Re: [gentoo-user] emerge --resume question
On Monday 03 April 2006 20:39, Peter Ruskin wrote: On Monday 03 April 2006 12:06, Peter Campion-Bye wrote: Hi, Trying to do an 'emerge -e world' I come up against the odd package that doesn't build without some intervention. I know I can do 'emerge --resume --skipfirst' and carry on, but sometimes the broken package could be built ok with a little intervention, for example, my build breaks on pilot-link, which can be simply fixed with 'FEATURES=-sandbox emerge pilot-link'. Problem is, I can't do a resume after this as the emerge list from the world build has been blatted when I emerged pilot-link. So, is there any way I can preserve the state of the world build, so that after building the breaking package I can do a --resume --skipfirst and carry on with the world build? Any other suggestions for workarounds? I'm just experiencing the same thing - with over 800 ebuilds this is an important question. I'm noting the problem ebuilds in a log file manually, so I can see to them when it's all over. This is resilient to the single merge between resumes case in 2.1_pre. For 2.0, you can manually back up and restore /var/cache/edp/mtimedb. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge --resume question
On Monday 03 April 2006 22:12, Peter Ruskin wrote: On Monday 03 April 2006 13:12, Jason Stubbs wrote: I'm just experiencing the same thing - with over 800 ebuilds this is an important question. I'm noting the problem ebuilds in a log file manually, so I can see to them when it's all over. This is resilient to the single merge between resumes case in 2.1_pre. For 2.0, you can manually back up and restore /var/cache/edp/mtimedb. Sorry Jason, I don't understand. How would I backup and restore /var/cache/edp/mtimedb and what would that do? It does as the opening poster asked: So, is there any way I can preserve the state of the world build, so that after building the breaking package I can do a --resume --skipfirst and carry on with the world build? The not very aptly named mtimedb file contains the resume list among other things. Hence, copying it somewhere else before fixing a broken build and copying it back after will preserve the resume list. Don't do it for more than single ebuild maintenance though as the file contains other things too. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge only pkg not dependdencies...
On Wednesday 29 March 2006 21:39, Meino Christian Cramer wrote: Hi, I looked through the manpage of emerge to find an option to tell emerge just to install a certain package and to skip all dependencies. --nodeps I only find the opposite of this: Skip the package and install all dependencies. --onlydeps Is there a way to install just the package and forghet all dependencies ? See above. ;) -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Aborting due to QA concerns
On Thursday 30 March 2006 02:17, JimD wrote: I did an emerge -vb gnome and I am almost finished. However I am now getting this weird error from gnome-netstatus: What makes portage issue an Aborting due to QA concerns 103 files installed in /var/tmp/portage/gnome-netstatus-2.12.0/image///var/tmp/portage/gnome-netstatus-2.12.0/image/ If you look at that path /var/tmp/portage/gnome-netstatus-2.12.0/image/ is duplicated which means the final directory that portage would install to would be the directory it's installing from. This is a bug that needs reporting (if hasn't been already). -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] [OT] Terminal formatting and colors escape sequences
On Thursday 02 March 2006 22:49, Bo Andresen wrote: My problem is that I am unable to locate a reference that defines the escape sequences. Google for xterm escape sequences yields many results. s/xterm/ansi/ for even more. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Cannot resolve a virtual package name to an ebuild.
On Saturday 25 February 2006 20:03, David Corbin wrote: trombone dcorbin # emerge -uav portage These are the packages that I would merge, in order: Calculating dependencies -!!! Cannot resolve a virtual package name to an ebuild. !!! This is a bug, please report it. (virtual/libintl-0) -- Is there someone way I can work around this? I don't seem to be able to emerge anything. Update portage. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Cannot resolve a virtual package name to an ebuild.
On Saturday 25 February 2006 20:03, David Corbin wrote: trombone dcorbin # emerge -uav portage These are the packages that I would merge, in order: Calculating dependencies -!!! Cannot resolve a virtual package name to an ebuild. !!! This is a bug, please report it. (virtual/libintl-0) -- Is there someone way I can work around this? I don't seem to be able to emerge anything. Hrmm.. Sorry. # export PORTDIR=$(portageq envvar PORTDIR) # mv $PORTDIR/virtual ./virtual.backup # emerge --oneshot portage # mv ./virtual.backup $PORTDIR/virtual -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge error
On Friday 24 February 2006 23:56, John Fawcett wrote: Calculating dependencies /!!! Cannot resolve a virtual package name to an ebuild. !!! This is a bug, please report it. (virtual/perl-Storable-2.13) Update portage. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage
On Monday 30 January 2006 02:57, Tony Davison wrote: Portage -2.1_pra4-r1 has a new dependency of dev-python/pycrypto-2.0.1, anyone know why? I only askbecausi it is new and nothing else (on this box) depends on it, so I wondered what it would break. It's to support non-MD5 hashing algorithms. If you have a look at portage's Manifest, you'll see MD5, RMD160 and SHA256 hashes for all files. These are generatable and verifiable from 2.1_pre4. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Strangeness on a fresh install
On Thursday 26 January 2006 11:49, Allan Spagnol Comar wrote: Hi all, I am having a real strange problem on a fresh install, I had used a stage3 tarball from internet and when I try to emerge syslog-ng I got this error checking for C compiler default output...configure: error: C compiler cannot create executables See config.log for more details. I had a look at config.log and do not understand nothing can someone help with that one configure:2328: i686-pc-linux-gnu-gcc -O3 -march=pentium3 -pipe -fomit-frame-pointer -msse -mmmx -mfpu conftest.c 5 cc1: error: invalid option `fpu' These are the important lines. By the look of it you have -mfpu in your CFLAGS which, as the following output says, is not a valid option (unless you are running SPARC). -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] [POLL] portage-2.1 USE flag ordering
On Wednesday 25 January 2006 03:43, Philip Webb wrote: 060124 Jason Stubbs wrote: I've started a poll on the specific question of USE flag ordering in portage-2.1_pre3 at http://forums.gentoo.org/viewtopic-t-426033.html 1) Create a login if you don't have one 2) Log in 3) View the above page 4) Click your choice 5) Click the submit button I have the above URL showing the Forum topic on Konqueror 3.5 there is no choice to click nor any submit button: yes, I logged in. Strange. Well, you should be able to find it near the top of the Portage Programming forum. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge files sometimes just abort without error
On Friday 20 January 2006 07:33, Mike Owen wrote: On 1/19/06, Daevid Vincent [EMAIL PROTECTED] wrote: I've noticed lately that when I do an emerge -Davu world (for example), that some of the ebuilds just abort part way through compilation. No error. No anything. I'm just sitting there at a command prompt like nothing happened. What gives? I've been seeing the same thing occasionally for the last 2-3 weeks I'd say. I'm still trying to track it down, so I haven't filed a bug report yet. About the only ~arch software on my box is modular X and it's deps, although I think it may have started before I did that upgrade. http://bugs.gentoo.org/show_bug.cgi?id=104705 -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] KDE 3.5 : Kicker other ebuilds fail
On Monday 02 January 2006 00:02, Philip Webb wrote: Before I submit a bug report, has anyone had a similar experience ? Does anyone have anything to suggest to try first ? I started to install KDE 3.5 (split) -- still ~x86 -- got Kdelibs done, then went on to Kdebase-startkde, which pulls in 17 dependencies. Some of them went thro' ok, but then Kicker refused as below; I tried others Khelpcenter Kdebase-kioslaves Kdesu Ksplashml failed at the same point. I took time out to get my Docbook pkgs upto-date, but no change. How do you define refused? Did it error out and drop you back to a prompt? If so, what were the last 20-30 lines of output before make stated there was an error? -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Saturday 31 December 2005 21:57, Holly Bostick wrote: If you look at the output [nomerge ] sys-libs/db-4.2.52_p2-r1 -bootstrap +doc +java -nocxx +tcltk the reason db is calling for java is because you have the java USE flag set for db. Do you really need db to use Java? If not, disable the flag (# echo 'sys-libs/db -java' /etc/portage/package.use); problem solved. All the way up until the next package which depends on java. Only the first package that is came across is listed as the parent. sys-libs/db doesn't call for any specific version of java. The complaint is that sun-jdk-1.5 is installed but emerge is wanting to install sun-jdk-1.4. This indicates that sun-jdk-1.5 is likely masked. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Saturday 31 December 2005 23:53, Holly Bostick wrote: Jason Stubbs schreef: On Saturday 31 December 2005 21:57, Holly Bostick wrote: Anyway, my latest emerge world failed because of sun-jdk-1.4.2.10 (!!!). The current one is 1.5 something. Meaning (to me) that Kevin is referring to the current *available* *version* of sun-jdk, not to any version he might have installed (and my impression is that he does not in fact have any version of sun-jdk installed), and he's just concerned that an out-of-date version will be installed rather than the latest. Okay, that could be taken either way. I'm sure that all possibly relevant information is now available though. :) -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Sunday 01 January 2006 02:39, Holly Bostick wrote: And iirc, the virtual is not going to be linked to a hard-masked package (or at least it most likely is not atm, or the hard-masked packages are listed after the stable packages). So what I would do is check /var/cache/edb/virtuals and see what the listing for jdk actually is. Ok, you're right off the track here. /var/cache/edb/virtuals hasn't been used (at all) since 2.0.50. Packages that provide virtuals are looked up when emerge is run and the internal list of those is essentially arbitrary. Virtuals listed in /etc/make.profile (and parents) are put after those in correct order. Virtuals listed in /etc/portage/profile/virtuals are put at the front. When there is a virtual dependency, these are checked in order for the first unmasked package that is able to satisfy the dependency. The fact that sun-jdk is being chosen means that it is the first package to have any unmasked versions available from the internal list. sun-jdk-1.4* being chosen means that sun-jdk-1.5* is still masked. sun-jdk-1.5* is in package.mask so adding dev-java/sun-jdk (or even =dev-java/sun-jdk-1.5*) to /etc/portage/package.unmask should fix that issue. It also has a KEYWORDS of ~x86 ~amd64 so adding the same to /etc/portage/package.keywords should fix that issue. Beyond that there shouldn't be a problem. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Sunday 01 January 2006 02:57, Kevin O'Gorman wrote:
Oh, and yes, java is ~x86 both in unmask and in keywords.
Can you list exactly what sun-jdk lines you have in
/etc/portage/package.{unmask,keywords} please?
--
Jason Stubbs
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Saturday 31 December 2005 13:58, Kevin O'Gorman wrote: Or should I say portage as a whole. Anyway, my latest emerge world failed because of sun-jdk-1.4.2.10 (!!!). The current one is 1.5 something. The weird part is that I cannot find any reason for that package to be emerged. 1) the world file contains just sun-jdk 2) equery depends =sun-jdk-1.4.2.10 shows nothing at all 3) sun-jdk does not appear at all in /etc/portage/package.* files How should I explore this further? Whatever command you used to update, add --tree --pretend to it. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge going weird
On Saturday 31 December 2005 13:58, Kevin O'Gorman wrote: Or should I say portage as a whole. Anyway, my latest emerge world failed because of sun-jdk-1.4.2.10 (!!!). The current one is 1.5 something. The weird part is that I cannot find any reason for that package to be emerged. On Saturday 31 December 2005 16:02, Kevin O'Gorman wrote: treat ~ # emerge -aDvu --tree world ... [nomerge ] sys-libs/db-4.2.52_p2-r1 -bootstrap +doc +java -nocxx +tcltk [ebuild NSF ] dev-java/sun-jdk-1.4.2.10 +X +alsa -browserplugin +doc -examples -jce +mozilla +nsplugin 35,592 kB sys-libs/db doesn't specifically rely on a version less than 1.5. Perhaps, the problem is: $ emerge -p =sun-jdk-1.5* ... !!! All ebuilds that could satisfy =sun-jdk-1.5* have been masked. !!! One of the following masked packages is required to complete your request: - dev-java/sun-jdk-1.5.0.06 (masked by: package.mask) # [EMAIL PROTECTED] # Lotsa things in the tree don't compile with 1.5 yet # 1.5 defaults too -target 1.5 making downgrading to a 1.4(/1.3) # impossible, see bug #69970 and bug 65937 for more information/discussion # http://www.gentoo-wiki.com/Java_FAQ Do you have sun-jdk in /etc/portage/package.unmask? -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] either emerge or portage is completely broken
On Friday 30 December 2005 14:27, Robert Persson wrote: I ran emerge sync earlier this evening and I got the following at the end of the output: Performing Global Updates: /usr/portage/profiles/updates/4Q-2005 (Could take a couple of minutes if you have a lot of binary packages.) .='update pass' *='binary update' @='/var/db move' s='/var/db SLOT move' S='binary SLOT move' p='update /etc/portage/package.*' .Traceback (most recent call last): File /usr/bin/emerge, line 10, in ? import portage File /usr/lib/portage/pym/portage.py, line 7480, in ? do_upgrade(mykey) File /usr/lib/portage/pym/portage.py, line 7385, in do_upgrade key=dep_getkey(line.split()[0]) File /usr/lib/portage/pym/portage.py, line 3609, in dep_getkey if mydep[-1]==*: IndexError: string index out of range I now get the same error messages when I try to emerge anything, including portage. In other words portage is completely broken. What should I do? One of your /etc/portage/package.* files contains an invalid atom. By the look of the traceback, you have a line with just a * perhaps? -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] /usr/bin/emerge gone!
On Friday 23 December 2005 15:33, Michael Sullivan wrote:
I'm not sure how, but /usr/bin/emerge is gone on my server box. How can
I get it back?
# tar jxf ${DISTDIR}/portage-2.0.53.tar.bz2
# mkdir -p /usr/lib/portage/{bin,pym}
# mv portage-2.0.53/bin/* /usr/lib/portage/pym
# mv portage-2.0.53/pym/* /usr/lib/portage/pym
# ln -sf ../lib/portage/bin/emerge /usr/bin/emerge
# emerge portage
--
Jason Stubbs
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Portage broken, can't load module cache_errors
On Sunday 18 December 2005 01:27, Graham Murray wrote: Bruno Lustosa [EMAIL PROTECTED] writes: Just synced and tried to update world. Portage updated itself, and then died when trying to exec itself. I updated from 2.0.53 to 2.1_pre1. The error is: This error has been fixed. I was going to post on the mailing list but found that the fix had made it's way here already. Luckily I did not get that error but emerge then took 'an eternity' to calculate the world dependencies before actually building any packages. I added a message about three hours ago that is displayed after portage is installed noting this. The default cache format has changed so you'll need to run `emerge metadata` to get the local cache back up to date. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] portage changelogs?
On Tuesday 11 October 2005 15:10, Graham Murray wrote: New versions of portage 2.0.X are showing up in ~arch, but the Changelog entries only mention 2.1.0_alpha versions. Where should I look for what has changed, eg between 2.0.52* and 2.0.53_rc? To prevent duplication of work, the physical ChangeLog is no longer being maintained. Instead, a ChangeLog will be generated from the subversion commit logs instead. A script will be ready and the/a ChangeLog added back to the tarballs again in time for the next _rc / final release. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge improvements... again
On Monday 10 October 2005 03:19, Jerry McBride wrote: I'm getting ready to make a post on bugs.gentoo.org to ask for a couple of feature enhancements for portage's emerge routine. Feature enhancements are on hold until the 227 currently open bugs are organized properly and the count lessened somewhat. Note, most existing feature requests have been marked as LATER so the above number is pretty much only real bugs. 1- In /etc/make.conf add a new key to FEATURES that would cause portage to cleanup the DISTDIR folder if everything emerges cleanly. Removing distfiles for packages when they are unmerged/upgraded will be added. When removing all distfiles, you'd end up downloading the same file several times when there are updates to the revision. 2- How about emerge apps that would ONLY cover things NOT in emerge sys? This has not been considered, so opening a bug for this would be okay. Adding an all target is planned, which would cover everything installed regardless of presence in system or world. With such a target, changing world to not automatically include system should be fine. 3- Make a list of hooks available to hackers, like myself. This is also planned. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] /usr/bin is such a mess
On Saturday 08 October 2005 02:01, Yrjö Hatakka wrote: kynsi bin # ls kmymoney2 tr xclock ar ar kmymoney2 tr xclock x-apps kde-apps console-apps all dumped to same directory This unix heritage or is it LSB stupidity never ceases to amaze me Perhaps something like gobolinux is for you? -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] /usr/bin is such a mess
On Saturday 08 October 2005 03:06, Yrjö Hatakka wrote: x-apps kde-apps console-apps all dumped to same directory This unix heritage or is it LSB stupidity never ceases to amaze me Perhaps something like gobolinux is for you? So you are happy with the mess? Like why is metacity in /usr/bin why should anyone have it in his path? It just pollutes directory Works for me. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] ccache taking way too long time
On Friday 30 September 2005 06:44, Fernando Meira wrote: drwxrwS--- 18 root portage 576 Sep 29 00:21 . Which portage version are you using? 2.0.51.22-r2? There was a bug in one of them that would cause ccache permissions to be updated everytime in exactly the above situation. What you have there is 2760 permissions. What you need is 2750. `chmod 2750 /root/.ccache` will fix it. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Size of portage tree
On Wednesday 28 September 2005 21:49, Michael Kjorling wrote: Don't forget that portage keeps all downloaded source code in /usr/portage/distfiles. The actual portage tree totals something like 100 MB plus losses due to inode/block size. A couple of years ago it not much more than 100MB, but nowadays... /mnt/archive/gentoo $ du -sh * 750Mdistfiles 516Mrsync -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] cdrecord-prodvd extrange behaivor
On Tuesday 27 September 2005 23:29, Mauro Sauco wrote: I run cdrecord as usual, then everything looks fine, then the LED of the drive goes on for a couple of seconds then goes off and everything continues perfectly till the end, the media is ejected, but it is not touched at all as if -dummy was active. No apparent error mesage. # cdrecord -dev=0,0,0 -speed=1 -v --eject -dao myimage.iso Not that the output would suggest that it's misinterpreting the passed parameters, but does the following yield the same? # cdrecord -v -dao -eject speed=1 dev=0,0,0 myimage.iso The man page SYNOPSIS states [general options] dev=device [track options] and has no hyphen on either speed or dev and only a single hypen on eject. As I said, it doesn't seem like it's misinterpreting from the output but it can't hurt to try... -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] deep-no, update-no, deep update-yes
On Sunday 25 September 2005 23:29, Mark Knecht wrote: Why is it that some updates are not found by --deep, not found by --update, but then found by --deep --update together? --deep doesn't work without --update. This will be addressed at some stage, although perhaps not with the functionality you'd expect. --deep without --update would only install any missing dependencies. -- Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Security Updates and Portage Trees
On Monday 19 September 2005 15:00, gentuxx wrote: does updating a package for a security fix using the --oneshot option update the same package that is housed in the world tree? There is no world tree. There is only a list. --oneshot has no affect on this list. If so, can I assume that the same package will be updated next time I update world? Meaning, if I run --oneshot for mozilla-firefox-1.0.6-r7 and mozilla-firefox-1.0.7-r1 comes out, will 1.0.6-r7 be upgraded to 1.0.7-r1? If it was in the world list prior to you running --oneshot, it'll still be in the world list afterward. Hence, it will be updated with world. If every security fix comes out with --oneshot being recommended, how do I know if it's a dependency of a package in world, or an entity in world? (This seems like an extension of the questioning above.) What does it matter in the context of a security update? Also, for the most recent firefox update, I would run the command as recommended with the -p flag, and it would see the package. If I run emerge -Dupv mozilla-firefox I only get a few of the (supposed) dependencies, and not the package itself, while the package installed (when I do emerge search mozilla-firefox) is 1.0.6-r5. If that is the case then 1.0.6-r5 is the latest version available for you with respect to your current snapshot of the tree. -- Jason Stubbs pgpgOHJHMeSrI.pgp Description: PGP signature
Re: [gentoo-user] dev-lang/python vulnerability
On Monday 19 September 2005 18:35, Paweł Madej wrote: I've done glsa-check -t new and i got that my python is vulnerable i got in world only dev-lang/python-2.4.1-r1 but it says that vulnerable is 2.3.5-r2 so i've searched for any vulnerable version on my system and i've found python-2.3.5 My question is: Could I unmerge 2.3.5 version which is not in my world tree and use only 2.4.1 or is it some dependency which have to be on system and i should only update that port to unvulnerable version? If you've run python-updater, it should be safe to unmerge python-2.3.x. -- Jason Stubbs pgpa36Z4aVCrM.pgp Description: PGP signature
Re: [gentoo-user] Security Updates and Portage Trees
On Monday 19 September 2005 13:16, gentuxx wrote: If I update firefox with the --oneshot option, I know that it won't update the world tree, but why? Why is that the recommended procedure? Does that give me any benefit? Also, why would a package be available as a --oneshot and NOT through a normal emerge -Dupv world? The package would be available through -Dupv as well, but not everybody likes to update all packages (especially on servers). I love how portage unifies the packaging system, and I feel like if I run all of these --oneshot updates for security fixes, that I'll have all of these stray programs running around on my system, that won't get updated next time I emerge world. --oneshot won't remove the package from world. It just prevents it from being added. If the package is installed but not in world, it is presumably there as a dependency from another package. Hence, updating world will still grab the package. Using --oneshot just keeps the world file clean. -- Jason Stubbs pgpJ1kBcYynH2.pgp Description: PGP signature
Re: [gentoo-user] why is Joe part of 'system' ?
On Monday 05 September 2005 04:23, Holly Bostick wrote: What I'm surprised at is that no one has said, Look in /var/cache/edb/virtuals, where you will likely see that Joe is set as (one of) the in-use virtual/editor(s) on your actual system. /var/cache/edb/virtuals is a relic from 2.0.50. In that big message displayed at the end of merging any of the 2.0.51 series that nobody reads, it states that virtuals are now calculated on the fly. -- Jason Stubbs pgpugT19rSd3S.pgp Description: PGP signature
Re: [gentoo-user] why is Joe part of 'system' ?
On Monday 05 September 2005 20:44, Holly Bostick wrote: Jason Stubbs schreef: /var/cache/edb/virtuals is a relic from 2.0.50. In that big message displayed at the end of merging any of the 2.0.51 series that nobody reads, it states that virtuals are now calculated on the fly. So I could delete the whole file without consequence? Yes. And OK, doesn't that mean that someone should submit an enhancement bug to b.g.o indicating that Portage should check whether you are emerging *one of* the packages that provides a given virtual, or whether you're unmerging *the only* package that provides a virtual (similar to the previous SLOTS issue)? Maybe it's time to start cleaning up 'big scary messages' that alarm users unnecessarily (because unmerging Joe in this instance is not going to break anything, but there are instances where it might, yet the message does not distinguish one way or another). I've already posted a patch to [EMAIL PROTECTED] for feedback as the issue came up there as well. Apparently there was another dev that was aware of the issue, but he never spoke up and I have not seen any bugs with regard to this either. Maybe it's time to start reporting? -- Jason Stubbs pgpAtTjDT7uN1.pgp Description: PGP signature
Re: [gentoo-user] Re: Following a well overdue world -u I have no network
On Wednesday 31 August 2005 18:42, Harry Putnam wrote: Nick Rout [EMAIL PROTECTED] writes: For the network problem: first thing to know - have you updated your config files with etc-update or dispatch-conf? I used dispatch-conf, because of the archiving behaviour but I see it doesn't really archive all of etc now. You've hit a bug in dispatch-conf. /etc/init.d/net.eth0 was previously a file. Now it's a symlink to net.lo. dispatch-conf fails to handle that change correctly. Hence, delete /etc/init.d/net.eth0 and recreate it as a symlink and all should be fine. -- Jason Stubbs pgp1817pdiPRP.pgp Description: PGP signature
Re: [gentoo-user] /home becoming readonly every night
On Thursday 25 August 2005 09:13, Matt Nordhoff wrote: On 08/24/05 15:59, Nick Rout wrote: For somewhere between 3 days and a week I have been rising to find that /home has become readonly overnight. Filesystem errors, maybe? That's why I was always getting read-only partitions... Run fsck on it. Also check dmesg. It should say exactly why the filesystem was made read-only. -- Jason Stubbs pgpC7yDfnCQPE.pgp Description: PGP signature
Re: [gentoo-user] PHP and files over 2 Gb
On Monday 22 August 2005 20:51, Jim Hatfield wrote: I just bought a Pinnacle ShowCenter and am using it with some Apache/PHP/MySQL code to play mpeg movie files to a TV. It appears that PHP as built can't handle files over 2Gb - I get warnings whenever the application peruses a directory with such files in it. I had a look at the USE flags for PHP but didn't see anything obvious. Is there an easy way to build PHP to have large file support? http://bugs.gentoo.org/show_bug.cgi?id=24373 -- Jason Stubbs pgpjzyHmgxJYA.pgp Description: PGP signature
Re: [gentoo-user]
On Saturday 20 August 2005 16:08, Zac Medico wrote: Ian K wrote: --- ACCESS VIOLATION SUMMARY --- LOG FILE = /tmp/sandbox-media-video_-_f4l-0.2-25033.log open_wr: /usr/qt/3/etc/settings/.qtrc.lock open_wr: /usr/qt/3/etc/settings/.qtrc.lock --- - Ian, I got the same result here. I will post a comment on bug 82154. If you don't care about the sandbox violation, you can temporarily disable the sandbox to work around the problem like this: FEATURES=-sandbox emerge f4l addpredict /usr/qt/3/etc/settings. It's a common problem with Qt apps. It's even in both qt.eclass and kde.eclass (although they both use addwrite rather than addpredict!?) -- Jason Stubbs pgpAC1ml0brap.pgp Description: PGP signature
Re: [gentoo-user] use flags (once more)
On Saturday 20 August 2005 15:59, Willie Wong wrote: On Sat, Aug 20, 2005 at 02:00:27AM -0400, John Dangler wrote: If I add a USE=whatever to my make.conf file, does that get added to the flags already in the default profile, or does make.conf override what's in the defaults file? yes, it is cascading. And your make.conf is read last. Which means whatever in the defaults will hold unless your make.conf gives a different one... Except for the environment, which is processed after make.conf. -- Jason Stubbs pgpLHBpnWzMdu.pgp Description: PGP signature
Re: [gentoo-user] use flags (once more)
On Saturday 20 August 2005 19:31, Neil Bothwick wrote: On Sat, 20 Aug 2005 16:24:29 +0900, Jason Stubbs wrote: yes, it is cascading. And your make.conf is read last. Which means whatever in the defaults will hold unless your make.conf gives a different one... Except for the environment, which is processed after make.conf. And /etc/portage/package.use, which is processed between the two? Yep. make.globals, make.defaults, use.defaults, make.conf, package.use, env. -- Jason Stubbs pgpXDuzRlgMUs.pgp Description: PGP signature
Re: [gentoo-user] record emerge messages
On Monday 08 August 2005 07:28, Rodrigo Lazo Paz wrote: Hi, I've been looking and I haven't found a way to record the emerge's messages like * Warning! Gentoo's GLIBC with NPTL enabled now behaves like the * glibc from almost every other distribution out there There's a log or a tool that records that? Bug 11359. There's also PORT_LOGDIR which will log everything. -- Jason Stubbs pgpjiSQIY951I.pgp Description: PGP signature
Re: [gentoo-user] how to archive all mail?
On Saturday 06 August 2005 01:05, Jarry wrote: Christopher Fisk wrote: Use MailScanner ( http://www.mailscanner.info ). It's a spam/virus filter in addition, but will also give you the functionality you are looking for. Is it not problem, if I use spamassassin clamav? Or do I have to switch? I would not like to mess things and have no mail delivered at all... Easy to do with procmail if that's already installed. You can just add a rule to both forward and continue delivery to /etc/procmailrc. -- Jason Stubbs pgpLDwENXjNB0.pgp Description: PGP signature
Re: [gentoo-user] how to archive all mail?
On Saturday 06 August 2005 01:31, Jarry wrote: Jason Stubbs wrote: Easy to do with procmail if that's already installed. You can just add a rule to both forward and continue delivery to /etc/procmailrc. I thought procmail takes care only for incomming mail. Am I wrong? Because I need to archive both incomming and outgoing mails... No your not wrong. Ever played with sendmail rules? They're not fun. ;) There's many other possibilities before you have to go down that road, though. Try googling for archiving outgoing mail with sendmail. -- Jason Stubbs pgpoepxAyjrQK.pgp Description: PGP signature
Re: [gentoo-user] pcmcia problems (cardmgr can't find device)
On Thursday 04 August 2005 09:17, Lucien D. wrote: phaedrus lu # zgrep CONFIG_PCMCIA /proc/config.gz # CONFIG_PCMCIA_DEBUG is not set CONFIG_PCMCIA=y CONFIG_PCMCIA_PROBE=y # CONFIG_PCMCIA_AHA152X is not set # CONFIG_PCMCIA_FDOMAIN is not set # CONFIG_PCMCIA_NINJA_SCSI is not set # CONFIG_PCMCIA_QLOGIC is not set # CONFIG_PCMCIA_SYM53C500 is not set You're missing some important information here. Here's mine: CONFIG_PCCARD=y # CONFIG_PCMCIA_DEBUG is not set CONFIG_PCMCIA=y # CONFIG_PCMCIA_LOAD_CIS is not set # CONFIG_PCMCIA_IOCTL is not set CONFIG_CARDBUS=y CONFIG_YENTA=y # CONFIG_PD6729 is not set # CONFIG_I82092 is not set # CONFIG_TCIC is not set CONFIG_PCCARD_NONSTATIC=y CONFIG_CARDBUS enables support for 32-bit cards (which you probably want) and you'll also need to enable at least one driver. Looking at your lspci output, the Yenta driver should work fine. -- Jason Stubbs pgpgJf3UteNZ2.pgp Description: PGP signature
[gentoo-user] Hold on portage feature requests
Hello all, The subject says it all: no more feature requests for portage (the package manager) until further notice. This does not include submitting of patches that add new features. Further notice will likely mean when the next major version of portage becomes stable. The reason behind this is that at approximately two thirds of bugs received are feature requests and they are drowning at the real bugs. More importantly, the critical bugs are becoming very hard to keep track of. This, at a time when we are focusing on fixing major and critical bugs only so as to get the next version completed quicker. Most of the current feature requests will be available at the time when the next version goes stable (and that which isn't should be relatively painless to implement) so don't worry that things will go stagnant. However, if you are worried, I'll be posting a weekly summary of portage bug activity to gentoo-portage-dev@gentoo.org from now on. If you'd like to join the portage team or just feel like giving a quick hand, have a browse through the bugs and see what fixes you can come up with. -- Jason Stubbs pgp0LZa8QYoih.pgp Description: PGP signature
Re: [gentoo-user] glsa
On Saturday 16 July 2005 00:56, Noah Roberts wrote: How long between when a GLSA is distributed and glsa-check will pick it up? I just got a message from announce that has a GLSA I believe my server will be vulnerable to but glsa-check isn't picking it up. In fact I just did a 'glsa-check -l | grep 200507' and nothing showed up. Does the portage db need to be updated? I figured it would use the rss. GLSAs are held in $PORTDIR/metadata/glsa -- Jason Stubbs pgpmKxKDEUGzX.pgp Description: PGP signature
Re: [gentoo-user] what's new with all those new packages?
On Tuesday 12 July 2005 19:35, Martin S wrote: On 11/07/05, Myk Taylor [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I think if you're using Gentoo, there is an assumption that if it's new (and acceptably stable), I want it. That's hardly in line with Gentoos choice policy. Gentoo isn't about imposing someone elses preferences on everyone. You're not forced to upgrade. You're not even forced to sync with the rsync servers. If you're happy with your machine the way it is, don't upgrade. Simple. Regards, Jason Stubbs pgp4nhBX1Gxq0.pgp Description: PGP signature
Re: [gentoo-user] revdep-rebuild badness
On Saturday 18 June 2005 21:18, fire-eyes wrote: I'm trying to do a revdep-rebuild -pv (pretend, verbose) however I get errors, and I can't complete it. What should I do? All prepared. Starting rebuild... emerge --oneshot --nodeps -pv =app-office/openoffice-bin-1.9.104 =media-sound/timidity++-2.13.0-r1 =media-video/xine-ui-0.99.3 =net-im/centericq-4.20.0-r1 =xfce-extra/xfce4-showdesktop-0.3.0 .. These are the packages that I would merge, in order: Calculating dependencies - emerge: there are no ebuilds to satisfy =media-sound/timidity ++-2.13.0-r1. emerge --oneshot -pv =app-office/openoffice-bin-1.9.104 media-sound/timidity++ =media-video/xine-ui-0.99.3 =net-im/centericq-4.20.0-r1 =xfce-extra/xfce4-showdesktop-0.3.0 Regards, Jason Stubbs pgpG6gf1ouQur.pgp Description: PGP signature
Re: [gentoo-user] Upgrade/Downgrade cycles
On Friday 17 June 2005 06:11, Christoph Gysin wrote: Richard Fish wrote: It is a long-standing bug in portage, and apparently not easy to fix. reference to bugzilla, anyone? http://bugs.gentoo.org/show_bug.cgi?id=13632 pgpFmqilciZ1D.pgp Description: PGP signature
Re: [gentoo-user] symbolic link to var tmp
On Saturday 04 June 2005 18:37, Ognjen Bezanov wrote: Philip Webb wrote: 050603 Andreas Vinsander wrote: Philip Webb wrote: IIRC you can make a symbolic link only within the same partition. others please correct me, if i'm mistaken. I will correct you now! You can make symbolic links across partitions, just try it! ok thanks (wry grin)! you are correct. The message above is all i got regarding my issue. Im guessing that my email account is messing up again regarding this list. Can anyone CC me the solution to my problem if its already been mentioned, and if any further replies could be CCd to me as well, thx PORTAGE_TMPDIR is the correct answer that you missed. Regards, Jason Stubbs pgpBol06jBR1x.pgp Description: PGP signature
Re: [gentoo-user] emerge busybox Why?
On Saturday 04 June 2005 23:02, Bill Roberts wrote: My usual emerge -uavDt world this morning wanted to emerge busybox. It doesn't show as a dependency for anything, and I've never emerged it for anything. Any idea why?? I believe it's been added to system in place of sash. It's similarly very small in size but has a much larger feature set that is generally very useful in emergencies. Regards, Jason Stubbs pgpH9BhbpSTCr.pgp Description: PGP signature
Re: [gentoo-user] IOError: [Errno 21] Is a directory
On Wednesday 01 June 2005 22:13, Kevin Bucknum wrote: This has happened before, but it usually clears up after a sync or two. It hasn't for the last couple weeks, and I'd like to get the box back up to date, and maybe give the new KDE a try. Every time I sync I get: Performing Global Updates: /usr/portage/profiles/updates/2Q-2005 (Could take a couple of minutes if you have a lot of binary packages.) .='update pass' *='binary update' @='/var/db move' s='/var/db SLOT move' S='binary SLOT move' p='update /etc/portage/package.*' [EMAIL PROTECTED] (most recent call last): File /usr/bin/emerge, line 10, in ? import portage File /usr/lib/portage/pym/portage.py, line 7306, in ? do_upgrade(mykey) File /usr/lib/portage/pym/portage.py, line 7197, in do_upgrade db[/][vartree].dbapi.move_ent(mysplit) File /usr/lib/portage/pym/portage.py, line 4593, in move_ent fixdbentries(origcp, newcp, pkgdir) File /usr/lib/portage/pym/portage.py, line 3875, in fixdbentries f = open(dbdir+/+myfile, r) IOError: [Errno 21] Is a directory Any idea what to look for? $ find /var/db/pkg -mindepth 3 -type d Any results that returns shouldn't be there. I'd be interested in what it does return though. Regards, Jason Stubbs pgptaZInT63E9.pgp Description: PGP signature
Re: [gentoo-user] IOError: [Errno 21] Is a directory
On Wednesday 01 June 2005 22:56, Kevin Bucknum wrote: -Original Message- From: Jason Stubbs [mailto:[EMAIL PROTECTED] Sent: Wednesday, June 01, 2005 8:25 AM To: gentoo-user@lists.gentoo.org Subject: Re: [gentoo-user] IOError: [Errno 21] Is a directory Any idea what to look for? $ find /var/db/pkg -mindepth 3 -type d Any results that returns shouldn't be there. I'd be interested in what it does return though. Regards, Jason Stubbs /var/db/pkg/game-utils/showeq-5.0.0.15/files It's nuked. Thanks. Would you know how it got there? If you didn't copy it yourself, some ebuild is doing naughty stuff. 'there are no ebuilds to satisfy showeq'. Third party? Regards, Jason Stubbs pgpe9EzvDEc3q.pgp Description: PGP signature
Re: [gentoo-user] IOError: [Errno 21] Is a directory
On Wednesday 01 June 2005 23:33, Kevin Bucknum wrote: /var/db/pkg/game-utils/showeq-5.0.0.15/files It's nuked. Thanks. Would you know how it got there? If you didn't copy it yourself, some ebuild is doing naughty stuff. 'there are no ebuilds to satisfy showeq'. Third party? Regards, Jason Stubbs I had it in overlay for a while, but I'm pretty sure I blew that away along time ago. It's showing at http://packages.gentoo.org/ebuilds/?showeq-5.0.0.18 as the current. I haven't run it since end of 2004 so don't really remember the history of when I did what. I had forgotten that I'm excluding 60% of the portage tree from syncing. I checked the current ebuilds though and they are a picture of perfection. Guess it's just one of those things that just isn't gonna be solved... Regards, Jason Stubbs pgpdi6Q1Say58.pgp Description: PGP signature
Re: [gentoo-user] It's baaaaack...! (Python 2.4.1)
On Wednesday 01 June 2005 00:26, Simon Maynard wrote: I am probably wrong, but I thought that portage would only install a new slot if either it was required by a package or the package is in your world file? Or if it's in system or if you specify --deep or it is a shallow upgradable dependency of whatever other targets are being chosen to upgrade or install. Regards, Jason Stubbs pgpa7BWiOFdQu.pgp Description: PGP signature
Re: [gentoo-user] kde 3.4.1?
On Monday 30 May 2005 21:56, Holly Bostick wrote: Volker Armin Hemmann schreef: On Monday 30 May 2005 12:11, Jan Meier wrote: Anyone have an idea when this show goes on the road? There is actually no release announcement at kde.org. Strange that there is a ebuild for that. Greets Jan well, it is usuall, that the ebuilds are ready, before a kde release is out - hardmasked. Maybe he unmasked completly kde3.4.x, or some dev unmasked 3.4.1 by accident, but I would not care too much about it. 3.4.1 should come soon ;) According to /usr/portage/profiles/package.mask, the tarballs aren't available yet (so I didn't bother unmasking). I thought maybe the ebuilds were nonetheless in Portage because the devs have some secret source for preliminary testing or something. The kde ebuild maintainers do, yes. This unfortunately does not include me. ;) Regards, Jason Stubbs pgp2Svpk5yVBV.pgp Description: PGP signature
Re: [gentoo-user] why it takes so long for emerge -B
On Monday 30 May 2005 22:46, askar ... wrote: Hello! I have installed mozilla from source. Now I wanted to build a package to use it in another PC. #emerge -B mozilla took so long time build the package. Is this OK? --buildpkgonly builds it from source again. What you wanted was quickpkg. Regards, Jason Stubbs pgpsMthg7bUAK.pgp Description: PGP signature
Re: [gentoo-user] When a dependancy built by hand
On Tuesday 31 May 2005 12:03, Harry Putnam wrote: I want to install emacs-w3m. I have already built an emacs install by hand from cvs emacs tar ball. I always build emacs myself. Running emerge -v -p emacs-w3m shows: These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild N] app-editors/emacs-21.4 +X -Xaw3d -debug +gnome -leim -lesstif +motif +nls -nosendmail 0 kB [ebuild N] dev-libs/boehm-gc-6.3-r1 -c++ 754 kB [ebuild N] media-libs/gdk-pixbuf-0.22.0-r3 -doc -mmx 388 kB [ebuild N] www-client/w3m-0.5.1-r1 +X -async -cjk +gpm +gtk +imlib -imlib2 -lynxkeymap -migemo +nls +ssl -xface 1,852 kB [ebuild N] app-emacs/emacs-w3m-1.4.3 708 kB Emerge wants to install another emacs. How can I tell it I already have emacs installed? One way would be to build it from scratch too, but I'd prefer to stay close to an emerge installed system when possible. /etc/portage/profile/package.provided Regards, Jason Stubbs pgp5WVvTZH0EC.pgp Description: PGP signature
Re: [gentoo-user] What's up with KDE 3.4?
On Saturday 28 May 2005 01:03, Radu Filip wrote: Hi, I was wondering what problem(s) might be with KDE 3.4 that prevents it to be in x86 after few months since the KDE team released it? See message entitled KDE 3.4 visibility support disabled posted just under 24 hours ago to gentoo-user@gentoo.org and [EMAIL PROTECTED] Regards, Jason Stubbs pgpHMmWd2g1nK.pgp Description: PGP signature
Re: [gentoo-user] Unmerging extra python
On Thursday 26 May 2005 23:36, Neil Bothwick wrote: On Thu, 26 May 2005 16:24:28 +0200, Holly Bostick wrote: Now of course I know that Portage depends on Python, and I certainly don't want to mess up Portage, so how do I get rid of this version of Python (or how do I recover if there's no way to get rid of it without breaking anything)? Portage depends on Python 2.3, unmerging that and leaving 2.4 could break things, the other way round shouldn't. Portage works with =python-2.2. Unmerging either 2.3 or 2.4 should not cause any troubles with portage at all as the libraries and scripts live outside the python directories. However, it is worth playing safe, so use quickpkg to build a binary package of any potentially important package before unmerging it. Even if you break portage, you can still install a binary package with tar xf /path/to/package.tbz2 -C / This is always good advice. Regards, Jason Stubbs pgpq1G4BpiwLq.pgp Description: PGP signature
Re: [gentoo-user] Unmerging extra python
On Friday 27 May 2005 09:36, Holly Bostick wrote: These are the packages that I would unmerge: !!! Trying to unmerge package(s) in system profile. 'dev-lang/python' !!! This could be damaging to your system. Waiting 10 seconds before starting... (Control-C to abort)... Press Ctrl-C to Stop in: 10 9 8 7 6 Exiting on signal 2 This is a bug. It will give the normal --pretend output after the countdown is finished, but there shouldn't be a countdown at all when doing --pretend. Portage checks the package you are about to unmerge against the list of system packages and then warns if it matches. Slots aren't taken into account at all but probably should be. That would likely reduce the mixed-message factor I seem to be sensing here. Okay. So no countdown during --pretend as well as using slot checks. Care to open a bug so I don't forget please? :) Regards, Jason Stubbs pgpNeB8WV6zN2.pgp Description: PGP signature
Re: [gentoo-user] package.* mess!
On Wednesday 25 May 2005 22:14, Matan Peled wrote: Mrugesh Karnik wrote: Hello all, My Gentoo system is up and running beautifully. Along the way, I've installed many software. But, now, I want to make it all systematic, ie, use the files like package.use, package.keywords etc to list out the options i've used for the different software. Now the problem is that I've installed the software with different use flags and keywords right now the command line. Now I want to make it systematic and proper. I'm just wondering if you people could help me find a good way as to how to go about this. Right now I can see myself doing a very very tedious job of using equery and such and listing out everything properly... a few hours' work. I'm prepared to do it of course, but I'd like any suggestions that might help me do it better. Thank you. Yours faithfully, Mrugesh Karnik emerge world -uDpv and fix all the packages that want to be downgraded. emerge world --newuse and fix all the packages that want to change their useflags. Mostly right. You'll need to add --deep and --update to the --newuse run as well. Regards, Jason Stubbs pgpB1YQYLmOVO.pgp Description: PGP signature
Re: [gentoo-user] Re: driver CardBus bridge Texas Instruments PCI1211
On Saturday 21 May 2005 17:36, askar ... wrote: In the kernel I enables option for 'CardBus yenta-compatible' - this seems the one I was looking for. After recompilation the kernel I rebooted the system. 3 lamps of the PCMCIA card was on: 1) Power 2) Act and 2) Link The pcmcia card I use is Planex ENW-3503-TX. This is the 10Base-T card. The cardbus seems working - lights of power and act are on. When I connect LAN cable the light for Link also switches on. In /etc/conf.d/net I set IP address for eth0. When I did # /etc/init.d/net.eth0 start, it complains 'no such device ...unknown interface'. I think the driver need to be installed. In the kernel settings I don't see the driver for Planex ENW-3503-TX. Could anybody help me? Google can. ;) Searching for Planex ENW-3503-TX linux gave a list of card types and what chipsets they contain on the first result. Your card has a Winbond W89C926. I then searched for Winbond W89C926 linux and got mostly similar results to the first search, but there was one that indicates that the card is NE2000 compatible. One of the PCMCIA network drivers in the kernel is NE2000 compatible PCMCIA support, so I'd suggest you give that one a try. Regards, Jason Stubbs pgpNb3YNuDceA.pgp Description: PGP signature
Re: [gentoo-user] emerge after kernel upgrade
On Saturday 21 May 2005 21:48, Dan Johansson wrote:
On Thursday 19 May 2005 22.27, Zachary Medico wrote:
--- Dan Johansson [EMAIL PROTECTED] wrote:
Hi,
After I've done an upgrade of my kernel I have to
re-emerge some kernelmodules
(madwifi-driver among other). BUT when I do this
emerge removes the module
from the /lib/modules/... from the old kernel
version. Is there a way to make
emerge keep the old version as well so I can boot
the old kernel again with
full functionality?
This should not happen.
You could add this to /etc/make.conf:
CONFIG_PROTECT=$CONFIG_PROTECT /lib/modules
Adding $CONFIG_PROTECT is not needed unless it already has some definition in
make.conf. Also, ${NAME} syntax should always be used in portage config
files.
Thanks, that did the trick.
What portage version are using?
Regards,
Jason Stubbs
pgpZABctvueO9.pgp
Description: PGP signature
Re: [gentoo-user] emerge after kernel upgrade
On Saturday 21 May 2005 22:08, Dan Johansson wrote: On Saturday 21 May 2005 14.55, Jason Stubbs wrote: What portage version are using? My portage version is sys-apps/portage-2.0.51.19 on x86. There's been a hack in portage all the way through 2.0.51 and I'm pretty sure 2.0.50 as well that prevents _anything_ from being removed from /lib/modules. Are you positive that things are being removed? Regards, Jason Stubbs pgp7xh3R2iJTF.pgp Description: PGP signature
Re: [gentoo-user] -fvisibility=hidden
On Sunday 22 May 2005 00:43, Julien Cayzac wrote: On 5/21/05, Jason Stubbs [EMAIL PROTECTED] wrote: You miss the point. Adding that flag to CFLAGS (or CXXFLAGS) is faulty in and of itself. It is not a general optimization flag. It is something that each package's codebase needs to be updated to support. When upstream updates their packages, they will also update whatever build system to specify that flag accordingly. From the GCC manpage: A good explanation of the benefits offered by ensuring ELF symbols have the correct visibility is given by ``How To Write Shared Libraries'' by Ulrich Drepper (which can be found at http://people.redhat.com/~drepper/) - however a superior solution made possible by this option to marking things hidden when the default is public is to make the default hidden and mark things public. This is the norm with DLL's on Windows and with -fvisibility=hidden and __attribute__ ((visibility(default))) instead of __declspec(dllexport) you get almost identical semantics with identical syntax. This is a great boon to those working with cross-platform projects. As I understand this, any shared object that relies on the fact that its symbols will get magically exported while not explicitly exporting them is broken. That said, Portage provides a way to mask broken c(xx)flags for those ebuilds. Yes, but there was not only no requirement up until recently, there was no way to specify visibility. This means that almost everything is broken at the moment. Or alternatively, you can consider telling the compiler that a piece of code does specify it's visibility when it actually doesn't to be a broken action. It's with the latter viewpoint that any bugs filed asking for -fvisibility=hidden to be filtered will be marked INVALID. Regards, Jason Stubbs pgpjJw3dMRM5R.pgp Description: PGP signature
Re: [gentoo-user] Re: driver CardBus bridge Texas Instruments PCI1211
On Sunday 22 May 2005 00:54, askar ... wrote: On 5/21/05, Jason Stubbs [EMAIL PROTECTED] wrote: On Saturday 21 May 2005 17:36, askar ... wrote: In the kernel I enables option for 'CardBus yenta-compatible' - this seems the one I was looking for. After recompilation the kernel I rebooted the system. 3 lamps of the PCMCIA card was on: 1) Power 2) Act and 2) Link The pcmcia card I use is Planex ENW-3503-TX. This is the 10Base-T card. The cardbus seems working - lights of power and act are on. When I connect LAN cable the light for Link also switches on. In /etc/conf.d/net I set IP address for eth0. When I did # /etc/init.d/net.eth0 start, it complains 'no such device ...unknown interface'. I think the driver need to be installed. In the kernel settings I don't see the driver for Planex ENW-3503-TX. Could anybody help me? Google can. ;) Searching for Planex ENW-3503-TX linux gave a list of card types and what chipsets they contain on the first result. Thanks. But when I searched with the above keyword, the search results in 2 pages, and all sites in japanese... I noticed that. I figured you'd probably be able to read seeing I thought Planex was a domestic-only brand. Luckily I can read Japanese. ;) Your card has a Winbond W89C926. I then searched for Winbond W89C926 linux and got mostly similar results to the first search, but there was one that indicates that the card is NE2000 compatible. One of the PCMCIA network drivers in the kernel is NE2000 compatible PCMCIA support, so I'd suggest you give that one a try. I use gentoo 2005.0, kernel 2.6.11. In my kernel I have only NE2100 in: Device drivers-Networking support-Network device support-Ethernet (10 or 100Mb)--AMD Lance and PCnet (AT1500 and NE2100) support. Is this the right place? Device Drivers --- Networking support --- PCMCIA network device support --- NE2000 compatible PCMCIA support One question more - usually we can know about connected devices with command lspci. But in the result from lspci no information about my pcmcia-card. As far as I know, lspci only shows you what's on the PCI bus. PCMCIA is a different bus which is generally connected to the PCI bus. Regards, Jason Stubbs pgpPLvBgsOzb4.pgp Description: PGP signature
Re: [gentoo-user] qpkg and etcat deprecated?
On Tuesday 10 May 2005 00:33, Willie Wong wrote: There has long been talk of deprecating etcat and qpkg in favor of equery, has that been finalized? I updated to portage-2.0.51.21-r1 yesterday and can no longer find the etcat and qpkg commands. And what is it with the changelogs? Maybe I've looked in the wrong place, but I didn't see that change mentioned at all in the changelog for portage (actually, I didn't even see a changelog when I issued `emerge -vlPu portage'). etcat and qpkg are not part of portage. They are part of gentoolkit. Regards, Jason Stubbs pgpCsVwosjVOA.pgp Description: PGP signature
Re: [gentoo-user] how to japanese intput
On Saturday 07 May 2005 16:44, askar ... wrote: Hello! I'm using Gentoo 2005.0, KDE and Fluxbox. I want to be able writing in japanese. Is there step-by-step setup guide for this. I looked at internet there some information, but each site has its own differences. I dont want to japanise everything - only japanese input needed. Also with Emacs. For KDE, skim is the easiest. Just install it as well as whatever input methods you require and run it. No messing with locale settings or anything. I'm currently using it with canna and skk. I'm not an emacs user so not sure about that one. I know that ddskk is easy to install and comes with a tutorial if you like SKK, although the tutorial is in Japanese. The other emacs package for Japanese is tamago, which has the description Emacs Backend for Sj3 Ver.2, FreeWnn, Wnn6 and Canna. Regards, Jason Stubbs pgpE8fch2hfKg.pgp Description: PGP signature
Re: [gentoo-user] Howl build error
On Saturday 07 May 2005 20:12, Allan Spagnol Comar wrote: I got the following error while trying to emerge howl i don have a clue of how I can make it right, does some one has ? ThankYou, Allan In file included from Posix/posix_interface.c:45: /usr/include/linux/ethtool.h:18: error: syntax error before __u32 ^^^ Likely the cause is incompatible Linux headers. After a quick search of bugs.gentoo.org for bugs with howl in the summary, I found bug #87513. Apparently, the bug with linux-headers has been fixed but there was no version bump (which is not the correct way for it to have been handled). That is, if you reemerge linux-headers the problem should be fixed and howl should emerge properly. Regards, Jason Stubbs pgp6GyBGLn3Eq.pgp Description: PGP signature
Re: [gentoo-user] gentoo.org Is Down?
On Sunday 08 May 2005 00:15, Drew Tomlinson wrote: Can anyone get to http://www.gentoo.org? I keep getting an Internal Server Error so I don't think it's just me. Works fine here. Could you check the IP address that www.gentoo.org is resolving to please? There's several servers so there might only be one with issues. Regards, Jason Stubbs pgpPRWbo3VzOB.pgp Description: PGP signature
Re: [gentoo-user] emerge sync - portage: Update type slotmove not recognized. - !!! No profile directory; system mode unavailable.
On Saturday 30 April 2005 17:40, Richard Watson wrote: I just tried to emerge sync an install I did off a stage 3 LiveCD I bought during 2004. I'm getting some strange results. Can anyone tell me what the output below is telling me. I tried running fixpackages but that did not fix it. Now when I try and emerge any package portage keeps trying to sync (again I think) and fails the package installation. The output I'm confused about is: portage: Update type slotmove not recognized. This would probably indicate portage-2.0.50. !!! No profile directory; system mode unavailable. Did you get a deprecation notice and change the /etc/make.profile symlink? Is the CD anything other than 2004.2? In either case, relink /etc/make.profile to /usr/portage/profiles/default-x86-2004.2. Then ignore the deprecation notice and above slotmove error until you've upgraded portage to 2.0.51. Regards, Jason Stubbs pgpll6ipA3lm4.pgp Description: PGP signature
Re: [gentoo-user] emerge sync - portage: Update type slotmove not recognized. - !!! No profile directory; system mode unavailable.
On Saturday 30 April 2005 22:18, Richard Watson wrote: Failed to write to mtimedb: 'module' object has no attribute 'HIGHEST_PROTOCOL' That would be due to running python-2.2. Although, the portage-2.0.51.19 ebuild is meant to detect the use of python-2.2 and patch accordingly. Perhaps you have both 2.2 and 2.3 installed? Regards, Jason Stubbs pgpwfpCa3MNBB.pgp Description: PGP signature
Re: [gentoo-user] resume emerge after reboot
On Thursday 28 April 2005 20:20, Tamas Sarga wrote:
Hi,
I'd like to do an emerge -e world. It would take 3 days, but I can't
sleep next to my machine, 'cause of the sound of the 7000RPM CPU cooler.
If I start an emerge, then C-c it, shutdown, start the PC at morning,
can I resume the emerge -e?
Sure. emerge --resume. --resume functionality is planned to be improved in the
future. However, at the moment emerging something else will kill the record
of the previous emerge. I usually do a quick manual loop for it like this:
# emerge -ep world | grep ebuild | awk '{print =substr($0, 17)}' world
# for x in $(world); do emerge --oneshot $x grep -v $x world world2
mv world2 world || break; done
The second and third lines are one line.
This allows me to stop and emerge something else first if I want to. If
something fails, it also means that I can play around with it until it is
merged without having to use --resume every time. Or you could remove the
|| break and have it continue on regardless of failures and you'll have a
list of problems at the end.
Regards,
Jason Stubbs
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] resume emerge after reboot
On Friday 29 April 2005 10:23, Daevid Vincent wrote: No, 'emerge --resume' is lame. It doesn't do what you're thinking it does. What you want is: ebuild path to last ebuild that was worked on merge should do the trick -- at least for the last package that was worked on. To find the last ebuild that was worked on, check /var/log/emerge.log... it'll be one of the last lines in the file. After that, you should be able to just issue the original emerge command, and it should pick up where it left off. This can be dangerous depending on the package you are trying to resume the compile of. There are also other side-effects of using ebuild (such as packages not being added to world) that will invariably lead to INVALID bugs. Regards, Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] emerge broken. HELP!
On Monday 25 April 2005 18:42, Andrew Gaydenko wrote: After yesterday updating I can not emerge anything or do sync-ing. The error is shown below. How to repair the error? http://bugs.gentoo.org/show_bug.cgi?id=90347 Patch available. Regards, Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] fetch then build
On Saturday 16 April 2005 21:47, Eric S. Johansson wrote: I frequently find myself fetching packages then building. Reading through the emerged documentation that does not seem to be any way to do both in one step fetch first, and then if successful, fetch second? I tried: emerge -fDva world emerge -uDv world which only mostly prefetched files (misssed a bunch). tried F instead of F and it missed more and changed what it installed. ideas? There's no way to quit a fetch run if a single fetch fails. ps. updating laptop that was idle for 6+ months: 320 packages... a good test of my gentoo skills. so far so good. you can bet your ass I'm backing up /etc before running dispatch-conf... Wait a couple of weeks and there'll be a couple of dispatch-conf releases that should make a little bit safer. Regards, Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge needs a --nocolor option
On Tuesday 12 April 2005 20:49, A. Khattri wrote: There really ought to be a --nocolor option for emerge, alongside the --nospinner option. (I know I could add it to FEATURES in make.conf but I dont want to do that). I run emerge --sync from cron on my servers - the email from cron contains a lot of control codes. Well, that's three people saying it should be implemented but no offerings off any code yet... Here's my solution: Add NOCOLOR=true to your crontab. Regards, Jason Stubbs -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Emerge needs a --nocolor option
On Tuesday 12 April 2005 23:24, Jason Cooper wrote: Jason Stubbs ([EMAIL PROTECTED]) scribbled: On Tuesday 12 April 2005 20:49, A. Khattri wrote: There really ought to be a --nocolor option for emerge, alongside the --nospinner option. (I know I could add it to FEATURES in make.conf but I dont want to do that). I run emerge --sync from cron on my servers - the email from cron contains a lot of control codes. Well, that's three people saying it should be implemented but no offerings off any code yet... Here's my solution: Add NOCOLOR=true to your crontab. ehh, challenge accepted. :) And applied to CVS. initial patch attached. This is against 'emerge' from sys-apps/portage-2.0.51.19. it adds '--nocolor' with the abbreviation '-r'. Looks like my tabs might be off, let me know if they are. I removed the '-r' short option. I've never looked at this code before, nor this language. So I'm sure I screwed it up... The spaces vs tabs in the options array is just a formatting thing. Other than that there were no problems. Regards, Jason Stubbs -- gentoo-user@gentoo.org mailing list
