > I suspect there will be a lot of other packages that will have trouble > with Python 3.2. As DJ Lucas said the other day, it's possible to fix some things with the 2to3 tool included with python 3. However, there are quite a lot of things out there that would need fixing so I personally install both Python-2.7.2 and Python-3.2.2 with 2.7.2 as the default /usr/bin/python. It's a bit like with gtk+-3, more things are starting to use gtk+-3 but I think it'll be a while before I stop installing gtk+-2 Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Python 3.2.2
On Fri, 28 Oct 2011 21:24:11 -0500 Bruce Dubbs wrote: > Are there any naming conflicts between python3 and python2? It looks > like I have: > > /usr/bin: > > pydoc3 python3python3.2 python3.2m > pydoc3.2 python3-config python3.2-config python3.2m-config > > /usr/lib: > > /usr/lib/libpython3.2m.so /usr/lib/libpython3.so > /usr/lib/libpython3.2m.so.1.0 /usr/lib/libstdc++.so.6.0.16-gdb.py > > /usr/lib/python3.2: > ... > > Are there any gotchas in building side-by-side? What build instructions > do you use in each case? I install pthon 2 and python 3 side by side. I'm not aware of any conflicts. As you've shown above, python 3 doesn't install /usr/bin/python, if you want it with python 3 you have to make the symlink yourself. As for build instructions they're both: ./configure --prefix=/usr --enable-shared make make install Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Linux-PAM and Bekkeley DB
On Sun, 30 Oct 2011 12:02:53 -0500 Bruce Dubbs wrote: > Wayne Blaszczyk wrote: > > On 30/10/11 09:43, Jeremy Huntwork wrote: > >> Try building db with --enable-dbm. > > > Thanks, that worked. As mentioned by DJ in the previous post, I think > > this should be included in the standard build. > > I don't generally use PAM, so I don't mind any changes to it. I think Wayne was actually suggesting a change to the way berkeley-db is installed > I'm curious though. What do others get from PAM? Personally I install PAM so I can use pam_faildelay.so with ssh to set an arbitrary amount of time between login attempts. Brute force attacks are not practical if the script has to wait 2 minutes for each password it tries. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Python2 vs Python3
On Sun, 30 Oct 2011 14:03:59 -0500 Bruce Dubbs wrote: > basicnet/netprogs/subversion.xml Subversion definitely fails to compile it's python bindings with python-3 as /usr/bin/python, I tested that earlier. > general/genlib/glib2.xml Glib2 fails in it's gio tests if /usr/bin/python is python3, that can be worked around by neutering the gio/tests/Makefile or by removing the /usr/bin/python => python3 symlink > general/genlib/libxml2.xml With a lot of work it's possible to compile and install libxml2's python bindings into python3 but the effort was all in vain, they don't work as they use python 2 isms > xsoft/graphweb/firefox.xml > xsoft/other/thunderbird.xml The Thunderbird and Firefox configure scripts explicitly check that /usr/bin/python is python2. The only thing that I know of that uses python3 is pycairo, and even with that they produce a py2cairo that works with python2. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Python2 vs Python3
On Mon, 31 Oct 2011 01:33:40 -0500 DJ Lucas wrote: > On 10/30/2011 04:55 PM, Andrew Benton wrote: > > With a lot of work it's possible to compile and install libxml2's > > python bindings into python3 but the effort was all in vain, they don't > > work as they use python 2 isms > Did you try using 2to3 on them? It *seems* to work pretty well, and > gives actual line numbers for things that it doesn't know how to fix. > I'm know very little of Python, but a few searches in Google gave me > what I believe to be the needed fixes for the few things that 2to3 > couldn't handle in the xcb bindings. 2to3 is a useful tool for converting python2 to python3. The problem with libxml2 is the C files: types.c:594:17: error: ‘PyInstanceObject’ undeclared (first use in this function) Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
libmpeg2-0.5.1 Segmentation fault
Hello Although libmpeg2-0.5.1 compiles fine with gcc-4.6 it segfaults when I try to run anything linked to it. The solution is to edit libmpeg2/idct_mmx.c and change all the "static const" into just static: sed -i 's#static const#static#' libmpeg2/idct_mmx.c I found the solution in Mandriva's bugzilla: https://qa.mandriva.com/show_bug.cgi?id=63279 There's a patch there if you prefer patches. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: glib1 and gtk1
On Fri, 4 Nov 2011 21:39:38 -0400 Jonathan Oksman wrote: > xmms was fun back in the day, but the reality is the codebase is not > going to be brought into a gtk2 future. There was a fork called Beep > Media Player which ported the codebase to gtk2, but the project > appears to have been abandoned. Audacious was forked from Beep Media Player. When I added Audacious to BLFS it was because xmms was the only thing that I used that still needed gtk1. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: glib1 and gtk1
On Sat, 5 Nov 2011 10:03:40 -0400 Jonathan Oksman wrote: > Nice! I'm going to have to give that a whirl. The current version of Audacious is 3.0.4 http://distfiles.atheme.org/ Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Request for comment
On Tue, 08 Nov 2011 22:21:27 -0600 Bruce Dubbs wrote: > If people just posted to blfs-dev that said "I just > built package-x.y.z and I had to change the instructions to..." I'd do > the XML. I just built libmpeg2-0.5.1. It compiles according to the instructions in the book but segfaults when I try to run anything linked to it. The problem is gcc-4.6 optimising away static constants. The solution is to edit libmpeg2/idct_mmx.c and change all the "static const" into just static: sed -i 's#static const#static#' libmpeg2/idct_mmx.c I found the solution in Mandriva's bugzilla: https://qa.mandriva.com/show_bug.cgi?id=63279 There's a patch there if you prefer patches. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox-8.0.source
On Wed, 09 Nov 2011 13:28:56 -0600 Bruce Dubbs wrote: > I'd appreciate knowing the instructions you used. I can't do it yet > because I haven't put X on my development system yet. I compile firefox with: tar xf firefox-8.0.source.tar.bz2 chown -R root:root mozilla-release cd mozilla-release # make a separate build dir: mkdir firefox-build-dir # now make firefox build in that dir: echo "mk_add_options MOZ_OBJDIR=@TOPSRCDIR@/firefox-build-dir" > mozconfig # If we have more than 1 cpu do a parallel build: [ $(getconf _NPROCESSORS_ONLN) = 1 ] || echo "mk_add_options MOZ_MAKE_FLAGS=\"-j$(getconf _NPROCESSORS_ONLN)\"" >> \ mozconfig echo "ac_add_options --prefix=/usr ac_add_options --enable-application=browser ac_add_options --with-pthreads ac_add_options --with-system-bz2 ac_add_options --with-system-jpeg ac_add_options --with-system-png ac_add_options --enable-system-lcms ac_add_options --enable-system-cairo ac_add_options --with-system-zlib ac_add_options --enable-system-sqlite ac_add_options --with-system-libvpx ac_add_options --enable-system-hunspell ac_add_options --enable-startup-notification ac_add_options --enable-shared ac_add_options --disable-static ac_add_options --disable-debug ac_add_options --disable-crashreporter ac_add_options --disable-installer ac_add_options --disable-updater ac_add_options --disable-tests" >> mozconfig make -f client.mk cd firefox-build-dir make install Obviously, if you've not installed lcms, sqlite, libvpx, hunspell or startup-notification change that mozconfig to suit. Hope that helps. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: firefox-8.0.source
On Thu, 10 Nov 2011 07:33:08 -0600 DJ Lucas wrote: > Guys, just in case you are not aware, we should probably consider using > external npapi. We can of course, fix anything that is using the > incorrect version scheme, but looking to the future, that might be a bit > of work (OpenJDK specifically will be a PITA, don't know about other > open source plug-ins). I'm not sure what you mean there. Do you mean compile Firefox to use an external npapi? Greping through the output of ./configure --help I can see no mention of npapi. When I compile gnash-0.8.9 I configure it --with-npapi-incl=/usr/include/firefox-8.0 I also had to add a small fix it so it could use the new firefox headers (firefox-7.0 worked fine): sed -i 's#^char#const char#' \ plugin/npapi/{mozilla-sdk/np_entry,plugin}.cpp Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: links-2.2 with --enable-graphics
On Sun, 13 Nov 2011 11:53:39 -0600 Bruce Dubbs wrote: > Todd Simpson wrote: > > I found that links-2.2 does not compile when it is configured with the > > --enable-graphics switch and the system has libpng-1.5.5. > > > > The reason is that recent versions of libpng do not expose the contents > > of the png_struct structure, and the png.c file in links-2.2 tries to > > directly access a field of the structure. > > Did you try links-2.3? That seems to be pretty recent. I've tried links-2.3. It compiles and runs fine with libpng-1.5.5. It didn't need patching. ldd shows it is linked to libpng. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: BLFS Progress
On Mon, 14 Nov 2011 12:52:48 + Ken Moffat wrote: > AFAIK, the current release is still 7.6. Of course, many of the > individual packages have had more releases. I guess the big > question (if you want to update it) is "do you want to stick with > the 1.9 server, or move to 1.10 ?" FWIW I've been using xorg-server-1.11.0 since the end of August with no problems. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Just thought of something
On Wed, 16 Nov 2011 04:55:13 + Ken Moffat wrote: > Those were good enough to build firefox-8.0 (with its other > dependencies, particularly yasm (1.1.0) and a newer libvpx (I used > v0.9.7-p1, I expect that is now old hat) Funny you should mention that. Current Firefox from mercurial needs to be fixed to build against current libvpx from git, but it's quite happy with libvpx-0.9.7-p1. The current Firefox configure script tests the libvpx headers for a variable that was renamed in September, it can be fixed with: sed -i 's#VPX_CODEC_USE_INPUT_PARTITION#VPX_CODEC_USE_INPUT_FRAGMENTS#' \ configure Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
ppp-2.4.5
Hello, I've been thinking of doing some work on BLFS. Would anyone mind if I updated ppp to 2.4.5? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: ppp-2.4.5
On Wed, 16 Nov 2011 20:35:31 -0600 Bruce Dubbs wrote: > LOL. Be my guest. How will you test it? Who uses it today? I was > thinking it should be dropped. I use it and can test it for pppoe and pppoatm, but there's some stuff on the page about dialup, GPRS and EDGE that I can't test so I'll just leave that unchanged. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: ppp-2.4.5
On Wed, 16 Nov 2011 20:51:16 -0600 Bruce Dubbs wrote: > If you really want a challenge, get nfs-utils to work with glibc-2.14{,.1} Sorry, I've never used nfs. There are enough things that I do use that need some attention. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
gtk+-2.24, gtk+-3.2 and gdk-pixbuf
Hello, I'd like to update Audacious to the current version, 3.0.4, but Audacious depends of gtk. It can build against either gtk-+2 or gtk-+3. The version of gtk-+2 that we have in the book, 2.20.1 is quite old so before I update Audacious I'd like to update gtk-+2 to the current version, 2.24.8. gtk+-2.24 depends on gdk-pixbuf which isn't in the book so before I can update Audacious or gtk I need to add gdk-pixbuf to the book. Audacious prefers gtk+-3 so I'd like to add that to the book as well. So my plan is: add gdk-pixbuf-2.24.0 update gtk+-2 to 2.24.8 update Audacious to 3.0.4 and build it against gtk+-2 add gtk+-3.2.2 update Audacious to recommend gtk+-3 Would that be OK? I know there is an issue about the book growing ever larger, but more and more things are using gtk+-3 so I think we need to add it to the book. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: guile 2.0.3
On Fri, 18 Nov 2011 22:56:34 +1100 Wayne Blaszczyk wrote: > Also, I could not build Guile with libunistring. > Should libunistring be a required dependancy? Do you mean without libunistring? Guile-2.0.3 builds fine with libunistring-0.9.3 installed here. ldd shows libguile-2.0.so is linked to libunistring.so.0. It also shows that libguile is linked to libffi.so.5. libffi isn't listed as a guile dependency. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gdk-pixbuf-2.24.0
On Sat, 19 Nov 2011 12:29:39 +1100 Wayne Blaszczyk wrote: > The dependency pkg-config is linked to the Configuring pkg-config > section rather than the pkg-config page. Thanks Wayne, fixed. It should have been instead of I'll have a go at gtk+-2 tomorrow. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gdk-pixbuf-2.24.0
On Sat, 19 Nov 2011 13:42:14 +1100 Wayne Blaszczyk wrote: > NP, I couple of more things I've just noticed. > gobject-introspection and jasper are optional dependencies. Indeed they are. Thanks again. Fixed (again). Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: cups and glib
On Sat, 19 Nov 2011 19:46:20 +1100 Wayne Blaszczyk wrote: > Based on the current build instructions, Cups creates files under > /usr/etc and /usr/var. > Glib2 creates bash completion scripts under /usr/etc. > Not directly related to BLFS, but it also looks like Coreutils has > installed a library (libstdbuf.so) under /usr/libexec, unless I did > something wrong there? I agree these things should be fixed. The coreutils thing is an LFS problem but can be fixed by passing --libexecdir=/usr/lib to configure. I've not installed cups-1.5.0 yet so I can't comment. Glib can be made to put it's bash completion scripts in /etc passing --sysconfdir=/etc to configure. Glib also puts a header file, glibconfig.h into /usr/lib/glib-2.0/include. Moving that into /usr/include requires seds to the glib/Makefile and the pkg-config pc file. It also affects some other things like gobject-introspection which has the path to glibconfig.h hard coded in 2 Makefiles. gtk+-2 does a similar thing with a header but moving that one doesn't cause any problems for anything else. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] GTK+-2.24.8
On Sun, 20 Nov 2011 14:32:07 +1100 Wayne Blaszczyk wrote: > On 20/11/11 12:53, Wayne Blaszczyk wrote: > > The md5sum does not match. > > I believe it should be 5408d4a70ac132f7c439d7de40e83aff > > Regards, > > Wayne. > I should wait until the build finishes before posting;) > The following two commands failed for me: > cp -v docs/faq/*.html /usr/share/doc/gtk+-2.24.8/faq > cp -v docs/tutorial/*.html /usr/share/doc/gtk+-2.24.8/tutorial > I believe it should be > cp -v docs/html/faq/*.html /usr/share/doc/gtk+-2.24.8/faq > cp -v docs/html/tutorial/*.html /usr/share/doc/gtk+-2.24.8/tutorial Thanks Wayne. I think I know what the problem is with the docs. I didn't test the instructions without docbook-utils installed (I'm guessing you don't have docbook-utils). If the Makefile regenerates the tutorial and faq it puts them in a different place to the prebuilt versions. I need to work on the sed to the Makefiles so that the html ends up in the same place. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Sun, 20 Nov 2011 18:01:38 -0600 Bruce Dubbs wrote: > I've been trying to get livetex properly built and installed today and > have had a lot of problems. > > First, the build issues are relatively minor. I do not have X installed > and have to pass --without-x --disable-xdvipdfmx, but other than that, > it seems to build OK. > > The problem is in the install phase. It wants to create the following > directories in /usr: > > bin include lib share texmf texmf-dist > > The include and lib directories are OK, but the bin directory has a lost > of entries pointing to ../texmf-dist/scripts/ and ../texmf/scripts/ > > Probably worst of all, it removes the /usr/bin/man executable and > replaces it with a symlink to ../texmf/doc/man. > > I can probably write instructions to move all this stuff around to the > proper locations, but I wonder if it's worth it. Do we really need TeX > in BLFS? We do have teTeX, but that has not been updated since 2006. > > Right now, I'm considering dropping all of Chapter 48: > > 48. Typesetting > * TeX Live-20110705 > * teTeX-3.0 > * JadeTeX-3.13 > > Does anyone use this? I note that it is not installed by default in > either ubuntu or RHEL. We do use in in about 28 places in the book for > optionally rebuilding documentation, but I don't know if those sections > are really valuable to users or not. I tried to install it about a year ago and gave up. It left me with the impression that it was steaming pile of unmaintained broken rubbish that nobody cared for enough to fix. I am very much in favour of dropping broken rubbish from the book. I'm sorry you lost a days work to bring us this information. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] atk, pango, cairo and ffmpeg
Hello, I've started work on a new page to add gtk+-3 to the book. It's getting lots of developer attention so it's well maintained and installing it is a lot more straightforward than some of the other things I've done lately so I expect to have it done tomorrow. I've noticed that atk, pango, cairo and ffmpeg are not up to date. If no one objects I'd like to work on them next. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] udev-175
On Sun, 20 Nov 2011 17:40:19 -0800 (PST) Fernando de Oliveira wrote: > Hi, > > RE: LFS 6.5, 6.7, 6.8 > > I have been able to upgrade udev to 173, but since then, the attempts to > upgrade to 174 or directly to 175 have not succeeded. > > Systems boot until the error message (I am reading and typing): > > "fsck.ext4: No such file or directory while trying to open /dev/sda1 > Possibly non-existent device? [ Fail ] > FAILURE: > It looks like udevd is not running. Udev 174 and 175 install udevd into /usr/libexec which is not in the ${PATH} for the bootscripts. mv /usr/libexec/udevd /sbin and see if that helps. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Mon, 21 Nov 2011 08:01:01 + Jeremy Henty wrote: > > Bruce Dubbs wrote: > > > I can probably write instructions to move all this stuff around to the > > proper locations, > > ??? The book already has instructions to do this. I assumed you had > written them. > > > Does anyone use this? > > I use TeX. I very much appreciate having instructions in the book. > > BTW, I had to hack texk/kpathsea/texmf.cnf some more to replace the > $PATH* variables with something that worked. Otherwise texconfig > falls over complaining about the lack of directories like "//texmf". > Here's the sed: > > sed -i \ > -e '/^[^%]/s|{\$SELF[[:upper:]]*\(,\$SELF[[:upper:]]*\)*}|/usr|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*/lib|/usr/lib|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*/etc|/etc|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*|/usr/share|' \ > texk/kpathsea/texmf.cnf > Maybe we should install it in somewhere like /opt so it doesn't interfere with /usr/bin/man? I still think that if it's overwriting /usr/bin/man it should be commented out of the book until it's been fixed. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] ffmpeg
On Mon, 21 Nov 2011 21:25:13 + Ken Moffat wrote: > Hi Andy, > > I see you're intending to update ffmpeg, and I remember that in the > past you've disagreed with the configure flags I added. You seem to > manage to keep up to date with so many projects that you probably > know a lot more about this than I do, but nevertheless I'll offer my > (edited) notes from when I upgraded to 0.8.4 - Are you not tempted to update the page yourself? I don't mind, if you want to do it, drive on. I only offered to do it 'cos the version in the book is really old and I know a lot has changed with ffmpeg since 0.6.1. I see the latest version is 0.8.6. I've found that the new API (the 0.8 releases) has broken a few things but they're mostly fixable with seds. > > 1. For uploading certain formats (.mov, .mp4) to youtube, I need to > flatten them with qt-faststart. This was why I looked at newer > versions, because this failed to build in the 0.6 versions on the > system I was using (although it did build in older systems). > >make -C tools qt-faststart >install -m 0755 tools/qt-faststart /usr/bin I think you should put that in the book :) > > 2. A version of aac is included in ffmpeg (using it needs an > override on the command line when ffmpeg is run using the aac > acodec, [ -strict experimental ] (or, maybe only for encoding) > so libfaac is not needed. When I tried to encode an .m4a with ffmpeg it sounded terrible so I always use faac to encode them. ffplay sounds great decoding an .m4a > > 3. The internal ffacdec replaced libfaad during 0.7. > > 4. --enable-version3 on top of --enable-gpl, to use gpl version 3 : > I'm not sure if this changes what is built, or only its licence. > > 5. --enable-pthreads to use threading. andy@eccles:~$ '/home/andy/save/src/ffmpeg/configure' --help | grep pthreads --disable-pthreads disable pthreads [auto] I think --enable-pthreads is the default. > > 6. --enable-libvpx adds the libvpx decoder and encoder (there are > other vp* decoders which might perhaps do this, unsure if that > applies to encoders, I gave up noting the full list of them). ffmpeg includes a libvpx decoder and the developers are quite pleased with their decoder. --enable-libvpx adds an encoder to ffmpeg, but (my attitude is) if you've installed libvpx, which includes vpxenc, why would you need to use ffmpeg to encode a vpx video? I suppose it would be used to allow ffmpeg to convert from one format to another without piping from one application to another. > 7. --enable-libvorbis adds the libvorbis encoder as well as the > native vorbis encoder, there are claims that the libvorbis encoder > is/was better (better-quality, ISTR). > > 7. --enable-libtheora adds the libtheora encoder. > > 8. --enable-libmp3lame adds the libmp3lame encoder. > > 9. --enable-libx264 adds the libx264 encoder. > > 10. --enable-libxvid adds the libxvid (xvidcore) encoder. > > 11. I'm still using --enable-rtmp, but I have doubts if it is giving > me anything extra. > > Hope this might be helpful, and not too out of date. Thanks, it's interesting to think it through. I just use ffmpeg as a dependency for other things (lots of things use it). For me, I just use it for watching or listening to things encoded by other people but I know from your emails to the lists that you use it to encode videos. I build ffmpeg in a separate object directory and configure it: --enable-shared --disable-static --enable-postproc --enable-gpl and that does everything I need. When I rip a cd I encode it to m4a with faac and mp3 with lame. I think if I update the the ffmpeg page in BLFS I will use the options that work for me but discuss the options you've outlined above in the "Command Explanations" section; so I you want all these encoder options to be the default in the book you should update the page yourself. I think you should. I think you know more about it than me. I can help fix the breakage that ffmpeg-0.8 will introduce to other things. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] A comment about xz
On Mon, 21 Nov 2011 14:28:55 -0600 Bruce Dubbs wrote: > I've been working a bit with tex-live. I now uses xz to compress their > very large files (126M and 1.2G). I note that it takes quite a bit more > time to extract these files than bz2. The tradeoff may be worth it, but > the better compression is not free. > > time tar -xf texlive-20110705-texmf.tar.xz > > real2m3.082s > > This is on a reasonably fast system. I don't think you can blame xz for this. 1.2GB is a huge file. I get: andy@eccles:~$ time tar -xf texlive-20110705-texmf.tar.xz real2m0.428s user1m30.334s sys 0m5.243s So far, the same as you. But look at this, if I just use xz to decompress it it still takes one and a half minutes: andy@eccles:~$ rm -rf texlive-20110705-texmf andy@eccles:~$ time xz -d texlive-20110705-texmf.tar.xz real1m33.782s user1m31.334s sys 0m1.523s However, then using tar alone and it's back to two minutes again: andy@eccles:~$ time tar -xf texlive-20110705-texmf.tar real2m0.242s user0m0.370s sys 0m4.563s It decompresses to 2.3GB which thrashes the disk a lot. I suspect that much of the time is taken doing disk input/output. As I understand it the LZMA2 algorithm is slow to compress (and uses a lot of RAM) but is moderately quick (and uses a moderate amount of RAM) to decompress. The smaller file size for downloads makes it worth the extra time it takes to compress the archive in the first place. FWIW if I'm backing up a partition I don't compress it at all as I'm not short of disk space, I just use tar to make it into a single file. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Mon, 21 Nov 2011 17:31:15 -0600 Bruce Dubbs wrote: > Jeremy Henty wrote: > > Bruce Dubbs wrote: > > > >> I can probably write instructions to move all this stuff around to the > >> proper locations, > > OK, this is what I have: > > ./configure --prefix=/usr \ > --disable-native-texlive-build \ > --enable-shared\ > --without-luatex \ > --enable-mktextex-default \ > --with-banner-add=" - BLFS"&& > make > > make check > > make DESTDIR=$PWD/livetex-tmp install && > rm -v $PWD/livetex-tmp/usr/bin/man&& > for FN in $PWD/livetex-tmp/usr/bin `find -type l`; do > if [ `readlink $FN | grep "\.\./texmf"` ]; then should that be grep -q > ln -svf `readlink $FN | sed 's|\.\./texmf|../share/texmf|'` $FN > fi > done && > unset FN > > As root: > > cp -dRv livetex-tmp//usr/include/* /usr/include && > cp -dRv livetex-tmp/usr/lib/* /usr/lib && > cp -dRv livetex-tmp/usr/bin/* /usr/bin && > cp -dRv livetex-tmp/usr/doc/man/* /usr/share/man&& I don't get any livetex-tmp/usr/doc directory, however, there are lots of man pages in livetex-tmp/usr/texmf/doc/man so this works for me: cp -dRv livetex-tmp/usr/texmf/doc/man/* /usr/share/man && > > tar -xf ../texlive-20110705-texmf.tar.xz -C /usr/share \ > --strip-components=1 && > > cp -dRv livetex-tmp/usr/texmf /usr/share/texmf && At this point I ran into problems: cp: cannot create regular file `/usr/share/texmf/texmf/web2c/fmtutil.cnf': No space left on device Oh dear. I'll have a look at the rest in a bit. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Tue, 22 Nov 2011 13:18:03 + Andrew Benton wrote: > At this point I ran into problems: > > cp: cannot create regular file `/usr/share/texmf/texmf/web2c/fmtutil.cnf': No > space left on device So I made a larger partition and untared a minimal LFS build onto it (no xorg) and on that system the build fails quite early: gcc -DPACKAGE_NAME=\"xpdfopen\ for\ TeX\ Live\" -DPACKAGE_TARNAME=\"xpdfopen-for-tex-live\" -DPACKAGE_VERSION=\"0.82\" -DPACKAGE_STRING=\"xpdfopen\ for\ TeX\ Live\ 0.82\" -DPACKAGE_BUGREPORT=\"te...@tug.org\" -DPACKAGE_URL=\"\" -DPACKAGE=\"xpdfopen-for-tex-live\" -DVERSION=\"0.82\" -I.-Wimplicit -Wreturn-type -Wdeclaration-after-statement -Wno-unknown-pragmas -g -O2 -MT sendx.o -MD -MP -MF .deps/sendx.Tpo -c -o sendx.o sendx.c sendx.c:21:22: fatal error: X11/Xlib.h: No such file or directory compilation terminated. make[2]: *** [sendx.o] Error 1 make[2]: Leaving directory `/home/andy/texlive-20110705-source/utils/xpdfopen' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/andy/texlive-20110705-source/utils' make: *** [all-recursive] Error 1 Perhaps libX11 is a required dependency? I've got to do some work now. I'll come back to this later. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Mon, 21 Nov 2011 17:31:15 -0600 Bruce Dubbs wrote: > for INFO in livetex-tmp/usr/texmf/doc/info/*.info; do >install -v -m0644 $INFO /usr/share/info >intall-info --info-dir=/usr/share/info $INFO Typo, s/intall-info/install-info/ Apart from that and the problem I posted about earlier with livetex-tmp/usr/texmf/doc/man it all went well and seems to install properly. However, when I try to do the "Configuring TeX Live" section from the book I get: root@eccles:/home/andy$ texconfig-sys font rw /usr/bin/texconfig: line 248: //texmf/web2c/mktexdir: No such file or directory texconfig: failed to set `font rw'; reason: directory `//texmf-var/fonts' does not exist. Which brings me to: On Mon, 21 Nov 2011 08:01:01 + Jeremy Henty wrote: > BTW, I had to hack texk/kpathsea/texmf.cnf some more to replace the > $PATH* variables with something that worked. Otherwise texconfig > falls over complaining about the lack of directories like "//texmf". > Here's the sed: > > sed -i \ > -e '/^[^%]/s|{\$SELF[[:upper:]]*\(,\$SELF[[:upper:]]*\)*}|/usr|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*/lib|/usr/lib|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*/etc|/etc|' \ > -e '/^[^%]/s|\$SELF[[:upper:]]*|/usr/share|' \ > texk/kpathsea/texmf.cnf I can't find texk/kpathsea/texmf.cnf root@eccles:/home/andy$ find /usr -name texmf.cnf /usr/share/texmf/web2c/texmf.cnf /usr/share/texmf/texmf/web2c/texmf.cnf /usr/share/texmf-dist/doc/latex/pgfplots/texmf.cnf /usr/share/texmf-dist/doc/generic/pgf/text-en/texmf.cnf should I run the sed on one of them instead? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] livetex
On Tue, 22 Nov 2011 21:14:46 + Andrew Benton <63n...@gmail.com> wrote: > On Mon, 21 Nov 2011 17:31:15 -0600 > Bruce Dubbs wrote: > > > for INFO in livetex-tmp/usr/texmf/doc/info/*.info; do > >install -v -m0644 $INFO /usr/share/info > >intall-info --info-dir=/usr/share/info $INFO > > Typo, s/intall-info/install-info/ > > Apart from that and the problem I posted about earlier with > livetex-tmp/usr/texmf/doc/man it all went well and seems to install > properly. However, when I try to do the "Configuring TeX Live" section > from the book I get: > > root@eccles:/home/andy$ texconfig-sys font rw > /usr/bin/texconfig: line 248: //texmf/web2c/mktexdir: No such file or > directory > texconfig: failed to set `font rw'; reason: directory `//texmf-var/fonts' > does not exist. > > Which brings me to: > > On Mon, 21 Nov 2011 08:01:01 + > Jeremy Henty wrote: > > > BTW, I had to hack texk/kpathsea/texmf.cnf some more to replace the > > $PATH* variables with something that worked. Otherwise texconfig > > falls over complaining about the lack of directories like "//texmf". > > Here's the sed: > > > > sed -i \ > > -e '/^[^%]/s|{\$SELF[[:upper:]]*\(,\$SELF[[:upper:]]*\)*}|/usr|' \ > > -e '/^[^%]/s|\$SELF[[:upper:]]*/lib|/usr/lib|' \ > > -e '/^[^%]/s|\$SELF[[:upper:]]*/etc|/etc|' \ > > -e '/^[^%]/s|\$SELF[[:upper:]]*|/usr/share|' \ > > texk/kpathsea/texmf.cnf > > I can't find texk/kpathsea/texmf.cnf > > root@eccles:/home/andy$ find /usr -name texmf.cnf > /usr/share/texmf/web2c/texmf.cnf > /usr/share/texmf/texmf/web2c/texmf.cnf > /usr/share/texmf-dist/doc/latex/pgfplots/texmf.cnf > /usr/share/texmf-dist/doc/generic/pgf/text-en/texmf.cnf > > should I run the sed on one of them instead? Doh! I see, texk/kpathsea/texmf.cnf in the texlive source. Running the sed on the first one got me past that problem, but now at updmap-sys I get: root@eccles:/home/andy$ updmap-sys Can't locate TeXLive/TLUtils.pm in @INC (@INC contains: /usr/share/tlpkg /usr/lib/perl5/site_perl/5.14.1/x86_64-linux /usr/lib/perl5/site_perl/5.14.1 /usr/lib/perl5/vendor_perl/5.14.1/x86_64-linux /usr/lib/perl5/vendor_perl/5.14.1 /usr/lib/perl5/5.14.1/x86_64-linux /usr/lib/perl5/5.14.1 .) at /usr/bin/updmap line 18. BEGIN failed--compilation aborted at /usr/bin/updmap line 18. I can see TeXLive/TLUtils.pm in the texlive-20110705-source but Bruce's install method doesn't install it. mkdir /usr/lib/perl5/site_perl/5.14.1/TeXLive cp utils/biber/TeXLive/*.pm /usr/lib/perl5/site_perl/5.14.1/TeXLive After that the rest of the "Configuring TeX Live" commands worked. However, when I come to try and use it: andy@eccles:~$ tex example1.tex This is TeX, Version 3.1415926 (TeX Live 2011 - BLFS) fatal: memory exhausted (xmalloc of 18446744071586067984 bytes). It appears it's trying to allocate 17179869182.02 gigabytes of memory. It could well be user error on my part but I still have the impression that this is a steaming pile of unmaintained rubbish. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Typo - libmowgli download package misspelled
On Fri, 25 Nov 2011 09:30:24 -0600 William Immendorf wrote: > Apparently the package name for libmowgli is misspelled - the package > name should be libmowgli-0.9.95.tar.bz2. The little 2 at the end is > missing in the name. Doh! Thanks for the heads up. Fixed. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] Firefox
Hello, Is anyone working on Firefox? Would it be Ok if I updated it? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] NSPR-4.8.9, pthreads and x86_64
Hello, When I try to install nspr-4.8.9 configure fails like so: checking for pthread_create in -lpthreads... no checking for pthread_create in -lpthread... no checking for pthread_create in -lc_r... no checking for pthread_create in -lc... no configure: error: --with-pthreads specified for a system without pthread support looking at config.log I see:configure:5934: checking for pthread_create in -lpthread gcc -m32 -o dummy dummy.c -Wall -lpthread In file included from /usr/include/features.h:382:0, from /usr/include/pthread.h:23, from dummy.c:2: /usr/include/gnu/stubs.h:7:27: fatal error: gnu/stubs-32.h: No such file or directory compilation terminated. The problem is the -m32 switch, it doesn't work on an x86_64 system. If I take out the -m32 that out then the command succeeds: sed -i 's# -m32##' configure With that sed the build succeeds and installs, but is it wise? There must be a reason that they make the CC command CC="$CC -m32" Either we sed configure to remove -m32, we stop using the --with-pthreads option or we only use the the --with-pthreads option on i386. I'll try installing it with the sed and build Firefox against it tomorrow. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] NSPR-4.8.9, pthreads and x86_64
On Mon, 28 Nov 2011 18:38:24 +1100 Wayne Blaszczyk wrote: > I didn't have a problem building this on a x86_64 system. You did add > the --enable-64bit parameter? No, I didn't notice it buried down in the command explanations. Shouldn't there be some text about it up above the configure command? It's Ok to put optional tweaks in Command Explanations but essential stuff should be discussed beforehand. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] NSPR-4.8.9, pthreads and x86_64
On Mon, 28 Nov 2011 22:47:23 +1100 Wayne Blaszczyk wrote: > Agreed. > Feel free to update the page. Thanks. I'd like to change the way we handle the scripts it installs. It seems odd to install them and then delete them. This sed prevents them being installed: sed -i 's#^\(RELEASE_BINS =\).*#\1#' pr/src/misc/Makefile Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Firefox and GTK Engines
On Tue, 29 Nov 2011 00:00:55 +1100 Wayne Blaszczyk wrote: > GTK Engines should be an optional dependency, without it I get the > warnings about clearlooks theme missing. > The highlighting of the menu does not great without it. Is that when you configure it or at runtime? If it's at runtime I could maybe add a note like the one about `which` on the docbook-utils page. I'm working on nspr, nss and xulrunner at the moment. I'll get back to firefox in a couple of days when I've got these other dependencies sorted out. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Firefox and GTK Engines
On Tue, 29 Nov 2011 07:08:53 +1100 Wayne Blaszczyk wrote: > runtime. Thanks Wayne, I'll look into it and add a note about it when I get back to Firefox in a few days time. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] GPL Ghostscript URL has moved
On Mon, 28 Nov 2011 16:22:00 -0800 Nathan Coulson wrote: > Looks like the ghostscript url's are in > http://downloads.ghostscript.com/public/, > BLFS uses http://ghostscript.com/releases/ghostscript-9.04.tar.bz2 Thanks. Fixed. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Firefox and GTK Engines
On Tue, 29 Nov 2011 07:08:53 +1100 Wayne Blaszczyk wrote: > On 29/11/11 00:58, Andrew Benton wrote: > > On Tue, 29 Nov 2011 00:00:55 +1100 > > Wayne Blaszczyk wrote: > > > >> GTK Engines should be an optional dependency, without it I get the > >> warnings about clearlooks theme missing. > >> The highlighting of the menu does not great without it. > > > > Is that when you configure it or at runtime? > > runtime. I think this is an issue that's common to all gtk2 applications so I'm not going to write about it on the Firefox page. It'd probably be best to talk about gtk themes on the gtk-engines page and make a note pointing to it on the gtk page. I'll do it in a day or two. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] Yasm
Hello Firefox requires yasm http://yasm.tortall.net/ to compile the libvpx code included in the Firefox source. I'm going to make yasm a recommended dependency of Firefox and Xulrunner. Should yasm be added to the book? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Firefox
On Sun, 27 Nov 2011 00:15:33 +1100 Wayne Blaszczyk wrote: > Could I make one suggestion. > Can we have the /usr/lib/mozilla/plugins as the directory, and symlink > to this rather than the other way around as it is now. > This way, the plugins will not be tie to a specific version of Firefox > and save the user trouble of copying them across every time Firefox is > upgraded. Fernando de Oliveira has just made a good point about this on BLFS Support. If we put plugins in /usr/lib/mozilla/plugins they'll be found by all Mozilla products including Thunderbird. This raises security concerns, if there is an exploitable bug in a plugin. I seem to recall that there were some problems with Adobe's pdf plugin. Certainly flash has its bugs. Whatever. The bugs are all there in the browser. If the user wants an email application that can render html they may well like to have plugins enabled too. Personally I prefer Sylpheed because it doesn't render html but each to their own. What do other people think? Should I change the Firefox/Xulrunner pages so they don't suggest putting the plugins in /usr/lib/mozilla/plugins? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg packages - building less, and minimal config
On Thu, 1 Dec 2011 03:44:35 + Ken Moffat wrote: > MesaLib-7.10.3 > I plan to test 7.11.2 but I bailed on this build - after sedding > out the requirement for libdrm_nouveau >= 0.6 (I only have ati > video cards) it told me I needed llvm : I think Andy posted about > how to tame that package the other day, but sitting in front of > a text screen I figured the best way forward was to go back to a > version I'd built recently on an older system which I new had fewer > dependencies. FWIW, I can build llvm without static libs like so: ./configure --prefix=/usr # Don't build static libraries: sed -i 's#BUILD_ARCHIVE :*= 1#NO_BUILD_ARCHIVE = 1#' \ $(grep -rl 'BUILD_ARCHIVE :*= 1' .) # Build shared libraries: find . -name Makefile -exec sed -i '/SHARED_LIBRARY/d /LIBRARYNAME/a SHARED_LIBRARY = 1' {} \; # Remove some hard coded static libs. Use shared libs instead: for thing in {utils/*,projects/sample/tools/sample}/Makefile Makefile.rules do sed -i 's#.* \([^ ]*[ert]\)\.a#LDFLAGS = -l\1#' ${thing} done make -j$(getconf _NPROCESSORS_ONLN) make install Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg packages - building less, and minimal config
On Thu, 1 Dec 2011 16:41:23 + Ken Moffat wrote: > Thanks for (re) posting this. Am I right in thinking llvm is > required for current Mesa, or do you have another sed to let it > build with gcc ? I don't know for sure (as I install llvm) but I think it's gallium that uses llvm. It may compile without llvm if you configure it: --disable-gallium-llvm Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg packages - building less, and minimal config
On Thu, 1 Dec 2011 17:46:53 + Andrew Benton wrote: > On Thu, 1 Dec 2011 16:41:23 + > Ken Moffat wrote: > > > Thanks for (re) posting this. Am I right in thinking llvm is > > required for current Mesa, or do you have another sed to let it > > build with gcc ? By the way, I think gallium uses llvm at runtime to do 3D stuff, Mesa compiles with gcc > I don't know for sure (as I install llvm) but I think it's gallium that > uses llvm. It may compile without llvm if you configure it: > --disable-gallium-llvm I've just done a test build without llvm. I configured Mesa with: ./configure --prefix=/usr --enable-gallium-egl --disable-gallium-llvm \ --enable-openvg --enable-xcb --with-gallium-drivers=i915,nouveau,r600,swrast It's compiled fine without llvm. glxgears runs. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Fri, 02 Dec 2011 10:33:55 -0600 Todd Simpson wrote: > I tried to build gtk+-3.2.2 on my LFS 7.0 system and the configure > script failed, with a message that I needed pango-1.29.0 or later. I > had pango-1.28.2 because that's what the BLFS book said was needed. > > I downloaded pango-1.29.4 and it compiled and tested OK, and after > installing it, gtk+-3.2.2 could be built. Yes, I think you're right. I'll update pango tomorrow > > I also found that in the gtk+-3.2.2 test suite, there were tests that > expected a directory called ~/.local/share (where ~ is the home > directory of the user running the tests). Nothing I had done before had > created this directory. The tests failed at first, but succeeded after > I manually created the directory. Thanks, I'll have another look at gtk+-3.2.2 in a couple of days. I don't think this is the biggest problem with BLFS at the moment. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Fri, 2 Dec 2011 17:08:52 + Andrew Benton wrote: > On Fri, 02 Dec 2011 10:33:55 -0600 > Todd Simpson wrote: > > > I tried to build gtk+-3.2.2 on my LFS 7.0 system and the configure > > script failed, with a message that I needed pango-1.29.0 or later. I > > had pango-1.28.2 because that's what the BLFS book said was needed. > > > > I downloaded pango-1.29.4 and it compiled and tested OK, and after > > installing it, gtk+-3.2.2 could be built. > > Yes, I think you're right. I'll update pango tomorrow I'm starting to have second thoughts about this. I looked at http://en.wikipedia.org/wiki/Pango which says that the current stable version of Pango is 1.29.5 but then I looked at http://www.gtk.org/documentation.php and the links there seem to suggest that pango-1.29 is unstable and pango-1.28 is the stable series. Personally I've been using pango-1.29.3 for months and I'm not aware of any bus in it, but then I'm English and Pango is for laying out international text. So what do people think, is it Ok to put pango-1.29 in the book? gtk+-3.2 requires pango-1.29. gtk+-3.2.2 is the current stable version of gtk+-3. It's a contradiction. Advice please. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Sat, 03 Dec 2011 06:51:01 -0600 DJ Lucas wrote: > Not much choice here. If it is required then it is required. I would > wonder if pango-1.30 is on the horizon though. As a quick gauge, what > was the pango requirement of 3.2.1? All the gtk+-3.2 versions require pango-1.29.0 Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Sat, 3 Dec 2011 15:41:43 + Ken Moffat wrote: > It's required for gtk+-3.2.2. 1.29.4 is good enough to build that, > but 1.29.5 needs a development glibc. I think you mean glib, pango-1.29.5 wants glib-2.31. So I'll update the book to pango-1.29.4 which will have to do until pango-1.30 is released. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Gnash 0.8.10dev
On Sat, 3 Dec 2011 09:59:10 -0800 (PST) Fernando de Oliveira wrote: > Hi, > > I have had a sour time with gnash, which (dev version) I had succesfully > compiled and installed, months ago, in LFS-6.8, but not in 6.5 nor 6.7. For > reasons I have already written, I believe some "flash" animation plugin could > be in the book. > > I had problems installing the stable 0.8.9, so I only can contribute with > 0.8.10dev. > Jolly good, glad you got it working. I'm still using gnash-0.8.9. It needs some of the headers from the firefox source: mkdir /usr/include/npapi cp dom/plugins/base/*.h /usr/include/npapi Then I can compile gnash-0.8.9 like so: # Patch needed for current FFmpeg: patch -p1 < ../gnash-0.8.9-ffmpeg.patch # sed fix needed for firefox-8.0: sed -i 's#^char#const char#' plugin/npapi/{mozilla-sdk/np_entry,plugin}.cpp ./configure --prefix=/usr --enable-gui=gtk --sysconfdir=/etc \ --enable-media=ffmpeg --with-npapi-incl=/usr/include/npapi \ --with-npapi-plugindir=/usr/lib/mozilla/plugins make -j$(getconf _NPROCESSORS_ONLN) make install make install-plugin Andy gnash-0.8.9-ffmpeg.patch Description: Binary data -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Sun, 04 Dec 2011 13:19:52 +1100 Wayne Blaszczyk wrote: > gnome-media is no longer part of Gnome3 (as of 3.2). > The next release of Gnome will be next March. > If anyone is interested, here is my current build for Gnome3: > http://www.linuxfromscratch.org/~wblaszcz/blfs-book-xsl/index.html Very good. I'd like to see vte-0.30.1 in the book (as I use it for roxterm) it can be installed in parallel with vte-0.28 as it uses gtk+-3 and calls itself vte2_90 and libvte2_90.so so there are no clashes with names. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] texlive
On Sat, 03 Dec 2011 20:20:45 -0600 Bruce Dubbs wrote: > Option 4. Yet another option is to completely remove TeX from the book. > It is usually referenced in other packages to rebuild documentation. > I really doubt many users do that. > > Right now I lean towards option 3, but need to get feedback. I vote for option 4. Instructions for how to use a precompiled installer aren't really the LFS way of doing things. If there were an editor who was keen on tex, who knew how to get it working, that would be great but 'til then, away with it. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Gnash 0.8.9 (was Gnash 0.8.10dev)
On Sun, 4 Dec 2011 06:08:45 -0800 (PST) Fernando de Oliveira wrote: > I do not know the meaning of make -j$(getconf _NPROCESSORS_ONLN). Please, > educate me. Oops, sorry, that was copy and pasted from one of my scripts and I didn't notice it. getconf is installed by glibc. getconf _NPROCESSORS_ONLN returns the number of processors your computer has, or appears to have (on a hyperthreading intel CPU with only one core it will appear to have 2 cores due to hyperthreading). The -j option to make runs multiple jobs in parallel, this is useful if you have multiple CPU cores as it can use all the cores and speed up the build. So make -j$(getconf _NPROCESSORS_ONLN) runs make and adjusts the number of jobs it runs according to the system it is running on. This is useful if the script is to be run on several different machines. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Gnome 3
On Mon, 5 Dec 2011 00:27:59 + Ken Moffat wrote: > Well, I certainly won't be following the full build. But it's > already *very* useful in looking at dependencies for the things I do > intend to build. > > I can't see that this will break anything outside gnome-2, so I > support dropping it in. What are we doing with gnome-2? Presumably it will be dropped at some point, what sort of timescale are we looking at? 6 months? a year? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Gnome 3
On Mon, 05 Dec 2011 20:09:19 +1100 Wayne Blaszczyk wrote: > I could only get gnome-shell to work with the nvidia drivers. The > nv drivers just didn't cut it, they kept going back to fallback mode. Have you tried Nouveau? I used to use it (I don't use any Nvidia hardware at the moment) and found it worked well. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Gnome 3
On Mon, 5 Dec 2011 02:05:11 + Ken Moffat wrote: > Anything different that you're concerned about ? I was just curious what was happening with gnome. I think gstreamer uses Gconf? I also use gksu (I think that uses gconf and orbit?) for gparted. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg update
On Mon, 05 Dec 2011 02:18:13 -0600 DJ Lucas wrote: > LLVM will need to be added to the book for Mesa (Andy posted a nice > build script the other day that links everything dynamic for 2.9, but > 3.0 fails to build for me). Thanks very much :) I've had a look at llvm-3.0 and this seems to work: touch tools/edis/EnhancedDisassembly.exports ./configure --prefix=/usr sed -i 's#BUILD_ARCHIVE :*= 1#NO_BUILD_ARCHIVE = 1#' \ $(grep -rl 'BUILD_ARCHIVE :*= 1' .) find . -name Makefile -exec sed -i '/SHARED_LIBRARY/d /^LIBRARYNAME/a SHARED_LIBRARY = 1' {} \; for thing in {utils/{F,f,n,Tab}*,projects/sample/tools/sample}/Makefile do sed -i -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ -e 's#^USEDLIBS#LDFLAGS#' ${thing} done make -j4 make install make check gives: Expected Passes: 5273 Expected Failures : 48 Unsupported Tests : 15 Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg update
On Mon, 5 Dec 2011 20:07:00 + Andrew Benton wrote: > I've had a look at llvm-3.0 and this seems to work: > > touch tools/edis/EnhancedDisassembly.exports > ./configure --prefix=/usr > sed -i 's#BUILD_ARCHIVE :*= 1#NO_BUILD_ARCHIVE = 1#' \ > $(grep -rl 'BUILD_ARCHIVE :*= 1' .) > find . -name Makefile -exec sed -i '/SHARED_LIBRARY/d > /^LIBRARYNAME/a SHARED_LIBRARY = 1' {} \; > for thing in {utils/{F,f,n,Tab}*,projects/sample/tools/sample}/Makefile > do sed -i -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ > -e 's#^USEDLIBS#LDFLAGS#' ${thing} > done > make -j4 > make install I've tidied that up a bit: sed -i 's#x)/docs#x)/share/doc#' Makefile.config.in touch tools/edis/EnhancedDisassembly.exports ./configure --prefix=/usr find . -name Makefile -exec sed -i -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1' \ -e 's#^BUILD_ARCHIVE :*= 1#NO_BUILD_ARCHIVE = 1#' \ -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ -e 's#^USEDLIBS#LDFLAGS#' {} \; make -j4 make check make install Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg update
On Mon, 05 Dec 2011 17:38:28 -0600 Bruce Dubbs wrote: > Here's something else that is even more tidy: > > find -name Makefile -exec\ >sed -ir -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1'\ >-e 's#^BUILD_ARCHIVE :*= 1#NO_' \ >-e '/^USEDLIBS/s# ([A-Za-z]*).a# -l\1#g' \ >-e 's#^USEDLIBS#LDFLAGS#' {} \; > > Using the -r switch you don't have to escape parens. Moving the > \newline gives a nice visual separation to differentiate from \1 and \; > The ampersand shortens up the second expression. > > The . in find is not required, it is the default. > > I didn't completely check it, but thought I'd pass some ideas. Unfortunately there were a couple of errors in what I posted earlier. This one has been checked and works: touch tools/edis/EnhancedDisassembly.exports ./configure --prefix=/usr find -name Makefile -exec\ sed -ir -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1'\ -e 's#BUILD_ARCHIVE :*= 1#NO_' \ -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ -e 's#^USEDLIBS#LDFLAGS#' {} \; sed -i 's#BUILD_ARCHIVE = 1#NO_' Makefile.rules sed -i 's#x)/docs#x)/share/doc#' Makefile.config make -j4 make check make install Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] Packages for BLFS
Hello, There are quite a few packages that I use every day that are not in BLFS. I think it's a shame that they're not in the book so I've started to knock up example pages that show what a BLFS page might look like for them. The first 4 are: http://www.linuxfromscratch.org/~andy/BLFS/vte2_90.html This is the gkt+-3 version of vte that's already in the book. It can be installed in parallel with vte-0.28 as it adds 2_90 to the ends of the names of everything it installs. This will have to go into the book when Gnome 3 is added. I've written a page for it as it's used by Roxterm: http://www.linuxfromscratch.org/~andy/BLFS/roxterm.html Roxterm is basically gnome terminal without the need to install Gnome. What's not to like? http://www.linuxfromscratch.org/~andy/BLFS/openbox.html Openbox is a lightweight window manager, much like Fluxbox. I like it 'cos I can launch all the programs I use from the right click menu which is configured with a simple xml file in ~/config/openbox http://www.linuxfromscratch.org/~andy/BLFS/rox-filer.html Rox is a simple, lightweight gtk+-2 file browser. Give them a try. I think they all deserve to be added to the book. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Packages for BLFS
On Tue, 06 Dec 2011 12:59:00 -0600 DJ Lucas wrote: > Checked out LXDE's packages? Also uses openbox for the recommended wm. I used > to use flux with the others as well, so I found it to be a nice inbetween > environment. I think around 350 packages for a fully functional desktop. I use some packages from LXDE. LXPanel (which depends on 2 other things) and LXAppearance. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gtk+-3.2.2 and pango
On Sun, 04 Dec 2011 18:15:51 +1100 Wayne Blaszczyk wrote: > On 04/12/11 15:30, Ken Moffat wrote: > > Impressive! Looks as if I might be putting pulse into this build, > > now that I've got something to work from. What's stopping you > > putting all this into the book ? > It's incomplete. Most of it, were you see the ??, need to have their > dependencies reviewed, and there is a lot of placeholders for the > description text with the new packages. I've found it a battle to keep > up with the minor revisions. I don't think I have the time to do all > this. I could just dump all the stuff as is, and let other people update > the individual pages as required? Let me know if there is any objections > to this approach. +1 For putting these Gnome 3 pages into the book. I think if we wait until someone has prepared perfect pages for the whole of Gnome 3 before we add it to the book then it will never happen. Put some sort of note at the top of all the pages that need improving/finishing and then we can update a page at a time and inch towards getting it done. Something like: This page is a work in progress and all the details have not been checked. Then as we work through we can remove the note from pages that have been checked. One page at a time. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xulrunner 8.0.1
On Thu, 8 Dec 2011 09:29:31 -0800 (PST) Fernando de Oliveira wrote: > 1. A small typo in mozconfig: > mk_add_options MOZ_OBJDIR=@TOPSRCDIR@/firefox-build-dir > but the command reads > make -C xulrunner-build-dir install. Thanks, I hadn't noticed. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Firefox 8.0.1 (against Xulrunner)
On Thu, 8 Dec 2011 09:29:42 -0800 (PST) Fernando de Oliveira wrote: > 1. I believe, in the hurry, a partial mozconfig text is in the book, because > prefix is intended to be /usr, not /usr/local, as I found after first > successful build. Thus, > > ac_add_options --prefix=/usr Yes, you're right. It's not needed if you untar it into /usr/lib but it is needed if you build against xulrunner. > > 2. The application to be built is browser, but it seem to be missing: > > ac_add_options --enable-application=browser I don't think this is needed, if you read the configure script it is the default. > 4. Again, because it is not necessary: > > -ln -sv ../mozilla/plugins /usr/lib/firefox-8.0.1 && It's not necessary, but it makes it clear that we're keeping the plugins in a common directory. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] libdrm and mesa
On Fri, 09 Dec 2011 23:58:04 -0600 Bruce Dubbs wrote: > I noticed that libdrm-2.4.28 has been released, but it doesn't build > with the current instructions: > > make[3]: Entering directory `/usr/src/libdrm/libdrm-2.4.28/tests/radeon' >CC rbo.o > rbo.c:34:17: fatal error: rbo.h: No such file or directory > > Using --disable-radeon allows it to build/test ok. > https://bugs.freedesktop.org/show_bug.cgi?id=43538 > It's been fixed upstream, but we may want to wait for the next release. I made a patch to add the 2 missing files: http://www.linuxfromscratch.org/patches/downloads/libdrm/libdrm-2.4.28-build-fixes-1.patch Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] libdrm and mesa
On Fri, 09 Dec 2011 23:58:04 -0600 Bruce Dubbs wrote: > Now on to Mesa-7.11.2. > It fails in make. There are evidently incompatible changes in llvm 3.0. > > gallivm/lp_bld_debug.cpp:30:40: fatal error: > llvm/Target/TargetRegistry.h: No such file > > http://lists.freedesktop.org/archives/mesa-dev/2011-September/011639.html > > but that's not enough. There are other failures. I'll try backing up > to llvm-2.9 tomorrow. I've made a patch that you could try: http://www.linuxfromscratch.org/patches/downloads/MesaLib/Mesa-7.11.2-llvm-3.0-1.patch It compiles but I've not tested it. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] gimp
Hello Ken, Thanks for updating the gimp, it was on my todo list but I wanted to update webkit first as the gimp uses it to view the help files. At least, it does for me, however, I don't see webkit mentioned on the gimp page as an optional dependency? Should I add it to the gimp page when I've updated webkit (in a couple of days) Also, the 2 patches on the page are a 404 page not found error for me at the moment. Should the URLs point to: http://www.linuxfromscratch.org/patches/downloads/gimp ? At the moment they point to: http://www.linuxfromscratch.org/patches/blfs/svn Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gimp
On Sun, 11 Dec 2011 04:53:45 + Ken Moffat wrote: > On Sun, Dec 11, 2011 at 01:33:15AM +0000, Andrew Benton wrote: > > Hello Ken, > > Thanks for updating the gimp, it was on my todo list but I wanted to > > update webkit first as the gimp uses it to view the help files. At > > least, it does for me, however, I don't see webkit mentioned on the > > gimp page as an optional dependency? Should I add it to the gimp page > > when I've updated webkit (in a couple of days) > > I think I put it in a comment in the xls, but I'm *surprised* > yours uses webkit - mine opens aurora (i.e. firefox). Anyway, it > tests for webkit-1.0.pc - in my experience that comes from old (<1.6) > versions of webkit-gtk built with gtk+-2. But I think you told me > 1.6 could be built with gtk+-2 so perhaps it installs a different > version of hte pkgconfig file in that case ? > > My build of 1.6.1 with gtk+-3 installed webkit-3.0.pc. > > I build the gimp long before I touch the gnome applications, but > even after webkit-gtk has been installed, the log from the gimp's > configure shows - > > checking for X11/xpm.h... yes > checking for WEBKIT... no > checking for SVG... yes > > PS - since I'm talking about the gimp now - PyGTK appears to be > broken, probably by glib-2.30. pygobject-2.28.6 crapped out with > > pygi-info.c: In function '_pygi_info_new': > pygi-info.c:165:14: error: 'GI_INFO_TYPE_ERROR_DOMAIN' undeclared > (first use in this function) > > pygobject-3.0.2 is fine, but it installs pygobject-3.0.pc - > pygtk-2.24.0 is looking for pygobject-2.0.pc. 3.0.2 is in Wayne's > gnome3 packages, I assume it gets used by something. > > At the time pygtk failed on me, I couldn't get to fedora gitweb to > see what they do. I checked PLD, but they don't install pygtk (nor > do they make check - there is much to be said for that attitude ;-). > I've since proved that I don't need pygtk for what I do in gnumeric, > so it doesn't bother me but I did get reminded by the (my?) > explanation on the --disable-python switch. > > ĸen > -- > das eine Mal als Tragödie, das andere Mal als Farce > -- > http://linuxfromscratch.org/mailman/listinfo/blfs-dev > FAQ: http://www.linuxfromscratch.org/blfs/faq.html > Unsubscribe: See the above information page -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] gimp
On Sun, 11 Dec 2011 04:53:45 + Ken Moffat wrote: > I think I put it in a comment in the xls, but I'm *surprised* > yours uses webkit - mine opens aurora (i.e. firefox). Anyway, it > tests for webkit-1.0.pc - in my experience that comes from old (<1.6) > versions of webkit-gtk built with gtk+-2. But I think you told me > 1.6 could be built with gtk+-2 so perhaps it installs a different > version of hte pkgconfig file in that case ? > > My build of 1.6.1 with gtk+-3 installed webkit-3.0.pc. Yes, if both gtk+-2 and gtk+-3 are installed, webkit defaults to building against gtk+-3. To build against gtk+-2 use the configure option --with-gtk=2.0 Everything it installs is suffixed with a -1.0 whereas with gtk+-3 everything is suffixed with a -3.0 so they can both be installed in parallel, like gtk+-2 and gtk+-3 themselves > > I build the gimp long before I touch the gnome applications, but > even after webkit-gtk has been installed, the log from the gimp's > configure shows - > > checking for X11/xpm.h... yes > checking for WEBKIT... no I think if you read the configure script it's looking for webkit-1.0 > checking for SVG... yes > > PS - since I'm talking about the gimp now - PyGTK appears to be > broken, probably by glib-2.30. pygobject-2.28.6 crapped out with > > pygi-info.c: In function '_pygi_info_new': > pygi-info.c:165:14: error: 'GI_INFO_TYPE_ERROR_DOMAIN' undeclared > (first use in this function) I think that this GI stuff is gobject introspection brokeness. pygobject-2.28.6 installs fine for me with --disable-introspection > > pygobject-3.0.2 is fine, but it installs pygobject-3.0.pc - > pygtk-2.24.0 is looking for pygobject-2.0.pc. 3.0.2 is in Wayne's > gnome3 packages, I assume it gets used by something. I don't use any gtk+-3 python stuff (yet) so I can't comment. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg
On Sun, 11 Dec 2011 00:13:08 -0600 Bruce Dubbs wrote: > Mostly for DJ, > > I finally finished going through Xorg. I always get a sense of > satisfaction when bringing it up for the the first time on a new system. > It's the same feeling when I start LFS for the first time after a major > update. > > DRI and glxgears/glxinfo worked right off the bat. glxgears was only > about 550 FPS or so (my main system is 9900 FPS with a proprietary > nvidia driver), but that's mostly a HW issue. If you use the new glxgears from the mesa-demos it syncs with the refresh rate of the monitor so it always gives me 60 fps > I reviewed your instructions and I think they are extremely well done. > I did make a couple of formatting changes and touchups here and there, > but I liked what you did. > > Some of the dependencies needed to be updated and I did that. I think > we are pretty much up to speed there. > > After the builds, we have 'Xorg-7.6-2 Configuration', but we don't ever > tell the user how to test xorg. I guessed at 'startx' and that brought > up twm. We need to say that somewhere. xinit works too. > > The default startx brings up three xterms and an xclock. We also need > to tell the user that the leftmost xterm is a login xterm and that > exiting from that xterm will terminate the X session. > > BTW, I built in /opt/xorg. I like to do that so I can build another > version of xorg without changing the current programs that would be the > case if installing in /usr. > > On my system, the 3rd xterm (xterm -geometry 80x20+494-0) is completely > covered by the other two. We probably should mention that possibility. > > Looking at the Xorg.0.log I had a couple of entries that I don't > completely understand. > > 1. (WW) Open ACPI failed (/var/run/acpid.socket) > > I've seen that before. I suppose we might want to consider adding apcid > to the book. > > 2. (II) NV: driver for NVIDIA chipsets: many > (WW) Falling back to old probe method for vesa > (WW) Falling back to old probe method for fbdev > > I'm not sure what these warnings mean. I'm sure they're harmless, but I > wonder why they're there. I think it's because you don't have kernel mode setting enabled in you kernel config. > Later it says: > > (II) UnloadModule: "vesa" > (II) Unloading vesa > (II) UnloadModule: "fbdev" > (II) Unloading fbdev > (II) UnloadModule: "fbdevhw" > (II) Unloading fbdevhw > > Perhaps I just need to set /etc/X11/xorg.conf.d/videocard-0.conf > > 3. (EE) open /dev/fb0: No such file or directory > > Is this a kernel configuration issue? I have CONFIG_FB=y, but not > CONFIG_FB_DDC or CONFIG_FB_BOOT_VESA_SUPPORT. I do think it's harmless, > but I'm just trying to run down the details. > > 4. It seems to ID the monitor and video card correctly: > (II) NV: driver for NVIDIA chipsets: ... > (--) NV: Found NVIDIA Quadro NVS 55/280 PCI at 01@00:00:0 > (II) NV(0): Monitor name: DELL IN1910N > > I'm not sure how to get NOUVEAU working. I think libdrm > --enable-nouveau-experimental-api needs to be set, but I'm not sure what > else. I think Mesa needs --with-gallium-drivers=nouveau, you need to install the driver xf86-video-nouveau which is only available through git, you need to recompile your kernel to enable the kernel part of the nouveau stack and you need to put driver=nouveau in your xorg.conf. It's a bit of hastle to get it all set up but once you do it makes recompiling the kernel a lot easier because the code is in the kernel. The Nvidia closed source blob makes recompiling the kernel such a pain! I think the nouveau people have done an excellent job. It's all open source. I stopped buying Nvidia hardware because they won't pay for a developer to work on nouveau. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg
On Sun, 11 Dec 2011 00:13:08 -0600 Bruce Dubbs wrote: > BTW, I built in /opt/xorg. I like to do that so I can build another > version of xorg without changing the current programs that would be the > case if installing in /usr. I think you're very brave to do that. I think reinstalling a newer version of xorg would break pango and cairo and everything that depends on them (ie, pretty much anything that uses xorg). Command line programs would work but trying to get any GUI things working again would be a nightmare. You'd have a mass of things in /usr that were compiled against the old xorg. Reinstalling the apps may work, but they may try to link against the old broken stuff. If you want to reinstall xorg it's easier to just redo the whole system than to try and deal with the brokeness. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg
On Sun, 11 Dec 2011 12:09:24 -0600 Bruce Dubbs wrote: > What you do is have a symlink of /opt/xorg -> xorg-7.6-2, build with a > different prefix and when ready to test, change the symlink. Works > great for kde, qt, gnome, fop, and and ant also. But what happens to anything in /usr which is linked to libX11 or whatever? How do they behave when you bump to a new version? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Xorg
On Sun, 11 Dec 2011 12:04:10 -0600 Bruce Dubbs wrote: > I thought it's primary use was a benchmark. I can't ever remember > running it at less than about 200 fps. http://wiki.cchtml.com/index.php/Glxgears_is_not_a_Benchmark > > >> After the builds, we have 'Xorg-7.6-2 Configuration', but we don't ever > >> tell the user how to test xorg. I guessed at 'startx' and that brought > >> up twm. We need to say that somewhere. > > > > xinit works too. > > startx is a script and the default ends up as > > client=/etc/X11/app-defaults/xinitrc > server=/opt/xorg/bin/X > clientargs="" > xserverauthfile=$HOME/.serverauth.$$ > serverargs="-auth ${xserverauthfile}" > display=":0 > > xinit "$client" $clientargs -- "$server" $display $serverargs So starx is a simple script that ends up calling xinit. > > I think it's because you don't have kernel mode setting enabled in you > > kernel config. > > I'm not sure which setting you mean. Kernel mode setting is enabled by default when you select the nouveau driver in you kernel config. With some drivers (eg the radeon ati one) you get a choice whether top have KMS enabled by default, but even then you can enable or disable it with a kernel command line option. KMS is incompatible with the Nvidia binary blob. > LOL. And you think the binary nvidia install is hard! Yes, I've always found the Nvidia driver to be a PITA because it makes recompiling the kernel much more compilcated. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] [blfs-book] r9100 - in trunk/BOOK: . general/genlib general/genutils introduction/welcome x/lib
On Sun, 11 Dec 2011 22:30:24 -0700 (MST) bdu...@linuxfromscratch.org wrote: > Author: bdubbs > Date: 2011-12-11 22:30:20 -0700 (Sun, 11 Dec 2011) > New Revision: 9100 > > Modified: >trunk/BOOK/general.ent >trunk/BOOK/general/genlib/libglade.xml >trunk/BOOK/general/genutils/rep-gtk.xml >trunk/BOOK/introduction/welcome/changelog.xml >trunk/BOOK/x/lib/cairo.xml >trunk/BOOK/x/lib/gtk+2.xml >trunk/BOOK/x/lib/hicolor-icon-theme.xml > Log: > Updated to rep-gtk-0.90.2. > > Modified: trunk/BOOK/x/lib/gtk+2.xml > === > --- trunk/BOOK/x/lib/gtk+2.xml2011-12-12 01:28:51 UTC (rev 9099) > +++ trunk/BOOK/x/lib/gtk+2.xml2011-12-12 05:30:20 UTC (rev 9100) > @@ -93,13 +93,16 @@ > Install gtk+-2 by running > the following commands: > > -sed -i '/db2html/i@HAVE_DOCBOOK_TRUE@\t mkdir gtk-tut; > \\ > -/gtk-tut html/i@HAVE_DOCBOOK_TRUE@\t mv *.html gtk-tut; \\' \ > -docs/tutorial/Makefile.in && > -sed -i '/db2html/i@HAVE_DOCBOOK_TRUE@\t mkdir gtk-faq; \\ > -/gtk-faq html/i@HAVE_DOCBOOK_TRUE@\t mv *.html gtk-faq; \\' \ > -docs/faq/Makefile.in && > -sed -i 's#^\( gtk_icon_view_\)l#\1queue_l#' gtk/gtkiconview.c && > +sed -i -e '/db2html/i@HAVE_DOCBOOK_TRUE@\t mkdir > gtk-tut; ' \ > + -e '/gtk-tut html/i@HAVE_DOCBOOK_TRUE@\t mv *.html gtk-tut; ' \ > +docs/tutorial/Makefile.in > && > + > +sed -i -e '/db2html/i@HAVE_DOCBOOK_TRUE@\t mkdir gtk-faq; ' \ > + -e '/gtk-faq html/i@HAVE_DOCBOOK_TRUE@\t mv *.html gtk-faq;' \ > +docs/faq/Makefile.in > && > + > +sed -i -e 's#^\( gtk_icon_view_\)l#\1queue_l#' gtk/gtkiconview.c > && > + > ./configure --prefix=/usr --sysconfdir=/etc && > make > > @@ -111,11 +114,12 @@ > > Now, as the root > user: > > -make install && > -sed -i 's/#man_MANS/man_MANS/' docs/reference/gtk/Makefile && > -make -C docs/reference/gtk install-man && > -install -v -m755 -d /usr/share/doc/gtk+->k2-version; && > -cp -v -R docs/faq/html /usr/share/doc/gtk+->k2-version;/faq && > +make install > && > +sed -i 's/#man_MANS/man_MANS/' docs/reference/gtk/Makefile && > +make -C docs/reference/gtk install-man && > + > +install -v -m755 -d /usr/share/doc/gtk+->k2-version; > && > +cp -v -R docs/faq/html /usr/share/doc/gtk+->k2-version;/faq > && > cp -v -R docs/tutorial/html /usr/share/doc/gtk+->k2-version;/tutorial > && > install -v -m644 docs/*.txt > /usr/share/doc/gtk+->k2-version; There doesn't seem to be any technical reason to make these changes. Don't you think it would have been polite to discuss this before you committed the changes? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Install llvm by running the following commands resulted in an error.
On Mon, 12 Dec 2011 13:08:58 -0600 Bruce Dubbs wrote: \ > sed -ir -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1' \ > -e 's#BUILD_ARCHIVE :*= 1#NO_' \ > -e '/^USEDLIBS/s# ([A-Za-z]*).a# -l\1#g' \ This is wrong, the brackets should be escaped from sed like so: > -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ You can't do a backreference if you don't escape the brackets. The brackets were escaped when I sent it to the list: http://linuxfromscratch.org/pipermail/blfs-dev/2011-December/021550.html Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Install llvm by running the following commands resulted in an error.
On Mon, 12 Dec 2011 15:34:42 -0600 Bruce Dubbs wrote: > Then why is is working for me? Note 'sed -ir' Sorry, I hadn't noticed the -r > $cat sedtest > USEDLIBS module.a > > $ cat sedtest | sed -r -e '/^USEDLIBS/s# ([A-Za-z]*).a# -l\1#g' > USEDLIBS -lmodule Works the same here, but I don't know why. According to: http://en.wikipedia.org/wiki/Regular_expression#POSIX_Extended_Regular_Expressions "support is removed for \n backreferences", but if that's true then why does the backreference work? Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] Install llvm by running the following commands resulted in an error.
On Tue, 13 Dec 2011 00:16:39 + Andrew Benton wrote: > On Mon, 12 Dec 2011 15:34:42 -0600 > Bruce Dubbs wrote: > > > Then why is is working for me? Note 'sed -ir' > > Sorry, I hadn't noticed the -r > > > $cat sedtest > > USEDLIBS module.a > > > > $ cat sedtest | sed -r -e '/^USEDLIBS/s# ([A-Za-z]*).a# -l\1#g' > > USEDLIBS -lmodule > > Works the same here, but I don't know why. According to: > http://en.wikipedia.org/wiki/Regular_expression#POSIX_Extended_Regular_Expressions > > "support is removed for \n backreferences", but if that's true then why > does the backreference work? > > Andy However, when I try the commands as they are in the book I get this: configure: creating ./config.status config.status: creating Makefile.common config.status: executing setup commands config.status: executing Makefile commands config.status: executing lib/Makefile commands config.status: executing lib/sample/Makefile commands config.status: executing tools/Makefile commands config.status: executing tools/sample/Makefile commands sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference \1 on `s' command's RHS sed: -e expression #3, char 35: invalid reference
Re: [blfs-dev] Install llvm by running the following commands resulted in an error.
On Mon, 12 Dec 2011 19:17:29 -0600 Bruce Dubbs wrote: > I think I made an error. What my script has is: > > find -name Makefile -exec \ > sed -ir -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1'\ > -e 's#BUILD_ARCHIVE :*= 1#NO_' \ > -e '/^USEDLIBS/s# \([A-Za-z]*\).a# -l\1#g' \ > -e 's#^USEDLIBS#LDFLAGS#' {} \; > > The parens should be escaped. That didn't make it into the book. I've figured it out in the bath. the -i and the -r need to be separated. This works without the brackets escaped: find -name Makefile -exec \ sed -i -r -e '/^LIBRARYNAME/a SHARED_LIBRARY = 1'\ -e 's#BUILD_ARCHIVE :*= 1#NO_' \ -e '/^USEDLIBS/s# ([A-Za-z]*).a# -l\1#g' \ -e 's#^USEDLIBS#LDFLAGS#' {} \; Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
[blfs-dev] Geany and Sylpheed
Hello, I've written a couple of new pages for 2 packages I think should be in the book. Both use gtk+-2 but have few other dependencies. I use them every day and think they're excellent. Geany is a text editor: http://www.linuxfromscratch.org/~andy/BLFS/geany.html Sylpheed is an easy-to-use email client: http://www.linuxfromscratch.org/~andy/BLFS/sylpheed.html An index for my growing collection of BLFS pages: http://www.linuxfromscratch.org/~andy/BLFS Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] seamonkey
On Tue, 13 Dec 2011 17:28:48 -0600 Bruce Dubbs wrote: > First, the mozconfig is missing EOF. I'll fix that. Doh! Sorry for that. > Second, I think we used to install enigmail with seamonkey. Has anyone > tried that? No. There were instructions for installing enigmail on the page but it was all commented out so it didn't render. I've never used enigmail. If I did want it I would download and enigmail.xpi extension. > Third, I think it would be better to be more verbose with the > instructions. That is: > > mkdir -pv > tar -xvf > ln -sfv > cp -v > > I'll go ahead and fix those too. Fair enough. Style issues again. > I was able to get webm working, but I'm working over an ssh connection. > Video works fine, but there is no sound. I've not tried sound over > ssh before. Does anyone know if that is supported? If so, what needs > to be done to get it working? No idea sorry. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] D-BUS instructions
On Thu, 15 Dec 2011 16:25:17 -0600 Bruce Dubbs wrote: > Also, the construct is not explained. I think it should be removed. I agree it should be explained, then people can make an informed choice. I don't mind that it's different from the other pages where useradd is used. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] D-BUS instructions
On Thu, 15 Dec 2011 19:10:04 -0600 Bruce Dubbs wrote: > More on D_BUS. The install creates /var/run/dbus, but that will go away > with a reboot. We need to figure out a way to create it upon boot. > > The first LFS boot script, mountvirtfs, does: > > mkdir -p /run > mount -n /run || failed=1 > mkdir -p /run/{var,lock,shm} > > The /etc/sysconfig/rc.site is sourced so we need to change rc.site with > something like: > >RUNSUBS="${RUNSUBS} dbus" > > and add to mountvirtfs: > >for D in ${RUNSUBS}; do mkdir -p $D; done > > For other subdirectories of /var/run we would only need to add > >RUNSUBS="${RUNSUBS} subdir" > > to the end of rc.site. > > A second option would be to just hard code any needed directories like > we do with {var,lock,shm}. A directory entry is negligible if it is not > used. > > A third option would be to add it to the dbus bootscript, but that is > not required if the user is just adding dbus-launch to the appropriate > line in the ~/.xinitrc file. > > Comments? Another option (which works for me) is to configure dbus: --with-system-socket=/run/dbus_socket --with-system-pid-file=/run/dbus.pid Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] D-BUS instructions
On Fri, 16 Dec 2011 00:46:48 -0600 Bruce Dubbs wrote: > This is getting complicated. I can build and install dbus-1.4.16.tar.gz > without noticing any problems. > > In order to run the tests, I need: > > D-Bus GLib Bindings (update to 0.98) > D-Bus Python Bindings (update to 0.84.0) >which now needs > PyGObject (update to 2.28.6) >which needs > gobject-introspection (update to 1.31.1) > pycairo (update to 1.10.0_ >which fails with "The python version is too old, expecting > (3, 1, 0)" (I do have python3 installed, but 'python' is 2.7.2). You need py2cairo: http://cairographics.org/releases/py2cairo-1.10.0.tar.bz2 Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] D-BUS instructions
On Fri, 16 Dec 2011 20:27:29 + Ken Moffat wrote: > Ah, I see, you build dbus but not the bootscripts. AFAIK, the > bootscript has been required for a long time. I see no obvious > reason to add dbus-launch to .xinitrc. I also see no reason to add dbus-launch to .xinitrc. On my system I start dbus with a bootscript, I can see it in ps aux as: dbus-daemon --system but I also see: dbus-launch --autolaunch 465105239381aefc38f1f5e2020d --binary-syntax --close-stderr and: /usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session so it seems that something that uses it (notification-daemon?) uses dbus-launch to spawn another dbus-daemon for the session even though there is one running for the system. I don't think the system one is needed. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] D-BUS instructions
On Fri, 16 Dec 2011 18:03:42 -0600 Bruce Dubbs wrote: > I'm not an expert in how dbus is used, but if you do not use the > bootscript and do use the dbus-launch, isn't that sufficient? Yes, I think so. I'm going to delete the dbus bootscripts on my systems and just start it from xinitrc > What I'd like to do is explain it, but I need to understand it better > first. I think this is a mechanism for different processes to > communicate without using shared memory or semaphores or pipes, but it > may just be a wrapper around those constructs. > > I suspect (but am not sure) that dbus is not very useful on a non-gui > desktop or server. If that's the case, then launching via dbus-launch > to .xinitrc makes sense. As I understand it dbus is an inter program communication system for desktop systems. I don't think it has a role (and is probably a potential security risk) on a server. Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: [blfs-dev] legacy gnome2 packages:wq and gnome-etc-dir
On Tue, 20 Dec 2011 20:10:48 +1100 Wayne Blaszczyk wrote: > My preference is option 4) /etc (Something that Fedora and Ubuntu uses, > just to name a few) +1 for this. I install my gconf stuff --sysconfdir=/etc Andy -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page