Re: [gentoo-user] KDE 3.5.10 in portage...
On Sunday 14 September 2008 22:37:21 Volker Armin Hemmann wrote: Making the 'official' overlay paludis-only was a BIG mistake. It's not paludis-only, it will work with any package manager whose developers care enough to support the useful features of the kdebuild-1 EAPI. and that was an official api? Yes? No? Official according to whom? It is a stable and well documented eapi that any package manager that regards those features to be useful can implement. Was it needed? No - proven by kdesvn-portage And noone ever claimed otherwise. [...] The KDE overlay isn't produced by the paludis-group. no, it was just made by vivid paludis fans. Hey, lets make an overlay a lot of user want - and make it paludis only. That way we can push paludis! Wow. What do you base this nonsense on? No, to provide ebuilds that make use of useful features that benefit both users and developers. which one? Which features 'benefit both users and developers' and are sooo important that the kde overlay had to be paludis only? Name them please. - '-scm' support (--dl-reinstall-scm for users) - use dependencies (no surprising interruptions mid-merge) - suggestions (you see them upfront rather than in elog messages afterwards) - sets The latter of these is not subject to eapi but incredibly useful when dealing with huge amounts of packages. It obsoletes meta packages and makes reinstalls or uninstalls of all packages in the sets trivial. For Paludis it also means that you can unmask/keyword two hundred packages just by addding the sets to your packages.{keywords,unmask} equivalents. Both Paludis and Portage 2.2 now has sets support although the details of their implementations vary greatly. Oh, does paludis support and equivalent to 'keep-going' or '--ignore-failures' or are people who wants this extremly usefull features still attacked and insulted? Paludis had --continue-on-failure long before --keep-going was implemented in Portage (which is months after the creation of the kdebuild overlay). This is also one of the advantages that users of live KDE ebuilds got by using Paludis (or get if you consider the additional flexibility when compared to --keep-going to be useful). On Monday 15 September 2008 00:38:18 Volker Armin Hemmann wrote: lets see - an overlay is setup to develop and test ebuilds for KDE4 that should some day go into the tree. Deciding to use a feature that the official pm does not provide - and only one of the three makes the 'testing' part and 'for the tree' pretty superfluos. And again I wonder what you base this nonsense on. Perhaps you never ever bothered to look at this overlay, what it provides or what the intentions behind it was? As one of the original decision makers behind it I can tell you, that we never intended to put any of these ebuilds in the tree. For this reason it also never contained any release of KDE. Releases were maintained separately in the tree using eapi 1. It only contains live ebuilds. Which is where '-scm' and sets support provide the biggest advantages. And we didn't do it to harass users. We did it because we wanted to get some real world experience with some of the features that Paludis had provided for years yet there were no indications Portage would support any time soon. Live KDE packages was deemed the place where adding this requirement made the most sense. Managing two hundred packages without those features is pain anyway. We decided that the monthly KDE releases that we were packaging and adding to the main tree using eapi 1 were frequent enough for those who didn't want Paludis for whatever reason. If anybody disagreed with that they could maintain their own overlay (which they did/do). We also announced it over three weeks before we actually made it happen so anybody who cared about the live KDE ebuilds can't really complaim about having been caught by surprise. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Speed up with pbzip2 or not!?
On Sunday 27 January 2008 18:29:56 Enrico Weigelt wrote: Why not an universal wrapper script ? maybe something like: magic-uncompress [-t format] [-c] [-o output] input This script could try to find out the input format automatically (the optional -t parameter allows to explicitly specifiy the format) Once we have this script, ebuilds could be rewritten step by step and no one (else than magic-uncompress) has to care about the actual commands behind. Some could be done w/ tar and zip, etc. You've just reinvented unpack(). It's been a part of portage for ages and is used in just about every ebuild in the tree... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Solid with Networkmanager Support
On Sunday 27 January 2008 20:03:32 Thomas Kahle wrote: i'm trying to get Networkmanager support in KDE, but i cannot compile kde-base/solid with networkmanager use flag. It seems like portage just filters the flag, as it is in braces... emerge output shows USE=bluetooth (-networkmanager) -test Can anyone confirm this or has a solution ? It means the flag is masked on your profile (currently it is on most profiles). Probably it's broken. If you really want to try to enable it `man portage` tells you how. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4.0?
On Saturday 26 January 2008 10:59:15 Neil Bothwick wrote: On Fri, 25 Jan 2008 21:19:48 -0800 (PST), maxim wexler wrote: [blocks B ] kde-base/kdebase-3.5.7-r6 (is blocking kde-base/kdelibs-4.0.0) From the gentoo-kde4-faq: Q: You said I could install KDE 3.5 and KDE 4.0, but they block each other! A: You will need to install the latest revision (not version) of kde-base/kdebase (if you use monolithic ebuilds) or kde-base/kdebase-startkde (if you use split ebuilds). These versions include patches to allow multiple versions of KDE to coexist. It's quite clear, you need the latest revision of KDE 3 to use it with KDE4. The latest is 3.5.8 qhich is why 3.5.7-r6 is blocking. Nah, =kde-base/kdebase-3.5.7-r6 isn't blocked either. Latest revision of either 3.5.7 or 3.5.8. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4.0?
On Saturday 26 January 2008 02:21:23 maxim wexler wrote: #emerge -avD kde-base/kde-meta:kde-4 eight times and everytime I needed to add something to package.keywords, like so: =x11-libs/qt-4.3.3 ~x86 =app-crypt/gpgme-1.1.6 ~x86 =dev-util/cmake-2.4.7-r1 ~x86 =dev-libs/rasqal-0.9.15 ~x86 =dev-cpp/clucene-0.9.19 ~x86 =dev-libs/redland-1.0.6 ~x86 =x11-apps/xinit-1.0.5-r2 ~x86 =kde-base/kdepasswd-4.0.0:kde-4 ~x86 I notice the entries from the site all end with '* *' but I've been using '~x86' all this time without any problem. Is that wrong? They end in '**'. This is no longer necessary for x86 or amd64. Other arches will still need it. I'm assuming this is correct since portage seemed to accept them in turn. Yep. Now when I came to last entry emerge kept repeating itself. I checked, package.keywords was saved after the last entry Yes, apparently the examples are incomplete after all. This will be fixed. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4.0?
On Friday 25 January 2008 19:45:39 maxim wexler wrote: Which is rather pointless given that the KDE docs provides example package.{keywords,unmask} files... http://www.gentoo.org/proj/en/desktop/kde/kde4.xml Thanks for this. The link from kde.org goes to a page dated Dec 11 2007 which doesn't mention v4.0. The link on kde.org is to kde.gentoo.org which does have a link to the KDE 4.0 Guide. I did what it said using the recommended(Split) method, then this: localhost heathen # emerge -avD kde-base/kde:kde-4 That's not the split method. Since you chose the split packages you need: kde-base/kde-meta:kde-4 See http://www.gentoo.org/doc/en/kde-split-ebuilds.xml for further info... [SNIP] Further down the page it suggests I emerge kdebase-startkde Which is for those who want a minimal desktop rather than everything in KDE (which is what kde-meta is for). localhost heathen # emerge -avD kde-base/kdebase-startkde [SNIP] - x11-libs/qt-4.3.3 (masked by: ~x86 keyword) Add that version to package.keywords. Stable users will need ~arch keywords for a few dependencies of KDE 4. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4.0?
On Thursday 24 January 2008 21:33:12 Dale wrote: I emerged autounamsk and used the command autounmask kde-base/kde-meta-4.0.0 without the quotes of course. There may be a block or two. I had one that involved qt. I also had to mask qt 4.4 since it does not have dbus support. Which is rather pointless given that the KDE docs provides example package.{keywords,unmask} files... http://www.gentoo.org/proj/en/desktop/kde/kde4.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Polymake Ebuild and general questions
On Sunday 20 January 2008 17:57:51 Thomas Kahle wrote: I've written an ebuild for a math-software called polymake (www.math.tu-berlin.de/polymake). You find it attached. You have to digest it yourself to use. (ebuild polymake-2.3.ebuild digest) Maybe some people could test it (on amd64 for instance...). Now some questions: 1.) The Program uses non-autotools self made configuration via make configure. Currently the Ebuild will ask questions to the user, which I want to prevent. How can i get around this? Is there some Bash trick to answer all questions with Enter ? 2.) The program needs to be rebuild after an upgrade of dev-lang/perl. How can i implement this in the ebuild ? 3.) Is there any chance that, after testing, this will land in the portage tree ? How can I do this, -email to dev-mailinglist? -bugzilla ? - ... ??? I know you've already gotten help on other mailing lists so I'm posting this for the benefit of other gentoo-user readers. A gentoo-devhelp mailing list was created last week. This is now the correct list for this kind of questions. :) http://www.gentoo.org/main/en/lists.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4 first impressions
On Monday 21 January 2008 09:42:17 Alan McKinnon wrote: Kontact works properly and so does the kmail component. kmail on it's own does not work. wtf is up with that anyway? erm... kdepim (including kontact and kmail) wasn't released with kde 4.0.0 at all... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Once again baffled by portage
On Friday 18 January 2008 15:36:29 Alan McKinnon wrote: Roll on KDE4 when this monolithic nonsense will go away and there will only be -meta ebuilds. Actually it was decided to keep monolithics in KDE 4.0.0. Splits are now the default which means they are listed first in any-of dependency blocks such as e.g. || ( kde-base/kompare:kde-4 kde-base/kdesdk:kde-4 ) in KDE 4. But monos are still around. There is a list of the monos in the url I posted in another mail to this thread. Interesting, I wasn't aware of that. Obviously my most current info is out of date. But why was that decision taken? I understand keeping it for kde-3.5.x users, but kde-4 is essentially an entirely different product, and the -meta ebuilds do everything the monolithic ones ever did. The configure steps do add time though, but other than that everything seems to work the same Well, it wasn't really my decision. ;) But I can say that a pro is that it provides the users with a choice and the maintainance overhead when compared to only doing splits isn't really all that big. The only real con in my opinion is that it confuses those who haven't read the kde-split document before installing kde. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] CFLAGS
On Thursday 17 January 2008 17:10:11 Alan McKinnon wrote: Itanium2. You don't want to go there. Trust me. It's a bastard evolution of one of Intel's worst ideas ever I thought that would be Pentium 4 EE. :P -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Once again baffled by portage
On Thursday 17 January 2008 16:28:24 Grant Edwards wrote: How does one figure out where these blocks are coming from? There are no other versions of kompare installed. kde-base/kdesdk-3.5.7 _is_ installed. http://www.gentoo.org/doc/en/kde-split-ebuilds.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Once again baffled by portage
On Thursday 17 January 2008 17:06:28 Alan McKinnon wrote: Roll on KDE4 when this monolithic nonsense will go away and there will only be -meta ebuilds. Actually it was decided to keep monolithics in KDE 4.0.0. Splits are now the default which means they are listed first in any-of dependency blocks such as e.g. || ( kde-base/kompare:kde-4 kde-base/kdesdk:kde-4 ) in KDE 4. But monos are still around. There is a list of the monos in the url I posted in another mail to this thread. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Once again baffled by portage
On Friday 18 January 2008 01:08:51 Neil Bothwick wrote: Actually it was decided to keep monolithics in KDE 4.0.0. Splits are now the default which means they are listed first in any-of dependency blocks such as e.g. || ( kde-base/kompare:kde-4 kde-base/kdesdk:kde-4 ) in KDE 4. But monos are still around. There is a list of the monos in the url I posted in another mail to this thread. Would it be possible to modify the kde-meta eclass so that a split vs monolithic block gave a more informative error message. Even a generic message pointing to that URI would be a great help. Since we only have the RDEPEND=!kde-base/kdesdk:kde-4 etc. dependency syntax which doesn't allow messages I don't see how (if you see a possibility I'm missing say so). The message you get is a generic resolver failure. If you have ideas for better wording of the blocker message from portage I guess you should file a portage bug. If you have any ideas for extending the ebuild format to make it possible to add custom messages with urls then I guess filing a PMS/EAPI bug is the way to go. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Is GWN dead?
On Monday 14 January 2008 10:48:08 Etaoin Shrdlu wrote: You can use Daniel Robbins' stage3s: Why should I? They're likely also built with different use flags than the ones I use. Thus I will end up recompiling anyway. Ah ok, I wrongly thought yours was more a problem of outdated stage3 tarballs rather than different USE flags. In this case yes, the best option is starting from stage1. Not really. emerge -e world from a stage 3 is still both considerably less effort than stage 1 and much more reliable. Furthermore stage 1 is completely unsupported and for a very good reason. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] SSH tunnel With Portage
On Monday 14 January 2008 14:03:52 Alan McKinnon wrote: Now I am confused. Why do you run Putty on a Linux machine? I used it once just to confuse and confound co-workers and have it runable in Wine just to impress some other people. You do not need wine to run putty on Linux. There's a UNIX port. Just emerge putty and run it. Not that I don't agree it's useless with all the alternatives that are available on Linux too.. ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] KDE4 overlay
On Friday 11 January 2008 22:26:34 Alan McKinnon wrote: Anyone know where this overlay disappeared to? It moved to git. Assuming dev-util/git is installed: # layman -f layman -d kde layman -a kde Having said that KDE 4.0.0 should show up as package.mask'ed in gentoo-x86 within a few days. Eclasses for this has just been submitted to -dev@ ... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Suddenly emerging unstable packages = why?
On Tuesday 08 January 2008 23:30:51 Kevin O'Gorman wrote: However, it is true that emerge sync and eix-update were in two separate jobs scheduled an hour apart. It is vaguely conceivable that they got out of step somehow. I've unified them, and hope things go better now. [...] Any particular reason you run two separate jobs and not just eix-sync (which does both in sequence)? Originally, because the output was hard to read I think. And I figured that starting them an hour apart would ensure sequence anyway. Surely you ran update-eix and possibly diff-eix rather than eix-sync then? Running eix-sync and emerge --sync would mean syncing twice a day (for no good reason)... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: besides package.use where to store info needed at emerge
On Thursday 03 January 2008 19:12:50 Neil Bothwick wrote: On Thu, 03 Jan 2008 12:01:07 -0600, [EMAIL PROTECTED] wrote: Sorry to keep nagging with this but I must still not have it right. I do not see that compile flag being set during emerge: It looks like the cvs ebuild doesn't use EXTRA_ECONF. You could file a bug report about this. It uses econf. econf uses EXTRA_ECONF. On Thursday 03 January 2008 19:01:07 [EMAIL PROTECTED] wrote: cat /etc/portage/env/dev-utils/cvs EXTRA_ECONF='--enable-rootcommit' You got the category wrong. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] removing X
On Monday 24 December 2007 18:31:16 [EMAIL PROTECTED] wrote: I'm attempting to remove X from a former desktop machine now going to see action as a semi-DMZ. What is the best way to go about removing X and all its files. Removing the basic x11-base/xorg-x11 is easy enough but there appears to be dozens of other X related pkgs installed. x11-proto/* has apparently dozens of relatives installed. emerge does not appear to accept globbing or maybe I'm just doing it wrong. Yeah, emerge does not accept globbing. bash does though so you could just cd to /var/db/pkg and take advantage of that. For paludis users there are arguments to help with this kind of thing. :) Uninstall options Options which are relevant for --uninstall. --with-unused-dependencies (--no-with-unused-dependencies) Also uninstall any dependencies of the target that are no longer used --with-dependencies (--no-with-dependencies) Also uninstall packages that depend upon the target Also paludis' equivalent for --depclean (--uninstall-unused) doesn't require you to upgrade everything. ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] No help on annoying `sandbox' error
On Friday 28 December 2007 22:58:27 [EMAIL PROTECTED] wrote: I'm having a continuing annoying problem from sandbox trying to write out of its crib. I've posted here twice before but caught no ones attention. Possibly this is something screamingly obvious and people just ignored the posts. `sandbox' doesn't like my root .bash_history. ACCESS DENIED open_rd: /root/.bash_history ACCESS DENIED open_rd: /root/.bash_history I see nothing really like this on bugzilla although their are other access violations there. I guess it needs to be turned in as a bug but fisrst tell me if its really a bug or something to do with my ill-informed setup. emerge --info ? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Paludis newbie questions [was: Excellent Paludis interview]
On Sunday 23 December 2007 18:56:40 David Relson wrote: I suspect a better worded message would have let me find the mistake Such as? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Paludis newbie questions [was: Excellent Paludis interview]
On Saturday 22 December 2007 20:09:40 David Relson wrote: As I'm a paludis newbie and as it allows a multitude of settings, I used portage2paludis.bash to create /etc/paludis/use.conf. Whatever is missing from use.conf is a combination of my ignorance and the script. A copy of use.conf is attached. From use.conf: */* [...] -ipv6 - isdnlog [...] ^^^ Right after -ipv6 there's a minus that disables no flag... ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Paludis newbie questions [was: Excellent Paludis interview]
On Saturday 22 December 2007 02:50:54 David Relson wrote: I'm experimenting with paludis. Seems fine, though a bit verbose and cryptic. Running paludis -i world produces: Unhandled exception: * In program paludis -i world: * When making environment from specification '': * When loading paludis configuration: * When reading use file '/etc/paludis/use.conf': * When adding source '/etc/paludis/use.conf' as a use file: * When validating use flag name '': * Name '' is not a valid use flag name (paludis::UseFlagNameError) [...] Apparently you have a use.conf entry with no use flag (which makes that entry invalid). Paludis is stricter than Portage so it won't just ignore invalid input. What am I missing that produces the unhandled exception and migrate to Paludis configuration messages? We've received rather few reports from people testing the portage environment so it hasn't had much testing. It also disables some features that are available when using the paludis environment. Hence the warning. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Rebuilding world system after -march change
On Friday 21 December 2007 17:39:50 Marzan, Richard non Unisys wrote: Recently, I've changed the -march from k8 to athlon64 and it failed to build packages. Am I doing something wrong here? I suspect that I am supposed to rebuild some packages first before rebuilding with emerge -eDNtv system and world. Any helpful info will be helpful. Thanks in advance. Huh? -march=k8 and -march=athlon64 are identical. The problem must be something else (a typo perhaps? - Maybe post emerge --info ?). -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] re-emerge glibc after linux-headers update?
On Thursday 20 December 2007 16:56:37 Grant wrote: The ELOG for my linux-headers update says: Kernel headers are usually only used when recompiling your system libc, as such, following the installation of newer headers, it is advised that you re-merge your system libc. Failure to do so will cause your system libc to not make use of newer features present in the updated kernel headers. Should I re-emerge glibc? Was that piece of advice really unclear? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Excellent Paludis interview
On Thursday 20 December 2007 10:08:13 Rumen Yotov wrote: Watch out for some scripts (perl-cleaner, claw-mail, etc.) in which the use of portage/emerge is embedded. Put 'paludis' as USE-flag. Unless you use a crappy, unsupported overlay no such use flag exists. [...] Regarding claws-mail there's a script to rebuild it's plugins - see elogs. Ok. Since i first tried paludis-0.2.1, may still have some old use info (laziness) about paludis USE-flag (IIRC revdep-rebuild portage-utils, etc.had it). [...] There's also a paludis-extras overlay, which is rather separated from official paludis but have some nice things (and problems sometimes ;-). That was actually the 'crappy, unsupported overlay' I was referring to. That also is where the packages with a paludis use flag came from. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Excellent Paludis interview
On Thursday 20 December 2007 09:43:20 Neil Bothwick wrote: flagedit will warn you if you have any unsupported USE flags set. That and eix-test-obsolete are useful for keeping make.conf and /etc/portage clear of cruft. And the config-decruft ruby script that can be used with Paludis can do the same (and more) in a much more verbose manner. :) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Excellent Paludis interview
On Wednesday 19 December 2007 21:37:38 Rumen Yotov wrote: Watch out for some scripts (perl-cleaner, claw-mail, etc.) in which the use of portage/emerge is embedded. Put 'paludis' as USE-flag. Unless you use a crappy, unsupported overlay no such use flag exists. Kind of curious how that relates to claw-mail anyway. Python-updater on the other hand supports all three package managers in the tree and just defaults to using portage. :P And yes, I'm a Paludis user (for over a year now). -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Persistent revdep-rebuild issues: sun-jdk-1.4.2.16
On Thursday 20 December 2007 02:43:04 Dale wrote: I have had mine showing it needed to recompile gcc for a very long time now. There is also a bug for it too. It has been there for a really long time. There are hacks to work around the bug but some don't like the hack and neither did I. I just omit gcc and do whatever else it says. It's not a perfect tool but it works, mostly anyway. ;-) It's not a hack. But anyway.. Do you actually need gcj? Otherwise just disable USE=gcj and be done with it... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Excellent Paludis interview
On Thursday 20 December 2007 02:46:20 Iain Buchanan wrote: And the recent addition of the option '--continue-on-failure' won me over all over again. :) I've been wondering for a long time why portage doesn't continue with building other packages when one fails - so long as deps are met why stop? Because the current resolver in Portage isn't powerfull enough to know if the deps are met. Acutally, while I'm on the subject of features - here's another one I'd like to see: parallel merges instead of parallel makes. It's planned (both for Portage and Paludis). For Portage there's even a bug with a patch which isn't considered suitable for inclusion in Portage. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Python vs C++ [was: Gentoo Rules]
On Monday 17 December 2007 14:38:30 Raphael wrote: So, even if Portage was recoded in C++, performance improvements would be marginal and the cost in man-hours would be too high. It would take months before reaching the maturity level Portage has now and all this time could be better spent trying to find solutions to its architectural bottlenecks. I believe that a good solution would be evolving Portage to use different forms of storage, like databases or even LDAP. In a home desktop, you could use SQLite, which is light weight. In a Office enviroment, you could use a larger database, like MySQL or PostgreSQL. In this second case, it would even make sharing the package list faster, since the only current method is sharing it over NFS. I understand that doing so could bloat Portage dependencies, but it is, IMHO, a good way to improve its speed. This post is hilarious for several reasons. Firstly there already exist a package manager for Gentoo which is written in C++. Paludis. And it has a lot of features that Portage has been missing for five years. And it's way more flexible than Portage. Secondly if you just put ebuilds in a database you gain nothing. I.e. other than the added bloat. I/O is still going to be the major bottleneck. :P -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Python vs C++ [was: Gentoo Rules]
On Monday 17 December 2007 16:14:24 Raphael wrote: Hey, I made someone laugh today. Good deed of the day: check! :P :) I was unaware of Paludis. Re-reading the thread now, I saw that someone mentioned it. After googling for it, seems a lot of people are fond of it. Why is it not the default package manager yet? It's not my decision so I'm not going to answer this. I can, however, say that Paludis still lacks a couple of features that Portage has (most notably binary package support and pkg/slot-move support). As for the second part, yes, using a database wouldn't get rid of the I/O problem, but could diminish it, since database data isn't spread across several directories and files. And I'm not proposing to store the entire ebuild within, but a representation of it that could be easily queried. Which would add an awful lot of complexity and require major design changes in order to gain anything. The beauty of the ebuild format is its simplicity. I don't really think it's worth it. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Gentoo Rules
On Saturday 15 December 2007 20:00:54 Grant wrote: The real blocker for features that I'd like Gentoo to support is Portage. There is only 1½ people working on it and changing anything in it is hard because Portage is a horrible mess. There's plenty of activity in the tree but new desired features cannot be used in the tree until Portage supports them. It also doesn't make matters better that over the years all sorts of weird hacks (that now have to be supported) have been added to the tree instead of waiting for proper solutions. Most people who are capable of helping to improve Portage just don't want to touch it. Would you say that portage is the main block in the way of Gentoo's continued progress? Actually I guess that's pretty much exactly what you said. I didn't realize portage is where the problem lies. In fact, I thought we were all still proud of portage I'm going to think about this some. It all depends on what kind of features you're interested in. The features I happen to be interested in requires ebuild changes (which means that me using another package manager doesn't help) *and* requires package manager support before said ebuild changes can happen. Therefore for me Portage is a blocker... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Python vs C++ [was: Gentoo Rules]
On Sunday 16 December 2007 20:32:58 Randy Barlow wrote: C++ is most certainly going to yield faster programs since it is a machine compiled language and python is interpreted. In this case it's not really significant. The biggest performance hit for a package manager for Gentoo remains I/O no matter which language you use... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Python vs C++ [was: Gentoo Rules]
On Sunday 16 December 2007 22:04:52 Randy Barlow wrote: Bo Ørsted Andresen wrote: In this case it's not really significant. The biggest performance hit for a package manager for Gentoo remains I/O no matter which language you use... Yeah, you are right - although there is one step of an emerge that seems to hit my P3 hard: right after it reads the data from disk to determine dependencies, it goes 100% CPU for a few seconds. Just think, we could save a few seconds! ;) But seriously, I totally think Python is the tool for the package manager job - I was just trying to outline when you might choose C++ over python... Heh, I don't think Python is the right choice. Performance just isn't the reason. ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Gentoo Rules
On Saturday 15 December 2007 03:35:51 Grant wrote: My ideas aren't really important unless they're everyone else's ideas too. What is it exactly you want to achieve by starting these pointless threads? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Gentoo Rules
On Saturday 15 December 2007 15:05:28 Grant wrote: Neil correctly translated my pseudo-English to what I actually meant. I don't want to make Portage binary based. I just want to make Portage's binary package support more conveniently usable on big networks. Even eclasses in the tree don't have any sort of checksums and they aren't even included in binary packages either... I don't think there is any shortage of great ideas here. Can we get into specifics on how projects are born and become successful? So, what would need to happen for one of these projects to take off would be one or more people to be in charge of it and organize it, and they recruit as many people as possible to work on the project along with them? The real blocker for features that I'd like Gentoo to support is Portage. There is only 1½ people working on it and changing anything in it is hard because Portage is a horrible mess. There's plenty of activity in the tree but new desired features cannot be used in the tree until Portage supports them. It also doesn't make matters better that over the years all sorts of weird hacks (that now have to be supported) have been added to the tree instead of waiting for proper solutions. Most people who are capable of helping to improve Portage just don't want to touch it. Does that recruitment generally take the form of volunteers finding the project as opposed to the project finding volunteers? Any light to shed on this process for me? If there's one thing we definitely don't need it's more clueless people who become developers just because they claim they want to do something. Being stalled is better than major screw ups that hurt everyone and than moving in the wrong direction. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Emerging virtual/editor installs nano - why?
On Friday 07 December 2007 16:21:31 Alexander Skwar wrote: So the expectation should be, that it's not going to be possible to preselect an mta the way I've shown, as soon as virtual/mta is converted to a new style virtual. Is that right? Is there work going on to change all the old style virtuals to new style ones? You are right except for the fact that virtual/mta will not be converted any time soon (if ever). Old style virtuals allow their providers to block all other providers by blocking the virtual. This isn't possible with the current format for new style virtuals so it isn't going to happen. editor didn't need that feature and thus could easily be converted... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Emerging virtual/editor installs nano - why?
On Friday 07 December 2007 18:58:57 Alexander Skwar wrote: You are right except for the fact that virtual/mta will not be converted any time soon (if ever). Old style virtuals allow their providers to block all other providers by blocking the virtual. This isn't possible with the current format for new style virtuals so it isn't going to happen. editor didn't need that feature and thus could easily be converted... Ah, okay. Pretty bad situation, though, isn't it? I mean, with the way it is now, we've got some virtuals which work with virtuals and some, which don't. Do you happen to know, if there's a bug which should rectify this broken situation? If by 'broken' you mean 'inconsistent' then I can assure you that there are plenty of much more important bugs open... ;) But I guess bug #148251 should be of interest in reference to virtual/mta. Personally I happen to hate old style virtuals and want them dead asap. As you can see on the bug there's not much indicating that'll happen anytime soon though.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Orphan libraries in my system
On Sunday 02 December 2007 10:24:50 Mick wrote: Should I remove the lot? (gcc-3.3.4 was unmerged from my system years ago). Yes. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Building all packages except gcc
On Monday 19 November 2007 16:35:17 Dan Farrell wrote: How would I go about rebuilding all installed packages, except gcc? I suppose I could do emerge --emptytree world, but that would also merge gcc, which I don't want, because I want to be sure that the whole system is rebuilt with the same compiler. Any help is appreciated... Maybe I missed something, but I believe the easy (and most correct) answere would be /etc/portage/package.mask. Mask out package versions greater than your current gcc version, and portage will refuse to update it (or proceed with any updates that depend on the newer version, so don't forget about having edited the file!). package.mask is useless when you use --emptytree. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] The generated cache was invalid...
On Tuesday 20 November 2007 00:40:16 Iain Buchanan wrote: I've been seeing this error during every emerge for a while now, and I don't know where it comes from, or what it means: * Updating desktop mime database ... * Updating shared mime info database ... * Updating icons cache ... The generated cache was invalid. [ !! ] This comes from the gnome2-utils eclass. It runs gtk-update-icon-cache -qf on the dirs in /usr/share/icons/. Running emerge with --debug should tell you which dir it's trying to update when it emits the error message.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] dev-libs/apr-0.9.12 pulled back by revdep-rebuild
On Saturday 17 November 2007 06:17:26 de Almeida, Valmor F. wrote: How do I find out why apr-0.9.12 and apr-util-0.9.12 are pulled back when using revdep-rebuild? What I have currently is: This could be bug #189720 which would mean you need to manually remerge slot 1 of apr and apr-util. [1] https://bugs.gentoo.org/show_bug.cgi?id=189720 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] How do I pass options to emerge
On Tuesday 13 November 2007 16:15:31 Neil Bothwick wrote: On Tue, 13 Nov 2007 16:37:02 +0200, Rumen Yotov wrote: The package netqmail allows me to specify the value QMAIL_CONF_SPLIT before it compiles. How do I actually set this value? How do I set optional flags when I build a package? Do i need to place them in a file somewhere or can I export QMAIL_CONF_SPLIT= in my shell prior to invoking emerge? -- There're two ways to make this happen: 1.EXTRA_ECONF=qmail_conf_split emerge qmail -av 2.Hack the ebuild in your local-overlay (/usr/local/portage/net-mail/netqmail). 3. In this case, where an env var needs to be set ENVVAR=blah emerge -av foo 4. mkdir -pv /etc/portage/env/net-mail \ echo 'ENVVAR=blah' /etc/portage/env/net-mail/netqmail That way it will be set at every emerge of that one package only.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] glibc unmerged by accident
On Tuesday 13 November 2007 15:35:37 Dirk Heinrichs wrote: Am Dienstag, 13. November 2007 schrieb ext Florian Philipp: busybox ash And then? I guess Python will still not work - no emerge. Then unpack the binpkg (which is needed either way as mentioned in other replies) on /. After that put the binpkg in $PKGDIR/All/ and `emerge --usepkgonly glibc` to tell portage about the change... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] python-2.4 is it safe to unmerge it ?
On Sunday 11 November 2007 18:28:15 Helmut Jarausch wrote: I'm using a Python-2.5 based system since a year now. (currently python-2.5.1-r3) I've run python-updater and revdep-rebuild. Is it safe to unmerge python-2.4.3-r1 (in my case)? Yes. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] CFLAGS for Pentium Dual Core E2160 ?
On Sat, Nov 10, 2007 at 02:18:46PM +, David W Noon wrote: If you change the CHOST, CFLAGS and CXXFLAGS variables, then do: emerge -e system emerge -e world you will have converted everything except your kernel to 64-bit. Clearly you are completely clueless. Do *NOT* do this. It will break your system. It is not possible to run any 64 bit code with a 32 bit kernel. There is a guide [1] you must follow very carefully if you ever need to change your CHOST between two x86 CHOSTs. Even that has potential to break your system if you are not careful. If you need to switch from an x86 CHOST to an x86_64 CHOST there is no sane method short of a reinstall. [1] http://www.gentoo.org/doc/en/change-chost.xml -- Bo Andresen -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Deluge of emerge failures (ACCESS DENIED open_wr: /etc/passwd)
On Friday 09 November 2007 21:55:20 [EMAIL PROTECTED] wrote: [ ok ] ACCESS DENIED open_wr: /etc/passwd ... ACCESS DENIED open_wr: /etc/passwd --- ACCESS VIOLATION SUMMARY --- LOG FILE = /var/log/sandbox/sandbox-dev-libs_-_glib-2.14.3-11618.log open_wr: /etc/passwd ... open_wr: /etc/passwd [SNIP] emerge --info Portage 2.1.3.19 (default-linux/amd64/2006.1, gcc-4.2.2, glibc-2.7-r0, 2.6.23-gentoo-r1 x86_64) [SNIP] sys-apps/sandbox:1.2.18.1 https://bugs.gentoo.org/show_bug.cgi?id=196720 ? Upgrade sandbox to -r2 ? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Deluge of emerge failures (ACCESS DENIED open_wr: /etc/passwd)
On Friday 09 November 2007 23:11:26 [EMAIL PROTECTED] wrote: On Fri, Nov 09, 2007 at 10:36:48PM +0100, Bo ?rsted Andresen wrote: https://bugs.gentoo.org/show_bug.cgi?id=196720 ? Upgrade sandbox to -r2 ? Any idea how to actually do that, since the bug prevents me from upgrading? Is there some temp hack I can edit or patch to get around the problem for now? The bug itself isn't described very clearly, not the comments with it. FEATURES=-sandbox emerge -1 sandbox -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Deluge of emerge failures (ACCESS DENIED open_wr: /etc/passwd)
On Saturday 10 November 2007 00:13:56 [EMAIL PROTECTED] wrote: I just tried and got this: checking for C compiler default output file name... configure: error: C compiler cannot create executables See `config.log' for more details. and the log file says gcc-config error: i686-pc-linux-gnu-gcc wrapper: Could not determine which compiler to use. Invalid CTARGET or CTARGET has no selected profile. This is unrelated. eselect-compiler really sucked.. :p https://bugs.gentoo.org/show_bug.cgi?id=135688 https://bugs.gentoo.org/show_bug.cgi?id=182361#c21 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Elibc GNU userland...
On Tuesday 06 November 2007 21:18:30 pk wrote: Can someone in the know explain what this means? I googled and saw that GNU userland is related to Gentoo/BSD. Not really. Gentoo/GNU/Linux uses a GNU userland. Gentoo/*BSD uses a BSD userland.. My guess would be that the Elibc is also BSD related. I'm running a Gentoo/GNU/Linux-system... Gentoo/GNU/Linux uses a glibc ELIBC. Gentoo/FBSD uses FreeBSD ELIBC. Other alternatives include uclibc.. Why would sed be emerged with -GNU and tar plus others be (+)GNU? (-GNU%*) means the conditional was removed from IUSE since the last time you installed the package. (GNU%*) means it was added to IUSE. IUSE records all conditionals that an ebuild can use. As you can read in the discussion zmedico refers to USERLAND, ELIBC, ARCH and KERNEL, however, gets treated specially, which means an ebuild can have conditionals on them without recording it in IUSE. Therefore the addition or removal of either of those variables may not change anything at all to the build which is why it's only a cosmetic change.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] emerge openssl fails.
On Monday 29 October 2007 12:28:01 David Harel wrote: After re-emerge glibc, same problem. Attached the log file in case I missed a change with the problem. Just too bad you attached a log for the wrong package... Anyway, maybe downgrading linux-headers just wasn't such a good idea... Also.. Stop top-posting. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Writing an ebuild, #gentoo-dev-help best place?
On Sunday 28 October 2007 19:55:13 Peter Alfredsen wrote: On Sunday 28 October 2007, Grant wrote: I'm writing an ebuild and need some help. Is #gentoo-dev-help the best place to get it? [...] I've been checking it all morning here: http://cgiirc.blitzed.org/ and it always seems to be empty. Is that a real client? If it show it's empty, then it's a non-functioning one. #gentoo-dev-help has 72 users right now. It's not exactly non-functioning. It's just on the wrong irc network. #gentoo-dev-help is on freenode.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Question re desktop-file-utils
On Friday 26 October 2007 22:19:23 Anthony E. Caudel wrote: On several emerges I see the message Install dev-util/desktop-file-utils, if you want to help to improve Gentoo. Can't really tell what this package does but how does it help Gentoo? It validates .desktop files. If installed Gentoo uses this to print an elog message for .desktop files that doesn't follow the specs. Obviously it only helps Gentoo if you make sure it gets reported at bugs.gentoo.org whenever you encounter messages about invalid .desktop entries in packages in the tree... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] libxml++config.h missing - what package should provide the file?
On Thursday 25 October 2007 09:14:15 Alexander Skwar wrote: I'm trying to compile gnome-extra/assogiate-0.2.0, and it fails, because it cannot find libxml++config.h: In file included from /usr/include/libxml++-2.6/libxml++/exceptions/parse_error.h:25, from ../libassogiate/mime-type.hh:30, from mime-package.hh:26, from mime-package.cc:24: /usr/include/libxml++-2.6/libxml++/exceptions/exception.h:28:28: error: libxml++config.h: No such file or directory Could somebody please tell me, which package should provide this file? Clearly it's supposed to be provided by the same package that provides /usr/include/libxml++-2.6/libxml++/exceptions/exception.h, dev-cpp/libxmlpp. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Huge problem
On Sunday 21 October 2007 18:23:16 Kenneth Prugh wrote: At the end running etc-update I found 26 config files to upgrade . Surely I did something wrong (to upgrade udev I had to unmerge coldplug) and the result has been a failure in booting gentoo. Here is the message: /sbin/rc: line 400: start: command not found Failed to start /etc/init.d/checkroot One or more critical startup scripts failed to start! Please correct this, and reboot ... Any suggestion? Should I reinstall everything? Try re-emerging sys-apps/baselayout. You may also want to have a look at --noconfmem in `man emerge`. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] memoir blocking tetex?
On Friday 19 October 2007 11:22:00 Jules Colding wrote: I'm trying to emerge memoir but it is blocking its own dependency - tetex: ## omc-2 ~ # emerge -va dev-tex/memoir These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild N] app-text/tetex-3.0_p1-r4 USE=X -Xaw3d -doc -lesstif -motif -neXt -tk 0 kB [ebuild N] dev-tex/memoir-20060211 1,720 kB [blocks B ] dev-tex/memoir (is blocking app-text/tetex-3.0_p1-r4) Total: 2 packages (2 new, 1 block), Size of downloads: 1,720 kB ## What can I do about that? Theoretically you could downgrade tetex to 2.x and mask tetex 3.x. I doubt you want to do that though since the reason for the block is that memoir is part of tetex 3.x... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] etc-udpate error after emerge --sync and portage update
On Wednesday 17 October 2007 16:57:59 Rafael Barrera Oro wrote: Since last emerge --sync and portage update i get the following when trying to attempt an etc-update. diff: %file1: No such file or directory diff: %/var/tmp/etc-update-19194/.diff-test-2: No such file or directory ERROR: 'diff -uN %file1 %file2' does not seem to work, aborting https://bugs.gentoo.org/show_bug.cgi?id=196278 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] portage sync error
On Monday 15 October 2007 11:00:03 Stefán István wrote: We have a pc that hasn't been updated for a long time ago, and now I tried to update it's portage with the emerge --sync command, but I got the following error: Updating Portage cache: 89%!!! Cannot resolve a virtual package name to an ebuild. !!! This is a bug, please report it. (virtual/blas-1.0) http://www.gentoo.org/proj/en/portage/doc/common-problems.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] portage sync error
On Monday 15 October 2007 12:29:42 Stefán István wrote: Thanks for the tip, the manual update was successful. Now I have another problem, when I try to update portage with the emerge command: # emerge -pv portage !!! Problem with sandbox binary. Disabling... !!! Problem with sandbox binary. Disabling... Start by fixing your sandbox. [...] [blocks B ] dev-lang/python-2.3.6-r2 (is blocking app-admin/python-updater-0.2) [...] How can I resolve this blocking? Make sure you read comment 1. https://bugs.gentoo.org/show_bug.cgi?id=192381 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] whoa, this new(ish) portage is nice
On Wednesday 10 October 2007 20:53:51 Christopher Copeland wrote: On 10 Oct 2007, at 13:57, Mark Shields wrote: I was thinking the same. I remember when I first started using Gentoo (2004?) when doing updates, I always wished it would spit out the notices at the end instead of every emerge. Imagine my surprise when I saw they had implemented that. Good show, Gentoo devs! Am I missing something, I don't notice anything new.. I configured the PORTAGE_ELOG settings many versions of portage ago, so maybe that has something to do with it? I already get all the elog stuff mailed to me for each ebuild which seems like an easier way to manage that information but I am still curious! Yeah, PORTAGE_ELOG_SYSTEM=save_summary echo is now the default when PORTAGE_ELOG_SYSTEM is unset in /etc/make.conf. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] which binutils?
On Tuesday 09 October 2007 02:56:06 W.Kenworthy wrote: Something that has bothered me ever since I had some version inconsistencies way back in the dark days ... now a new binutils update is in (not listed below) bunyip ~ # binutils-config -l [1] i686-pc-linux-gnu-2.16 [2] i686-pc-linux-gnu-2.16.1 * [3] i686-pc-linux-gnu-2.17 [4] i686-pc-linux-gnu-2.18 [5] i686-pc-linux-gnu-2.18.50.0.1 On a mixed stable/unstable x86, which binutils should one generally use? So why do you use multislot at all? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] which binutils?
On Tuesday 09 October 2007 04:12:10 W.Kenworthy wrote: On a mixed stable/unstable x86, which binutils should one generally use? So why do you use multislot at all? On a previous job (long ago now) I was switching gcc's and needed multislot. Now I dont. I presume if I remove the flag, and 'emerge world -NuDv' it will clean up after itself properly - or is it better to leave as is (critical machine :( If needed you can enable multislot for gcc in package.use without enabling it for binutils.. I'm not sure it cleans it on it's own. You may have to do it manually. Basically you just disable multislot for binutils, remerge it and then unmerge all other versions than the one you remerged... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] libusb upgraded failed with USE doc
On Saturday 06 October 2007 11:32:31 Dale wrote: Do you need the doc USE flag? Unless you are coding with the package, you normally do not need it. The flag enables extra documentation, not the standard man/info pages. Is this normally true for other packages? I have doc set in my USE line and I thought it was for the man/info pages. Yes. man and info pages must be installed regardless of USE=doc. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] libusb upgraded failed with USE doc
On Saturday 06 October 2007 21:01:06 Neil Bothwick wrote: Do you need the doc USE flag? Unless you are coding with the package, you normally do not need it. The flag enables extra documentation, not the standard man/info pages. It was a mistake when I started use gentoo at the very beginning, now if I remove the doc USE flag globally, most of the packages in my system will need a rebuild, and that's take a very long time. You could set it in /etc/portage/package.use for individual packages as they come up for upgrade, then change it globally and rebuild when you have the time. It's not as though the system can't be used while emerging and not all packages use this flag anyway. Or he could stop using --newuse for a while... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] How does module-rebuild know what's installed?
On Saturday 06 October 2007 00:13:20 Randy Barlow wrote: Howdy, every time I try to run module-rebuild on one of my systems it tries to emerge nvidia-legacy-drivers, which is no longer in portage. It's no big deal because I can obviously emerge the packages that need to be rebuilt manually, but I wanted to try and remove whatever is causing it to try to emerge this package. How can I fix this? Thanks! Eh.. `module-rebuild del x11-drivers/nvidia-legacy-drivers` ? But to answer the question in the subject field packages that inherit linux-mod get added to the moduledb (which happens to reside in /var/lib/module-rebuild/moduledb) by that eclass. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] insatiable revdev-rebuild
On Thursday 04 October 2007 05:37:19 Allan Gottlieb wrote: An emerge (of openssl, I believe, but am not sure) a few days ago triggered a request for me to run # revdep-rebuild --library libcrypto.so.0.9.7 # revdep-rebuild --library libssl.so.0.9.7 I have done so. The revdep-rebuild for libssl found nothing, but the one for libcrypto rebuilt openssl. However rerunning the command again again rebuilt openssh. A msg had explained that this is possible but didn't suggest that the request would never end. I have run the revdep-rebuild for libcrypto 4 times and it keeps rebuilding openssl What should I do to fix this problem? It also told you to remove lib{crypto,ssl}.so.0.9.7 after running those revdep-rebuild commands. revdep-rebuild finds that libssl.so.0.0.7 links against libcrypto.so.0.9.7 and the hack (preserve_old_lib from eutils.eclass) that the openssl ebuild uses to preserve those libraries until you've done this makes it look like they belong to the new version of openssl (even though they really don't).. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] undefined reference to `LINUX_TARGET_OS_CPP_BUILTINS'
On Wednesday 03 October 2007 23:55:49 Robert Szentmihalyi wrote: Building gcc for ARM with # crossdev --target arm-softfloat-uclinux-gnu Which means the you are more likely to get help on the gentoo-embedded mailing list. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] insatiable revdev-rebuild SOLVED
On Thursday 04 October 2007 16:07:16 Norman Rieß wrote: It also told you to remove lib{crypto,ssl}.so.0.9.7 after running those revdep-rebuild commands. [SNIP] There was no remove command It used to tell you. ;) https://bugs.gentoo.org/show_bug.cgi?id=159245 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] per-ebuild compil options
On Wednesday 03 October 2007 02:38:42 Iain Buchanan wrote: On Tue, 2007-10-02 at 09:56 +0100, Neil Bothwick wrote: mkdir -p /etc/portage/env.d/sys-devel [SNIP] hey that looks cool... except that it didn't work! I should be doing this to dev-libs/glib and sys-libs/glibc right? $ cat /etc/portage/env.d/dev-libs/glib It should have been in /etc/portage/env/ (no .d). -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Upgrading the kernel
On Wednesday 03 October 2007 17:44:52 Jed R. Mallen wrote: Thank you to all who responded. `make oldconfig` works as usual without the worries. I was a bit apprehensive because of the gentoo kernel upgrade guide warning about using oldconfigs but turns out it's safe afterall. I was kind of surprised to see that the gentoo kernel upgrade guide does indeed warn about `make oldconfig` (which isn't the same as warning about reusing old configs). So after digging a bit it turns out dsd has an explanation in his devspace.. ;) http://dev.gentoo.org/~dsd/make_oldconfig.htm -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Re: star
On Wednesday 26 September 2007 10:59:00 Neil Bothwick wrote: Pardon? tar xf somefile doesn't do any compression at all. I don't get what you mean. No, but it does do whatever decompression is required. Of course, you do have to specify a compression method when creating a compressed archive. Heh. unpack() doesn't even take advantage of that. It also doesn't use -j but it does use z to decompress tar.gz files.. ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] shared portage tree
On Monday 24 September 2007 13:36:34 Daniel Iliev wrote: Yes. This setup works here also and I have no breakages, but since I've let the x86 machies use shared portage tree I have the feeling the performance of portage dropped down notably. I don't know what happens at the end of emerge --sync, what the metadata consists of and if it is the same for each arch and installation. Should I do emerge --regen after syncing on each machine or something like that? If anything you should do emerge --metadata. If transfers the metadata/cache to /var/cache/edb. If you never modify eclasses in /usr/portage you can alternatively use the metadata_overlay cache module... http://www.gentoo.org/proj/en/portage/doc/faq.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] gcc 4.2.0 with java
On Saturday 22 September 2007 00:35:22 Andrew Gaydenko wrote: After adding 'sys-devel/gcc gcj' string to 'package.use' file and reemergeing (twicely) the gcc package 'revdep-rebuild' still shows: * Checking dynamic linking consistency [ 49% ] * broken /usr/lib64/gcc/x86_64-pc-linux-gnu/4.2.0/libgij.la (requires /usr/lib/../lib64/libgcj.la) [ 50% ] * broken /usr/lib64/gcj-4.2.0/libjvm.la (requires /usr/lib/../lib64/libgcj.la) [ 96% ] * broken /usr/lib/gcc/x86_64-pc-linux-gnu/4.2.0/libgij.la (requires /usr/lib/../lib64/libgcj.la) [ 100% ] A packages claiming to use 'gcj' was emerged (and used) without problems. Have I missed something? https://bugs.gentoo.org/show_bug.cgi?id=125728#c29 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Changing CHOST
On Sunday 23 September 2007 02:13:46 David Relson wrote: Now that my old AthlonXP mobo has been replaced by an AMD 64 X2 mobo, it's time for upgrading CHOST :- According to http://www.gentoo.org/doc/en/change-chost.xml after a couple of changes to /etc/make.conf, i.e. from: USE=x86 ... CHOST=i686-pc-linux-gnu CFLAGS=-O2 -march=athlon-xp -pipe to: USE=amd64 ... CHOST=amd64-pc-linux-gnu CFLAGS=-O2 -march=x86-64 -pipe [SNIP] Changing CHOST is valid when you have e.g. an i386 CHOST and want to change it to i686. It is not an option for going from 32 bit to 64 bit. You need to reinstall. Also.. don't set x86 or amd64 in USE manually! And finally x86-64 is not a valid march... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] glibc-2.3.6 fails to compile (in a vserver)
On Thursday 20 September 2007 13:49:14 Alan McKinnon wrote: http://www.ecos.sourceware.org/ml/libc-alpha/2006-04/msg00090.html https://bugs.gentoo.org/show_bug.cgi?id=131108 ? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Meaning of bold/regular font in output from emerge
On Thursday 20 September 2007 15:07:20 Markus Schönhaber wrote: Since portage 2.1.3.9 I sometimes see package names printed with different colors and fonts in the output in emerge --ask ... Example: # emerge --ask --update --deep world [SNIP] In the above output ntp, git and openoffice are shown in light green and boldface while all other packages are shown in a darker green and a regular font face. What do the differences in font weight and colour tell me? The packages in bold are in your world file. `man 5 color.map` is probably the best reference.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] continue emerge on failure
On Thursday 20 September 2007 17:28:14 Pawel K wrote: How to force emerge to continue build with the next package in case previous build failed. Paludis = 0.25 now has support for the ideal solution to this.. :) # paludis --help | grep continue -A 4 --continue-on-failure Whether to continue after a fetch or install error if-fetch-onlyIf fetching only (default) neverNever if-satisfied If remaining packages' dependencies are satisfied always Always (UNSAFE) # paludis --version paludis 0.25.0 [...] Until 0.26.x gets released this requires paludis-scm from trunk. 0.26.x shouldn't be too far away though.. http://paludis.pioto.org -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Playsound . . . won't!
On Monday 17 September 2007 12:29:20 Mick wrote: Checking dynamic linking consistency... broken /usr/bin/playsound (requires libFLAC.so.7) broken /usr/lib/gstreamer-0.10/libgstflac.so (requires libFLAC.so.7) broken /usr/lib/gstreamer-0.8/libgstossaudio.so (requires libgstinterfaces-0.8.so.0) broken /usr/lib/libSDL_sound-1.0.so.1.0.0 (requires libFLAC.so.7) broken /usr/lib/libakode_xiph_decoder.so (requires libFLAC.so.7 libOggFLAC.so.3) broken /usr/lib/gstreamer-0.8/libgstossaudio.la (requires /usr/lib/libgstinter faces-0.8.la) broken /usr/lib/libakode_xiph_decoder.la (requires /usr/lib/libOggFLAC.la) done. (/root/.revdep-rebuild.3_rebuild) Assigning files to packages... /usr/bin/playsound - media-libs/sdl-sound /usr/lib/gstreamer-0.10/libgstflac.so - media-plugins/gst-plugins-flac /usr/lib/gstreamer-0.8/libgstossaudio.so - media-plugins/gst-plugins-oss /usr/lib/libSDL_sound-1.0.so.1.0.0 - media-libs/sdl-sound /usr/lib/libakode_xiph_decoder.so - media-libs/akode /usr/lib/gstreamer-0.8/libgstossaudio.la - media-plugins/gst-plugins-oss /usr/lib/libakode_xiph_decoder.la - media-libs/akode done. [SNIP] Dynamic linking on your system is consistent... All done. What version of gentoolkit is that? Clearly revdep-rebuild is doing something wrong. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Playsound . . . won't!
On Monday 17 September 2007 14:41:37 Alan McKinnon wrote: broken /usr/lib/libakode_xiph_decoder.la (requires /usr/lib/libOggFLAC.la) done. [SNIP] More like revdep-rebuild doesn't know how to build libOffFLAC anymore as the ebuild that put it there isn't in portage anymore or masked or keyworded or whatevered. I've seen this myself a time or three on my own machines. I assume you meant libOggFLAC. It doesn't need to know how to build that. What it needs to know is how to rebuild /usr/lib/libakode_xiph_decoder.la (and clearly it does)... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Playsound . . . won't!
On Monday 17 September 2007 23:05:29 Mick wrote: What version of gentoolkit is that? Clearly revdep-rebuild is doing something wrong. Fri Mar 30 18:51:11 2007 app-portage/gentoolkit-0.2.3-r1 I'd try 0.2.4_pre7 then.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Dual failures in emerge -aDNvu world
Maybe attaching the snippet of the bochs log as a file would have been better. At least the wrapping of it is rather annoying... On Tuesday 18 September 2007 03:16:33 Kevin O'Gorman wrote: jade:/usr/share/sgml/docbook/dsssl-stylesheets-1.79/html/../common/../commo n/dbl1sr.dsl:3:5:E: reference to entity lat2 for which no system identifier could be generated jade:/usr/share/sgml/docbook/dsssl-stylesheets-1.79/html/../common/../commo n/dbl1sr.dsl:2:0: entity was defined here make[1]: *** [user/index.html] Error 8 make[1]: Leaving directory `/var/tmp/portage/app-emulation/bochs-2.3 /work/bochs-2.3/doc/docbook' # emerge -1 sgml-common https://bugs.gentoo.org/show_bug.cgi?id=192231 [SNIP] exceed LITLEN (240) /usr/share/sgml/docbook/utils-0.6.14/backends/pdf: line 15: pdfjadetex: command not found make[2]: *** [Xserver-spec.pdf] Error 9 make[2]: Leaving directory `/var/tmp/portage/app-doc/xorg-docs-1.4-r1 # emerge -1 jadetex https://bugs.gentoo.org/show_bug.cgi?id=191661 More docbooks/sgml rebuilds might be required.. :p -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Apache 2.2: missing KEEPENV (with solution)
On Sunday 16 September 2007 15:51:28 Michael Mauch wrote: the recent Apache 2.2.6 ebuild brought an /etc/init.d/apache2 that doesn't honour the KEEPENV variable anymore. Formerly one could preserve some of the environment variables (while all others would be unset to keep Apache's environment tidy). To resurrect that KEEPENV variable, I put in /etc/conf.d/apache2: Maybe you should post this at bugs.gentoo.org ? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Playsound . . . won't!
On Sunday 16 September 2007 00:11:24 Mick wrote: I am using playsound in KDE to play system sounds, but it has stopped working: $ /usr/bin/playsound chimes.wav /usr/bin/playsound: error while loading shared libraries: libFLAC.so.7: cannot open shared object file: No such file or directory I ran revdep-rebuild -X, but the problem persists. Any ideas? revdep-rebuild --ignore -X ? If it still persists then maybe you should show how revdep-rebuild fails.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] ERROR : sys-libs/db-4.5.20_p2 failed - econf failed
On Friday 14 September 2007 11:14:20 Jean-Philippe Caruana wrote: /opt/sun-jdk-1.6.0.02/bin/javac: error while loading shared libraries: libjli.so: cannot open shared object file: No such file or directory If might be worth bringing this up on the gentoo-java mailing list. https://bugs.gentoo.org/show_bug.cgi?id=165330 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 05:58:16 Dan Farrell wrote: berkdb = sys-libs/db. apr-util is slotted and 0.9.12 is in a different slot than 1.2.8. this might give you some idea of what packages might be effected by the nontrivial update of db: [EMAIL PROTECTED] ~ $ equery depends db [ Searching for packages depending on db... ] Heh. I'm sorry but that command will produce more noise than useful info. For instance on the laptop I'm writing this from I have 6 different versions of sys-libs/db installed. The above command will show all installed packages that *may* [1] depend on any installed version of sys-libs/db... [1] http://thread.gmane.org/gmane.linux.gentoo.user/179649/focus=179814 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 05:41:59 Mark Knecht wrote: This latest emerge pass seems to be working better. I did have to skip a couple of packages though. I'll go back and see if I can catch them with revdep-rebuild when it hopefully finishes up in the next hour. [...] db2omf: Could not construct the OMF maintainer element. Add an author, corpauthor, editor, othercredit, or publisher element with the role attribute set to maintainer to epiphany.xml. make[2]: *** [epiphany-C.omf] Error 10 make[2]: Leaving directory `/var/tmp/portage/www-client/epiphany-2.16.3/work/epiphany-2.16.3/help' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/www-client/epiphany-2.16.3/work/epiphany-2.16.3' make: *** [all] Error 2 !!! ERROR: www-client/epiphany-2.16.3 failed. I think you either need a later version of epiphany (2.18.x) or an earlier version of gnome-doc-utils (0.8.x). Since epiphany-2.18.2 is latest stable this is probably one of those cases where the -X for revdep-rebuild would make the whole difference... -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] I can't build pango
On Wednesday 12 September 2007 18:20:19 Jean-Philippe Caruana wrote: I try to build gnome on my system, but I can't build the pango dependence : [SNIP] I put the build.log here : http://perso.enstimac.fr/~caruana/perso/build.log [SNIP] here is my make.conf :# cat /etc/make.conf CHOST=i686-pc-linux-gnu [SNIP] here is my gcc config : # gcc-config -l [1] i486-pc-linux-gnu-4.1.1 * [2] i686-pc-linux-gnu-3.3.6 I'd tell you to remerge fontconfig and run revdep-rebuild but it looks to me like you have much more serious problems than the expat breakage... Your make.conf claims you're on an i686 CHOST and your gcc 3.3.6 compiler is but your current compiler apparently isn't... http://www.gentoo.org/doc/en/change-chost.xml -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] How to manage the upgrade.
On Wednesday 12 September 2007 20:59:50 econti wrote: Hi all, it's me again. Above all, thank you everybody for the help . . .but now I am a little confused. Well, here is what I understood: 0. Resolve blockages.. :p 1 - first of all: to upgrade gcc following what explained here - http://www.gentoo.org/doc/en/change-chost.xml This guide isn't for upgrading gcc. Do you really need to change your CHOST? If you do be careful when following the above guide since it has the potential to break your system bad enough that you'll have to reinstall if you aren't careful.. This is for gcc upgrades only: http://www.gentoo.org/doc/en/gcc-upgrading.xml 2 - to upgrade xorg (but I have not been able to find any usefull doc. Could you give any link, please?) http://www.gentoo.org/proj/en/desktop/x/x11/modular-x-howto.xml 3 - to upgrade the remaining pakages 4 - to install the new kernel 5 - to change profile How many mistakes did I do? P.S. I did not understand the expat matter. ;-( expat had an ABI bump which requires all programs that link against it to be rebuilt. That's a lot of programs... revdep-rebuild is supposed to take care of this but since you're going to rebuild everything anyway (due to the gcc upgrade and possibly the chost change) this won't really matter to you.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 01:36:56 Mark Knecht wrote: libtool: link: cannot find the library `/usr/lib/libgnomevfs-2.la' or unhandled argument `/usr/lib/libgnomevfs-2.la' # ls -l /usr/lib/libgnomevfs-2.la # equery check gnome-vfs ? -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 04:54:08 Mark Knecht wrote: OK, the emerge -DuN world finished up and said everything was good. I then ran revdep-rebuild and have a list of things that need to be rebuilt. One has failed - dev-libs/apr-util. checking gdbm.h presence... yes checking for gdbm.h... yes checking for gdbm_open in -lgdbm... yes configure: error: Berkeley DB not found. [...] '/var/tmp/portage/dev-libs/apr-util-0.9.12/temp/build.log'. [...] Sector9 ~ # emerge -pvD apr-util [...] [ebuild R ] dev-libs/apr-util-1.2.8 USE=berkdb gdbm ldap [...] I'm wondering what package should supply berkdb and why this ebuild doesn't force that package to get installed? berkdb = sys-libs/db. apr-util is slotted and 0.9.12 is in a different slot than 1.2.8. So try.. # emerge -va1 sys-libs/db:4.5 dev-libs/apr-util:0 If it still fails look at the config.log from apr-util to try to find why it's failing.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 05:23:45 Mark Knecht wrote: gda-firebird-recordset.c:517: error: dereferencing pointer to incomplete type make[2]: *** [gda-firebird-recordset.lo] Error 1 [...] !!! ERROR: gnome-extra/libgda-1.2.3 failed. You really ought to learn to search bugzie.. :p https://bugs.gentoo.org/show_bug.cgi?id=183465 [...} Is firebird yet another database? Yep. If you don't actually need it, turning off that use flag might work.. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Immensely disappointed in emerge -DuN world these days
On Thursday 13 September 2007 05:32:07 Mark Knecht wrote: That's very interesting. What about slotting issues? What if it's the old version in a slot that needs to be rebuilt? (FYI - I don't really understand slotting that much since I don't program. I've guessed it's because some code needs old libraries, etc., and somehow slotting takes care of that.) Slotting just allows two or more versions of a package to be installed at the same time. It's up to the maintainer to ensure that they don't interfere with each other so there's no magic in it.. Does -X install the latest version in the specific slot? It should. Unfortunately revdep-rebuild has a few bugs at the moment. So in the case of apr-util the latest stable version of revdep-rebuild will remerge the wrong version of apr-util (see bug #189720 for details). And of course latest ~arch version has a few other bugs at the moment (most of which are fixed in svn but not yet released).. :p Anyway, I've wondered at times about just removing all the specific revision numbers but I'm wary of going beyond my comfort zone and then ending up in a state that's more difficult to fix. -X will certainly do a better job than just removing all specific revision numbers manually. Maybe the revdep-rebuild guys should (could?) include -X if it's the right thing to do? The latest ~arch version does have -X as default. Once the bugs in that gets fixed it'll do a much better job than the current stable version. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Upgrade
On Tuesday 11 September 2007 16:53:58 Gary Rickert wrote: A quick question / addition to this. It's not exactly related. In the future start a new thread instead and please don't top-post.. When I am finished with an update I usually run etc-update. I was told last week that this has been phased out and dispatch-conf is better to use. Whoever told you that is just wrong. What is the difference? Read it in the handbook.. http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=3chap=4 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] why unmerge acpid ?
On Wednesday 12 September 2007 00:34:51 Jesús Guerrero wrote: [ebuild R ] gnome-base/gnome-applets-2.18.0-r2 USE=acpi apm gnome hal ipv6 -debug -doc -gstreamer 0 kB [...] if use acpi ! use hal ; then [...] That message appears when you emerge gnome-applets. And it is telling you from first hand that you *need* to emerge acpid _by hand_ if you want all the acpi based applications to behave correctly. As you can see from the conditional it shows up if you have acpi enabled and hal disabled. Clearly this is not the case for the OP (he has hal enabled too). -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] rpm-4.4.6-r3 install error
On Wednesday 05 September 2007 14:25:09 emerge.gentoo wrote: rpmdav.c:526: error: `ne_propfind_handler' undeclared (first use in this function) rpmdav.c:526: error: `pfh' undeclared (first use in this function) # emerge -1 neon https://bugs.gentoo.org/show_bug.cgi?id=187740 -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] rpm-4.4.6-r3 install error
On Thursday 06 September 2007 10:14:17 Alan McKinnon wrote: I tried to emerge the r4 version of the same package and got same error as before. [SNIP] I suggest you file a bug As said in another reply a bug already exists (#187740). so the gentoo maintainer can help get to the bottom of this. What gentoo maintainer? ;) -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] rpm-4.4.6-r3 install error
On Thursday 06 September 2007 12:25:10 Alan McKinnon wrote: so the gentoo maintainer can help get to the bottom of this. What gentoo maintainer? ;) Ah yes, I kinda just assumed that *someone* was maintaining the ebuild. Then I realised that we are talking about rpm. So, did I make a really really stupid assumption? If you look in metadata.xml in /usr/portage/app-arch/rpm/ (or use e.g. herdstat to query it for you) you'll see that the maintainer is 'maintainer-needed'. So yes, currently it has no maintainer. In my latest sync 498 out of 11901 packages have that status. -- Bo Andresen signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] portage update problem
On Thursday 06 September 2007 17:22:15 Rafael Barrera Oro wrote: Hello, just installed from gentoo LiveCd 2006.0 Why are you using such an ancient stage? after an emerge --sync, i upgraded portage, everything was running smooth until the unmerging of the previous portage version started, this ended abruptly with the error: *[portage-2.0.54] bash: /usr/lib/portage/bin/ebuild.sh: No such file or directory *after that, whenever i invoke the emerge command in any manner, emerge package, emerge sync, etc i get: Traceback (most recent call last): File /usr/bin/emerge, line 37, in ? from output import blue, bold, colorize, darkblue, darkgreen, darkred, green, \ ImportError: cannot import name colorize i am really baffled by this, and since what is failing its portage itself i am really power less to solve this since i am a little short of ideas i cry to you all for help Bug #172015 ? # cd /usr ln -s lib64 lib As per comment 4 on bug #170178 there are hardened amd64 multilib stages from 2007.0. -- Bo Andresen signature.asc Description: This is a digitally signed message part.