10 Year Old KDE Bug Finally Gets Fixed

[orzel]

It seems to me that you're not a hobbyist open source developer.

I guess this can always be debatable ("what you do is crap"), but I actually do consider myself an OS dev. I was even once very close to KDE (I only commited few patches that I'm not especially proud of so i wont say 'contributing').

But I do work on several other projects, and I do consider that I kinda "owe" something to my users. Throwing some (ugly, untested) code on git is not what i consider "being an OS dev". It's just this.. throwing code. As an OS dev, i do consider that I should either listen to users, especially when they are dev carefully reporting bugs. Or if I can't I should say it clearly and publicly.

I've stopped reporting bugs to kde after years of totally unsuccessful reporting to bugs.kde.org. All what i've seen was bugs pilling up on the bugzilla, a very lot of old duplicate, and most importantly, developers basically denying the bug or arguing as you do ("this is opensource, so you dont have to cry"). Fortunately this was not often, but this oh-so... you know... not what i'm used to as a dev.

[halfmanhalfamazing]

Really cool, they fixed a bug that probably nobody noticed for 10 years, while there are billions of tickets in the bugtracker, lot of them confirmed and untouched for years. They can keep going releasing "Major" 4.x releases whose biggest feature is "fixing bugs" and "Minor" 4.x.y releases with lot of changes breaking things even more.
Come on KDE, do some %$%!!#% QA, show that you care at least a little bit for your users. I'd love to go back to KDE...

Did you notice why Thiago fixed this bug ? Not because it was breaking 10k people or was voted the most in bugzilla, but because HE was hit by the bug.

As open source users, we(you and I) are the QA team.

If you really want to see a bug fixed, here's what you do:

Verify the bug in multiple distros, and file the bug with multiple distros' bugzillas. Particularly those with either A) fantastic programming talent, or B) Large money bags behind them.(By this I mean, but not limited to distros like Gentoo and Fedora)

Especially if you can get the bug to pop off one of the automatic bug reporting tools like Fedora's ABRT or Ubuntu's Apport.

This is how I got rid of a few bugs that were bugging me.

Do you want to complain? Or do you want the problem solved?

[Adaptee]

I've stopped reporting bugs to kde after years of totally unsuccessful reporting to bugs.kde.org. All what i've seen was bugs pilling up on the bugzilla,

Not really. Visit https://bugs.kde.org/weekly-bug-summ...ps=70&days=730 :

Total: 21026 bugs and 15778 wishes
44753 bugs opened, 48727 bugs closed in the last 730 days
(4993 wishes opened, 7107 wishes closed)

See, the total number of bugs & wishes are decreasing during the past 2 years. Sure, it always has the tendency of piling up, but the result is not always the same as you would have expected.

As comparison, visit https://bugzilla.gnome.org/page.cgi?...0&products=100 :

Total Reports: 43412 (51445 reports opened and 50858 reports closed. Including enhancement requests)

a very lot of old duplicate,

That would be true for bug tracker of every big FOSS projects, because developers are always minorities compared with reporters, and often there are no dedicated QA people helping developers to take care of the bug tracker.

Also don't forget KDE makes it extremely easy for average users to report bugs: "Help -> Report Bugs" for normal bugs, DrKonqi for crashes. One side effect, though, is bugs.kde.org often receives bug reports that are of low quality from users who are not experienced in writing useful bug reports .

[Elv13]

Today, I had a problem with Marble. Did I report the bug right away? No. I went to IRC, talked about it, then made a bug and guess what: it is fixed. I always do that and most of the time, it get fixed right away*. It work for most software I use, not only KDE ones. I reported some very serious bugs in other (mostly GTK apps, as I don't often IRC with them) and it took years to get a fix. If you don't push a bug upward in their priority list, it usually wont get fixed.

That said, it sometime work. I usually fix bugs reported to me on a mailing list withing a week or 2 and on the tracker in less than 2 months. 2 years ago, I also tried to report a crash a day to help Calligra get better (for the sake of doing some QA) and to my suprise, only one stayed open most then 24 hours. So it always come down to the project and the people maintaining it. It also come down to the size of the codebase. Fixing a bug in a 20k line project is simple as the whole code can fit in our head, if the design is clean, you don't even have to open a computer to have a clear idea of the problem. However, it is impossible for bigger apps or libraries.

*I am a dev too, but I usually consider diving in other projects rather inefficient and do it only if I really care

[Ovocean]

The Free software community needs a way to gather to pay developers for fixing bugs and developing features that no one is willing to fix/develop on their free time.
We need a crowdfunding platform specialized in this! I believe it would greatly help the libre ecosystem as whole and offer payed job opportunities for libre software devs.

I don't have the skills nor the time for this these days, only to spread the idea. Hopefully it gets picked up by someone.

[Awesomeness]

The Free software community needs a way to gather to pay developers for fixing bugs and developing features that no one is willing to fix/develop on their free time.
We need a crowdfunding platform specialized in this! I believe it would greatly help the libre ecosystem as whole and offer payed job opportunities for libre software devs.

Duh. You can't order someone to fix a bug nobody even noticed. This bug was never reported.
Not only that but your idea in general hurts FOSS communities. It would encourage an atmosphere of “give me money or I won't release bug fixes” (bugs may already been fixed by the dev in private but not committed into git).
If you just looked closer at GSOC you'd see that very often someone appears to develop a feature, gets accepted into GSOC, implements the feature fast and therefore buggy, and after collecting the money leaves without ever being heard of again. A case that comes to mind is KWin’s tiled window management. The code had to be removed for 4.10 because the GSOC developer did not maintain it.

[asdx]

How about if the taskbar gets some love too? It's been broken for a long time.

https://bugs.kde.org/show_bug.cgi?id=224447

[Ovocean]

Duh. You can't order someone to fix a bug nobody even noticed. This bug was never reported.

My post wasn't in relation to this specific bugfix, but to the discussion above with people arguing over the fact that a lot of bugs won't get fixed just because there's no developer willing to, even though they (the bugs) annoy many users.

Not only that but your idea in general hurts FOSS communities. It would encourage an atmosphere of “give me money or I won't release bug fixes” (bugs may already been fixed by the dev in private but not committed into git).

Believe me, I love to see all this work done for free, but at some point, there's work that should be done that nobody wants to do for free. There must be ways to get money in libre software development, and crowdfunding is a much better solution than corporate funding. If you have better ideas, let us know.

[orzel]

That would be true for bug tracker of every big FOSS projects, ... .

Of course. This as well as lot of other remarks ("it's foss, what did you expect", "do it yourself", "pay someone to do it") is obviously true. I dont have time to fix the bugs, i dont really care. I've spent way too much time carefully reporting bugs and i just gave up. I dont use KDE anymore, but i still do use some KDE apps which are really great (konvi, okular, gwenview) while hiding away with fear from the
ugliest of them (kmail, kmail, akregator, did i say kmail was crap?).

When i say "bugs are pilling and are ignored", this is the point of view of a developer, comparing to other FOSS. Yes, QA in FLOSS projects is very very often "not as great as it could". But i'm saying this is far worse here with KDE. Worse than other projects, and worse than in KDE before. This is not just another random "bouh, they did not fix my bug".

My point is rather: this is soo much a pity. KDE used to be so good, and they ruined it all. It was technically very solid, and this is no more. It has really few bugs and those used to get fixed quickly, and now this is quite the opposite. They were able to understand critics and to recognize errors, change their mind and such. My favourite example is the switch away from corba. And at last.. they used to care about users, now their mostly selfishly care about themselves. They are all about "us, as a communauty", watching their navel and doing "fun stuff to code because this is what foss is all about". Whatever fun stuff they do they merge half-done and pretend for years that they are just 'fixing minor pb' while the whole stuff is completely buggy, fat and imposed to the user. Wants name ? Akonadi, strigi/nepomuk and, yes, plasma.

Despite not using KDE anymore, I still have to mess with it because lot of family/friends use linux. That means different KDE versions and different distributions. The most important problem i have to fix is always to remove akonadi/strigi/nepomuk binaries and reduce plasma to its bare minimum, so that the computer start to get usable.

[Ancurio]

The Free software community needs a way to gather to pay developers for fixing bugs and developing features that no one is willing to fix/develop on their free time.
We need a crowdfunding platform specialized in this! I believe it would greatly help the libre ecosystem as whole and offer payed job opportunities for libre software devs.

I don't have the skills nor the time for this these days, only to spread the idea. Hopefully it gets picked up by someone.

http://www.freedomsponsors.org/