KDE Quality (problems)

[Duncan]

John Layt posted on Thu, 14 Apr 2011 17:17:12 +0100 as excerpted:

> On Thursday 14 Apr 2011 15:29:17 Billie Walsh wrote:
>> On 04/14/2011 08:59 AM, Anne Wilson wrote:
>> > On Thursday 14 April 2011 13:23:33 Billie Walsh wrote:
>> >> I suppose we all have our favorite parts of KDE. For me it's Quanta.
>> >> For whatever reason the devs at wherever have pretty much decided it
>> >> was not worth the effort to bring it along with KDE 4.x. There is
>> >> one person trying to bring it along and I hope someday he manages to
>> >> do it.
>> >> 
>> >> Why they decided to kill off Quanta is beyond me, well maybe "kill
>> >> off" is a bit of an over statement, but it is crippled very badly.

"They"...  Why THEY decided to kill...

>> > This thread is a perfect example of how basic facts about KDE are
>> > misunderstood.  It sounds, for instance, as though KDE consists of a
>> > body of people who sit in a room and make decisions about what will
>> > or won't be developed.  That couldn't be further from the truth.
>> > 
>> > There are thousands of developers, most of which code for fun after
>> > doing a day-job.  Once you understand this, it makes a lot more
>> > sense.

KDE, like free/libre and open source software (FLOSS) in general, is 
pretty much all volunteer, either of the individual devs doing the work, 
or as John Layt mentions below, corporations volunteering to sponsor 
individual devs to do that work.

>> > I would like Quanta to be ported to KDE4, too, but the truth
>> > is that it's a job that doesn't appeal to anyone capable of doing it.
>> > enough to spend their evenings on it.

Volunteers volunteer on what they're interested in.  Just because 
something needs doing to fill a hole in the system as it is, doesn't mean 
there's going to be someone that cares _enough_ about it to volunteer 
either their scarce time that could be spent elsewhere, or their hard 
earned money that could be spent elsewhere, to fill that hole.

This BTW is one of the big reasons documentation tends to be such a 
problem in FLOSS.  The devs doing the actual coding don't find it 
particularly interesting, and the users either don't understand it well 
enough to document, or once they do, their need is filled, so they're no 
longer particularly interested.  This is a problem the community has been 
aware of for awhile, filled in part by all the mainly user-peer lists and 
groups (such as this one) around, and more recently, by various wiki 
efforts.  But it remains a problem, never-the-less.

>> > Then you have things like Telepathy, that are ambitious, and will
>> > eventually be very useful, but again, the developers are going as
>> > fast as their time will allow.  The developers of Kopete lost
>> > interest in taking it further,

What did they do?  They lost interest.  They're volunteers, and when the 
interest level no longer caused them to prioritized development of the 
specific project in question above all the other things they could be 
doing with their time instead...

>> > and the new team wanted to do more, so
>> > a new project was born.  That's how it goes. Developers come and go. 
>> > Life happens to them.  They leave college, get a job, get a
>> > girlfriend, get married, have a family etc., etc., etc..  Their work
>> > slows and sometimes has to stop altogether.  They are simply human.

Again, this is a challenge for FLOSS.  Beyond a certain level, there's 
little appeal in the mundane task of supporting a reasonably mature 
project, continuing to keep it working with updated libraries, fixing 
bugs, etc.  This is where the distributions step in, providing that 
service for a time, but after a period of years, their efforts fail to 
keep up with the increasing demands of supporting software that has become 
more and more stale as the rest of the system evolves around it, and 
eventually they drop support as well.  (The enterprise level distributions 
tend to provide such support for longer, but at a cost of keeping up with 
current changes, and often at some monetary cost, as well, because paying 
someone to do it is the only way to sustain interest, at that point.  
Again, interest being the key.)

Rather, all the interest is in the new super-features.  Particularly since 
may of the volunteers are college students or the like.  If a larger share 
were people actually using the software in their work, the focus of 
interest would likewise shift toward stability, etc, from features.  But 
we deal with reality...  Unfortunately, once people's focus turns to use 
and they're interested in stability over features, that's exactly what 
happens, they're now users, not so much developers.  They have other 
commitments that take their time, a family, a day job, etc, and far less 
time and energy left to spend on what /would/ now interest them, if they 
had the time, actually focusing on stability and reliability.

Of course, that's where corporate voluntarism in the form of sponsorship 
comes in, that being a major reason why the software stack we have today 
is as stable and functional as it is.  But where businesses don't see it 
in their best interest to sponsor, and they can't see that if they're not 
aware of the situation, they won't do so.

>> > And that's why in FOSS we will always have some beautiful fireworks
>> > and at the same time some things that make us weep.
>> > 
>> > Anne

Thanks, Anne.  As should be obvious from my inlining above, I fully agree, 
and was in fact ready to post a similar comment, had you not beat me to 
it! =:^)

>> I realize how the "system" works. And I do applaud their efforts, they
>> do a great job. And, Quanta probably fills a "niche market" so it's
>> priority is low, except for those that "need" it.

FWIW, one of the projects I contribute to (the gtk-based pan, an nntp-news 
client that more or less directly competes with knode, but at least back 
when I started with pan, there were features pan had that knode didn't, so 
pan remains one of the few gtk apps I use in preference to the parallel 
kde app) is similar niche market.  I'm not a dev so can't help with that 
end, but am I believe the senior regular on the pan user and dev lists, 
helping other users where I can, answering questions and, I might even say 
even if I am crediting myself, helping to keep the community together and 
viable thru the loss of the primary developer to this same "lost interest" 
phenomenon.

While web development is at least an ongoing concern, news clients are a 
definite niche now fading farther and farther from the general public 
awareness, so pan and knode both have a much worse situation to deal with 
in some ways than any web developer tool does.

It's a tough situation to be in, and were it not for me and a few other 
stubborn die-hards, pan would be on its way toward being a small footnote 
in the history of net-news, itself on its way toward being a footnote in 
the history of the net, much as the gopher protocol, for example, already 
is.

But recently, some light at the end of the tunnel appeared.  Thru the 
cooperation of two (other) people, one a developer interested in pan but 
not in the politics involved with the gtk/gnome family to which it 
belongs, the other already a gnome translator with access to the gnome 
repository tho he's not a developer and thus couldn't by himself do much 
for pan besides provide translations for his particular languages, pan 
recently released the first real new version in years, where the last 
version released before that was only a minor bugfix release, to keep pan 
building on more modern systems and with an important security fix. 

As I said, while I didn't directly contribute to this, I do like to think 
I played a big part, as my efforts as the now-senior and most active 
member of the pan lists helped keep them together and active enough for 
these two people to discover pan and find each other on the list, in 
ordered to move pan forward! =:^)  Without me, it's very possible those 
lists would have been abandoned by now, and this would have never 
happened.  (Not that I want to break my arm patting myself on the back for 
it, but it really is scary, to one who uses pan for hours a day, to 
realize how close it came to disappearing, and it's really something to 
realize how major a part even a non-dev like myself very realistically 
played in keeping that from happening.)

The future still remains a bit uncertain.  I'm sure the two guys in 
question are still feeling out their new what-amounts-to co-dev 
relationship a bit.  But for the first time in /years/ it's not as if the 
list is a death-watch any more.  Along with the new release came the first 
update to the website (which the former lead was gracious enough to hand 
over to the control of the new guys), and we're actually talking about the 
possibility of new features once again!  It looks like pan might be around 
for awhile, after all! =:^)

Anyway, all that to basically say that there's hope, even for niche 
products, even after the original developer moved on, if there's enough of 
a community around the product to keep it going, and said community has at 
least one "anchor" who while he may not have the skills to develop the 
project itself, refuses to abandon the lists, etc, to death.  As long as 
there's still someone actively responding to questions (and enough 
questions so there's evident activity), the lists survive, and as long as 
the lists survive, there's a fighting chance for the project itself.

This I now know from personal experience!

So you Quanta folks, if you really believe in the project as much as it 
appears you do from the comments here, don't give up yet!

>> It's just a shame that such a great program is allowed to drift off
>> into Never-Never Land. You can spend several hundred dollars and not
>> find a better program.

Witness the above personal experience with pan.  Just because it's 
drifting doesn't mean it's yet dead and buried, especially with kde3 
itself, in the form of trinity, now an ongoing project.  As long as the 
old Quanta version can work on top of trinity (or be reasonably ported to 
do so), and as long as trinity remains viable, especially if it's base 
libraries gets included in distributions once again (I honestly don't know 
the status there), if the Quanta community itself remains viable, there's 
hope!

The question is, are any of you here involved with that community, helping 
it remain viable?  If not, why not?  It could very well be you that makes 
the difference between quanta ultimately returning and its drifting into 
slow death, just as I myself very possibly did for pan.

OTOH, again, it's all down to voluntarism.  If you aren't interested 
enough to keep the community active, and everybody else finds other stuff 
higher priority than the quanta community as well, it will certainly die.  
There's no condemnation for all the people that /might/ have kept that 
interest alive.  After all, they'd be volunteers doing so.  And volunteers 
can and do often simply find other things more important to do with that 
time.  There's nothing wrong with that.  It just happens, much like the 
death of the project will just happen, if nobody finds it sufficiently 
interesting to spend their time on, any more.

> We were lucky with Quanta for a long time that there was a commercial
> sponsor of the work, an American small business owner who liked it so
> much he sponsored an Eastern European developer to work on it. 
> Unfortunately that person can no longer afford to sponsor development,
> the developer has moved on, and few people in KDE are interested in web
> development work so no-one was available to port Quanta to KDE4.  I
> believe the future plans for Quanta is to use the KDevelop framework,
> but it's still a lot of work.
> 
> John.

Thanks, John, as well.  It's useful to know the story behind things.  Of 
course in this case it does serve to reinforce the point I made above, 
too.  Here it was a businessman volunteering his money, not a developer 
his time.  But the effect of that volunteering and the loss of it remains 
the same.

Hopefully, as with pan, something comes together for quanta.  And 
hopefully both pan and quanta continue healthy into their own long and 
prosperous futures.  But it's the volunteers that make it so.  If they go, 
so, ultimately, does the project.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman

___________________________________________________
This message is from the kde mailing list.
Account management:  https://mail.kde.org/mailman/listinfo/kde.
Archives: http://lists.kde.org/.
More info: http://www.kde.org/faq.html.