Re: statically relinking
Alberto Hernando wrote: As a part of my transition to LFS, I want to use it too, but there are no suitable packages for me, and they refuse to provide more packages. It isn't possible to relink a dynamically linked executable as a static. And if it were a static you would not have a problem. What I would do is this: 1) In debian, run ldd (or readelf) on the executable in question. This will list the libraries it needs and their current versions. 2) Now look for the versions of the libraries that support those versions. A common problem is libstdc++, many many proprietary binaries need libstdc++.so.5 - this can be built using the BLFS instructions for gcc-3.6 - but you could probably also copy your debian version across (unless they depend on ABI that are changed in, say, glibc). You can also try adding a symlink from the name required to the version you've got, but this will fail if the bit of the ABI that your app uses has changed. 3) Check the binary on the LFS system with ldd again, to ensure that all it's libs are satisfied. None of this is a precise science, and you may well get problems like memory leaks and buffer overruns that make the application unstable, but it's worth a try. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: To Indiana LFS'ers
John Gay wrote: > On Sunday 02 April 2006 08:12, Eric Stout wrote: As fellow Hoosiers probably know, our fine government has again decided > >>> I think we should all go back to the good old days before Railroad >>> Timetables (the original excuse), when every village had it's own time >>> zone. The computers can handle it! >> I'm all for having a single time. One time for the entire planet. >> > There is, it's called Zulu time. The U.S. Military uses it for anything that > needs co-ordination across time-zones. Just for information: The UK forces also use Zulu time operationally - even when not supporting the USA in their foreign wars. Now we have a ball rolling. How about the LFS project(s) going over to Zulu time completely? Of course, this is the same as UTC (BST - 1). R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: To Indiana LFS'ers
Archaic wrote: > As fellow Hoosiers probably know, our fine government has again decided > to change timezone laws. With a glorious history of changes being made > not just at the state level, but on a county-by-county level, and with > the addition of several changes that were implemented, reverted, > re-implemented, etc., it has been interesting to keep up with. No one in > central Indiana has had to think about changing their clocks since 1970, > but those days are gone. Now we get to enjoy the fun that the outlying > counties have been having for the last several decades. Goodness! I didn't know there were Scottish school children in Indiana! (Well that's the excuse our so-called government uses to justify the insane British Summer Time rules - Scottish school children can't find their way home in the dark!) I think we should all go back to the good old days before Railroad Timetables (the original excuse), when every village had it's own time zone. The computers can handle it! R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: (no subject)
Dan Nicholson wrote: > On 3/30/06, Richard A Downing FBCS CITP <[EMAIL PROTECTED]> wrote: >> Gmail can be so configured. > > Really? I completely missed that. Please enlighten me how to get > bottom-posting.by default. > > -- > Dan My mistake, it can't. I am absolutely damn sure it could when I last used it though! I gave it up as it couldn't handle threading of mailing lists with own-contributions. I feel a bit of a prat having niggled Henry on two points - both wrong. Apologies. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: (no subject)
Richard A Downing FBCS CITP wrote: > > Henry, > You are a dirty Top Poster :-) Please use our netiquet when > contributing on the LFS list. See FAQ. Gmail can be so configured. Sorry about that, I see you are already a reformed character. Blessed are the sinners come to repentance. :-) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: (no subject)
Henry christenson wrote: > A simple grammer change would reslove the issue. here is how it > appears in the book. > Required Patch (if compiled using GCC-3.4.x) > > This tells readers If you are compileing with any varient of gcc 3.4 > you need this patch. > > all that would need to be done to reslove this from being posted again is. > Required Patch (if compiled using anything over GCC-3.4.x) Henry, You are a dirty Top Poster :-) Please use our netiquet when contributing on the LFS list. See FAQ. Gmail can be so configured. How about: Required Patch (Yes, really, whatever damn version of gcc you have!) Although it is strictly not true if you have an antediluvian gcc (pre 3.4), but the number of those will be almost nil in LFS circles, particularly amongst new users, who are the ones that can't work it out for themselves. Bending over backwards for the backward isn't a good use of editor-resources. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Kdebase Please Fix
tom wrote: > I think the book should mention you need to add this --with-kio-smtp to > kdebase in order for kmail to authenticate otherwise smtp wont work. I suggest you offer to pay the editors to put your bug on the front of the queue. About $1000 should do it, and I'll happily collect it for them :-) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Feauturerequest: IPSec-tools, Squid, Privoxy
Randy McMurchy wrote: > Richard A Downing FBCS CITP wrote these words on 03/25/06 08:11 CST: > >> However, the hint format is so restrictive (no formatting at all, >> newlines actually inserted to force line breaks, yadayada...) that you >> might like to consider a wiki page instead. This is what I do. > > I am wondering if there shouldn't be some sort of disclaimer on > the BLFS Wiki "Notes" page along the lines of: > > The information contained herein may not have been prepared, or > reviewed by the BLFS staff. The information may be wrong, incorrect, > untested or may not work as it is advertised. Use of the > information in these notes is at your own risk." > > I believe it is important that users know that the material is not > officially sponsored by the BLFS staff. > Good idea. Although if it WAS reviewed and approved, presumably it would be in the book. On the other hand if it is expansion of what is in the book (as opposed to my disconnected jottings about packages that no-one wants :), it might well be approved already, but just not considered suitable for the book for some reason. I still don't like the Hints format though, and won't use it again. I have considered putting up my own wiki for the packages that I'm interested in. Then you can link to it or not as you will. But I'm not interested in wiki's, per se, so I probably won't. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Feauturerequest: IPSec-tools, Squid, Privoxy
Randy McMurchy wrote: > Matthias Berndt wrote these words on 03/25/06 05:14 CST: > >> is there anyone interested in adding IPsec-tools, Squid and/or Privoxy >> to the book? > > I believe standard protocol is to first submit a hint (one for each > package), and then request that the hint(s) be incorporated into the > book if there is interest. > However, the hint format is so restrictive (no formatting at all, newlines actually inserted to force line breaks, yadayada...) that you might like to consider a wiki page instead. This is what I do. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: startx automatically with LFS
Satish Chebrolu wrote: > Try setting the default init level to 5 in /etc/inittab. Please don't top post. It's also helpful to know what your advice means before giving it. You only explained half the story. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: startx automatically with LFS
Rik Potts wrote: > sorry, I did intend on saying in my forst post that I have tried this. > > On the command line it boots to it just says it has entered init level 5 > instead of 3. No other change. > Of course not. In order to make init level 5 do something you will need to add some level 5 instructions to inittab. Try man inittab and consider what you want to run instead of getty to get KDM or GDM to run. Then start looking at the symlinks in /etc/rc5.d to decide what else you need starting and stopping. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Digital Camera with a pure UDEV (no Hotplug) setup.
Alexander E. Patrakov wrote: > Richard A Downing wrote: > >> Does anyone know how to set up such a system for gphoto2 (not >> usb-storage)? Any hints would be very welcome. Udevmonitor output >> below. > > I have no digital camera, and thus can't provide the information you > need immediately. Please answer my question: > > Does it "just work" for root when the "usb-storage" module is not loaded > and /proc/bus/usb is mounted? > Thanks, Alexander. I had forgotten to mount /proc/bus/usb. Thanks Sebastian too. Of course the group set by devgid=14,devmode=0660 will do it! I need to decide which set of GID's to use though, BLFS's or CLFS's. I'll write a GPhoto2 page for the wiki later. Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Digital Camera with a pure UDEV (no Hotplug) setup.
Does anyone know how to set up such a system for gphoto2 (not usb-storage)? Any hints would be very welcome. Udevmonitor output below. Thanks, R. udevmonitor udevmonitor prints the received event from the kernel [UEVENT] and the event which udev sends out after rule processing [UDEV] UEVENT[1141302412.409920] add@/devices/pci:00/:00:02.1/usb3/3-2 UDEV [1141302412.410120] add@/devices/pci:00/:00:02.1/usb3/3-2 UEVENT[1141302412.416310] add@/devices/pci:00/:00:02.1/usb3/3-2/3-2:1.0 UEVENT[1141302412.416330] add@/class/usb_device/usbdev3.5 UDEV [1141302412.417974] add@/devices/pci:00/:00:02.1/usb3/3-2/3-2:1.0 UDEV [1141302412.421100] add@/class/usb_device/usbdev3.5 -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Thunderbird-1.5 Spelling (SOLVED)
Randy McMurchy wrote: > Dan Nicholson wrote these words on 02/09/06 15:20 CST: > >> Randy, what do you think about adding info about spellcheck >> dictionaries to Thunderbird/Mozilla? FYI, I don't know if the >> spellcheck is implemented the same way in Mozilla-1.7.x. > > I'm not sure that doing anything other than maybe a small mention > of dictionaries and providing the following link (thanks, Dan for > making finding this so easy) is necessary: > > http://dictionaries.mozdev.org/installation.html > > What did you have in mind, Dan? Hopefully not using the instructions > you provided to Richard. I'm not sure we want to go there. Of course, > mine is just one opinion, but using dictionaries from OpenOffice > and instructions contrary to the method prescribed by the Moz folks > just seems to be opening up a Pandora's box. > > Sounds like it would be good Wiki material however. > I'm content that this thread is in the archives. But a wiki page would perhaps be worthwhile, Dan? On the other hand, the book as at present only really serves the English/US community. So some mention of alternative dictionaries might be in order. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Thunderbird-1.5 Spelling (SOLVED)
Dan Nicholson wrote: > On 2/9/06, Richard A Downing <[EMAIL PROTECTED]> wrote: >> OK, for the record: >> >> chmod 644 /usr/lib/thunderbird-1.5/components/myspell/* >> >> somehow this fixes it. The perms were apparently screwed by installing >> the extension: spell-en-GB.xpi. > > This makes sense to me. If you don't have read permissions on the > dictionary (en-US.dic or whatever), then how could you properly use > the spell checker? > > Through the mozdev site (http://dictionaries.mozdev.org/), I was able > to find the OpenOffice page that contains all the dictionaries in .zip > format. I downloaded > > http://ftp.services.openoffice.org/pub/OpenOffice.org/contrib/dictionaries/en_GB.zip > > and inflated it. It came out with 644 permissions. > > I believe all the dictionaries on > > http://lingucomponent.openoffice.org/spell_dic.html > > are the same ones used by the myspell component in Mozilla. > > Richard, maybe you could download that zip and check if the en_GB.aff > and en_GB.dic files are the same as those from the .xpi. If so, maybe > it would be as easy as inflating the zip files into the appropriate > directory during the build to have all the dictionaries you want in > the global directory. Dan, Thanks for the info. The files from http://ftp.services.openoffice.org/pub/OpenOffice.org/contrib/dictionaries are not identical to the .xpi versions. They are different sizes to start with, and a diff shows numerous differences in, what I assume are, flags concerning plurality and other endings, hyphenation and the like. The word lists are the same however, so they are completely usable - in fact I'm using them now. This could be worth a note in the book? Richard. P.S. Wrote this before SWMBO called me to dinner (then I went to the pub...), just seen your other email. Your way should work for a new install. Thanks. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Thunderbird-1.5 Spelling (SOLVED)
Randy McMurchy wrote: > Richard A Downing wrote these words on 02/09/06 11:19 CST: > >> chmod 666 /usr/lib/thunderbird-1.5/components/myspell/* >> >> I'd appreciate knowing what perms you have on those files, Randy? > > [EMAIL PROTECTED]: ~/build > ls -l /usr/lib/thunderbird-1.5/components/myspell > total 688 > -rw-r--r-- 1 root root 2731 Jul 16 2005 en-US.aff > -rw-r--r-- 1 root root 695748 Jul 16 2005 en-US.dic > > [EMAIL PROTECTED]: ~/build > ls -ld > /usr/lib/thunderbird-1.5/components/myspell > drwxr-xr-x 2 root root 4096 Jan 20 21:09 > /usr/lib/thunderbird-1.5/components/myspell > OK, for the record: chmod 644 /usr/lib/thunderbird-1.5/components/myspell/* somehow this fixes it. The perms were apparently screwed by installing the extension: spell-en-GB.xpi. I don't know of any way of installing dictionaries during the build, but it would be nice not to be dependant on binary downloads. But, hey, I can type Real English (tm) without red lines now! R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Thunderbird-1.5 Spelling (SOLVED)
Randy McMurchy wrote: > Richard A Downing wrote these words on 02/09/06 01:34 CST: >> The spellchecker underlines EVERY word as I type. >> >> I installed the English(British) dictiionary, no change. If I 'ignore' >> a word that doesn't take the underline off. If I 'ignore-all' a word it >> doesn't change the other instances. Seems well broken. >> >> Any ideas? > > None. It works great here. If you have your mail-client ID display > on, you'll see I'm typing this from a Thunderbird 1.5 (built 20060120) > client and I don't have anything underlined except for a couple of > things in my sig. > > Oops, sig becomes underlined. And Richard, I am one that *really* > appreciates the on-the-fly spell checker, so it was one feature I > checked out first thing. > > Wish I could help. > You did. Just knowing that it works, and knowing I was using your method of building, then it had to be something local. I checked what root could do, and... The problem is the permissions for the files in /usr/lib/thunderbird-1.5/components/myspell The instance of Tbird seems to need to have rw access to them. ro access isn't enough. How these got installed with the wrong perms I don't know, but they were all there, the US dictionary had 600 perms - so root could write American, but ordinary users couldn't. The GB dicts were 400, so no-one could write proper English! chmod 666 /usr/lib/thunderbird-1.5/components/myspell/* I'd appreciate knowing what perms you have on those files, Randy? R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Thunderbird-1.5 Spelling
I built Tbird-1.5 according to the book, on a recent LFS-by-jhalfs-svn system. The spellchecker underlines EVERY word as I type. I installed the English(British) dictiionary, no change. If I 'ignore' a word that doesn't take the underline off. If I 'ignore-all' a word it doesn't change the other instances. Seems well broken. Any ideas? R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Compiling OpenOffice 2.0.1
On Fri, 20 Jan 2006 21:36:44 -0600 DJ Lucas <[EMAIL PROTECTED]> wrote: > Richard A Downing wrote: > > > > An editor, probably DJ Lucas, will assign this bug to himself > > Thanks for putting me on the spot Richard! :-) I had the opposite intention. I would say, 'take your time and make it good'. I'm sure Tor Olav can use the binary for a few days if he really needs an update. ATB R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Compiling OpenOffice 2.0.1
On Thu, 19 Jan 2006 11:47:39 -0600 Randy McMurchy <[EMAIL PROTECTED]> wrote: > [EMAIL PROTECTED] wrote these words on 01/19/06 11:40 CST: > > On 1/19/06, Tor Olav Stava <[EMAIL PROTECTED]> wrote: > >>Are there any plans to upgrade the SVN book to include this > >>version? It supposedly contains a few bugfixes, so it sounds like a > >>good idea to me.. > > > > I would also very much appreciate that. > > http://blfs-bugs.linuxfromscratch.org/show_bug.cgi?id=1758 > It's clear from the bug that noone has yet accepted the task of upgrading BLFS to OO-2.0.1 (the bug is unassigned). That's all that anyone can say, and this is what Randy's terse reply means. An editor, probably DJ Lucas, will assign this bug to himself when he starts the process of upgrading the book. You can monitor this on the blfs.book mailing list (or news:gmane.linux.lfs.beyond.book ). Otherwise: your guess is as good as anyone elses. You could always do the work yourself and submit a patch :-) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: hotplug
On Wed, 04 Jan 2006 12:34:37 -0200 Rebelde <[EMAIL PROTECTED]> wrote: > Hi, I am trying to understand the Hotplug System, but I can not see > where are the files that run when I plug a usb device for example. > Can anyone help me? I recollect a fairly good explanation of advanced udev in an email by Alexander Petrakov on one of the lists a few months ago. Check the list archives. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: can i use root user to build blfs
On Thu, 22 Dec 2005 01:58:04 -0800 Craig Jackson <[EMAIL PROTECTED]> wrote: > Wow, I must've missed that one too. I guess I will be unprivileged > as much as I can from now on. I never read the BLFS book straight > through like I did with LFS. Does this have to do with security or > does it actually change the functionality? > > Epitome > [EMAIL PROTECTED] > > > On 12/21/05, Archaic <[EMAIL PROTECTED]> wrote: > > > > On Wed, Dec 21, 2005 at 11:16:34PM -0800, Amu wrote: > > > iam using user=root for building whole blfs > > > can i use root? > > > > What does the book say? It affects security, but not functionality if you get you permissions right - which is worth getting right from the beginning. But, Safety mainly - it's too easy to screw up your precious new LFS system if you run as root. Look at the difference between 'rm -rf ~*' and 'rm -rf *' - not a mistake to make when PWD=/ - others are more subtle. I don't think there are any instructions in BLFS that would cause problems run as root, but then they have not been checked for that ;-) Finally Craig, Please read the netiquet section of the LFS FAQ. You are a dirty top-poster and a damned HTML mailer. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: From Where to start Building Binaries
On Tue, 20 Dec 2005 11:25:26 +0100 "Feldmeier Bernd" <[EMAIL PROTECTED]> wrote: > Hi all, > > Well, if have a little bit different question. > > As I dont want to install any development tools > in chapter 6 final LFS to keep the rootfs small for embedded purpose, > is it okay to always use the /tools toolchain in > temp chroot system for e.g. BLFS creation?? > > l also need to change the PATH then in > chapter 5 end to avoid any usage of final rootfs > libs usage?? > > regards Bernd > Bernd, No, you shouldn't use the /tools toolchain. It may have dependencies on the build-host's libraries. Your best bet is to build the chapter 6 toolchain passing something like --prefix=/opt so that they are built out of the normal /usr hierarchy (then use CC="/opt/bin/gcc" - see the recommendations for gcc-3.3.6 in BLFS for ideas), then delete /opt when you have all the applications built. I'm assuming you're building the system on one machine to embed in another - if you take my approach you just don't copy /opt. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: From Where to start BLFS
On Tue, 20 Dec 2005 06:24:06 - "IraqiGeek" <[EMAIL PROTECTED]> wrote: > > We have to start BLFS now ... We are in confusion from whr we have > > to start .. > > Where to start with BLFS is totally up to. It depends on what extra > features you want to have on your LFS box. > > > Whether we have to do BLFS on login to LFS or over the host system > > (Mounting LFS and chrooting it and working over tht )... > > Since LFS has its own development toolchain, I dont see any point in > working on a chrooted environment over Debian. You can boot LFS, and > build any extra packages you need right from the LFS system. However, unless you make notes or print it out or have another system, you need something to read BLFS with, at least. I usually build wget and lynx on the chrooted LFS, so that I can (a) use wget to get sources, and (b) use lynx to read BLFS (I know that lynx can download thanks). Then, for a workstation, I build (on the booted system) in this order to get xorg up and running: expat, freetype, fontconfig, libpng, xorg. After that I go for firefox, to read BLFS as rendered as intended, with: pkg-config,cairo,glib2,pango,atk,libjpeg,libtiff,libIDL,zip,firefox. I have that much scripted now. If I was building a server (without X11) I'd do it completely differently though: make a list of the services and their dependecies and build those. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: fsam7400 and ipw2200 on LFS 6.1.1
On Sun, 11 Dec 2005 16:47:39 +0100 Michael Weller <[EMAIL PROTECTED]> wrote: > Resolved (sometimes you just need to send an email..). I'd like a five-dollar bill for every time that's happened to me! I could buy a new monitor, at least. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: OpenOffice: it's horrible!
On Sun, 11 Dec 2005 13:06:24 +0100 Jeremy Monnet <[EMAIL PROTECTED]> wrote: > On 12/11/05, Randy McMurchy <[EMAIL PROTECTED]> wrote: > > Jean-Philippe Mengual wrote these words on 12/10/05 17:56 CST: > > > > > checking for required Perl modules... Can't locate > > > Compress/Zlib.pm in @INC > > > > This means you'll need to download and install the Compress::Zlib > > module > > > > http://www.cpan.org/authors/id/P/PM/PMQS/Compress-Zlib-1.38.tar.gz > > > > Download from the URL shown above and install it using conventional > > Perl Module instructions: > > > > perl Makefile.PL > > make > > make test > > make install > > > > For future reference, keep the following in mind. > > > > The Perl Modules are listed with the dependencies below the listed > > module. So, for the Archive::Zip module, you need to install the > > module shown below it, the Compress::Zlib module. > > > > Maybe an easy way (when you have internet connection available) is to > use the alternate install solution for perl modules > http://www.linuxfromscratch.org/blfs/view/svn/general/perl-modules.html > . When you use it it manages alone the dependencies, so it asks you to > install Compress::Zlib when you want Archive::Zip . Possibly. However, be aware that this method (that I wrote for BLFS some years ago) only works well if the packages all apply cleanly. Sometimes, when the package runs it's test, an on-line requirement isn't met, and the message: 'Package won't install without force' appears. It is always safer to use the Unpack/Perl Makefile.PL/make/make test/make install method, as you can then keep good logs and check them. I don't use the on-line method any more. Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: ohphone
On Thu, 8 Dec 2005 21:55:27 +0100 Jean-Philippe Mengual <[EMAIL PROTECTED]> wrote: > Hi, > Richard, thanks, thanks, thanks very much!! Jean-Philippe, Glad it worked out for you. When I have time, I'll investigate using it. I think YOU should write a Hint about building and using it (since you are a user) - but as English isn't your first language, I'm happy to edit it for you if you would like that. Feel free to email me direct. Once there is a Hint, Bruce will look more favourably on putting it in the BLFS book. VoIP is a coming technology. Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: openH323, does someone know?
On Wed, 7 Dec 2005 18:51:31 +0100 Jean-Philippe Mengual <[EMAIL PROTECTED]> wrote: > Hi, > On my computer (LFS and BLFS system), I'd like to install ohphone, a > program to speak via IP to an other person, which uses NetMeeting > (Microsoft) or a software with this protocol. I hope someone knows, > because I can't nor install it, neither find information about what I > did bad and I have to do. I try also on the mailing list of openH323, > but getting information is hard, they nearly never answer. Since we > can use ohphone on Linux and Windows, and it's a general code, maybe > few users use Linux. > In fact, ohphone has dependencies. > Dependencies: pwlib, openH323 I suspect the versions of pwlib and Openh323 have to match in some way. At least this is my memory from the one time I tried this - the versions are critical - I had to use CVS versions too. The BLFS book suggests that pwlib-1.8.7 compiled with OpenLDAP support and OpenH323-1.15.6 work together, at least the editor that last tested Gnomemeeting found this to work. Maybe try these versions? Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: about the nfs-server script
David Fix wrote: > You know, I know this is late for a comment... But I just thought of > something... Could it be as simple as doing something like this: > > kill `pgrep nfsd` That's just an inelegant equivalent of: rpc.nfsd -- 0 I think Tor Olav is trying to confirm that the kernel threads have stopped before he shuts down. Hence his script. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: about the nfs-server script
Tor Olav Stava wrote:
> Tor Olav Stava wrote:
>
>> I'm trying to wrap it in some if() statements to see if I can get a
>> nice solution to this, but if someone have already fixed it I would
>> appreciate if you share it, and maybe update the script in SVN.
>
>
> After a while of trial and error, I've come up with a solution that
> seems to work almost perfect. In the nfs-server script I've changed the
> relevant section to:
>
>if pidof nfsd 2>&1 > /dev/null ; then {
>rpc.nfsd -- 0
>sleep 1
>if pidof nfsd 2>&1 > /dev/null ;
>then echo_failure
>else echo_ok
>fi
>}
>else echo_warning
>fi
>
> Notes: It will only issue a [WARN] if the process is not running, not
> the additional "Not running" line.
>
> If anyone has any comments to this, I'd be happy. I haven't done any
> shell programming, so there's bound to be prettier ways to do this.
>
Tor Olav
I think that your solution, while not seemingly elegant, is a good
workaround. rpc.nfsd is just a userland program to tell the kernel
module how many threads to run. As such, it reports success if it told
the kernel OK - what the kernel does with this info is a different
matter. Your solution correctly tests 'what the kernel did'. So I
think it's right.
The only real question left is 'Why do you want to know?'. Do you abort
the shurdown if the nfsd threads don't die, or carry on with a worry
that some data wasn't safely written? Not sure what the right answer is
here.
R.
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: xfree86 help
Mike Hernandez wrote: > IMHO backticks are evil and should be avoided at all costs, but > everyone else seems to like them, so ;) +1, at least as far as books (for reading with the Mark-1 eyeballs) are concerned. For some reason almost all fonts make them impossible to detect unless you know they are there already (which as script authors we all do :-). Same does with lower case l and digit 1. Perhaps we should remove all those too. ;-) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gcc 2.3.5
Randy McMurchy wrote: > Richard A Downing wrote these words on 10/11/05 15:52 CST: > > >>Take tongue out of cheek. We are FAR better at keeping our >>documentation up to date than the Kernel Developers who would rather >>introduce a 'really neat bit of new code' than document the bloody >>important stuff they wrote last year. > > > Simply for the sake of discussion, let me pose a question. > > Certainly there is limited amount of time the kernel dev's have to > spend on the project. > > Would you rather that time be spent in documenting their work, or > developing new code and fixing bugs in old code? > > In a perfect world, or when you work for an employer, you must do > both. But in an open-source environment, this might not be feasible. > > Just a thought. > I agree, you have a good point. However, it's real difficult to use undocumented stuff - e.g. udev last year - without at least some docos. I guess some people can read the code (or monitor all the various mailing lists)- but much of that is uncommented and hence quite hard to understand. When I was in control of software development at ICL, we said: 'Code without User Documentation doesn't exist' - docs first, then code was the rule. I recognise we can't insist on that in the volunteer-written software world - hell, WE can't insist on anything! (quite rightly) I think your penultimate para should read: 'when you work for an employer, he must pay for both'. Otherwise I hope he goes rapidly out of business. Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gcc 2.3.5
Matthew Burgess wrote: > Andrew Benton wrote: > >> True, but strangely the linux-2.6.14-rc4/README says >> >>> COMPILING the kernel: >>> >>> - Make sure you have gcc 2.95.3 available. >> >> >> >> Serves me right for looking. I haven't installed gcc-2.95 for a long time > > > Andrew, you do realise that us LFSers know far better than the kernel > developers don't you :) (tongue firmly in cheek!). Take tongue out of cheek. We are FAR better at keeping our documentation up to date than the Kernel Developers who would rather introduce a 'really neat bit of new code' than document the bloody important stuff they wrote last year. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: printing in heterogenous network
Alexander E. Patrakov wrote: > Richard A Downing wrote: > >>> Alexander E. Patrakov wrote: >> Hey, now that (all your post) MUST become a hint, at the very least! >> Ideally a page in the BLFS book - it's vital information. >> >> I thought I understood it, but clearly I didn't, since I didn't know >> that it was possible to offer printer services to Windows without samba! > > > Sorry, I am not sure that it fits nicely on > http://www.linuxfromscratch.org/ because it is not a Linux hint. It > mainly discusses Windows configuration. If others think that it is not a > problem, I will turn this into a hint. My view is that it relates to using a Linux printer server on a network containing Windows Workstations. Lots of us have those. To use LFS to build a printer server/service is a legitimate, one might say ideal, use of LFS. Despite the hint needing to be mostly about Windows configuration, it's extremely relevant. I've also become very out of date with Windows since Windows95 (last one I used in anger, nay, Rage!), although I have installed Windows 2000 for my wife, I insist she does all the configuring, saying 'that's Windows, it's written for people who don't understand computers - so it's not my problem'. I suspect there are a few of us like that too. To give her a hint and say: 'here's how to use my printer' would be nice. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: printing in heterogenous network
Alexander E. Patrakov wrote: > Alexander E. Patrakov wrote: > >> Alternatively, you may want to use the stock Adobe PostScript driver >> (although I don't know any reason for such preference). In such case, ... Hey, now that (all your post) MUST become a hint, at the very least! Ideally a page in the BLFS book - it's vital information. I thought I understood it, but clearly I didn't, since I didn't know that it was possible to offer printer services to Windows without samba! Now what would be really neat is to have the printer (P) on a linux box (A), and the driver on Windows box (B). Then print to P from A using the driver on B. This would be a boon to those with closed source windows drivers. :-) Thanks Alex. Richard. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Workspace Managers, Window Managers and Desktop Environments
I use OpenBox3, possibly the most minimalist WM that uses XML for it's configuration files. Then I add PyPanel for the 'where the hell is that app I started a few hours ago?' feeling. I use mrxvt instead of xterm for the tabbed virtual screens. I like OpenBox3's menus, although you need a good XML editor to construct and edit them - they have a nice obstruse syntax. I use coreutils (ls,mv,cp,etc..) as my file manager. I try KDE and Gnome occasionally, but they are just too huge, and I don't like the look&feel of either. Why do they need so many background processes? X's cut&paste works as well as their drag&drop for me. UNIX's great step forward was the ability to string simple applications together to do complex useful things - Windows, Gnome and KDE seem to promote the monolithic applications. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: A question about house keeping
Archaic wrote: > On Thu, Sep 15, 2005 at 10:50:46PM +0800, Stephen Liu wrote: > >>On the line "xterm -g 80x40+0+0 &", what is "-g" for? TIA > > > This question troubles me. In your massive bombardment of the lfs and > blfs support lists for several weeks, has no one mentioned the use of > "man" for reading manual pages? Or has it never occurred to you that -h > or --help will give quick answers for most programs? Please do some of > the research on your own. > Well said, Archaic. I've been thinking about how to hint broadly to Stephen that he is expecting far too much 'hand holding'. Seven or so new support request threads in recent days. RTFM. (Yes, all of them) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: user package hint & gnome-panel
Randy McMurchy wrote: > David Fix wrote these words on 08/31/05 14:12 CST: > > >>Just kidding. :P I actually have no idea, I'm just hoping to at least put >>a smile onto your face. I've thought of doing some kind of package >>management myself, but seeing stuff like this always makes me cringe... >>Anyone else have any smart ideas or experience? > > > I use package management via little scriptlets I put into my build > scripts. Putting it simply, it takes a picture of the filesystem > before and after the installation of a package, then compares those > pictures for differences. The differences are what is installed by > the package. > > Works well, however, it won't include files that are installed by > a package that overwrite an existing file. But that is a pretty > rare occurrence. > There is a released package (it's a script really - but quite sophisticated) that does this: Guarded Installation Tool. I used it for some time - it's drawback is that as the filestore fills up it gets progressively slower. There used to be a LFS hint about it, but I don't think it's maintained. http://home.wtal.de/ib/freisoft/ (Although it's German on the website, the package has English docs.) R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Сочи - осенние скидки
Ken Moffat wrote: > On Mon, 29 Aug 2005, Richard Snow wrote: > >> Санаторий «Морска звезда» wrote: > > [...] > >> Can someone translate this? >> > It's spam. Set your filters to reject character sets you can't read, > then you won't see it. > > Ken > A hint on message filtering would be good, one day :-) Look for 'Content-Type:' headers containing: 'charset=windows-1251;' for cyrillic, I guess ? This get you to reject cyrillic, but it's a big list to reject devanagri, Chinese, Kanji, Hiragana, Katakana,... R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: gcc 2.95.x
Andrew Benton wrote: > Dan Osterrath wrote: > >> Hi, >> >> I need to install libstdc++.so.5 on my LFS system > > > Isn't libstdc++.so.5 installed by gcc-3.3? Yes, gcc-3.3.6 in the BLFS svn. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: GCC4 build with OpenSSH-4.1p1 missing -ldl
Randy McMurchy wrote: > [cc'd to BLFS-Support as it appears it should be directed there] > > William Harrington wrote these words on 08/21/05 11:28 CST: > > >> Has anyone experienced the problem where the dynamic loader >> library is not included in the list of LIBS? > > > I am not seeing this issue. I am including the entire output from I built openssh as per BLFS on my gcc4 today. No problem building it at all, but the make -k check fails. This is the configure - I have very few optional deps configured: OpenSSH has been configured with the following options: User binaries: /usr/bin System binaries: /usr/sbin Configuration files: /etc/ssh Askpass program: /usr/sbin/ssh-askpass Manual pages: /usr/man/catX PID file: /var/run Privilege separation chroot path: /var/lib/sshd sshd default user PATH: /usr/bin:/bin:/usr/sbin:/sbin Manpage format: cat PAM support: no KerberosV support: no Smartcard support: no S/KEY support: no TCP Wrappers support: no MD5 password support: yes libedit support: no IP address in $DISPLAY hack: no Translate v4 in v6 hack: yes BSD Auth support: no Random number source: OpenSSL internal ONLY Host: i686-pc-linux-gnu Compiler: gcc Compiler flags: -g -O2 -Wall -Wpointer-arith -Wno-uninitialized Preprocessor flags: Linker flags: Libraries: -lresolv -lcrypto -lutil -lz -lnsl -lcrypt and the error in the tests: ... run test reconfigure.sh ... ok simple connect after reconfigure run test dynamic-forward.sh ... skipped (no suitable ProxyCommand found) run test forwarding.sh ... ok local and remote forwarding run test multiplex.sh ... test connection multiplexing: envpass test connection multiplexing: transfer scp: failed copy /bin/ls cmp: /bsources/openssh-4.1p1/regress/ls.copy: No such file or directory scp: corrupted copy of /bin/ls test connection multiplexing: status 0 test connection multiplexing: status 1 test connection multiplexing: status 4 test connection multiplexing: status 5 test connection multiplexing: status 44 Master running (pid=14738) Exit request sent. failed connection multiplexing make[1]: *** [t-exec] Error 1 make[1]: Target `tests' not remade because of errors. make[1]: Leaving directory `/bsources/openssh-4.1p1/regress' make: *** [tests] Error 2 281.9 Elapsed Time - openssh-4.1p1 However, ssh appears to work fine so far - investigating further. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Summer Time & Winter time adjustments
Declan Moriarty wrote: > > Right now the time is close to 20:00, the BIOS says 20:00 but the lfs > installations say 21:00. I'm on GMT about half of the year (to be > precise, 2/5ths to 3/5ths). The time command is for somebody else - > somebody who can dial up on pppd directly - that sort of guy. How do I > shift it back an hour or forward. > > But the various time programs seem broken at best, and linux seems > confused about clocks, having turf wars over this function & that one. > > [EMAIL PROTECTED] ~]# hwclock --show > Wed 17 Aug 2005 20:28:09 IST -1.008866 seconds > [EMAIL PROTECTED] ~]# date > Wed Aug 17 21:28:19 IST 2005 > [EMAIL PROTECTED] ~]# gettimeofday > -bash: gettimeofday: command not found > > I presume that's what inspired ddate :-) > > What's the story? How do I sort that one? /preamble Long ago, when even Sligo was unified as part of Great Britain, we had a little tiff with Germany's Kaiser Bill over who was top nation (America won, by the way) called by some WW1. During the tiff, the munitions workers kept skiving off down the pub, so the Gorvenment brought in Licensing Hours to make the pubs close when the factories were working and to make the girls who put HE in shells a little more sober - and also decided to make the working day longer in the summer by messing about with the clocks and so invented British Summer Time. Personally, I think it was the drinking hours that made the Irish revolting, but there you are. Later, after Ireland had done the right thing and got a divorce so they could have their own licensing hours, we had another little shenanigan with Germany over who was in charge in Europe (America won again!). During that little affair the Government introduced Double Summer Time, just to make even more bombs. /end preamble The problem you have with Linux and Clocks comes about this-wise: Sensible Operating Systems, like UNIX, keep their hardware clocks synchonised to Univeral Coordinated Time (UTC) - which just happens, for excellent political reasons, to coincide with GMT. This allows them to 'leave it alone' which is an excellent thing to do with Hardware anything. This is known as a Standard. Sensible Operating Systems then report date and time by applying the correct offset to the system clock, which might be the hardware clock, or, more likely a software clock initialised to the hardware clock, or even a network clock of greater accuracy. Bill Gates, when writing DOS didn't like standards, because people wouldn't pay for them (and , of course, because he was a ). So DOS kept it's hardware clock in Local time. When DOS grew old, bloated and senile and changed it's name to Windows, it kept its hardware clock in local time. When Linus and friends were writing Linux, they tried to write a Sensible Operating System, and so followed the UNIX custom, but knowing that most Linuxes were going to run on PC's and possibly even dual boot with DOS/Windows, they came to an 'accomodation' and allowed users to decide if the hardware clock should be UTC or Local Time. The decision you make is recorded in /etc/sysconfig/clock. Now if you don't run dual boot, and have your /etc/sysconfig/clock saying so, your BIOS (aka the hardware clock) will be an hour out (in Ireland, and UK) during summer time. hwclock reads the hardware clock. date and most other software use a system call to get the date and hence gets the one corrected to local time. Matt is also right, but I'm more entertaining :-) Richard. (Irish on my Paternal Grandmothers side) -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Firefox error to install
mlij wrote:
> Hi everyone!
>
> There was not any error to install it.
> When the BLFS 6.0 asks to run firefox in the /usr/bin, there is a failure:
>
> [EMAIL PROTECTED]:/usr/bin# firefox
> *** nsExtensionManager::_disableObsoleteExtensions - failure, catching
> exception so finalize window can close
> /usr/lib/firefox-1.0.6/run-mozilla.sh: line 159: 25954 Segment Fault
> "$prog" ${1+"$@"}
>
>
> I downloaded the version 1.06 (I changed the commands from 1.0 to 1.06)
>
> What happened?
>
> Mlij
>
This happens on the initial run of firefox as root if the font system is
not properly installed. The firefox initialization system is trying to
put up a message to say that it is deleting your old extensions (from
the previous installation, if you have one), but can't, as the font
files are not accessible.
Make sure you have the X11 font system set up and re-read the BLFS pages
on X Window System Components.
Richard.
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Re: setting up firewall
DJ Lucas wrote: Do what the book tells > you and execute '/etc/rc.d/init.d/iptables start' else you are calling > the iptables binary in /sbin, and start is an invalid argument. > Once-apon-a-time the firewall script was called 'firewall' and IMO this was a better name for it. Iptables (the executable) is a program that sets up the kernel's ip tables, it's not a daemon like, say, cupsd. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Evolution-2.0.2
I think I've got my pkgconfig right, and built the right versions of the dependencies, and yet Evolution's configure says this: checking for glib-2.0 libbonoboui-2.0 >= 2.4.2 gnome-vfs-2.0 libgnomeui-2.0 libglade-2.0 libgnomecanvas-2.0 libxml-2.0 gconf-2.0 gal-2.2 >= 2.2.3... Package gal-2.2 was not found in the pkg-config search path. Perhaps you should add the directory containing `gal-2.2.pc' to the PKG_CONFIG_PATH environment variable No package 'gal-2.2' found configure: error: Library requirements (glib-2.0 libbonoboui-2.0 >= 2.4.2 gnome-vfs-2.0 libgnomeui-2.0 libglade-2.0 libgnomecanvas-2.0 libxml-2.0 gconf-2.0 gal-2.2 >= 2.2.3) not met; consider adjusting the PKG_CONFIG_PATH environment variable if your libraries are in a nonstandard prefix so pkg-config can find them. PKG_CONFIG_PATH=/usr/X11R6/lib/pkgconfig:/usr/local/lib/pkgconfig: \ /opt/gnome-2.10.1/lib/pkgconfig $ ls /opt/gnome-2.10.1/lib/pkgconfig bonobo-activation-2.0.pc libedataserver-1.2.pc camel-1.2.pc libedataserverui-1.2.pc camel-provider-1.2.pc libegroupwise-1.2.pc cspi-1.0.pc libgail-gnome.pc eel-2.0.pclibgnome-2.0.pc epiphany-1.6.pc libgnomecanvas-2.0.pc evolution-data-server-1.2.pc libgnomecups-1.0.pc gail.pc libgnome-menu.pc gal-2.4.pclibgnomeprint-2.2.pc gconf-2.0.pc libgnomeprintui-2.2.pc gedit-2.10.pc libgnomeui-2.0.pc gnome-desktop-2.0.pc libgtkhtml-2.0.pc gnome-doc-utils.pclibgtkhtml-3.1.pc gnome-keyring-1.pclibgtkhtml-3.6.pc gnome-mag-1.0.pc libgtop-2.0.pc gnome-mime-data-2.0.pclibloginhelper-1.0.pc gnome-vfs-2.0.pc libnautilus-burn.pc gnome-vfs-module-2.0.pc libnautilus-extension.pc gnome-window-settings-2.0.pc libpanelapplet-2.0.pc gswitchit.pc libspi-1.0.pc gtksourceview-1.0.pc libwnck-1.0.pc gucharmap.pc ORBit-2.0.pc libbonobo-2.0.pc ORBit-CosNaming-2.0.pc libbonoboui-2.0.pcORBit-idl-2.0.pc libebook-1.2.pc ORBit-imodule-2.0.pc libecal-1.2.pcsystem-tools-backends.pc libedata-book-1.2.pc xml2po.pc libedata-cal-1.2.pc I tried adding symlinks to the versions I had from the versions it wanted, which works, but it isn't what the book says. R. -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
