Re: The case for a kdelibs 4.8
On Fri, Sep 30, 2011 at 9:54 AM, David Faure fa...@kde.org wrote: On Thursday 29 September 2011 20:01:00 Kevin Kofler wrote: But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. But who is (or would be) working on features in kdecore | kdeui | kio | kfile? Well this is over a month too late, but I have a enhancement change for kcookiejar that needs to go into kdelibs/kioslave for KDE 4.8. The patch has actually been pending for a merge since KDE 4.6. See https://bugs.kde.org/show_bug.cgi?id=54300. Unfortunately, I do not know how to proceed with committing the kdelibs portion of the patch without actually polluting the kdelibs 4.7.4 version. Waiting for the final December release of the final KDE 4.7.4 is one approach, but then the commit would be too late for the feature freeze. We either need to exempt kdelibs from the upcoming feature freeze or I need an exemption to commit these changes past the freeze times currently established for KDE 4.8
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011, Kevin Kofler wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. The rule so far has always been that the kdelibs version must be the same as the KDE SC version. Changing this will also break all our Fedora KDE SC specfiles, which all have a: BuildRequires: kdelibs4-devel = %{version} We will release a kdelibs 4.8.0 tarball for this purpose (as other distros have the same issue). Greetings, Dirk
Re: The case for a kdelibs 4.8
Am Samstag, 1. Oktober 2011, 08:27:02 schrieb Martin Gräßlin: One of the main reasons for the rebranding was to realize that KDE is not one product, but a community that produces multiple products among them a desktop environment (Plasma). What you just try to tell us is that the complete rebranding is nonsense and we should go back to talking just about KDE for everything bringing the users back to assuming they need Plasma in order to use KMail. I think that runs short. “KDE” is what gets a happy KMail user to try Konqueror or Plasma. In my opinion, it is important to make sure that people realize that these aren’t just random development frameworks, but the proven and powerful KDE frameworks. Even better: These are the KDE Frameworks 4.8, the same version the rock stable KDE Workspace 4.8 uses, so it is really a good idea to use them as a base for your application. But for that you need the release to be clearly correllated: This is version 4.8 as released and tested by the KDE community. I really like the fact that finally we have a release where it will be clear that KDE is the community and not one large product. All those who did not get it yet, might finally understand it if we write a really good release announcement. I don’t like that “if”… And besides: As much as KDE is a community and not one single product, many people will still want to just run the KDE software collection on the KDE workspace and code using the KDE frameworks. In short: They want to get the full package and not have to worry about details. Best wishes, Arne signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
Aaron J. Seigo wrote: the features that got into the 4.7 branch to date have been things that were already worked on before the Frameworks decision was made. it's was an odd cas were features had been worked on while 4.7 was frozen with the expectation of a 4.8 ... and that left us with the choice of having a 4.8 release which we didn't want for those few features, leaving those features out and screwing up the planning of the applications depending on those changes or fudging a little bit. Not all those features are merged in yet, see e.g. my Plasma PackageKit GSoC work. note that some of the more actively changing and developed code in kdelibs have moved to separate git repositories already to make this issue moot for them (kactivities, nepomuk) That is also a nightmare to packagers. Especially kactivities is a big problem because kde-workspace 4.7 (not 4.8!) is now reported to rely on a new kactivities which is not in kdelibs 4.7. That makes no sense whatsoever. Kevin Kofler
Re: Re: The case for a kdelibs 4.8
Martin Gräßlin wrote: One of the main reasons for the rebranding was to realize that KDE is not one product, but a community that produces multiple products among them a desktop environment (Plasma). What you just try to tell us is that the complete rebranding is nonsense and we should go back to talking just about KDE for everything bringing the users back to assuming they need Plasma in order to use KMail. That's what some of us have been trying to tell you for months… Kevin Kofler
Re: The case for a kdelibs 4.8
Aaron J. Seigo wrote: so some features will indeed have to wait .. and that's not a horrible thing because it means that frameworks will get more developer attention and the attention it is getting already will not be slowed even further by having to deal with bringing over features from 4.7.x into the re-arranged libraries in frameworks. So you're arguing that by doing this, you FORCE developers to work on 5.0? Well, I'm sorry, but you cannot really force volunteer developers to do anything. I don't believe that actively preventing us from working on the branch we want to work on is going to solve any problem. the result will be that the time between now and 5.0 libs being available will be shortened, which is exactly what we, as application developers, need. I also strongly doubt that yet another binary-incompatible break is what application developers really want or need. Some applications are still not ported to the KDE Platform 4. (And I know that Qt is also breaking the ABI. That's something I also can't agree with, especially considering that Qt 4 was advertised as the big ABI break which will make new ABI breaks unnecessary for a very long time and we've been continuously told that there are no plans whatsoever for a Qt 5 any time soon, almost up to the day before the Qt 5 announcement. But I realize that this is out of KDE's control.) there is apparently a fundamental misunderstanding here: KDE Platform 4.7 is being maintained. but only maintained. if such a spell checker thing happened right now, we could enable that in the 4.7 branch. many of us are still committing bug fixes and other improvements to that branch, which are then merged regularly into the frameworks branch. what we aren't doing it focussing new feature development efforts in the 4.7 branch. that is happening in frameworks instead, but 4.7 is still very much maintained. as such, there should be zero reason for downstream packaging efforts to require patching bugfixes into their builds themselves. Back when the Hunspell patches for kdelibs 3.5 and for the K3Spell compatibility API were discussed, the Sonnet maintainer argued that adding support for a new spell checker is a feature, not a bugfix, and to be honest, I think he's right. which begs the very good question: why aren't they? they should be the same people. we need the 5.0 release, and that will happen faster if we don't split our resources. remember how long the 4.0 development took, in part because we kept equivocating between more 3.x releases and getting on with 4.0? I really don't buy that doing 3.x releases is what made 4.0 take so long. I think 4.0 took so long because 4.0 was just a gigantic change. Having 3.5 made the long wait for 4.x bearable, and I think we should have done at least a 3.6 (of the whole SC), too (before 4.0). And probably a kdelibs 3.7 (after 4.0, maybe even after 4.1 or 4.2, those releases aren't needed every 6 months) for all those applications not ported yet at that time, to improve integration into KDE Platform 4 (e.g. a style which looks as much like Oxygen as technically possible would have been helpful, or a backport of the Oxygen icon theme to the old icon names) and backport features where possible without breaking compatibility. OTOH, I'm very much interested in working on 4.8, and in fact I already did (see my Plasma PackageKit GSoC work), but my patches are now stuck in reviewboard and cannot be committed due to the deep freeze. they can be committed to the frameworks branch. in my eyes, those patches are not 4.x efforts at all, but things that belong in frameworks. I have to strongly disagree with that. I developed those for 4.x and 4.x only. Of course I don't want that feature to get lost with the move to 5.x, so I will forward-port it, but I don't see why this should not be available in 4.x, just because it was developed a couple months after your arbitrary cutoff date (the 4.7 feature freeze) which was not announced (as the final cutoff for kdelibs 4.x, as opposed to just 4.7, features) in advance. Refusing the change in 4.x is also going to make me LESS motivated to do the forward-port for frameworks. Do you really want to encourage wild patching by distributions, adding or backporting a patchwork (pun intended) of features? of course not. and imho collaborative packaging efforts would mean not threatening upstream development in this way. We aren't threatening anybody by offering features to our users. I remember Aaron Seigo complaining on his blog about the feature backports done by distributions in 4.0, 4.1 and 4.2 era, but if you aren't going to allow any new features upstream, you will be leaving us no choice. the choice that packagers have is to actually work with us instead of against us. the idea that it sucked when we did it before, and now we're going to do it again! is a little twisted. :) the features you offered as needing
Re: The case for a kdelibs 4.8
PS: Aaron J. Seigo wrote: the choice that packagers have is to actually work with us instead of against us. We would very much love to work with you. In fact, this is why I submitted my patches to KDE ReviewBoard before even getting them into Rawhide. I really WANT these to be upstream. Fedora doesn't like carrying downstream patches as a general principle. However, working with you is only possible if you are also interested in working with us, which implies listening to our needs, concerns and wishes. By closing down the branch where our current development is necessarily focused on since that's what we will be shipping in the near future, you're already starting down the wrong road. For those Plasma PackageKit features, sure, they're not strictly required, which is why we got away without them until now. But to our users, they mean that installing a widget through download new widgets… will actually install the required dependencies, so the widget they downloaded will actually WORK. What, to us developers, is a feature, actually fixes a bug from a user's perspective. We, the distributions, interact directly with users, and so are receptive to their needs, concerns and wishes, and tend to base ours on them. And finally, I strongly doubt that my features are the only post-4.7 kdelibs features running into the freeze. (In fact, what's now going on with kactivities and nepomuk proves quite the opposite.) Kevin Kofler
Re: The case for a kdelibs 4.8
PPS: I wrote: However, working with you is only possible if you are also interested in working with us, which implies listening to our needs, concerns and wishes. By closing down the branch where our current development is necessarily focused on since that's what we will be shipping in the near future, you're already starting down the wrong road. For a prime example of what happens if you close down the version distributions are actually shipping in favor of long-term development, just look at GRUB 1. Fedora's GRUB 0.97 package is now at patch level 75! And it has its own git repository, i.e. essentially a fork (grub-fedora on git.kernel.org, it's now down because all of git.kernel.org is currently down). GRUB 2 is now finally becoming production-ready, and Fedora will be switching to it in Fedora 16, but it was nowhere near that state when upstream discontinued GRUB 1. GRUB 0.97 was released on May 8, 2005 (and IIRC that was only a bugfix release and they stopped accepting new features even earlier). Distributions have been forced to ship patched versions for over 6 years. (GRUB 0.97 as released doesn't support many needed features, e.g. ext4.) As a result, it has become a morass of all-different forked versions. You can't tell a priori what any given GRUB 0.97 package actually supports. Sure, focusing on the long term to some extent is needed, but we distributors need something we can ship NOW, not months or years from now. Kevin Kofler
Re: The case for a kdelibs 4.8
On Saturday, 1 de October de 2011 12:58:01 Kevin Kofler wrote: (And I know that Qt is also breaking the ABI. That's something I also can't agree with, especially considering that Qt 4 was advertised as the big ABI break which will make new ABI breaks unnecessary for a very long time and we've been continuously told that there are no plans whatsoever for a Qt 5 any time soon, almost up to the day before the Qt 5 announcement. But I realize that this is out of KDE's control.) Between the 4.0 and 5.0 release, it will have been 7 years. I think it's long enough. KDE could take Qt 4.8, fork it and continue developing it. Any show of hands? As for the announcing of the plans, there were no plans to do Qt 5 until there were plans. Once there were plans, they were announced. -- Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org Software Architect - Intel Open Source Technology Center PGP/GPG: 0x6EF45358; fingerprint: E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358 signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
On Saturday, October 01, 2011 08:27:02 AM Martin Gräßlin wrote: On Saturday 01 October 2011 00:12:05 Arne Babenhauserheide wrote: Am Freitag, 30. September 2011, 10:07:27 schrieb Aaron J. Seigo: will say Platform 4.7, Plasma Workspaces 4.8 and application updates (or something along those lines). that was not just a marketing ploy, but an attempt to align our public communication with the realities that already existed in KDE development. I hope I am not the first to note that this sounds really horrible. Take this message: KDE releases 4.8: Platform and Workspaces got some spectacular changes, while applications received much polish. In a blog, this becomes YAY! KDE releases 4.8! Take your example. It becomes: Plasma 4.8 released! Well, where is KDE in that? Suddenly the community it is all about becomes a sidenote. And a newspaper will likely only see “hm, some stuff our readers won’t understand” instead of “new version of the tools from KDE!” There is a reason why Apple releases MacOSX 10.8 and not “Xcode 4.1, Apple Mach 1.3, Quartz 4.7, …” One of the main reasons for the rebranding was to realize that KDE is not one product, but a community that produces multiple products among them a desktop environment (Plasma). What you just try to tell us is that the complete rebranding is nonsense and we should go back to talking just about KDE for everything bringing the users back to assuming they need Plasma in order to use KMail. I really like the fact that finally we have a release where it will be clear that KDE is the community and not one large product. All those who did not get it yet, might finally understand it if we write a really good release announcement. I understand it and I think it's great. At the same time, I think this idea that from a development perspective it's OK to mix and match and release any piece of the SC at any given time is technical nonsense. As long as we leave the marketing to the marketeers and the developers don't get distracted by it, then I think it's fine. I think users ought to be able to run KMail (to pick your example) on any desktop environment and have it work well. I completely agree with this idea, but I don't agree with the idea to throw away the coordinated development schedule and release process. Scott K
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 18:11:12 Kevin Kofler wrote: Wait and see the chaos that will come up when users open their Help/About in Konqueror and it tells them that they're using Konqueror 4.8.0 under KDE 4.7.6. (And yes, it still says only KDE in 4.6.5, I haven't checked 4.7 or 4.8 for whether that's fixed there.) it says KDE Development Platform x.y.z. The rule so far has always been that the kdelibs version must be the same as the KDE SC version. There is no rule – at most a common practise A common practice which had been enforced by minimum required kdelibs versions in CMakeLists.txt. for, e.g. kde-workspace, that minimum version for kdelibs is 4.7. it may be even lower for other modules (many apps probably compile with rather older kdelibs than that). this really isn't an argument for or against a 4.8 .. -- Aaron J. Seigo humru othro a kohnu se GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA EE75 D6B7 2EB1 A7F1 DB43 KDE core developer sponsored by Qt Development Frameworks signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 23:57:53 Scott Kitterman wrote: I don't like the fact that KDE developers decided to ignore their own policy on maintenance updates. I think it breaks your contract with your downstreams. In the case of what's been done so far, it doesn't have an impact on us. The changes were in before our release and are technically the features that got into the 4.7 branch to date have been things that were already worked on before the Frameworks decision was made. it's was an odd cas were features had been worked on while 4.7 was frozen with the expectation of a 4.8 ... and that left us with the choice of having a 4.8 release which we didn't want for those few features, leaving those features out and screwing up the planning of the applications depending on those changes or fudging a little bit. it was a compromise choice made to mimize downside. as with all such compromises, it isn't perfect (ergo the label of 'compromise'), but after looking at the issues and looking at different scenarios we decided this was the best option available to us compared to the others. note that some of the more actively changing and developed code in kdelibs have moved to separate git repositories already to make this issue moot for them (kactivities, nepomuk) -- Aaron J. Seigo humru othro a kohnu se GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA EE75 D6B7 2EB1 A7F1 DB43 KDE core developer sponsored by Qt Development Frameworks signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 23:57:56 Kevin Kofler wrote: https://git.reviewboard.kde.org/r/102175/ https://git.reviewboard.kde.org/r/102291/ https://git.reviewboard.kde.org/r/102350/ none of these are critical feature additions. they elevate the usability of libplasma and are valuable, but we've lived just fine without these features to date. if we were working on and releasing a 4.8 version of kdelibs, our users would have to wait for these features until 4.8 was released (e.g. in january) and distributions made packages available to their users (often much later than that). instead, they'll have to wait until frameworks is ready. and this code can, and should, go into libplasma in the frameworks branch today. -- Aaron J. Seigo humru othro a kohnu se GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA EE75 D6B7 2EB1 A7F1 DB43 KDE core developer sponsored by Qt Development Frameworks signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 14:01:50 Kevin Kofler wrote: 1. This puts kdelibs 4 into maintenance mode even before KDE Frameworks 5 is anywhere near a release, let alone versions of the workspace and applications actually using it. As a result, we will fail to deliver new features to our end users for months if not years. we will deliver no new features in kdelibs, yes. but certainly we will be shipping new features in applications (including the workspaces) using those libraries. so our users will indeed be getting new features and we will indeed be improving the software our users get. bug fixes and performance improvements also continue, and those are things that our users really do care about as well. features are valuable, but they aren't the only thing of value. as 4.7 is still being actively maintained, even kdelibs is very much still being improved from a user's POV. And no, kdelibs features do not only target developers: Some application or workspace features require kdelibs changes, or in some cases, even consist of kdelibs changes only, e.g. my Plasma PackageKit integration work is entirely in kdelibs (libplasma). many application and workspace features do not require changes in kdelibs. though, as you note, some do. there are already many improvements in libplasma2 that the workspaces will only get to take advantage of when frameworks is released. in the meantime, we're still rolling out new features and even whole new shells (the Contour work and plasma-device shell, for instance) using the 4.7.x platform libraries. so some features will indeed have to wait .. and that's not a horrible thing because it means that frameworks will get more developer attention and the attention it is getting already will not be slowed even further by having to deal with bringing over features from 4.7.x into the re-arranged libraries in frameworks. the result will be that the time between now and 5.0 libs being available will be shortened, which is exactly what we, as application developers, need. 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. The rule so far has always been that the kdelibs version must be the same as the KDE SC version. that's not a rule anymore, and hasn't been for a while now. other than the packaging specfile changes that seem to be required in the Fedora packages, i'm not sure what the real world confusion would be. when we release apps and workspace 4.8, we'll also do a platform 4.7.x release. the release communication will not say SC 4.8 it will say Platform 4.7, Plasma Workspaces 4.8 and application updates (or something along those lines). that was not just a marketing ploy, but an attempt to align our public communication with the realities that already existed in KDE development. backported to the compatibility libraries as well. For example, when we switched our default spell checker in Fedora from aspell to hunspell in Fedora 9 (i.e. 4.0 era), I had to add support for hunspell to kdelibs3, or our users would have had to install 2 spell checkers to use KDE apps! (Even several apps in the default KDE installation hadn't been ported to kdelibs4 yet in Fedora 9.) That change never got upstreamed because kdelibs3 is no longer maintained, yet it would have been useful to many distributions. there is apparently a fundamental misunderstanding here: KDE Platform 4.7 is being maintained. but only maintained. if such a spell checker thing happened right now, we could enable that in the 4.7 branch. many of us are still committing bug fixes and other improvements to that branch, which are then merged regularly into the frameworks branch. what we aren't doing it focussing new feature development efforts in the 4.7 branch. that is happening in frameworks instead, but 4.7 is still very much maintained. as such, there should be zero reason for downstream packaging efforts to require patching bugfixes into their builds themselves. 4. The core developers, for whose convenience this change was made, are not the only ones working or willing to work on kdelibs. it wasn't convenience, which sort of makes it sound like we weren't trying hard enough ;), but necessity. the necessity is driven by resource constraints and the need for changes in kdelibs that require changes to the ABI (and which will happen anyways with Qt5) The main reason that was given for dropping kdelibs 4.8 is that this means one less branch to worry about for the people working on kdelibs, but the people who'd work on 4.8 and 5.0 are NOT the same people! which begs the very good question: why aren't they? they should be the same people. we need the 5.0 release, and that will happen faster if we don't split our resources. remember how long the 4.0 development took, in part because we kept equivocating between more 3.x releases and getting on with 4.0? I understand that the developers who are
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011 14:01:50 Kevin Kofler wrote: Hi, as you probably already know, a decision was recently made that kdelibs 4.7 would be the last 4.x release series of kdelibs, and work would be ongoing in the 5.0 (frameworks) and 4.7 (KDE/4.7) branches only. I think this is a huge mistake, for several reasons (the TL/DR crowd can stop right here, but if you want to know why, please read on!) [..] From what I remember from the desktop summit the picture you draw here is quite an exaggeration of what is actually happening. kdelibs 4.7 is meant to be frozen for new features, but not for bugfixes. Bugfix releases of kdelibs-4.7 happenend and I'm sure will continue on happening. As for the versioning I don't see why one of those bugfix releases couldn't be rebranded as 4.8.0 if that makes things easier (that was even briefly mentioned at the release team BoF). It does not solve feature backports of course. The KDE Frameworks 5.0 development is not meant to take forever. In fact I think it's meant to be finished around early 2012, which would leave us with a frozen kdelibs for one KDE SC release, maximum two. It's also not a huge *we will break everything! Kittens will die!* release, but basically just a restructuration of the code, with little to no adjustments necessary for applications. (that was how it was marketed, anyway). The way I understood it was that there could very well be a KDE SC 4.9 release shipping a 4.9 workspace on top of 5.0 frameworks. As for the version numbers, I don't really see the problem. As long as the integrity of the KDE SC releases remains, different version numbers in its components are not really /that/ important. I do see how it could help though. Grs, Heinz signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8 (fwd)
Forwarding my answer to kde-core-devel in order to give us packagers a voice outside our own closed mailing list. Cheers, Eric -- Eric Hameleers al...@slackware.com Jabber: al...@jabber.xs4all.nl -- Forwarded message -- Date: Thu, 29 Sep 2011 10:37:20 -0700 (PDT) From: Eric Hameleers al...@slackware.com To: Wulf C. Krueger philant...@exherbo.org Cc: kde-packa...@kde.org Subject: Re: The case for a kdelibs 4.8 On Thu, 29 Sep 2011, Wulf C. Krueger wrote: On 29.09.2011 14:01, Kevin Kofler wrote: [Lots of excellent explanations deleted] So I am urging you to reconsider your decision and reopen master for those people willing to work on 4.8. It's not too late yet, there's a month left until the soft feature freeze for KDE SC 4.8. I totally agree with Kevin. This would lead to another packaging nightmare and would be confusing to users as well. Kevin's plea has my fullest and strongest support. Cheers, Eric -- Eric Hameleers al...@slackware.com Jabber: al...@jabber.xs4all.nl ___ Kde-packager mailing list kde-packa...@kde.org https://mail.kde.org/mailman/listinfo/kde-packager
Re: The case for a kdelibs 4.8
On Thursday, 2011-09-29, Rolf Eike Beer wrote: The reason to stop master was (as far as I understand) to make the frameworks branch easily to maintain. If someone is working on 4.8 (bugfixing, features) all this has to be ported to frameworks, too. So you develop a moving target on a moving target. What about a more stricter than usual policy for additions to 4.8? Like every new feature wanted in 4.8 has to go through reviewboard for 4.8 and frameworks 5. That way the developer doing the feature work for 4.8 will also take care of the forward porting task. Cheers, Kevin -- Kevin Krammer, KDE developer, xdg-utils developer KDE user support, developer mentoring signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
On Fri, Sep 30, 2011 at 10:07 AM, Aaron J. Seigo ase...@kde.org wrote: On Thursday, September 29, 2011 14:01:50 Kevin Kofler wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. The rule so far has always been that the kdelibs version must be the same as the KDE SC version. that's not a rule anymore, and hasn't been for a while now. other than the packaging specfile changes that seem to be required in the Fedora packages, i'm not sure what the real world confusion would be. when we release apps and workspace 4.8, we'll also do a platform 4.7.x release. the release communication will not say SC 4.8 it will say Platform 4.7, Plasma Workspaces 4.8 and application updates (or something along those lines). that was not just a marketing ploy, but an attempt to align our public communication with the realities that already existed in KDE development. I am not saying this is the best solution, I am just putting out as a possible compromise. Would it be possible to just call kdelibs 4.7.8 or whatever it ends up being kdelibs 4.8, without adding any significant new features? I know you can, by KDE policy, add new features at a x.x release, but is there any requirement that you have to? I know there are numerous other issues involved so this may be a terrible idea. -Todd
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011 20:01:00 Kevin Kofler wrote: But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. But who is (or would be) working on features in kdecore | kdeui | kio | kfile? Feel free to prove me wrong, but with most kdelibs developers working on either bugfixing or frameworks, I don't see who's left to work on features in kdelibs. I agree that e.g. plasma and nepomuk might be in a different situation though. Maybe with a proper git merge support for kdelibs master we could re-evaluate the decision to freeze it, it's less work than having to cherry-pick every fix into 3 branches. -- David Faure, fa...@kde.org, http://www.davidfaure.fr Sponsored by Nokia to work on KDE, incl. Konqueror (http://www.konqueror.org).
Re: The case for a kdelibs 4.8
On Donnerstag 29 September 2011 18:11:12 Kevin Kofler wrote: Wait and see the chaos that will come up when users open their Help/About in Konqueror and it tells them that they're using Konqueror 4.8.0 under KDE 4.7.6. (And yes, it still says only KDE in 4.6.5, I haven't checked 4.7 or 4.8 for whether that's fixed there.) True but only in one line ans the explanation that the Dev Platform is meant, is written in About KDE. That said, I'm toying with ideas about a complete redesign of the About window for KF5. When/If I come up with a somewhat presentable mock-up, I'll post it and hopefully a programmer implements something based on that. A completely new About window will also fix legacy issues like that string. and strangely shipping Kontact 4.4.11 with SC 4.5+ was also not confusing to most users. Actually, that was (and still is) quite confusing to many users. Look at the mailing lists and forums. I really don't think it was the majority and the Kontact case is user-visible software unlike Frameworks/Platform which normal users usually do not care about. MS Office 2003 also ran on Windows 2000 and the majority of users had no problems understanding that. ;-) And we better continue to educate our users that Platform version and Workspace version does not need to match. I have yet to see any ideas for KDE Plasma Desktop 5.0 which means that maybe KPD 4.9 in summer 2012 will require KF 5.0. So unless someone completely rewrites KPD for the summer release (maybe basing it on Plasma Active) and therefore justifying a major version jump to 5.0, as a promo person I'd rather educate our users now before we see people (at worst: reviewers) complaining that KDE 5 looks just like KDE 4.8. (As a side note I also think that KDE Applications should completely lose their version number and get date-based versioning because any application can get major new features at any time – see Dolphin 2.0 in SC 4.8.) Fedora 15 actually has KDE SC 4.6.x, not 4.7. Just a few days ago someone on your mailing list claimed otherwise but whatever... that's not the main topic here.
Re: The case for a kdelibs 4.8
On Friday, September 30, 2011 04:15:51 PM Markus Slopianka wrote: (As a side note I also think that KDE Applications should completely lose their version number and get date-based versioning because any application can get major new features at any time – see Dolphin 2.0 in SC 4.8.) Today, for Kubuntu, we work very hard to package all of the KDE SC because we think it's an important value to deliver this upstream set of packages as a complete set to our users (IIRC, except for bindings we accomplished that for 4.7 even with all the package splits). If there is no more integrated release of a KDE SC, then we really don't need to worry about that and so we'll probably focus more just on stuff we use (I know the Debian KDE packagers are already planning this due to all the package splits). So I expect that one side effect of doing away with the traditional KDE release process will be less KDE software packaged in distributions. Scott K
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011, Andras Mantia wrote: On Thursday, September 29, 2011 21:43:34 Stefan Majewsky wrote: Without judging on the other arguments which look very reasonable to me... On Thu, Sep 29, 2011 at 2:01 PM, Kevin Kofler kevin.kof...@chello.at wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. [...] ...what exactly stops you from advertising kdelibs 4.7.x as version 4.8? (And I mean advertise, so only the user-visible parts, not version numbers in .so files or whatever.) Now that would add a lot of confusion. Can the version number of kdelibs 4.7.x simply be increased to 4.8 when 4.8 is branched ? It would have only bugfixes, but it seems it would make life simpler for some people. Alex
Re: Re: The case for a kdelibs 4.8
A Dijous, 29 de setembre de 2011, Scott Kitterman vàreu escriure: On Thursday, September 29, 2011 11:47:22 PM Albert Astals Cid wrote: A Dijous, 29 de setembre de 2011, Scott Kitterman vàreu escriure: On Thursday, September 29, 2011 08:01:00 PM Kevin Kofler wrote: On Thursday 29 September 2011, Heinz Wiesinger wrote: From what I remember from the desktop summit the picture you draw here is quite an exaggeration of what is actually happening. kdelibs 4.7 is meant to be frozen for new features, but not for bugfixes. Bugfix releases of kdelibs-4.7 happenend and I'm sure will continue on happening. As for the versioning I don't see why one of those bugfix releases couldn't be rebranded as 4.8.0 if that makes things easier (that was even briefly mentioned at the release team BoF). It does not solve feature backports of course. But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. Even worse, features have already creeped into the 4.7 branch because they are needed and there's no 4.8 branch, so this isn't a theorectical point. Why is that bad, that is what was agreed when the freeze took place. Agreed on by who? Read this list Plan to transition to KDE Frameworks. Albert
Re: Re: The case for a kdelibs 4.8
A Divendres, 30 de setembre de 2011, Aaron J. Seigo vàreu escriure: On Thursday, September 29, 2011 23:57:56 Kevin Kofler wrote: https://git.reviewboard.kde.org/r/102175/ https://git.reviewboard.kde.org/r/102291/ https://git.reviewboard.kde.org/r/102350/ none of these are critical feature additions. they elevate the usability of libplasma and are valuable, but we've lived just fine without these features to date. if we were working on and releasing a 4.8 version of kdelibs, our users would have to wait for these features until 4.8 was released (e.g. in january) and distributions made packages available to their users (often much later than that). instead, they'll have to wait until frameworks is ready. and this code can, and should, go into libplasma in the frameworks branch today. I sincerely think you are being unfair here. The stuff you needed was merged in, and probably was more invasive than those features. You justify yourself in that it was developed before we knew 4.8 was not going to be there. I'd say that can easily qualify for Kevin too. Albert -- Aaron J. Seigo humru othro a kohnu se GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA EE75 D6B7 2EB1 A7F1 DB43 KDE core developer sponsored by Qt Development Frameworks
Re: Re: The case for a kdelibs 4.8
A Divendres, 30 de setembre de 2011, Alexander Neundorf vàreu escriure: On Thursday 29 September 2011, Andras Mantia wrote: On Thursday, September 29, 2011 21:43:34 Stefan Majewsky wrote: Without judging on the other arguments which look very reasonable to me... On Thu, Sep 29, 2011 at 2:01 PM, Kevin Kofler kevin.kof...@chello.at wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. [...] ...what exactly stops you from advertising kdelibs 4.7.x as version 4.8? (And I mean advertise, so only the user-visible parts, not version numbers in .so files or whatever.) Now that would add a lot of confusion. Can the version number of kdelibs 4.7.x simply be increased to 4.8 when 4.8 is branched ? It would have only bugfixes, but it seems it would make life simpler for some people. As I already said in this thread, that is Dirk's plan as he stated in the Release Team BoF in Berlin. Albert Alex
Re: The case for a kdelibs 4.8
Am Freitag, 30. September 2011, 10:07:27 schrieb Aaron J. Seigo: will say Platform 4.7, Plasma Workspaces 4.8 and application updates (or something along those lines). that was not just a marketing ploy, but an attempt to align our public communication with the realities that already existed in KDE development. I hope I am not the first to note that this sounds really horrible. Take this message: KDE releases 4.8: Platform and Workspaces got some spectacular changes, while applications received much polish. In a blog, this becomes YAY! KDE releases 4.8! Take your example. It becomes: Plasma 4.8 released! Well, where is KDE in that? Suddenly the community it is all about becomes a sidenote. And a newspaper will likely only see “hm, some stuff our readers won’t understand” instead of “new version of the tools from KDE!” There is a reason why Apple releases MacOSX 10.8 and not “Xcode 4.1, Apple Mach 1.3, Quartz 4.7, …” Technically it is cool to have a kmail 5.8 which depends on frameworks =5.1, but as a User I am interested in the changes KDE did for 5.8: All the changes, to Frameworks as well as to the Workspaces and the Applications. If I then decide to not use the testing frameworks, it makes me happy when an application only needs an earlier version. But do you actually think that most people can follow the info-input of having different version numbers in a release - and really remember the news? Due to 3 hours daily travel time, 40h work per week and wife and son, I won’t be able to do that in the near future, so there are likely many more people who can’t. Best wishes, Arne -- singing a part of the history of free software: - http://infinite-hands.draketo.de signature.asc Description: This is a digitally signed message part.
Re: The case for a kdelibs 4.8
For example, when we switched our default spell checker in Fedora from aspell to hunspell in Fedora 9 (i.e. 4.0 era), I had to add support for hunspell to kdelibs3, or our users would have had to install 2 spell checkers to use KDE apps! (Even several apps in the default KDE installation hadn't been ported to kdelibs4 yet in Fedora 9.) That change never got upstreamed because kdelibs3 is no longer maintained, yet it would have been useful to many distributions. Please keep the life cycle of your software for distributors and end users in mind when you decide your schedules, not just the convenience of a few core developers. Given that it is now proven and tested code, who stops you committing it into KDE/3.5 branch? 4. The core developers, for whose convenience this change was made, are not the only ones working or willing to work on kdelibs. The main reason that was given for dropping kdelibs 4.8 is that this means one less branch to worry about for the people working on kdelibs, but the people who'd work on 4.8 and 5.0 are NOT the same people! I understand that the developers who are pushing forward towards the new frameworks are not interested in 4.8, but you should understand that many other KDE contributors are not interested at all in 5.0 at this time. I, for one, will start caring about 5.0 the day some application (be it a Plasma workspace or a regular application) using it will enter Rawhide (Fedora's trunk), and I guess other distro folks are feeling the same. OTOH, I'm very much interested in working on 4.8, and in fact I already did (see my Plasma PackageKit GSoC work), but my patches are now stuck in reviewboard and cannot be committed due to the deep freeze. Do you really want to encourage wild patching by distributions, adding or backporting a patchwork (pun intended) of features? I remember Aaron Seigo complaining on his blog about the feature backports done by distributions in 4.0, 4.1 and 4.2 era, but if you aren't going to allow any new features upstream, you will be leaving us no choice. The reason to stop master was (as far as I understand) to make the frameworks branch easily to maintain. If someone is working on 4.8 (bugfixing, features) all this has to be ported to frameworks, too. So you develop a moving target on a moving target. Not that I do not understand your point (I'm not sure if I have an opinion on this yet), just to get a common understanding about the reasons. I would have wished more than once if we could have done a new release on an older branch, e.g. a KDE 4.6.6 where the KLineEdits in Konqueror would be fixed. Some sort of long-term branch, or at least one maintenance release more when the .0 or .1 is out. You only get severe end user testing on the new major release when the .0 is out, so those who don't want to break their system will stay on the older branch, which has bugs long fixed in the new one. Some sort of chicken and egg problem, of course, but it would be helpful for those people to get an additional release with bugfixes after the next .0 is out. This could be e.g. in the middle between .0 and .1 to not have the release workload for 2 releases at once. And: yes, I know, this is even more work for the release team. Poor Dirk. Eike
Re: The case for a kdelibs 4.8
On Donnerstag 29 September 2011 14:01:50 Kevin Kofler wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. Since almost exactly 2 years we (esp. the promo team) are communicating that Platform/Frameworks, Applications and Workspaces are three separate products that just happen to be shipped on the same date. One of the reasons why the rebranding still didn't get to everybody is that some distributions (mostly Fedora!) still spread the wrong message about some “KDE 4.7” to its users. (see eg. https://fedoraproject.org/wiki/SIGs/KDE/KDE47Packaging#KDE_4.7_Packaging ) Users of other distributions do not have such problems and strangely shipping Kontact 4.4.11 with SC 4.5+ was also not confusing to most users. The rule so far has always been that the kdelibs version must be the same as the KDE SC version. There is no rule – at most a common practise and that was broken as well after Fedora 15 has packages for SC 4.7 with Kontact 4.4. Changing this will also break all our Fedora KDE SC specfiles Then fix your specfiles. AFAIK Fedora is also the last big distribution that still has monolithic packages like kdemultimedia, kdenetwork, and so on. Take the opportunity to fix that as well. openSUSE doesn't seem to have problems packaging 4.8 already: https://build.opensuse.org/project/show?project=KDE%3AUnstable%3ASC And what will kde4-config --kdeversion output? The version of kdelibs? Of course. What else? Want to see the Plasma version, enter plasma-desktop --version If you want the version of Dolphin, enter dolphin --version and so on. Users are going to be confused: Hey, I installed KDE SC 4.8, why does it say 4.7? Then you refer them to Wikipedia which holds that info since quite some time: https://secure.wikimedia.org/wikipedia/en/wiki/KDE_Software_Compilation_4#KDE_Plasma_Workspaces_and_Applications_4.8 you will break Fedora packages even further. Then – again – fix Fedora. The main reason that was given for dropping kdelibs 4.8 is that this means one less branch to worry about for the people working on kdelibs, but the people who'd work on 4.8 and 5.0 are NOT the same people! Do you hereby declare that Red Hat will provide resources to create KDE Platform 4.8 and an cherrypick useful developments from the frameworks branch to Platform 4.8?
Re: The case for a kdelibs 4.8
Rolf Eike Beer wrote: (re the support for spellchecking with hunspell) Given that it is now proven and tested code, who stops you committing it into KDE/3.5 branch? What for? There are no plans to do a 3.5.11 or 3.6.0 release, ever, and the one major distribution known to sometimes ship release branch snapshots from after the last release of the branch (Debian) just dropped the KDE 3 libs from their distribution (a decision I also don't agree with, FWIW, but that's off topic here). In addition, that patch is not exactly a bugfix… The 2 patches against 3.5.10 (one for KSpell and one for KSpell2) can be found in Fedora dist-git, and also in KDE Bugzilla. If you think they should be in the 3.5 branch, feel free to commit them. But AFAIK, we aren't really expected to commit anything to the 3.5 branch anymore, even though AFAIK there's no official freeze like the one being enforced on master. By the way, I also have a patch for K3Spell in the KDE 4 libs. We also ship that patch in Fedora. It's basically the same as the KSpell patch for the KDE 3 libs, and attached to the same bugs.kde.org report. But I don't know whether anything actually still uses K3Spell. (The reason I didn't commit that was because the Sonnet maintainer wasn't happy with the idea of adding a new feature to that compatibility API, even though it's needed for integration. And of course now I couldn't commit it anymore even if I wanted because kdelibs master is frozen, which brings us back to the topic…) The reason to stop master was (as far as I understand) to make the frameworks branch easily to maintain. If someone is working on 4.8 (bugfixing, features) all this has to be ported to frameworks, too. So you develop a moving target on a moving target. You'd just have to do regular merges from master instead of regular merges from KDE/4.7. It wouldn't change much apart from being more consistent with common best practices for SCM use. I would have wished more than once if we could have done a new release on an older branch, e.g. a KDE 4.6.6 where the KLineEdits in Konqueror would be fixed. Some sort of long-term branch, or at least one maintenance release more when the .0 or .1 is out. You only get severe end user testing on the new major release when the .0 is out, so those who don't want to break their system will stay on the older branch, which has bugs long fixed in the new one. Some sort of chicken and egg problem, of course, but it would be helpful for those people to get an additional release with bugfixes after the next .0 is out. This could be e.g. in the middle between .0 and .1 to not have the release workload for 2 releases at once. And: yes, I know, this is even more work for the release team. Poor Dirk. That also makes a lot of sense (so +1 to this suggestion), but it is a different issue from the one at hand. Kevin Kofler
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 03:27:58 PM Markus Slopianka wrote: On Donnerstag 29 September 2011 14:01:50 Kevin Kofler wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. Since almost exactly 2 years we (esp. the promo team) are communicating that Platform/Frameworks, Applications and Workspaces are three separate products that just happen to be shipped on the same date. One of the reasons why the rebranding still didn't get to everybody is that some distributions (mostly Fedora!) still spread the wrong message about some “KDE 4.7” to its users. (see eg. https://fedoraproject.org/wiki/SIGs/KDE/KDE47Packaging#KDE_4.7_Packaging ) Users of other distributions do not have such problems and strangely shipping Kontact 4.4.11 with SC 4.5+ was also not confusing to most users. We did this in Kubuntu and it was confusing. It was also technically challenging. Speaking as someone investing a lot of time in trying to do a high quality job of distributing KDE to end users: Please. Never, ever, do this to us again. I get what you want to communicate, but it's more complicated that that. The integrated release process of KDE has value. Just happen to be shipped on the same date isn't true. Marketing materials saying this are exactly that. Marketing materials. I agree with the goal of trying to make it easier to use pieces of KDE in different contexts, but the integrated delivery process that marketing wants to dispense with has real technical value that I hope KDE will not discard. The transition from kdepim 4.4 - 4.6/4.7 was a special case. I hope it stays that way. While there wasn't another way to do what had to be done in that case, it should not server as a general success story others should model themselves after.
Re: The case for a kdelibs 4.8
Markus Slopianka wrote: On Donnerstag 29 September 2011 14:01:50 Kevin Kofler wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. Since almost exactly 2 years we (esp. the promo team) are communicating that Platform/Frameworks, Applications and Workspaces are three separate products that just happen to be shipped on the same date. But now you're changing exactly that last part, or at least what will be released will no longer have a common and consistent version number. And in any case, even if this particular argument doesn't convince you, that does not invalidate all the other reasons why there should be a kdelibs 4.8. One of the reasons why the rebranding still didn't get to everybody is that some distributions (mostly Fedora!) still spread the wrong message about some “KDE 4.7” to its users. (see eg. https://fedoraproject.org/wiki/SIGs/KDE/KDE47Packaging#KDE_4.7_Packaging ) This is not a page targeted at our users at all, it's an internal KDE SIG TODO list. We sometimes use shorthand when we discuss things, it is obvious to us that what's really meant here is the KDE SC. But I changed it to KDE SC 4.7 Packaging to make you happy. And no, we cannot use a more specific term than KDE SC here, because this is really about packaging all the stuff that happens to be shipped on the same date, with a 4.7.n version number, by the KDE Project. The modules are released together, so they get packaged together. Users of other distributions do not have such problems Wait and see the chaos that will come up when users open their Help/About in Konqueror and it tells them that they're using Konqueror 4.8.0 under KDE 4.7.6. (And yes, it still says only KDE in 4.6.5, I haven't checked 4.7 or 4.8 for whether that's fixed there.) and strangely shipping Kontact 4.4.11 with SC 4.5+ was also not confusing to most users. Actually, that was (and still is) quite confusing to many users. Look at the mailing lists and forums. I think this was a mistake that shouldn't be repeated. What should have been happened when the Akonadi stuff was found not ready would have been a: svn rm branches/KDE/4.5/kdepim svn cp branches/KDE/4.4/kdepim branches/KDE/4.5/kdepim (and likewise for kdepim-runtime. And yes, SVN was still used at that time), i.e. mass-revert the unfinished stuff and do 4.5 releases of what works. (But the updated KAddressBook from 4.5 should have been used. The 4.4 one was already Akonadi-based and unfinished, and we ended up stuck with that temporary solution for over a year. It was a mistake to put that into 4.4 in the first place, but with that done, it should have been updated. Renumbering the releases to 4.5 instead of continuing with 4.4.x would have allowed inserting such a new feature.) The rule so far has always been that the kdelibs version must be the same as the KDE SC version. There is no rule – at most a common practise A common practice which had been enforced by minimum required kdelibs versions in CMakeLists.txt. and that was broken as well after Fedora 15 has packages for SC 4.7 with Kontact 4.4. Fedora 15 actually has KDE SC 4.6.x, not 4.7. The upcoming Fedora 16 has KDE SC 4.7.x including kdepim 4.7.x, so the version numbers are finally consistent again. Changing this will also break all our Fedora KDE SC specfiles Then fix your specfiles. AFAIK Fedora is also the last big distribution that still has monolithic packages like kdemultimedia, kdenetwork, and so on. Take the opportunity to fix that as well. If you had actually read that KDE47Packaging page you linked to and not just the title, you'd know that we're actually shipping some modules split in the upcoming Fedora 16. Not only are the modules released as split tarballs now all also packaged in split form, but we also split e.g. kdeutils into subpackages. The main reason that was given for dropping kdelibs 4.8 is that this means one less branch to worry about for the people working on kdelibs, but the people who'd work on 4.8 and 5.0 are NOT the same people! Do you hereby declare that Red Hat will provide resources to create KDE Platform 4.8 and an cherrypick useful developments from the frameworks branch to Platform 4.8? I cannot declare that because I don't work for Red Hat (let alone in some management position). I'm just a community packager for Fedora. Ask Red Hat directly. Kevin Kofler
Re: The case for a kdelibs 4.8
Scott Kitterman wrote: We did this in Kubuntu and it was confusing. It was also technically challenging. Speaking as someone investing a lot of time in trying to do a high quality job of distributing KDE to end users: Please. Never, ever, do this to us again. +1 The transition from kdepim 4.4 - 4.6/4.7 was a special case. I hope it stays that way. While there wasn't another way to do what had to be done in that case Actually, there would have been another way: revert the Akonadi merge by copying the 4.4 branch of kdepim to 4.5 and release that as kdepim 4.5. As I wrote in the other mail, that would also have allowed updating KAddressBook (which was already Akonadi-based in 4.4) to the much improved version from the 4.5 branch (while discarding all the other, unstable changes from 4.5). And it would have prevented the version number confusion, the chaos with translations etc. But IMHO the best solution would have been to postpone all this Akonadi stuff to 5.0 right from the start (especially considering that 5.0 is seriously being worked on now, to the point of stopping kdelibs 4.x releases). The original plan was for 4.0, that train was missed. 4.5/4.6/4.7 is way too late in the 4.x cycle for that kind of major change, it belongs into a major release. We (the KDE community) are now in the paradoxical situation where we won't ship even small uninvasive features in kdelibs, yet we happily replaced the PIM apps, which handle data many users consider essential, with what was essentially a rewrite. Yet both those apparently contradictory decisions are symptoms of the same problem: We really need to get used to working with 3 branches: next major feature release, next minor feature release and bugfix release. It is needed in some cases. It would have been needed for kdepim (next major feature release = KMail 2 etc. for 5.0, next minor feature release = small, incremental improvements for 4.n+1) and it's needed now for kdelibs. Each time we try to save a branch, the conflicting requirements of doing long-term refactoring vs. delivering new features to our users lead us straight to a disaster. it should not server as a general success story others should model themselves after. Big +1 to that! Kevin Kofler
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011, Heinz Wiesinger wrote: From what I remember from the desktop summit the picture you draw here is quite an exaggeration of what is actually happening. kdelibs 4.7 is meant to be frozen for new features, but not for bugfixes. Bugfix releases of kdelibs-4.7 happenend and I'm sure will continue on happening. As for the versioning I don't see why one of those bugfix releases couldn't be rebranded as 4.8.0 if that makes things easier (that was even briefly mentioned at the release team BoF). It does not solve feature backports of course. But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. The KDE Frameworks 5.0 development is not meant to take forever. In fact I think it's meant to be finished around early 2012, which would leave us with a frozen kdelibs for one KDE SC release, maximum two. It's also not a huge *we will break everything! Kittens will die!* release, but basically just a restructuration of the code, with little to no adjustments necessary for applications. (that was how it was marketed, anyway). The way I understood it was that there could very well be a KDE SC 4.9 release shipping a 4.9 workspace on top of 5.0 frameworks. I don't think a date of early 2012 is realistic at all. With upstream already working on Qt 5, I think it doesn't make any sense whatsoever to break everything twice, once for KDE Frameworks 5 and once for Qt 5. Yet I strongly doubt that Qt 5 will be out so soon. Even if the changes required for applications will be small, they will be needed (e.g. deprecated stuff will be dropped, and some applications are still using it), and I don't think it is friendly to application developers at all to have 2 flag days. Plus, it would mean that the KDE Frameworks and Qt major versions would get out of sync for the first time in KDE's history. We should also learn from the past: Things not meant to take forever end up taking longer than expected anyway, and each time, we're stuck with the temporary kludges for much longer than initially planned: * KDE 4 picked up delay after delay, and the long-lived 3.5 branch ended up becoming a mixed feature and bugfix branch (which IMHO is not necessarily a bad model, and which could even work for kdelibs 4.7, but it was still an exception). * The Akonadi-based kdepim picked up delay after delay as well, and so first its merge was postponed release after release, and only small pieces merged each time, and then we got stuck with a long-lived 4.4 branch, including a temporary and incomplete Akonadi-based KAddressBook for which we were initially promised that it'd be much better in 4.5. And now kdepim 4.4 is already discontinued and many users still consider 4.7 too unstable for their use. We must plan for major developments to take longer than the initial optimistic estimate. Kevin Kofler
Re: The case for a kdelibs 4.8
On Thursday 29 September 2011, Kevin Kofler wrote: On Thursday 29 September 2011, Heinz Wiesinger wrote: ... The KDE Frameworks 5.0 development is not meant to take forever. In fact I think it's meant to be finished around early 2012, which would leave us with a frozen kdelibs for one KDE SC release, maximum two. It's also not a huge *we will break everything! Kittens will die!* release, but basically just a restructuration of the code, with little to no adjustments necessary for applications. (that was how it was marketed, anyway). There will be changes necessary, but they will be small and/or scriptable. The way I understood it was that there could very well be a KDE SC 4.9 release shipping a 4.9 workspace on top of 5.0 frameworks. That's the idea. I don't think a date of early 2012 is realistic at all. With upstream already working on Qt 5, I think it doesn't make any sense whatsoever to break everything twice, once for KDE Frameworks 5 and once for Qt 5. Yet I strongly doubt that Qt 5 will be out so soon. That's right. Alex
Re: The case for a kdelibs 4.8
Without judging on the other arguments which look very reasonable to me... On Thu, Sep 29, 2011 at 2:01 PM, Kevin Kofler kevin.kof...@chello.at wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. [...] ...what exactly stops you from advertising kdelibs 4.7.x as version 4.8? (And I mean advertise, so only the user-visible parts, not version numbers in .so files or whatever.) Greetings Stefan
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 21:43:34 Stefan Majewsky wrote: Without judging on the other arguments which look very reasonable to me... On Thu, Sep 29, 2011 at 2:01 PM, Kevin Kofler kevin.kof...@chello.at wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. [...] ...what exactly stops you from advertising kdelibs 4.7.x as version 4.8? (And I mean advertise, so only the user-visible parts, not version numbers in .so files or whatever.) Now that would add a lot of confusion. Andras
Re: The case for a kdelibs 4.8
A Dijous, 29 de setembre de 2011, Kevin Kofler vàreu escriure: Hi, as you probably already know, a decision was recently made that kdelibs 4.7 would be the last 4.x release series of kdelibs, and work would be ongoing in the 5.0 (frameworks) and 4.7 (KDE/4.7) branches only. I think this is a huge mistake, for several reasons (the TL/DR crowd can stop right here, but if you want to know why, please read on!): 1. This puts kdelibs 4 into maintenance mode even before KDE Frameworks 5 is anywhere near a release, let alone versions of the workspace and applications actually using it. As a result, we will fail to deliver new features to our end users for months if not years. And no, kdelibs features do not only target developers: Some application or workspace features require kdelibs changes, or in some cases, even consist of kdelibs changes only, e.g. my Plasma PackageKit integration work is entirely in kdelibs (libplasma). Please describe the features you need. 3. Libraries, due to their nature, have a much longer shelf life than the workspace or applications. In order to keep existing applications working, distributions ship old (sometimes very old) libraries as compatibility libraries. For example, Fedora was probably the first distribution to drop the KDE 3 workspace, but we STILL ship the KDE 3 libraries! And we have no plans to drop them at this time. And RHEL is supporting them until at least 2017 (the scheduled end of life for RHEL 6). Fedora also still ships GTK+ 1. So it is just totally backwards to stop developing the libraries before the workspace and the applications. In fact, it should be just the opposite: We should continue doing kdelibs 4.x releases even after the rest of KDE SC has moved on to 5.x! And bugfix releases alone don't cut it: For consistency between old and new applications, some features should be backported to the compatibility libraries as well. For example, when we switched our default spell checker in Fedora from aspell to hunspell in Fedora 9 (i.e. 4.0 era), I had to add support for hunspell to kdelibs3, or our users would have had to install 2 spell checkers to use KDE apps! (Even several apps in the default KDE installation hadn't been ported to kdelibs4 yet in Fedora 9.) That change never got upstreamed because kdelibs3 is no longer maintained, yet it would have been useful to many distributions. Please keep the life cycle of your software for distributors and end users in mind when you decide your schedules, not just the convenience of a few core developers. What you are describing is actually fine means that old unmaintained libraries are still useful, so I do not see that it helps to prove your point. So I am urging you to reconsider your decision and reopen master for those people willing to work on 4.8. It's not too late yet, there's a month left until the soft feature freeze for KDE SC 4.8. Kevin Kofler
Re: Re: The case for a kdelibs 4.8
A Dijous, 29 de setembre de 2011, Scott Kitterman vàreu escriure: On Thursday, September 29, 2011 08:01:00 PM Kevin Kofler wrote: On Thursday 29 September 2011, Heinz Wiesinger wrote: From what I remember from the desktop summit the picture you draw here is quite an exaggeration of what is actually happening. kdelibs 4.7 is meant to be frozen for new features, but not for bugfixes. Bugfix releases of kdelibs-4.7 happenend and I'm sure will continue on happening. As for the versioning I don't see why one of those bugfix releases couldn't be rebranded as 4.8.0 if that makes things easier (that was even briefly mentioned at the release team BoF). It does not solve feature backports of course. But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. Even worse, features have already creeped into the 4.7 branch because they are needed and there's no 4.8 branch, so this isn't a theorectical point. Why is that bad, that is what was agreed when the freeze took place. Albert Scott K
Re: Re: The case for a kdelibs 4.8
A Dijous, 29 de setembre de 2011, Andras Mantia vàreu escriure: On Thursday, September 29, 2011 21:43:34 Stefan Majewsky wrote: Without judging on the other arguments which look very reasonable to me... On Thu, Sep 29, 2011 at 2:01 PM, Kevin Kofler kevin.kof...@chello.at wrote: 2. It will be confusing to our users, and even to some packagers, to have a KDE SC 4.8 on kdelibs 4.7. [...] ...what exactly stops you from advertising kdelibs 4.7.x as version 4.8? (And I mean advertise, so only the user-visible parts, not version numbers in .so files or whatever.) Now that would add a lot of confusion. That is actually Dirk's plan (or at least that is what i remember from the Release Team BoF in Berlin). Albert Andras
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 11:47:22 PM Albert Astals Cid wrote: A Dijous, 29 de setembre de 2011, Scott Kitterman vàreu escriure: On Thursday, September 29, 2011 08:01:00 PM Kevin Kofler wrote: On Thursday 29 September 2011, Heinz Wiesinger wrote: From what I remember from the desktop summit the picture you draw here is quite an exaggeration of what is actually happening. kdelibs 4.7 is meant to be frozen for new features, but not for bugfixes. Bugfix releases of kdelibs-4.7 happenend and I'm sure will continue on happening. As for the versioning I don't see why one of those bugfix releases couldn't be rebranded as 4.8.0 if that makes things easier (that was even briefly mentioned at the release team BoF). It does not solve feature backports of course. But one of my points is that we need features too, not just bugfixes. Continuing 4.7.x releases solves the problem of bugfixes just fine, but entirely fails to address the issue of features. Even worse, features have already creeped into the 4.7 branch because they are needed and there's no 4.8 branch, so this isn't a theorectical point. Why is that bad, that is what was agreed when the freeze took place. Agreed on by who? As a distributor of KDE, we rely (to meet the commitments we make to our users) on post-release updates of KDE to be bug fixes and not introduce new features. Once we release, we want consistency with the only change being resolution of problems and few/no regressions. I don't like the fact that KDE developers decided to ignore their own policy on maintenance updates. I think it breaks your contract with your downstreams. In the case of what's been done so far, it doesn't have an impact on us. The changes were in before our release and are technically low risk. I was aware of it, but becauase of the timing relative to our release schedule, didn't object. That isn't the same as agreeing. Today was our final freeze for non-critical uploads for our 4.7 based release. Elsewhere in this thread there has been discussion about allowing additional feature creep into the 4.7 branch because there's no 4.8 branch. From here on out it's a problem for us. I know most KDE developers don't pay a lot of attention to anything but the development release. For us integrators, it's mostly the one before that we're paying attention to, so when you mess with it, we tend to not appreciate it. Scott K
Re: The case for a kdelibs 4.8
On Thursday, September 29, 2011 11:47:55 PM Albert Astals Cid wrote: ... That is actually Dirk's plan (or at least that is what i remember from the Release Team BoF in Berlin). ... Are the results of this BoF published anywhere? Scott K