Re: [gentoo-user] openofficeq
John Dangler wrote: I know this isn't really the Gentoo (read - Linux) way, but, since I'm using this for business apps, I need to take the end users point of view for a moment. Open Office or the "Ximianized" Open Office. I read that Ximian was bought by Novell, so I'm wondering which of these two portage packages to get. www.ximian.org is no longer, apparently, so I'm wondering when this will turn into a commercial (read - for sale) product. Google for ximian openoffice to find the new homepage, please. I really doubt that this will become a commercial product (and it might not even be possible due to the licensing, i don't know which license they use), though i'm not really informed what's going on in the openoffice world. I read that the ximianized version is faster to load and run, and I also read that open office takes a long time to compile, so I'd hate to put the Ximianized version together and then find out somewhere down the road that I'd have to dump it and go to the other version. Any insight here is greatly appreciated. I'm finally (after 3 days and nights) getting to the end of having a "basic functioning laptop". Don't know about the merits of openoffice-ximian vs openoffice. Use openoffice-bin / openoffice-ximian-bin (precompiled binaries), and save yourself a lot of time, the compile takes _ages_. You can see your options by issuing a emerge -s openoffice to search for ebuilds containing the words openoffice. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] gaim emerge error
John Dangler wrote: I just tried emerging gaim and got this error: configure: WARNING: Could not find Tcl which is needed for the kadm5 tests configure: error: Could not find Tcl !!! Please attach the config.log to your bug report: !!! /var/tmp/portage/mit-krb5-1.4.1-r2/work/krb5-1.4.1/src/config.log !!! ERROR: app-crypt/mit-krb5-1.4.1-r2 failed. !!! Function econf, Line 485, Exitcode 0 !!! econf failed This is not an error in the gaim ebuild but in the app-crypt/mit-krb5 ebuild (look at the line that says ERROR: ...) which probably got pulled in as a dependency of gaim. Everything seems to work ok now, except a message I got during the emerge said to do this: make sure to run revdep-rebuild . What exactly does this do? http://www.gentoo.org/doc/en/gentoolkit.xml This is the first hit when googling for revdep-rebuild, please try google before asking the list. Should I still send the config.log file in as a bug? (I copied the log file to another location since I figured /tmp... anything would get overwritten) You seem to have maketest in FEATURES (have a look at the output of 'emerge info' to see what is set through your configuration and the profiles). When trying to reproduce this bug, it said that the maketest feature had been disabled for debugging, maybe you need to do an emerge sync (but i use amd64, so that might be different here)? anyways, aside from your solution, you could have also done FEATURES="-maketest" emerge foo too disable the maketest feature just for one merge. Hope that helps, Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] cannot load libc, permission denied
David Busby wrote: I cannot open /etc/ld.so.cache! What? See: cdrtx cdr_tx # ls -l /etc/ld.so.cache -rw-r--r-- 1 root root 18311 Aug 27 21:13 /etc/ld.so.cache Any other ideas? I reproduced your bug and diagnostics on a vmware install of gentoo exactly by doing a chmod a-x / So issuing a chmod a+x / should fix it if this is the cause. But it's very weird that the execute bit should not be set on / if it's this what is causing your errors. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Get rid of PAM?
Jerry Turba wrote: Thanks Willie and Marco for the ideas. I got the HOWTO and will read it and try it out. I wasn't aware that there was a gentoo wiki. Looks like lots of info there that I need to read. Thanks for the help. Hi, just for clarification so there is no confusion, my suggestion to pass init=/bin/bash to the kernel from grub/lilo on boot was only meant as an alternative to using a live-cd in case you brake your system authentication while removing pam. Using init=/bin/bash is probably a very bad way to replace any sort of authentication system, it's just useful as a way of starting up the system without running any sort of init scripts, mounting extra filesystems, loading kernel modules, etc... In case anyone reading this is interested in more info about the normal startup of the system (which is the implicit init=/sbin/init used if you don't supply an init parameter), there's more info here: http://www.tldp.org/LDP/intro-linux/html/sect_04_02.html The rest of the site www.tldp.org (The Linux Documentation Project) has great info on many things in the linux universe as well. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Get rid of PAM?
Willie Wong wrote: 2. I already have pam installed. What is the cleanest way to remove it without having any residual hiccoughs. http://gentoo-wiki.com/HOWTO_Remove_PAM Follow it exactly. If you miss a step, you might have to whip out a liveCD the next time your reboot to get into your systems. FWIW, you can always pass the kernel init=/bin/bash in your bootloader to bypass authentication completely. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Starting xfce4-panel automatically
Grant wrote: The xfce4-panel has stopped loading automatically for me and the calendar has started. Where can I specify that I want the panel started with xfce4 but not the calendar? panel: I believe it should be sufficient to start the panel (xfce4-panel) manually once via terminal or alt+f2 and then save the session on exit, which should then make the panel get started automatically on startup. calendar: Go to xfce settings manager -> xfce calendar and set start visibility = hide Or did you want to not run the calendar at all? Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] why gentoo doesn't have long description?
W.Kenworthy wrote: Lastly, top/bottom posting is a religion so you are unlikely to change either groups opinion, but just cause angst. [This is not meant as a flame, so please read this message to a lighthearted tune, sipping a glass of wine in the comfort of your favourite easy-chair :) -- and everyone else, please excuse this message's total offtopicness] A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail? Yeah, that one's a classic, but I think it quite nicely illustrates the point. I agree with you that it is annoying to scroll through a whole 10 pages of quoted e-mail conversation just to find the two-line answer hidden at the bottom, and that top-posting in this case would seem more pleasant. But I think this is an indication that the poster could not be bothered to actually quote the relevant passages that he is answering to instead of just unnecessarily leaving the whole conversation in, which other readers will have read previously anyway. The purpose of a quote IMHO should be a short reminder to other readers which part of the conversation you are referring to in order to prevent misunderstandings, not a complete ISO-9000 compatible documentation of the whole thing. When people include everything, they actually make it harder to tell which part they are referring to. So, I hope i haven't annoyed you too much with my ramblings. Maybe it's best to think about the whole issue under the aspect of better communication than some sort of holy "netiquette that must be strictly adhered to lest burning in the depths of nethell be avoided". I really do think that this is not a religious issue between textmode/graphical clients (i use thunderbird myself) but that there are actually rational arguments that can be made, and I hope you consider my post in this light. Thanks, Marco Disclaimer in case anyone is going to call me on it: Yes, I used to top-post myself till not so long ago (and also got annoyed when people got righteous about it), as e-mails on this list dating back a mere 3 months or more will illustrate... -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] VMWare?
Raphael Melo de Oliveira Bastos Sales wrote: As for comments on VMWare, it isn't free software and you need a licence key. Some purists don't like the fact that it isn't free. I never tested it, but I hear it is a great emulator. Plex86, the free software counter-part, is still under development. So it is really young and probably has less funcionalities. I don't know any other options. You may also want to try qemu (LGPL) and the acceleration kernel module kqemu, which makes qemu quite competetive with vmware (but is sadly not open-source). http://fabrice.bellard.free.fr/qemu/ Plex86 apparently can only run a patched linux guest os at the moment: http://plex86.sourceforge.net/ http://sourceforge.net/mailarchive/forum.php?thread_id=7762667&forum_id=26580 Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Finer grained date output.
Willie Wong wrote: The nanosecond option has always puzzled me, if only because on my box I never get any finer grained output than microseconds... i.e. date +%N always gives 0 as the last three digits. Now technically this computer runs at 2 GHz... so presumably it is possible to hit the enter key "not exactly on the microsecond". Is there something in the kernel? or the clock? I'm also getting the last three digits zeroes. I started digging a bit, this is what I found: gettimeofday(2) says it returns microseconds clock(3) apparently returns processor ticks, though naturally there will be some hard limit to it's accuracy due to scheduling, etc. The FreeBSD man page has some more about different clocks: http://www.freebsd.org/cgi/man.cgi?query=clocks Under Linux (couldn't find anything in fbsd) there is also /sbin/hwclock, apparently an interface to the hardware clock. It seems to take some while though to actually read it out. So I don't know where date(1) is getting it's nanoseconds from, but I suppose that nanoseconds is just there for upwards-compatibility and it calls gettimeofday(2) at the moment. Does anyone know some more details? Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Re: Finer grained date output.
Harry Putnam wrote: Ha .. Thanks.. has that been there a long time? Dunno, I noticed it a few weeks ago -- probably has been there since processors have had frequencies above a GHz i suppose. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] multiple kernel configs
John Dangler wrote: Does the moving of the config files have any effect on the already compiled kernel? (I was under the impression that a .config _went with_ a specific build) or is that a throwback to too much time in a m$ environment? the .config file determines how and especially which parts of the kernel are built by the kernel build environment -- the .config has no effect after the kernel has been built. a windows kernel will probably be built similarly, though who knows what happens in the kernel caverns of redmond... :) Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] USE flags
John Dangler wrote: The list of possible flags is somewhat overwhelming. And many of them, I wouldn't really know if I need them or not. So far, since I only have the base system running, I'm trying to get everything I want to have sans a graphic environment going, so I'm doing USE="-X" with the system level apps (although I don't know if that's the best plan for something like cdrtools/dvdrtools). just add the ones you need -- you're not missing out on anything :) you can always add them later on when you need them I really appreciate the feedback!!! no problem Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] updates
John Dangler wrote: That makes some sense. (Gentoo is all about choices) So, basically, I emerge the new 'slot' and then re-compile the new kernel version according to the handbook, giving me both the existing kernel version and the new version... Exactly, installing the new kernel sources does not automatically clean out the old ones. After you've compiled the new kernel and copied it to /boot, you can then add another entry to grub.conf so you can choose between the two kernels when booting. Hint: If you're happy with your old kernel config, copy it over from your old kernel dir (or from /boot if you've saved it there) to your new kernel directory. you can then run 'make oldconfig' which will only prompt you for new configuration options that have been added between the releases so you don't have to select everything again by hand. One further note about cleaning out old kernels: running make modules_install will install the kernel modules to a subdirectory in /lib/modules/. Once you're not *using* an old kernel anymore, you can remove the subdirectory specific to that kernel from there. emerge -C "=some-kernel-version" will only remove the kernel sources from /usr/src Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Finer grained date output.
Harry Putnam wrote: I know about perl scipting for this but wanted something like the `date' command that is its own dedicated program. Is there a unix tool that outputs a finer grain of time segments? man date look for nanosecond format, e.g.: date +"%N" Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] updates
you might also want to try man emerge most commands on your system will have a manpage, and they will probably give you helpful hints as to how they function. you can also do it with any other command: man ls man gcc man ldd man man ... Hope that helps, Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] updates
John Dangler wrote: I just finished the base install of the 2005.1 system (2.6.12-r6). When I run emerge -sync, and then emerge -uDvp system, I get a short list updates that portage wants to emerge, but there aren't any kernel updates. However, if I run emerge -uDvp world, there is a new version of the gentoo-sources (2.6.12-r9). Why wouldn't kernel updates be included in a system emerge? The packages in your system come from /etc/make.profile/packages (Note: do *not* edit this file) The packages considered for world are the ones from system plus the ones in /var/lib/portage/world The world file is where portage records the packages you emerged, e.g. when you typed emerge gentoo-sources it recorded it there. I suppose the reason the kernel is not in the system file is that this file is a kind of 'factory-defaults', which you shouldn't be normally changing. But there is more than just one kernel source tarball avilable in portage, e.g. gentoo-sources, vanilla-sources and some more. Putting this in the system file would unnecessarily constrain your choice as to which kernel to run. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] USE flags
John Dangler wrote: I have just installed a basic 2005.1 system (2.6.12-r6) on my laptop. I'm trying to get my arms around the USE flags. I found a set of 'default' settings (I think) under /usr/portage/profiles/base/use.defaults . From what I've read in the gentoo documentation, this seems to be a list of default USE= flags. What I'd like to try and get to is, a difference between what's there and the 'total' list, and why would I add others to my own make.conf file? The relevant part in the docs: http://www.gentoo.org/doc/en/handbook/2005.1/handbook-x86.xml?part=2&chap=2 A list of all USE flags: http://www.gentoo.org/dyn/use-index.xml Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] multiple kernel configs
John Dangler wrote: I changed some options to the menuconfig (trying to get that splash to work), and saved the changes to an alternate config file. I'd like to make a kernel with _that_ config file and keep it separate from my default 2.6-r12 kernel, since, when the splash causes the panic, I have some way to get back in. How do I compile a new kernel that I can add to the grub.conf with the alternate config ? The kernel config file used for compiling the kernel is called .config So, to compile a kernel with that config file, just replace the .config with your alternate file, saving the old .config of course if you want to keep it: mv .config my-old-config mv my-new-config .config (do this in the kernel source dir) You might have to call 'make clean' before you compile your new kernel, i'm not 100% sure on that -- it won't harm though. You can add as many kernels to your grub config as you like -- you might want to check with the install guide how to do that exactly: http://www.gentoo.org/doc/en/handbook/2005.1/handbook-x86.xml?part=1&chap=10 Basically, your second kernel gets the same kind of entry just as your first, but you will want to change the title and of course the kernel (the root line should be the same as in the first block). Move your kernel under the name you specified to the same location where your other kernel is, using the name you specified in the config. (you will probably have to mount your /boot partition to do all this naturally) I.e. add an extra block something like this to your grub.conf: title=Gentoo Linux 2.6.11-r11 root (hd0,0) kernel /kernel-2.6.11-gentoo-r11 Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Depclean
dont inaverdently remove them? For instance I am using udev for some time now, can I let depclean remove devfsd? yes you can, at least i had no problems with that Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Depclean
And then I'm quite surprise, 41 packages should be removed, and among them, a lot of usefull lib or tool (perl-ldap, xinetd, and so on ...) Is this result normal ? What did I miss ? 1. it might be old stuff that really isn't needed anymore, as already mentioned by others 2. portage up to a short while ago had a bug (or feature?!?) that when you emerged a package for the first time and used -u to do this, it would not be added to the world file. Nowadays it does the right thing and adds it to the world file if you use -u or not. This means that possibly some packages that you emerged yourself and weren't pulled in as dependencies by other packages in the world file are now regarded by --depclean as unnecessary, as they don't appear in the world file themselves and aren't dependencies of anything in there. This means that you'll have to go through the list resulting from a '--depclean -p' and have to see if you truly wanted that package, and if yes, re-emerge that package or add it to the world file yourself. Once you're done remerging the stuff you want to keep and have cleaned out the rest, it would probably be best to check that everything is ok with a revdep-rebuild In the future though, everything should work fine just as you were doing it. Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Re: OT: bash details
Well this is an excellent resource, BUT it seems devoid of any examples where a custom device driver, say for the serial port on a linux system, inserted as a module or is part of the kernel, and the associate software that allows users to access some of the hardware(features) and not other hardware/firmware/kernel features(code), unless they are root, or have a special (encrypted)key or another form of chicanery (biometric generated key). If you are interested in device drivers and kernel stuff on linux, you might want to check out these fine books: Linux Device Drivers, 3rd ed. (available online) http://lwn.net/Kernel/LDD3/ Linux Kernel Development 2nd ed. http://rlove.org/kernel_book/ Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] USE flags...
Vincent A. Primavera wrote: Hello, Just looking for some opinions here. What is a good approach to installing applications with a minimal amount of optional USE flags enabled? For example, if one were to run `emerge -pv kde-base/kde` you Hi, regarding kde, you might want to try the split ebuilds if you haven't heard of them before: http://www.gentoo.org/doc/en/kde-split-ebuilds.xml About the USE flags, sorry i have no better sane solution to your problem than to disable all at the beginning and then incrementally enabling more when you need them... Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Subversion 1.2
Steve [Gentoo] wrote: > Hmmm - that all sounds sane, but what is this default period of time? > What criteria must be met in order for a masked package (and > specifically for Subversion) to become unmasked? I *think* it is something along the lines of "30 days without a bug", not 100% sure though. >> Here, i just did it myself by putting this in my package.keywords >> (create this file if it doesn't exist) : >> >> =dev-util/subversion-1.2.1 ~x86 >> =dev-libs/apr-util-0.9.5~x86 >> =dev-libs/apr-0.9.5 ~x86 >> >> > In one way this looks better than my fiddling with USE - however I'm > reluctant to choose specific versions in a durable configuration file. > Ideally I'd like to follow the natural upgrade cycle in future. > Wouldn't putting those lines in my package.keywords file prevent me > getting, say, version 1.3 automatically when I do an "emerge -uD world" > in another few months? the line: =dev-util/subversion-1.2.1 ~x86 means that you tell portage that you'll accept subversion, version 1.2.1 exactly, with a keyword of '~x86'. You can use '>=' instead of '=', which means any version equal or greater than 1.2.1. the two following lines were the two dependencies i found by trying 'emerge -uD subversion'. once 1.3 or any version higher than 1.2.1 becomes stable (marked 'x86'), it will be considered by portage as well and will be merged. Hope that helps, Marco -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Subversion 1.2
Steve [Gentoo] wrote: > I'd have thought lots of people in the gentoo crowd would have been > eagerly awaiting subversion 1.2.x with its substantial new "reserved > checkout" - but nothing seems to have moved forward. you must have missed this link from the gentoo homepage (on the left): http://packages.gentoo.org/ a search yields this: http://packages.gentoo.org/search/?sstring=subversion also, for a command-line version, read: man equery You might also want to read up on the portage section in the gentoo handbook: http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=3&chap=3 sorry for such a canned answer, but i would only repeat what's been written there. > I've tried using ~x86 as my USE flag - but the 1.2 ebuild still won't > install reporting a "Problem in dev-util/subversion-1.2 dependencies"... > I'm reluctant to use an unstable subverison port as it would cost me a > fair bit of time if it scrambles my version controlled files. Does > anyone know what the problems are and why its taking so long to get > 1.2.x into the default portage tree? Don't put ~x86 in your USE flags just for that - use /etc/portage/package.keywords (see the above mentioned portage guide). I'm not exactly sure what you want - Gentoo leaves packages in unstable for a default period of time to make sure they work allright. If you want the newest version of a package, you must tell portage to do so by putting the appropriate stuff (subversion and it's dependencies) in /etc/portage/package.keywords. Here, i just did it myself by putting this in my package.keywords (create this file if it doesn't exist) : =dev-util/subversion-1.2.1 ~x86 =dev-libs/apr-util-0.9.5~x86 =dev-libs/apr-0.9.5 ~x86 You just add one package, ask portage to merge, then put in the next dependency, and so on... Tried this on amd64 (with ~amd64 instead of ~x86, naturally), it's happily compiling away... This was just info about portage, it is in no way any form of endorsement on the new version of subversion, as I haven't used it at all - and I don't know if you should be so impatient with a new version of a package that seems to be important to you and your data... Hope this helps, Marco -- gentoo-user@gentoo.org mailing list