A way for average users to influence development.

[Kvark]

If you as an indivdual user want software that doesn't exist yet. Then you need a way to speed up the development of such software.

When a company wants software that doesn't exist. They hire someone to write it. This is not possible for an individual user unless you are rich.

For gnu/linux users the traditional method is "Want a new program? - Code it yourself!". But average computer users can't write a single line of code. Almost the entire population in modern countries uses computers. To educate almost the entire population to programmers would be a terrible overkill and waste of everyones time, not to mention impossible. So "write it yourself" is not an option.

Personally I believe it is vital that users, not only developers and companies, can have a say about development. To be powerless on this issue is just as bad when it comes to software as having no right to vote is when it comes to politics.

There is a lot of users around. If a couple hundred users placed a donation each then it would add up and they could afford to hire someone to make it. The problem is all those users who want the same thing are spread all over the world and don't know about eachother. So it is very hard for them to team up like this.

There should be a bounty center where anyone can place a bounty of any size and a specification of demands that needs to be filled to claim it. Then anyone else who has the same demands can add to that bounty. This would give individual users a way to drive development of the things they want. If a few users want it then the bounty would give a signal and a small encouragement. If many users want it then the bounty would add up to a considerable amount with a considerable gravity on developers.



...Is this something that should be done? That can be done? Who is the right organization to make the center, ubuntu, a cross-distro organization, fsf? ...Is the idea great, reasonable, flawed or even bad?

[phen]

how will control the quality of the software? I think one big advantage of open source is, that its programmers usually use the software. They start writing it often for themselves i think. Am i wrong?

someone sees a big amount of money available for the developement of a new mobilephone connect tool (for example), and writes it because of the money. he releases a very buggy version, takes the money and is never seen again.

on the other hand, i do like your idea, because i would donate money for several programs myself. But the system has to have a protecion against abuse.

cheers,

kai

[majikstreet]

rentacoder.com ???

[NoTiG]

I think its a cool idea... except that bounties are usually paid after the work is done... and with a system where alot of people add tiny bounties to make one big bounty... i can see problems in collecting them all. Assuming that everyone was honest there wouldn't be a problem.........

[az]

rentacoder.com ???


Is that the place for open source development? I know there is a site where many people can add to the pot to finance a project. There are conditions that you can set up like if there is not enough money by such a date, everybody gets refunded and the project goes away...

I think that is an excellent tool for the community.

Anyone who wants to contribute can still directly improve the software by donating some money. Someone scratches your itch and they feed their kids.

[majikstreet]

Is that the place for open source development? I know there is a site where many people can add to the pot to finance a project. There are conditions that you can set up like if there is not enough money by such a date, everybody gets refunded and the project goes away...

I think that is an excellent tool for the community.

Anyone who wants to contribute can still directly improve the software by donating some money. Someone scratches your itch and they feed their kids.
I'm not sure if it's open source.. I don't know the details about it.

[Kvark]

Yes, there could be problems both ways (crappy solutions, or bids not payed). The center would need clear balanced rules, secure transactions and anti fraud insurance. Very specific demands and some beta previews along the way could cut down dissagreements. ...It is a tricky question exactly how to set all that up, therefore WHO sets it up is a very important question, it must be a very competent organization.



rentacoder.com seems cool except for some problems. It doesn't look like many users could add $5 each so it doesn't address the many users team up thing. As far as I know it is not well known in the gnu/linux community, which would be a requirement to fill the function we are talking about. Also I think to be effective as a gnu/linux bounty center it needs to be strictly GPL or dual licence. Cause to mix in propriarity software on the same site would make a mess.

[UbuWu]

I know there exists a website that does exactly this, and it was recently mentioned on some news websites, but I don't know its name anymore...

[poofyhairguy]

...Is this something that should be done? That can be done? Who is the right organization to make the center, ubuntu, a cross-distro organization, fsf? ...Is the idea great, reasonable, flawed or even bad?

I'm gonna kinda disappoint you...but I don't think it can be done. Why? Three reasons:

A. The not made here syndrome. If you install a lot of Linux's you start to notice a lot of repeated stuff. Debs and RPMs. All the repeated GUI config tools. ETC. the problem with your plan is that even if we did hire someone to make something, the distro developers might not touch it because it wasn't made in house. It sucks, but its true.

B. We already have bounties:

http://www.ubuntulinux.org/community/bounties/document_view

And they don't always attract new workers. Hell, Crossover has a system like what you propose and Access 97 has been on top for a while (97!):

http://www.codeweavers.com/compatibility/browse/name?app_id=186

It seems that OSS developers care about more than money. Thats cool for a free OS, but it sucks if you want new features. Commercial developers add features their customers want. OSS people add what they want. A cornerstone...no scrap that..a cardinal rule of OSS is "do it yourself." To change that would be to change how OSS works. You might have a better chance of making demands of Linspire (or another lowend mixed company).

C. I must admit I kinda want something like this as well. I would pay big bucks to have stable xcompmgr NOW, or Luminocity stuff NOW. Notice the NOW? Thats part of my society- instant gratification. The biggest problem with your system is that it lacks that. Bounties would have to sit for months....maybe even years to get big enough to be worth messing with or to find someone that wants to mess with it. When you donate the money, suddenly there are expectations. "If I pledge this it better be done by next release." That won't work. Then you'll have people that want to take their money away because "I didn't get it NOW" and the bounty system falls apart.

When you add money, you decrease patience. I bet the biggest draw to OSS some some developers is that they ARE NOT paid for their work so they DON'T have to work in someone else's time frame. If everyone donated $20 to my cause (stable xcompmgr) it might take MONTHS for it to build to a level when its actually worth doing the work. By then, the original developer might have done it anyway (social pressure) and I would be pissed that I spent the first money and got no return.



The truth is that the Gnu model is just not that compatible with the Western ideal of instant gratification. Money is not the primary motivator (good software is) and therefore bolting some capitalism on won't work. It sucks to say "it won't work" but I don't think it will work. Yes it sucks that there are BASIC features that are lacking that many people would want. Yes it sucks that most people aren't developers. But if you think about it, the system we have now is better than the one before it.

At least in the OSS system you can ask (nicely) the developers to add stuff. Or you can do it yourself. What happens if you asked ol MS for a feature? "Since you are a lowend customer (aka less than 10000 licences) we laugh at you." Think if you asked them for the code to do it yourself "Ha Ha Ha Ha Ha."

Sorry to be a party crasher, but thats how I (not Ubuntu or the Forum) feels after carefu thought on the issue.

[poofyhairguy]

Quick note:

I think that this system WOULD work GREAT for documentation. I personally would pay $20 for a "compile Metacity without its compositor howto." Since documentation has a lower barrier to entry, it could be cheaper and easier.

I would like that for documentation.

[darkmatter]

If you really want to 'influence' development, the best way (if you are not a hacker/programmer) is to make reasonable, staged suggestions to developers (as in no "I wan't a new app for this and I want it now!").

Join an irc channel where you can discuss your idea(s), post your suggestions to the devs (most project sites have an area for feature requests/bugs), build mockups of your 'improvement', join mailing lists, etc.

The way to influence development is to convince the devs (in a reasonable manner) that you feature is actually worth developing.


EDIT: ignore this edit. My keyboard hates me... :roll:

[Kvark]

All the negatives aside, I think it would be nice to have a say in what software I want, or how I would like to have a piece of software behave. Is there any way to "influence" the developers of Ubuntu currently?
You can try the options darkmatter mentions.

I have tried a couple times to discuss tools or features that would be nice to have in IRC channels. The conversations usually end with...

geek: thats a nice idea, good luck with making it
me: I can't do it
geek: learn then
me: no, everyone can't learn the same proffesion
geek: well, the feature won't just magically appear...
me: I was trying to suggest that someone makes it
geek: wow, you expect someone else to do it for you
geek: you are a leech and you are on my ignore list now

At this point I do feel like a leech. If I take a cab then I pay the driver, expecting a free ride would be wrong. So when I want something programmed, but can't pay the programmer on my own, then it feels very wrong.



I'm pretty good with code so I can do what I'm told, learn more about gnu/linux (right now I don't know anything about under-the-hood stuff) and then program it myself. But unlike driving a car as in the cab example. Programming is something most people can't understand. I have tried to teach the basics to friends and relatives and it seems like many simply can't think in terms of "first do this, then do that 5 times, then if x is more then 11 then do something, otherwise do something else".

[poofyhairguy]

Is there any way to "influence" the developers of Ubuntu currently?

Bugzilla.

[Kvark]

Is this a really enourmous bug, or is it some sort of reporting system? :)

It is a reporting system, found from ubuntulinux.org, at bug reporting under support.


Don't know more then that. I filed a bug that is solved by including GPLed hardware drivers from sourceforge. Nobody reacted to it.

[az]

Is this a really enourmous bug, or is it some sort of reporting system? :)
Both.

*edit* bugzilla is not that bad. Canonical is developing Malone for Launchpad. It is going to be more sophisticated than bugzilla.

[az]

It is a reporting system, found from ubuntulinux.org, at bug reporting under support.


Don't know more then that. I filed a bug that is solved by including GPLed hardware drivers from sourceforge. Nobody reacted to it.


What's the number? When did you file it? Was it assigned?

[Kvark]

What's the number? When did you file it? Was it assigned?
Bug#: 12341
Opened: 2005-07-02 22:27 UTC
Assigned To: Debian Bug Importer <debzilla@ubuntu.com>

Dunno what the debian bug importer thingy is. I can't see any signs of that anyone even looked at the bug.

[Stormy Eyes]

Personally I believe it is vital that users, not only developers and companies, can have a say about development.

I may be wrong, but I thought that users could have a say in development by frequenting mailing lists and forums, helping out, and in turn making polite requests. You've probably seen this before, but let's pretend for a moment that I'm writing an audio-player program (like XMMS, or something). I'm developing this app with the following opinions in mind:

1. I don't like Gstreamer, and think that a GUI app ought to wrap around commandline tools.
2. I do not use MP3, and do not want to deal with the legal issues surrounding MP3; I don't have time for such idiocy.
3. I don't want to be another cook stirring somebody else's stew. I want to make my own stew, in my own kitchen, and make it my way. So I'm writing my own app instead of contributing to an existing project.

User Alice, who loves Rhythmbox and thinks that Gstreamer is a cool idea, sends me a polite email asking me why I'm not contributing to Rhythmbox since it "needs developers". I politely explain that I would rather do my own project and see it through from start to finish; I have a resume to pad.

User Bob, a teenager with 69GB of bootleg MP3s downloaded from the P2P service du jour, sends me a couple K of poorly written flamage that boils down to the idea that Linux isn't ready for the desktop because my little app doesn't do what he wants, and that if he wants me to keep using Linux I had better put aside everything I want to do and cater to him. I reply to him, saying quite simply that I do not tolerate such rudeness from paying customers, and he is welcome to go back to using Windows.

And then there's user Charles, who likes my little tool (and keep your wisecracks to yourselves, please) but doesn't like having to use a separate app to rip CDs to Vorbis so he can have my player pump them through his speakers. Given that I can use commandline tools to rip Vorbis as well as play it, his suggestion is reasonable and fits with the rest of my project. I tell him that I can't make any promises, but I'll probably have rough CD ripping support in the next release.

The moral of this story? Be polite. I won't speak for other developers, but if I was developing a F/OSS project, and a user asked me nicely for a feature, I would either politely decline and explain my reasons, or plan to implement the new feature in an upcoming release. If users want a say in development, they have to remember that they are not paying customers, that they are making requests of volunteers, and that if they are too rude the developers might just make like Atlas and shrug.

To be powerless on this issue is just as bad when it comes to software as having no right to vote is when it comes to politics.

I don't agree with this analogy: having software that isn't just right is certainly inconvenient; I spent years wrestling with LaTeX and never quite grokking it because there wasn't a decent word processor available for Linux until OpenOffice.org got rolling. But while imperfect software without having a say in the development of said software is inconvenient, being unable to vote is likely to cost you your property, your liberty, and your life.

[az]

Bug#: 12341
Opened: 2005-07-02 22:27 UTC
Assigned To: Debian Bug Importer <debzilla@ubuntu.com>

Dunno what the debian bug importer thingy is. I can't see any signs of that anyone even looked at the bug.
Hmmm.

1. Look to see if the module shtat this build is not already in breezy. Go to packages.ubuntu.com and use the lower search tool, the one which looks for individual files. See if the kernel module is included in the Breezy kernel (example, it may be named acecad.ko - You would know more than I). If the driver is there, you are done, if not, go on to step two:

2. Debian bug reported goes the other way, I think. Like a bug from debian which also applies to ubuntu gets automatically imported into ubuntu. Why this is assigned to that is wrong, I think. Don't be shy to ask about it on the devel mailling list. I am sure that you will get a quick answer, especially since it has been a few weeks now.

Usually, bugs get assigned after a few days.



http://ubuntuforums.org/forumdisplay.php?f=29

I stongly reccommend you ask about your bug there.

[Kvark]

Thanks for the advice azz. I found a package in breezy that doesn't exist in hoary and that can't be anything else then that specific fix. :)

Awesome, it is solved. My bad I was confused by that the status hadn't been changed to solved or something.


So... bugzilla = good way to point out problems.

[UbuWu]

2. Debian bug reported goes the other way, I think. Like a bug from debian which also applies to ubuntu gets automatically imported into ubuntu. Why this is assigned to that is wrong, I think. Don't be shy to ask about it on the devel mailling list. I am sure that you will get a quick answer, especially since it has been a few weeks now.

Usually, bugs get assigned after a few days.


Actually it hasn't got anything to do with Debian, and it is just there for historical reasons (either it is some kind of a tribute to Debian, or somebody is just too lazy or too busy to fix it...)

When a bug is not yet assigned to anybody, it is listed as assigned to debian bug importer.

[poofyhairguy]

So... bugzilla = good way to point out problems.

bugzilla= official feedback channel

[az]

When a bug is not yet assigned to anybody, it is listed as assigned to debian bug importer.

Actually, it is listed as "unassigned". I think that if the bug reported manually changes it to something other than unassigned (for example, by accident when first reporting the bug), the bug may go unnoticed. Which is why I jokingly say that bugzilla is both a bug and a bug reporting tool.

And the debian bug importer is quite important. It automagically imports relevant debian bugs into the ubuntu versions of the packages.