competition is good? [LWN.net]

competition is good?

Posted Mar 25, 2007 1:09 UTC (Sun) by dlang (subscriber, #313)
In reply to: competition is good? by dion
Parent article: Linux and flash

you say we should have one desktop with config options for every possible user

but you have honest disagreement between different people about how much configurability there should be (and not just the gnome folks)

haveing underlying mechanisms that can be re-used by different projects to implement the same functionality is good, but if you make absolutly everything configurable you just create the need for multiple configuration mangement tools to implement sane configuration.

as an example X is fully configurable, but X by itself doesn't make a useable desktop, as a result there are a LOT of projects out there to implement the management of windows (not just gnome and kde, but also windowmaker, fvwm, icewindows, etc)

the other thing to remember is that there are a LOT of reasons why someone may not want to work on one project, there can be personal and technical disagreemnts with how something is done. if competition is bad then they have to either work with the existing project, or not work in that space at all. competition is good becouse it allows someone who disagrees to go off elsewhere and implement it their own way.

getting back to the issue at hand (a flash player), while the end goal is the same (the ability to play flash), there is still a lot of room for disagreement on how to get there. Even if both teams decide to use the same language for programming the project (I don't know if these two do or not), there's still a lot of room for disagreemnt on how the code and project should be orginized. even seemingly simple things like what codeing style a project requires could be enough justification for someone to not want to work on it.

Opensource programming isn't a zero-sum game. if someone works on project A it doesn't mean that they would work on project B if A went away (and it also doesn't mean that they would reduce the amount of effort that they choose to put into project C by working on A)

(Log in to post comments)

competition is good?

Posted Mar 25, 2007 14:27 UTC (Sun) by dion (subscriber, #2764) [Link]

I completely get your point that there are good reasons for not wanting to work on a project and of course it's good that people can go off and do their own thing.

I don't buy the argument that gnome is needed because it's better to have less configurability than what KDE offers, if that was a valid reason then gnome would consist of a very limited configuration tool for KDE.

While I think it's good for the individual to be able to choose what project they work on, I still don't think that it's better for users to have several competing implementations of the same solution, unless the point of those different projects is to figure out what the best solution is before EOLing the others.

Your example with the flash players is a good one, clearly it would suck to have to work on a project that has horrendous code, so you might be happier to just work on your own project in stead, but that doesn't mean that you will make progress any faster than the first project or that the resulting product will be any better, it just means that you will be happier.

I think it would have been better if the first project didn't have horrendous code and if you had chosen to work on the first project in stead.

I'm not saying that a hacker should be banned from starting on a competing project, just that it would be better if he chose to help the first one.

competition is good?

Posted Mar 29, 2007 18:05 UTC (Thu) by TRauMa (guest, #16483) [Link]

You're a fan of intelligent design? Cause that's not how evolution works. You just can't tell which approach to solving a problem is better at any point in time. The first mammals were a failure, yet no one decided to scrap the duplicated effort and just go with the dinosaur design.

And what gives you the idea we should just scrap GNOME? Why not KDE? Because KDE was there first? What kind of reason is that?

competition is good?

Posted Mar 30, 2007 8:26 UTC (Fri) by dion (subscriber, #2764) [Link]

I take your "intelligent design" suggestion as the strongest possible insult, but we're talking about software development, not a biological process.

Evolution is not a good development method for software, except for some very specialized areas where you can run thousands of generations pr. second.

Sure it's nice to think of the plethora of OSS projects as evolution in action, but behind every single one of those projects is at least one designer,

Natural selection doesn't really work with software either, the only thing an OSS project needs to live is to attract developers, that's why you have ridiculous things like desktop applications written in C that just refuse to go away or evolve into using a higher level language.

To get back to your question; yes, I think that GNOME should have been scrapped around the time when Qt was released under GPL or maybe even before that, even RMS conceded that the QPL was a Free license.

The reason to start GNOME in the first place was that Qt wasn't Free, so once that reason went away so should GNOME.

That being said, GNOME has improved and is a useful desktop now, the only thing lacking is a proper high level language, but I guess the up coming C# rewrite will take care of that.

Once GNOME has been rewritten in C# it will be an ideal desktop for former windows developers to get started on, so there is merit in that approach.

competition is good?

Posted Mar 29, 2007 19:24 UTC (Thu) by Duncan (guest, #6647) [Link]

> you say we should have one desktop with
> config options for every possible user
>
> but you have honest disagreement between
> different people about how much configurability
> there should be (and not just the gnome folks)

As a KDE user that appreciates the level of configurability it offers,
this is what I like about GNOME. There's an honest disagreement on UI
goals and styles, and GNOME is somewhere for those who disagree to go, so
they don't bother the KDE folks and try to limit the options some of us
like to use! I /hate/ trying to use GNOME, because it makes me feel
limited and locked down. However, I'm /very/ glad it's there, because I'd
a whole lot rather those in favor of such a thing have their own sandbox
to play in than try to impose their limitations on mine.

Additionally, the point made that FLOSS isn't a zero-sum game definitely
applies. Many coders are volunteers. You can't force volunteers to work
together to create the "ideal" (how ever you want to define it) project,
because it just doesn't work that way. If they like where it's headed,
they'll naturally put more hours into getting it there. If they don't,
they'll go elsewhere if possible, or simply find something else to do if
not. Thus, KDE/GNOME/XFCE/whatever isn't wasted effort, but simply those
with the talent finding the project that best matches their ideal, that
best stimulates them to contribute more towards its progress, and forcing
a one size fits all wouldn't get everyone working on the same thing to
make it advance faster, it would cause many to stop working entirely (or
again start branch projects if the could, which is in practice what
happens, because there's simply no way to force it in FLOSS).

The same of course applies to all the other "wars", vi/emacs, etc.

The first point doesn't so much apply in the flash case, due to the spec
already being laid out such that our implementations either work or don't.
However, the second point continues to apply. Trying to force just a
single project won't have the desired effect of speeding things up,
because being volunteers, some now contributing to the other project will
simply find other things to do with that time instead. In fact, due to
the cross-proliferation of ideas and friendly competition between the
projects, forcing a single project would likely slow things down
dramatically, because that would be lost.

Meanwhile, I have gnash merged here, but haven't found it all that
practical. Generally, I see the thing load, but it won't play anything.
I guess I'll have to try swfdec and see if I have any better luck. To my
knowledge, given I'm running amd64 in 64-bit, even if I was able to
legally run slaveryware (in general, I legally can't, since EULAs are
potentially binding here and I can't agree to the EULA, as they restrict
freedoms I'm no longer willing to sign away), even then there'd be serious
hoops to jump thru in ordered to get the proprietary flash player working,
since AFAIK there's no native 64-bit version.

Duncan