Free is too expensive (Economist) [LWN.net]

Free is too expensive (Economist)

Posted Apr 2, 2012 13:27 UTC (Mon) by anselm (subscriber, #2796) [Link]

You must know you're wrong, and by now you must know there were pretty good reasons for the regressions in printer support and you must know that there are people working very hard with Qt to get back all that functionality.

So? KDE 4 came out in 2008. You don't mean to tell me that this job has taken more than four years and isn't finished yet? I was under the impression that Qt was meant to make things easier for developers. I don't really care about what kind of internal mess keeps KDE from providing features that it used to provide. All I can see is that stuff that used to work well no longer works at all. I'm really an understanding sort of person but I can't fault other people – especially the ones whose KDE-3.5-based Linux systems I support – for thinking this is not the way things ought to go. (The stock reply here is, of course, »If it is that important to you then effing do it yourself«. But that only reinforces my original point.)

It's not just printer support, anyway. We're having this discussion in the first place because ABI stability in Linux is problematic especially with respect to the desktop environments (the kernel, libc, and X11 ABIs, to move farther down the stack, are doing reasonably well there). If the desktop environment projects were more dedicated to maintaining backwards compatibility, the problems touched on in this thread would be a lot easier to solve.

This includes issues such as making it difficult to have several versions of the software installed at the same time (which I understand is a problem with GNOME 2 vs. GNOME 3), or dropping all support for KDE 3.5 basically immediately after KDE 4.x came out (which at the time was barely usable). As things stand today, even if something like LSB stipulated KDE SC 4.8 as the official »stable« base for desktop applications, nobody knows what things will gratuitously be broken in KDE SC 4.9 because, like in so many cases before, somebody thought they desperately needed to be reimplemented from scratch. The only obvious way out of this is to keep a complete 4.8-vintage KDE around on the off-chance, and to hope that it (and applications running on it) will be able to interoperate with newer versions of KDE.

Free is too expensive (Economist)

Posted Apr 2, 2012 13:41 UTC (Mon) by boudewijn (subscriber, #14185) [Link]

Yes, I'm going to tell you that. Anyway, read for yourself:

http://www.layt.net/john/blog/odysseus/kf5_localization_a...

details the current situation, and

http://www.layt.net/john/blog/odysseus/the_good_the_bad_a...

tells you about the earlier work. Let me quote:

"Just to recap, the branch at http://qt.gitorious.org/~odysseus/qt/odysseus-clone/commi... contains the Advanced Page Selection features such as Current Page, Odd/Even Pages, Server Side page selection, and non-continuous Page Ranges."

That was in 2009.

So just stop spouting offensive nonsense like "KDE developers primary mission in life is providing entertainment for KDE developers". That simply isn't true, and you know it.

You can make your point in a way that shows that you're "really an understanding sort of person". and there's no reason for you to be a jerk on behalf of other people who aren't.

Free is too expensive (Economist)

Posted Apr 2, 2012 15:20 UTC (Mon) by anselm (subscriber, #2796) [Link]

That was in 2009.

Indeed. It is now 2012. There are various adjectives that one might think of when hearing this. The one that comes to my mind, considering that we have gone through 9 releases of KDE and several of Qt since when this stuff used to work, is »pathetic«. I'm deliberately not dissing John Layt here, who AFAICT did all he could to advance the issue and more, but the observation that the development processes of KDE and Qt apparently managed to stifle any sort of useful progress on an important regression like this for years, while at the same time taking on board loads and loads of completely new stuff, speaks volumes.

So just stop spouting offensive nonsense like "KDE developers primary mission in life is providing entertainment for KDE developers".

That's not what I said. I said »The KDE project's primary mission …«, which is something completely different.

The fact that there are a few developers who do feel called upon to fix these long-standing issues does not detract from the picture as a whole. The problem is really that there are apparently too few such developers. For every single KDE developer like John Layt, who I will be more than happy to ply with good German beer if he ever shows up in the Mainz area, there are probably several who are only in it for the fun. What we see as a result is a project which seems deeply in love with everything that is new and shiny and cool while not-quite-so-cool things like feature regressions from five years ago are left to a small minority of people who feel responsible enough to address them, if they are ever addressed at all. In a properly governed infrastructure-type (as opposed to fun-type) project, regressions like these would not have been allowed to arise in the first place. We're not talking about random bugs here, after all, but about things that can be documented and planned for in advance.

Let me quote from the first comment on one of the blog posts you mentioned which said

»Perhaps 2010 will become the year of the kde desktop, with complete printing options, (hopefully) network management, supported video cards and FINALLY a decent standard browser.«

It is now 2012 and at least two of these items remain on the to-do list.

Free is too expensive (Economist)

Posted Apr 2, 2012 20:33 UTC (Mon) by sfeam (subscriber, #2841) [Link]

I was ready to be sympathetic, despite being one of the people who has been suffering from the lack of decent print support in KDE4. Then I read this bit in the link you provided:
The CPD solves a very specific problem, that is the broken state of print dialogs on Linux using Cups. It doesn't solve the problem we and Qt have of also supporting printing on Windows and OSX using the same API.
So there is actually a decent module to handle printing, but linux users must wait 4 years to see it, just because the same solution wouldn't work on Windows? Heck, the Windows + KDE users, if there are any, probably have other ways to deal with printing. "We don't like #ifdef" hardly seems like an adequate justification for this decision.

Free is too expensive (Economist)

Posted Apr 2, 2012 15:30 UTC (Mon) by renox (subscriber, #23785) [Link]

>> "We must remember that, from what we can see, the KDE project's primary mission in life is providing entertainment for KDE developers, not convenience or stability for KDE users."
>
> Can you just stop gratuitously insulting us? It's extremely tiresome.

That his point of view which is not said in an insulting way, you disagree so you feel insulted, but FWIW I agree with his view: if the KDE project's goal was primarily the convenience and stability for KDE users, then things would have been handled very differently, for example not activating by default unstable technology, not making "big bang" release such as KDE4.0, etc.
But as it makes things less fun for developers, I think that only small projects can work this way, IMHO KDE is way too big to be developed like this.

Free is too expensive (Economist)

Posted Apr 16, 2012 12:06 UTC (Mon) by nye (guest, #51576) [Link]

>"We must remember that, from what we can see, the KDE project's primary mission in life is providing entertainment for KDE developers, not convenience or stability for KDE users."

> Can you just stop gratuitously insulting us? It's extremely tiresome.

> You must know you're wrong

To be honest, I for one wasn't aware that that was disputed and was hence under the impression that his statement was an objective fact. I would never have guessed that it would be found offensive, so I doubt there was any insult intended.