Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, October 2, 2006 19:15, Martijn van Oosterhout wrote: Best translated are required/important/standard, but those descriptions will be the least relevant to Joe Average, since these packages are already installed for him. The goal should be making sure that most of the packages people have installed are translated, as well as the most popular packages. That is a much more acheivable goal, and I think that's attainable in the near future... I'm not sure what your assertion about that goal is based on. Package translations are used by people trying to make a decision whether or not to install a given package, not to read up about already installled packages. Since req/imp/std packages are already installed at the majority of systems, users will much more likely be confronted with package descriptions for things they don't have, i.e. optional, than the things they already have installed. I hope you're not suggesting we need to get all languages covering all of optional before you think it is worthwhile? Indeed, I can be clear about this: you need to have some significant proportion of languages have a significant proportion of translations. Significant means not necessarily 100% but also not 5%. On Mon, October 2, 2006 20:48, Javier Fernández-Sanguino Peña wrote: This is very much related to the fact that the language teams see no gain for translating the DDTP right now since end-users would not be able to see them, as there is no support in apt or its frontends. Ah, a circle which needs to be broken. It can however be broken by any of the two parts in this situation; and I believe that being very close to the release, breaking the circle at the apt side is not highly desirable. It's easy to argue the other way around: translators should show that there's an actual use for the function before it gets added, especially in this late stage of the game. I'm not at all proposed at implementing the feature. I just think that the request for inclusion was very late and thus high-risk, if you want to justify that, you need to bring more than future expected usefulness. The latter is fine in itself, but suitable for the beginning of a release cycle. Thijs -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, Oct 02, 2006 at 12:54:13AM +0200, Michael Vogt wrote: BTW, I count 18 binary packages that would need a rebuild for this. This is a decent-sized library transition in its own right. We may have to recompile the rdepends of libapt anyway because of http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=390189 (recent g++ upload 4.1.1ds1-14 has a g++ regression) sigh This version of g++-4.1 hasn't been accepted into etch yet, and there's been no request from Matthias that we do so. Letting it into etch as a freeze exception suggests that we might have *other* packages fail to build as a result of similar ABI regressions in other libraries. That doesn't sound like a good idea to me unless someone is offering to do a full regression-test of testing using g++ 4.1.1-15. Upstream gcc bugreport: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29289 From this report, there's nothing to suggest the reverse-deps need to be rebuilt, only that the lib needs to be rebuilt so that the reverse-deps don't FTBFS. Is there something I'm missing? Matthias is still waiting for a comment from upstream on this. It maybe enough to recompile apt with the current g++, but it maybe that the only save option is to change the soname and recompile a rdepends. If there really is reason to believe this requires an soname change, I think we should instead consider backing this patch out of g++-4.1 in unstable until after the etch release, as compiler-induced ABI changes are clearly *not* supposed to be happening during a toolchain freeze. There's no API changes from APT side so just binary NMUs are enough AFAIK. So what is this ABI change that doesn't involve API changes? There is a API change involved. But it is backwards compatible so a recompile will be good enough. To make use of the translated descriptions the applications needs to be changed though. Patches are available for aptitude, python-apt, synaptic, libapt-front (0.3). I hope this helps and I'm sorry for the bad timing with this request :/ FWIW, this didn't answer the question what is the ABI change? :) Cheers, -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. [EMAIL PROTECTED] http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, 2006-10-02 at 11:54 +0200, Jens Seidel wrote: Consider how many people whould profit from it! I'm missing the following practical note a bit in this discussion: are there actually a significant number of translations to take the non-trivial venture of a very late apt update? I value the importance of the DDTP project, but the translating effort has only recently seriously started. Looking at the statistics[1], I see that the best language has yet only one third of descriptions translated in optional and extra, and steeply dropping to only a couple of percentpoints for those after that. Best translated are required/important/standard, but those descriptions will be the least relevant to Joe Average, since these packages are already installed for him. Concluding, we will not be able to claim that Debian has translated package descriptions, except for a very small number of languages. I think it's not worth the effort to risk delay or trouble for this; let's focus on other areas in etch now and make sure etch+1 has a really comprehensive set of translated descriptions. Thijs signature.asc Description: This is a digitally signed message part
Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, Oct 02, 2006 at 12:54:13AM +0200, Michael Vogt wrote: On Wed, Sep 27, 2006 at 11:42:35PM -0700, Steve Langasek wrote: On Thu, Sep 28, 2006 at 02:35:19AM -0300, Otavio Salvador wrote: On Wed, Sep 27, 2006 at 07:13:21PM +0200, Luk Claes wrote: Would you still accept an ABI change of apt to support description translations into etch? That's why I considered it so late for uploading to unstable. I didn't wanted to upload it without real-world testing because of the risk of having to break the ABI yet again to fix mistakes in the code. We may have to recompile the rdepends of libapt anyway because of http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=390189 (recent g++ upload 4.1.1ds1-14 has a g++ regression) Debian was always proud because of many software packages. Nevertheless it's a real drawback that package descriptions are only available in English. Many person with no English skills do not know the packages Debian provides and ignored these because of this. Once I installed a system with translated package descriptions for a friend I remember first time users of Debian browsing description just for fun, testing these packages, comparing with other systems, ... Without they never touched aptitude and complained about the usability. Consider how many people whould profit from it! Ten thousands, hundred thousands of users?! Please compare this with possible disadvantages and choose the proper solution! Once it enters testing I would also ask additional users from various lists (not only developers) to properly use and test it and would be willing to help these users to report possible problems. I'm sure many other people (translators and other) would do the same once you consider the changes for Etch. I now subscribed also to the APT bugs and will try my best ... Jens -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, Oct 02, 2006 at 01:17:18PM +0200, Thijs Kinkhorst wrote: On Mon, 2006-10-02 at 11:54 +0200, Jens Seidel wrote: Consider how many people whould profit from it! I'm missing the following practical note a bit in this discussion: are there actually a significant number of translations to take the non-trivial venture of a very late apt update? I value the importance of the DDTP project, but the translating effort has only recently seriously started. Looking at the statistics[1], I see See http://ddtp.debian.net/ for this link [1]. that the best language has yet only one third of descriptions translated in optional and extra, and steeply dropping to only a couple of percentpoints for those after that. Concluding, we will not be able to claim that Debian has translated package descriptions, except for a very small number of languages. I think it's not worth the effort to risk delay or trouble for this; let's focus on other areas in etch now and make sure etch+1 has a really comprehensive set of translated descriptions. Right. Nevertheless there are currently already at least 4 languages with (partly many more than) 1000 package descriptions. Also consider that the translation effort is independent of the Etch release (external database, no package upload are required, except of course for apt). Up to the release of Etch (for CD based installations) or even after it (network connection) users could profit from it. I can also guarantee that the effort will increase dramatically once we know that APT would support it. Currently the matra is: Let's ignore package descriptions as these will probably not be usable in Etch at all. Once users see a incomplete project they want to help, right!? Jens -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On 10/2/06, Thijs Kinkhorst [EMAIL PROTECTED] wrote: I value the importance of the DDTP project, but the translating effort has only recently seriously started. Looking at the statistics[1], I see that the best language has yet only one third of descriptions translated in optional and extra, and steeply dropping to only a couple of percentpoints for those after that. Not sure which statistics you're looking at, perhaps these? [1] Sure, maybe the top language has only 33% of optional done, but several languages cover the entire base install. Additionally, the numbers are not fixed. As many (most?) descriptions are shared between etch and sid, even after etch is released the descriptions will keep getting updated and improved. Best translated are required/important/standard, but those descriptions will be the least relevant to Joe Average, since these packages are already installed for him. The goal should be making sure that most of the packages people have installed are translated, as well as the most popular packages. That is a much more acheivable goal, and I think that's attainable in the near future... I hope you're not suggesting we need to get all languages covering all of optional before you think it is worthwhile? As for whether it's enough to make it happen for etch, that's not my call. But the vast majority of descriptions for etch+1 will be usable for etch also, so the decision should *not* be based on whether the descriptions are ready now. Have a nice day, [1] http://svana.org/kleptog/temp/ddts-stats.html -- Martijn van Oosterhout [EMAIL PROTECTED] http://svana.org/kleptog/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On Mon, Oct 02, 2006 at 01:17:18PM +0200, Thijs Kinkhorst wrote: On Mon, 2006-10-02 at 11:54 +0200, Jens Seidel wrote: I value the importance of the DDTP project, but the translating effort has only recently seriously started. Looking at the statistics[1], I see The first DDTP translation effort started years ago (over 4, actually) and it was quite serious at the time for some languages. It was stalled due to gluck's compromise and recently restarted. that the best language has yet only one third of descriptions translated in optional and extra, and steeply dropping to only a couple of percentpoints for those after that. This is very much related to the fact that the language teams see no gain for translating the DDTP right now since end-users would not be able to see them, as there is no support in apt or its frontends. A change in apt to make those visible when users run 'apt-cache search|show' even if apt frontends (aptitude, synaptic) do not use them yet would certainly make translation teams shift their efforts over to the DDTP. Just my 2c. Javier signature.asc Description: Digital signature
Re: Would an ABI change of apt for DDTP support still be accepted?
Javier Fernández-Sanguino Peña [EMAIL PROTECTED] writes: On Mon, Oct 02, 2006 at 01:17:18PM +0200, Thijs Kinkhorst wrote: On Mon, 2006-10-02 at 11:54 +0200, Jens Seidel wrote: I value the importance of the DDTP project, but the translating effort has only recently seriously started. Looking at the statistics[1], I see The first DDTP translation effort started years ago (over 4, actually) and it was quite serious at the time for some languages. It was stalled due to gluck's compromise and recently restarted. And my first APT patch was release in 2003[1]. As anyone can notice, it's not new. 1. http://lists.debian.org/debian-devel-announce/2003/04/msg00015.html -- O T A V I OS A L V A D O R - E-mail: [EMAIL PROTECTED] UIN: 5906116 GNU/Linux User: 239058 GPG ID: 49A5F855 Home Page: http://www.freedom.ind.br/otavio - Microsoft gives you Windows ... Linux gives you the whole house.
Re: Would an ABI change of apt for DDTP support still be accepted?
On Wed, Sep 27, 2006 at 11:42:35PM -0700, Steve Langasek wrote: On Thu, Sep 28, 2006 at 02:35:19AM -0300, Otavio Salvador wrote: On Wed, Sep 27, 2006 at 07:13:21PM +0200, Luk Claes wrote: Would you still accept an ABI change of apt to support description translations into etch? I gather that ABI change means an soname change? In that case, no, sorry, I think it's too late in the release cycle to be changing this for apt. I would like to ask you to review again your position. This code is around since 3 years ago and in use on Ubuntu too. Are too few packages that will need recompile. And yet the request comes as we should be preparing to feature-freeze apt *completely* for etch, not thinking about changes that require a recompile of all reverse-deps. Right. I'm to blame here that I was overly cautious with putting new code into libapt in unstable. There were no translations available on ftp.debian.org until end of July (when aj did a one-time import) and without those the code was not really testable for real-world use. When the translations were importet and I asked for testing on debian-devel I got little feedback on the actual code in experimental. That's why I considered it so late for uploading to unstable. I didn't wanted to upload it without real-world testing because of the risk of having to break the ABI yet again to fix mistakes in the code. BTW, I count 18 binary packages that would need a rebuild for this. This is a decent-sized library transition in its own right. We may have to recompile the rdepends of libapt anyway because of http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=390189 (recent g++ upload 4.1.1ds1-14 has a g++ regression) Upstream gcc bugreport: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29289 Matthias is still waiting for a comment from upstream on this. It maybe enough to recompile apt with the current g++, but it maybe that the only save option is to change the soname and recompile a rdepends. [..] There's no API changes from APT side so just binary NMUs are enough AFAIK. So what is this ABI change that doesn't involve API changes? There is a API change involved. But it is backwards compatible so a recompile will be good enough. To make use of the translated descriptions the applications needs to be changed though. Patches are available for aptitude, python-apt, synaptic, libapt-front (0.3). I hope this helps and I'm sorry for the bad timing with this request :/ Cheers, Michael -- Linux is not The Answer. Yes is the answer. Linux is The Question. - Neo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hey! :-) On 09/28/2006 04:13 AM, Andreas Barth wrote: * Otavio Salvador ([EMAIL PROTECTED]) [060928 08:47]: I would like to ask you to review again your position. This code is around since 3 years ago and in use on Ubuntu too. Are too few packages that will need recompile. How does it come that the code isn't promoted at the beginning of a release cycle, but at a time where the base freeze has happened? The upload of apt with support to translated package description is recent, but the code is old. What happens is that during i18n Extremadura (Sept. 7th-10th) we discusses that having apt in etch should be a good idea to have it supported and tested for etch+1, but I remember that our discussions in Extremadura always consider the Release Team opinion on the matter. After DebConf, Christian requested to have support for translated packages descriptions as a pet release goal, some time later the experimental upload took place, the DDTS was revived, a web interface was created and we thought that asking Release Team for apt transition is the right thing to do. It is sad that we don't have time to include apt in etch? Yes, it is. But the i18n/l10n teams will understand the reasons. If the transition is not suitable, ok, we will work hard to get it just after etch release, and to support other ideas for i18n in Debian. :-) I don't think we need to fight in that matter, as I said, it is OK to not have apt with translations in etch (it's sad but it's OK, nobody will get hurt and no SuperCows will die). And I do think it's possible to workaround it, at least for CDDs using backports.org or something in that matter. We are late at the release cycle, if it is still possible, I'm sure that everybody involved in i18n will do their best to help release team with regards to the apt. If Release Team thinks we should wait etch+1, so let's work hard to have etch on December and we can have apt supporting translations for package descriptions in unstable by the end of the year. ;) [...] Cheers, Andi Kind regards, - -- Felipe Augusto van de Wiel (faw) Debian. Freedom to code. Code to freedom! -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.5 (GNU/Linux) Comment: Using GnuPG with Debian - http://enigmail.mozdev.org iD8DBQFFG+57CjAO0JDlykYRAhJ5AKCEV76NYO4QrzCNdueCH+b9j0trdQCbBV74 Rg6XnODDy0A7kjO6dCUPQM0= =DeO4 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Would an ABI change of apt for DDTP support still be accepted?
On Wed, Sep 27, 2006 at 07:13:21PM +0200, Luk Claes wrote: Would you still accept an ABI change of apt to support description translations into etch? I gather that ABI change means an soname change? In that case, no, sorry, I think it's too late in the release cycle to be changing this for apt. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. [EMAIL PROTECTED] http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]