PDA

View Full Version : Steering the developers, rather than be steered



Pasdar
November 2nd, 2009, 03:36 PM
The Linux scene is such that developers in the Linux scene are all over the place, going in every direction into infinity and this results in minimal progress in infinity directions. Furthermore, the products are so-so, because they're looking at it from a developer point of view and not from a user pov.

Many of these developers are doing what they do because they like doing it, they do it for fun rather than work. It is also difficult for them to make money from their written source.

In the Linux scene, the developer is at the forefront of development. He decides what is best for the user and works on it in his free time. If you tell the developer to do something (go in a certain direction), you'll most likely get the virtual middle finger from him (he will tell you to do it yourself). This is the very opposite of the paid closed-source market where the user is king. He gets what he wants because he paid for it and you better deliver... otherwise they will not pay anymore.

The only way to change this and for us to become the people who are steering the developers in a certain direction, is by becoming their financiers. A donation here, and one there doesn't help. Donations should be regular and in return for something specific. It should be organized. For example (just an example):
Group A, X number of people want the pitvi developer(s) (a video editing tool) to become top-notch... so they together finance the developer(s) to work full time on it.

Groupd B, X number of people want GIMP to have every function available in Photoshop and they're willing to pay the developers to specifically work on achieving that.

What do you guys think? Can it be done, should it be done?

Npl
November 2nd, 2009, 03:44 PM
Most core linux developers arent doing this for "fun", but beeing employed by companies which get their money from maintaining running servers. The "hobbiest" and "free" linux is some weak marketing argument.

linux people are mostly paid so the product can sell services, means apart from all the evil money corporations there is also zero interest for improving usability. Good luck changing that.

(I know FOSS hobbiest projects exist, but the notable ones are pretty small and not in any way bound to linux)

Pasdar
November 2nd, 2009, 03:46 PM
Most core linux developers arent doing this for "fun", but beeing employed by companies which get their money from maintaining running servers. The "hobbiest" and "free" linux is some weak marketing argument.

linux people are mostly paid so the product can sell services, means apart from all the evil money corporations there is also zero interest for improving usability. Good luck changing that.

(I know FOSS hobbiest projects exist, but the notable ones are pretty small and not in any way bound to linux)
I don't think thats true, because at a gathering of Mark Shuttleworth some developers blamed him of wanting to change hobby into work. They literally said that most developers do it for hobby and don't want it to get turned into work. While they work at Intel, etc, etc, they do this on the side as hobby or to help the community because they care about it for one reason or another...

Npl
November 2nd, 2009, 03:52 PM
Read that page (http://www.linuxfoundation.org/news-media/announcements/2009/08/linux-foundation-updates-study-linux-development-statistics-who-wri).

More than 70 percent of total contributions to the kernel come from developers working at a range of companies...

Sorry, but the time you have for hobbies is not nearly enough to do significant amounts of work, and that counts twice for better developers which get important positions at their workplace. Or let alone have the time to do huge rewrites.

Screwdriver0815
November 2nd, 2009, 03:56 PM
The Linux scene is such that developers in the Linux scene are all over the place, going in every direction into infinity and this results in minimal progress in infinity directions. Furthermore, the products are so-so, because they're looking at it from a developer point of view and not from a user pov.

Many of these developers are doing what they do because they like doing it, they do it for fun rather than work. It is also difficult for them to make money from their written source.

In the Linux scene, the developer is at the forefront of development. He decides what is best for the user and works on it in his free time. If you tell the developer to do something (go in a certain direction), you'll most likely get the virtual middle finger from him (he will tell you to do it yourself). This is the very opposite of the paid closed-source market where the user is king. He gets what he wants because he paid for it and you better deliver... otherwise they will not pay anymore.

The only way to change this and for us to become the people who are steering the developers in a certain direction, is by becoming their financiers. A donation here, and one there doesn't help. Donations should be regular and in return for something specific. It should be organized. For example (just an example):
Group A, X number of people want the pitvi developer(s) (a video editing tool) to become top-notch... so they together finance the developer(s) to work full time on it.

Groupd B, X number of people want GIMP to have every function available in Photoshop and they're willing to pay the developers to specifically work on achieving that.

What do you guys think? Can it be done, should it be done?
this shows that you have not thought about the whole Linux-ecosystem.
There are some big companies (IBM, Red Hat, Novell, Intel... and many more) out there who actually pay developers to develop. They develop things which schould fit to the demands of the customers of these companies.
As an addition, these companies release the result of this work as open source.
The same goes for the DE's: Gnome and KDE. There some developers are paid by above mentioned companies to fullfill targets, to get a Linuxsystem which is usable for enterprises.

And then there are developers who do it in their sparetime, just for fun or because they need a certain functionality. They release their stuff under the GPL too so you can use it.

If you want a feature, you are always free to ask for it. The most developers don't show you the finger. If it is possible for them, they will most likely implement the feature you requested. You can not steer all of them with your money. Because then you run into something like a competition with the big companies (mentioned above).
The culture in open source is: you engage yourself and you get back.
As seen in the KDE development for example, they are really open minded and they listen to the users. Otherwise KDE 4.3 would not be in this state after the 4.0 desaster. Gimp: they now do a GUI which has one single window instead of three... why? Because of users who have asked for that.

But sure, if you want to pay some developers for fulltime work - just do it.

How is it done in closed source? Lets have a look on it: a company director or a marketing guy decide which features are implemented into the software. If you as the customer like it or not: its done. You can send E-mails to them and you can say that you pay them... if they don't want to or if it does not fit into their strategy, they don't do it.

Then there are different points of view at the users: there is one group who wants to have feature abc in there. But you want to have feature 123, which is the opposite... whats now? The one who pays more, gets it, or what?
so, sure you can pay who ever you want but you will not be able to steer the whole development of the Linux-ecosystem.

Screwdriver0815
November 2nd, 2009, 03:57 PM
I don't think thats true, because at a gathering of Mark Shuttleworth some developers blamed him of wanting to change hobby into work. They literally said that most developers do it for hobby and don't want it to get turned into work. While they work at Intel, etc, etc, they do this on the side as hobby or to help the community because they care about it for one reason or another...
Ubuntu != Linux!

Simian Man
November 2nd, 2009, 04:01 PM
Here's a hint. The vast majority of Linux development doesn't have anything to do with Mark Shuttleworth or Ubuntu :).

Pasdar
November 2nd, 2009, 04:04 PM
Read that page (http://www.linuxfoundation.org/news-media/announcements/2009/08/linux-foundation-updates-study-linux-development-statistics-who-wri).


Sorry, but the time you have for hobbies is not nearly enough to do significant amounts of work, and that counts twice for better developers which get important positions at their workplace. Or let alone have the time to do huge rewrites.
Are you even reading what I wrote? Companies actually do contribute to the kernel development because their servers run on it. They don't give a **** about your desktop or what programs you have on it, etc. Those are done by people in their free time.

Keyper7
November 2nd, 2009, 04:07 PM
I don't think thats true, because at a gathering of Mark Shuttleworth some developers blamed him of wanting to change hobby into work. They literally said that most developers do it for hobby and don't want it to get turned into work. While they work at Intel, etc, etc, they do this on the side as hobby or to help the community because they care about it for one reason or another...

(citation desperately needed)

What exactly Mark said to prompt them to say that, and what exactly was the response? I'm inclined to think that they were talking about not wanting their work in Linux to be in control of a CEO or something along these lines. But I cannot say anything conclusive (and neither can you) without citing the exact quote.

Screwdriver0815
November 2nd, 2009, 04:09 PM
Are you even reading what I wrote? Companies actually do contribute to the kernel development because their servers run on it. They don't give a **** about your desktop or what programs you have on it, etc. Those are done by people in their free time.
are you actually informing yourself about what you write?

There are loads of developers employed at companies who do nothing else than developing the GUI, networking stuff, programs... do you think that Evolution for example is developed by a bunch of hobbyists??

Wrong! Its done at Novell.

networkmanager: Red Hat

openoffice: Sun

the list goes on

Pasdar
November 2nd, 2009, 04:12 PM
this shows that you have not thought about the whole Linux-ecosystem.
There are some big companies (IBM, Red Hat, Novell, Intel... ....They develop things which schould fit to the demands of the customers of these companies.

These companies do jack **** on getting you a better functioning gnash, making pitvi a fully functioning video editor, making you a fully functional audio editor, etc. They do exactly as you said, contribute to things that would serve their own purposes and we're feeding off on that. Practically all of it is kernel development, with the exception of programs like OpenOffice that are made to hurt MS.

Screwdriver0815
November 2nd, 2009, 04:14 PM
These companies do jack **** on getting you a better functioning gnash, making pitvi a fully functioning video editor, making you a fully functional audio editor, etc. They do exactly as you said, contribute to things that would serve their own purposes and we're feeding off on that. Practically all of it is kernel development, with the exception of programs like OpenOffice that are made to hurt MS.
and? what is the actual statement, which you want to express with that?

Pasdar
November 2nd, 2009, 04:16 PM
Ubuntu != Linux!

They were not Ubuntu developers. They were linux developers from all over the place, all kinds of projects and teams... Mark has been busy on getting them do what I am talking about at a lower stage... namely to get them all directed in one direction and on a time schedule. That's why he was blamed by some NON-ubuntu developers.

Pasdar
November 2nd, 2009, 04:18 PM
and? what is the actual statement, which you want to express with that?

Haven't figured it out yet? I'm talking about the consumer market and their needs and the developers working in their free time for that specific market. Not about Intel helping intel cards work better because they have their own distro and we're feeding off on that or IBM that works on everything to hurt MS and we feed off those results.... specific programs that are made for the consumer market that users care about... not kernel development.... get it?

Simian Man
November 2nd, 2009, 04:23 PM
These companies do jack **** on getting you a better functioning gnash, making pitvi a fully functioning video editor, making you a fully functional audio editor, etc. They do exactly as you said, contribute to things that would serve their own purposes and we're feeding off on that.

Practically all of it is kernel development, with the exception of programs like OpenOffice that are made to hurt MS.

Not true at all. Did you know that, for example, Red Hat is the driving force behind most of the Gnome desktop? They develop and maintain Gtk+, metacity, nautilus, gnome-panel, evince and others? Not to mention NetworkManager and PulseAudio. Then there is Novell who support Banshee, F-Spot, and much KDE development.

Those software packages are hardly low-level, server stuff that you seem to think.

BTW I am a desktop Linux user and I don't give a flying **** about any of the software you listed.

Screwdriver0815
November 2nd, 2009, 04:25 PM
They were not Ubuntu developers. They were linux developers from all over the place, all kinds of projects and teams... Mark has been busy on getting them do what I am talking about at a lower stage... namely to get them all directed in one direction and on a time schedule. That's why he was blamed by some NON-ubuntu developers.
I hate to quote myself - just shows that you don't read anwers...


Then there are different points of view at the users: there is one group who wants to have feature abc in there. But you want to have feature 123, which is the opposite... whats now? The one who pays more, gets it, or what?

so we have it: different opinions, different directions. Linux and also the whole world of GPL and opensource is not a company, its an ecosystem.


Haven't figured it out yet? I'm talking about the consumer market and their needs and the developers working in their free time for that specific market. Not about Intel helping intel cards work better because they have their own distro and we're feeding off on that or IBM that works on everything to hurt MS and we feed off those results.... specific programs that are made for the consumer market that users care about... not kernel development.... get it?
and what is your problem? Do you want to play some "awesome gamez"? Install Windows then. Linux has a different targeted audience. If you want something else than what is at offer in the Linuxworld, just use another system - thats also the freedom, Linux is about.

If you want to have specific programs, they will be done when there is market for them. If there is no, then the program is not done. Its the same as in the Windows- or Apple world.

I am very satisfied about all the available programs in Linux. I don't give a damn about anything else

Viva
November 2nd, 2009, 04:25 PM
Why does it need to be changed?:confused:

RiceMonster
November 2nd, 2009, 04:26 PM
These companies do jack **** on getting you a better functioning gnash, making pitvi a fully functioning video editor, making you a fully functional audio editor, etc. They do exactly as you said, contribute to things that would serve their own purposes and we're feeding off on that. Practically all of it is kernel development, with the exception of programs like OpenOffice that are made to hurt MS.

I don't blame them. This is because these companies don't develop Linux with the consumer in mind (and to be honest, I wouldn't either). It's mainly for server, embedded space, and maybe enterprise workstations. This is where Linux is useful to them, and it wouldn't make sense for them to put money into gnash or a video editor.


Not true at all. Did you know that, for example, Red Hat is the driving force behind most of the Gnome desktop? They develop and maintain Gtk+, metacity, nautilus, gnome-panel, evince and others? Not to mention NetworkManager and PulseAudio. Then there is Novell who support Banshee, F-Spot, and much KDE development.

Those software packages are hardly low-level, server stuff that you seem to think.

BTW I am a desktop Linux user and I don't give a flying **** about any of the software you listed.

Well said. I find myself agreeing with you a lot.

Simian Man
November 2nd, 2009, 04:27 PM
They were not Ubuntu developers. They were linux developers from all over the place, all kinds of projects and teams... Mark has been busy on getting them do what I am talking about at a lower stage... namely to get them all directed in one direction and on a time schedule. That's why he was blamed by some NON-ubuntu developers.

The point is that Canonical and Ubuntu haven't actually *done* practically anything for Linux development. Besides their warm fuzzy marketing and free CDs getting a few more people to try Linux. What right in hell does Mark Shuttleworth have to "get them all directed in one direction and on a time schedule"?

Oh wait they developed upstart, that's a start I guess :).

Keyper7
November 2nd, 2009, 04:41 PM
All I can say about this proposal is that:


User: I made a patch to add feature X. The reasoning for including it is A, B and C. It would benefit several users besides myself.

Dev: I don't think it's a good idea to include that patch because of D, E and F.

User: I understand the need for D, E and F, but I do believe A, B and C are equally important because of G, H, and I.

(polite discussion continues and a conclusion is reached)

or


User: I would like feature X to be implemented. I'm not a developer myself, but I do think it's important because of A, B and C. It would benefit several users besides myself. I'm not a coder, but I can make mockups and provide testing if it helps.

Dev: I don't think it's a good idea to implement that feature because of D, E and F.

User: I understand the need for D, E and F, but I do believe A, B and C are equally important because of G, H, and I.

(polite discussion continues and a conclusion is reached)

usually work better than


User: I give you one million dollars to implement feature X.

Dev: I don't think it's a good idea to implement that feature because of D, E and F.

User: I didn't ask what you think.

(dev implements X, and users affected by D, E and F are screwed unless they pay too)

and certainly work better than


User: Feature X MUST be implemented because of A, B and C.

Dev: I don't think it's a good idea because of D, E and F.

User: WHAT? A, B and C are OBVIOUSLY MORE IMPORTANT than D, E and F. Don't the devs care about the users.

Dev: "Obviously" is not an argument. I'm not convinced yet that A, B and C are more important.

User: But it's obvious. I see. You devs are all geeks who don't care about what normal users think.

Dev: I never said that. Back on topic, could you please provide more reasoning for A, B and C?

User: That's it. I'm installing Windows and I'm going to rant in a forum. Goodbye, you nazis. Linux sucks. FOSS sucks. And since you devs are from Planet Earth, I guess Planet Earth sucks too.

But unfortunately, the last one is the most used.

FOSS is not a democracy, or a moneycracy, it's a meritocracy. You want something? Either do it yourself or convince the developers it's worth doing for the good of all users. It requires patience, reasoning and politeness. Usually not even the first post of feature requests has those.

GeneralZod
November 2nd, 2009, 04:42 PM
making pitvi a fully functioning video editor

I'm sure I heard that there are at least two full-time pitivi developers employed to work on it.

Edit:

http://en.wikipedia.org/wiki/PiTiVi#History

Viva
November 2nd, 2009, 04:44 PM
This sounds like just another variation of the master distro argument.

koenn
November 2nd, 2009, 04:52 PM
The point is that Canonical and Ubuntu haven't actually *done* practically anything for Linux development. Besides their warm fuzzy marketing and free CDs getting a few more people to try Linux. What right in hell does Mark Shuttleworth have to "get them all directed in one direction and on a time schedule"?

Oh wait they developed upstart, that's a start I guess :).

They've succeded in what numerous distros failed at for 20 years : put a Linux Desktop system in the market. Computer magazines have been proclaiming "the year of linux on the desktop" for years on end - but it wasn't until Ubuntu that mainstream media, for the general population, started paying attention to it and present it as an alternative.

The open source community has plenty of engineers and developers. It needs a few more people who can handle the business end.
M. Shuttleworth is one of them.

BuffaloX
November 2nd, 2009, 05:02 PM
What do you guys think? Can it be done, should it be done?

Wow you got a lot of kicking for this.
Maybe because your description of programmers is a bit offensive.
I have actually successfully on 4 separate programs had changes made, that I thought would improve them, so this is very possible.
To do this, you have to be very specific, very polite, and pointing to how your feature can be made relatively easy may help.
(This forces the programmer to think about the problem, and when you have got them thinking, results can be surprisingly quick.)

I actually agree, but you have to implement an easily accessible ecosystem.
Now only very few projects do this on their own, like Ardour.
I believe Joomla had/has something exactly like what you describe, a reward system which allow you to issue rewards for implementing features you want.

Someone mentioned possible inconsistencies between wishes, resulting in potential conflicts.
I suppose the Gimp one/multiple windows discussion is a good example of this, which seems to be easily solved by giving users both options.

I definitely agree with OP that this could be a very good thing for OSS.
An OSS feature reward system, that is easily accessible, for both users and programmers.
Hobby programmers could make some extra money, and maybe even go full-time if this took off.
For complex features, more people/organizations could join the same request, and increase the reward.

Next comes the problem with inclusion in main tree and maintenance.
Not any code/programmer will necessarily be welcomed into any project.

Pasdar
November 2nd, 2009, 05:17 PM
All I can say about this proposal is that:



or



usually work better than



and certainly work better than



But unfortunately, the last one is the most used.

FOSS is not a democracy, or a moneycracy, it's a meritocracy. You want something? Either do it yourself or convince the developers it's worth doing for the good of all users. It requires patience, reasoning and politeness. Usually not even the first post of feature requests has those.

If it stays like that it will fail miserably on the consumer market. Red Hat understood that and quit the consumer market, Canonical understood that and attempts to change the developers and they're having great success at doing that. At every gathering Mark tries to teach these developers that they need to have users in mind.

papangul
November 2nd, 2009, 05:34 PM
FOSS is not a democracy, or a moneycracy, it's a meritocracy.


FOSS ecosystem is a big, complex thing, it can assimilate and can consist of a variety of different models, systems or ideas in varying doses during different phases of it's evolution.

OPs idea is not exactly new, people(individuals) have been announcing bounties online for getting things done(software related) on websites for a long time. I have no idea whether it is practically possible be do similar thing in an organised basis.

Screwdriver0815
November 2nd, 2009, 05:36 PM
If it stays like that it will fail miserably on the consumer market. Red Hat understood that and quit the consumer market, Canonical understood that and attempts to change the developers and they're having great success at doing that. At every gathering Mark tries to teach these developers that they need to have users in mind.
okay, if you say that it fails, then it will fail... :rolleyes:

and when it fails, then it fails... so what?

Pasdar
November 2nd, 2009, 05:46 PM
okay, if you say that it fails, then it will fail... :rolleyes:

and when it fails, then it fails... so what?
I fail to understand why people like you even waste your energy in replying to such topics? If you don't care then go download slackware and have fun on your own. I don't care about you or your opinion. This topic is meant for people who have something to add the the question I added in the first post.

I think you are the type of user who uses Ubuntu because it works for you with your hardware. You do not care about the product more than that. Other than that you either don't have the money to buy windows, or find it a financial burden to some extent.

For me, I can give windows 7 to everyone living in my area and it wouldn't do a thing to me. Me using Ubuntu has very different reasons. Reasons I don't care to tell you. However, since I use it and there is no better alternative other than windows, that leaves me stuck on this OS. Which means I want to see it develop. What I see is a scene very much stuck closed-minded ideas of how things have to be/should be. This while if developers work together there is no entity on this planet that could compete.... its a shame really... but Im happy people like Mark try to change that...

koenn
November 2nd, 2009, 05:49 PM
At every gathering Mark tries to teach these developers that they need to have users in mind.
Not really. Developers write source code, and it's the distros that configure, compile, package and polish it and deliver it to the users. What I see M. Shuttleworth do is try to organize, coordinate and streamline this process, which is probably benificial to both the developers and all distros.

It's also of an entirely different order and on an entirely different level that this proposal of offering bounties for features, and has nothing to do with telling devs they should listing to (the endless whining, demanding and complaining of) users.

Pasdar
November 2nd, 2009, 05:53 PM
Not really. Developers write source code, and it's the distros that configure, compile, package and polish it and deliver it to the users. What I see M. Shuttleworth do is try to organize, coordinate and streamline this process, which is probably benificial to both the developers and all distros.

It's also of an entirely different order and on an entirely different level that this proposal of offering bounties for features, and has nothing to do with telling devs they should listing to (the endless whining, demanding and complaining of) users.

Actually, lemme quote him from memory of what he said during Linuxcon.... developers need to bring in someone from the street, someone who doesn't know or love open-source and let him use the program, etc, etc... and the developer should "shut the **** up"... from that he should learn what the user wants, etc, etc... one of the developers then said they don't have the money to do that like canonical, etc, etc.

PS: what he says doesn't have to do with a reward system, but it has to do with the reason I say a reward system is needed... to get the developers organised...

Simian Man
November 2nd, 2009, 06:07 PM
I don't blame them. This is because these companies don't develop Linux with the consumer in mind (and to be honest, I wouldn't either). It's mainly for server, embedded space, and maybe enterprise workstations. This is where Linux is useful to them, and it wouldn't make sense for them to put money into gnash or a video editor.

Well said. I find myself agreeing with you a lot.
Likewise :).


They've succeded in what numerous distros failed at for 20 years : put a Linux Desktop system in the market. Computer magazines have been proclaiming "the year of linux on the desktop" for years on end - but it wasn't until Ubuntu that mainstream media, for the general population, started paying attention to it and present it as an alternative.
I don't know about you but the mainstream media and general population I am a part of don't know about Ubuntu any more than they knew about Red Hat or Mandrake.


The open source community has plenty of engineers and developers. It needs a few more people who can handle the business end.
M. Shuttleworth is one of them.
No, it really doesn't.



Actually, lemme quote him from memory of what he said during Linuxcon.... developers need to bring in someone from the street, someone who doesn't know or love open-source and let him use the program, etc, etc... and the developer should "shut the **** up"... from that he should learn what the user wants, etc, etc... one of the developers then said they don't have the money to do that like canonical, etc, etc.

The thing is that on Linux the developers *are* users. Nobody works on a project that they don't intend to use. So if you take another user who has a different idea about how a project should work, who do you think will win: the user who knows very little about the project or the one who is actually developing it?

Pasdar
November 2nd, 2009, 06:19 PM
The thing is that on Linux the developers *are* users. Nobody works on a project that they don't intend to use. So if you take another user who has a different idea about how a project should work, who do you think will win: the user who knows very little about the project or the one who is actually developing it?
Correct, many cases yes. However, some would like to have their work valued (e.g. they would like to see financial reward for what they do). Some would like to make it (their hobby) their job, etc, etc. There are examples of that.

The user does not need to know anything about development or programming. The weird example that Keyper7 gave, makes no sense at all unless you're talking about things like Xorg or Kernel development. Consumer programs are very different... which functions are added to an already normally function program is just a question of choice, what it looks like is just a question of choice, whether time is spent to remove bugs or to add functions is just a question of choice.

Any developer who also knows something about marketing and the financial end of things would know that if he created his own reward system... meaning giving users the choice on what he should do next for money... in the end it would make his work more profitable and more popular... an example of such a reward system is used by codeweavers with such success that codeweavers even gives money back to the wine project...

koenn
November 2nd, 2009, 06:20 PM
Actually, lemme quote him from memory of what he said during Linuxcon.... developers need to bring in someone from the street, someone who doesn't know or love open-source and let him use the program, etc, etc... and the developer should "shut the **** up"... from that he should learn what the user wants, etc, etc... one of the developers then said they don't have the money to do that like canonical, etc, etc.
Nope, that was an explanation about some usability testing that Canonical (distributor !) did, and they'll take their conclusions back to their 'upstreams'.
And the message there was that developers should pay attention to usability, not that devs should listen to users and do as they're told.

Keyper7
November 2nd, 2009, 06:26 PM
If it stays like that it will fail miserably on the consumer market.

Define "like that", please. Because what I said in the post you've replied to was basically "when you are polite, things walk forward". Are you saying that politeness should NOT be a requirement when requesting features?

Screwdriver0815
November 2nd, 2009, 06:27 PM
I fail to understand why people like you even waste your energy in replying to such topics? If you don't care then go download slackware and have fun on your own. I don't care about you or your opinion. This topic is meant for people who have something to add the the question I added in the first post.

I think you are the type of user who uses Ubuntu because it works for you with your hardware. You do not care about the product more than that. Other than that you either don't have the money to buy windows, or find it a financial burden to some extent.

For me, I can give windows 7 to everyone living in my area and it wouldn't do a thing to me. Me using Ubuntu has very different reasons. Reasons I don't care to tell you. However, since I use it and there is no better alternative other than windows, that leaves me stuck on this OS. Which means I want to see it develop. What I see is a scene very much stuck closed-minded ideas of how things have to be/should be. This while if developers work together there is no entity on this planet that could compete.... its a shame really... but Im happy people like Mark try to change that...
maybe you haven't read my comments - why should you? You don't care anyway!

I told you my sight of the things and other did too. I for one fail to understand, whether you have now (after reading the posts in this thread or not) understood how the Linux and open source ecosystem works.
Mark has understood this and he does what is necessary to run his business. But even when some - some not all - statements by Mark seem like they coincide with your opinion, it is still a different approach.

the post of Keyper7 shows the things like they are. You will not change them, neither does Mark. Its an grown ecosystem.

I use Ubuntu because its a nice operating system. And I care about it, otherwise I wouldn't be member of the german translators team. Do you care about it? Do you contribute anything, beside your thoughts about some strategy?

I fail to understand why some people, like you for example, always come up and make statements how to "improve" something which is grown over years and which can not be changed like the monitor resolution or anything like that.
And moreover: I do not undertand why people like you for example then always come up with the statemen "okay, then it will fail", just because the facts are not in line with your opinion.

Linux exists since 1991. It will exist, no matter if you steer anything or if nothing is steered. The big players will continue to produce good products like Evolution, OpenOffice, Gnome, KDE all this amazing stuff, we like.

This is the same as in closed source: do you think, that Microsoft just does anything in Windows for the Home-user? No, they do it for enterprises. Thats the reason why Vista was such a failure. Because next to zero enterprises have switched to Vista.

The same way goes for Linux. Money makes the world go around - where is the money? In the enterprises. When there is a market for anything, some guys will show up and produce it.

koenn
November 2nd, 2009, 06:27 PM
I don't know about you but the mainstream media and general population I am a part of don't know about Ubuntu any more than they knew about Red Hat or Mandrake.
I've heard my boss, a manager with no technical background or interest whatsoever, explain the benefits of linux and open source to her boss, without any prompting from me. And she got most of it right.
I can hardly imagine such a thing happening 5 years ago.

Have you not seen the threads here about ubuntu in the news, BBC and so ?

nmccrina
November 2nd, 2009, 06:27 PM
I think this whole user vs. developer debate is missing the point. The point of open source software (or free software, or whatever) is that you (the "user") can modify the program if you need extra/different functionality. This kind of implies that the user is also a developer. If your modification isn't accepted into the main codebase, you're free to fork it to make it what you want. (I've just read TCATB though, so I have to say of course that this should only be done rarely :p ). There just aren't supposed to be this many non-programmers sitting around waiting for the next version and trying to dictate what the next version looks like. To me, this is the downside of Linux gaining (desktop) marketshare.

Pasdar
November 2nd, 2009, 06:29 PM
Nope, that was an explanation about some usability testing that Canonical (distributor !) did, and they'll take their conclusions back to their 'upstreams'.
And the message there was that developers should pay attention to usability, not that devs should listen to users and do as they're told.
You like to give your own twist to things don't you? Did you listen to the whole speech and the feedback from the room too?

No one said anything about orders. go back and read my PS again... he tries to get across that they need to have the user in mind and not develop from a developers point of view... that was also what his story about design was... that developers don't know what a program should look like... (what the end user will like, etc).

Keyper7
November 2nd, 2009, 06:31 PM
Nope, that was an explanation about some usability testing that Canonical (distributor !) did, and they'll take their conclusions back to their 'upstreams'.
And the message there was that developers should pay attention to usability, not that devs should listen to users and do as they're told.

Exactly. Mark was talking about the usability goals of the Ayatana project, but the development process itself of Ayatana is pretty much the same as of any other FOSS project.

As a matter of fact, Mark himself recently proposed reducing user participation (https://lists.launchpad.net/ayatana/msg00718.html) in the Ayatana discussions.

Bodsda
November 2nd, 2009, 06:47 PM
Just adding my 2 cents here.

I would hate to see average users start influencing developers by the size of their wallets. Things are going reasonably well at the moment. We have huge corporations who know roughly what people want and those are the right people to be influencing development. As soon as John Doe starts throwing his wage packets at devs he completely ruins everything. The program now is steered in a direction that fits 'his' needs, not the needs of the masses.

As far as I am concerned, if you want a feature, request it. If it gets rejected, learn how to program and write the feature yourself and then release it under the GPL. That way, the development focus is still for the masses and you get what you want.

If you don't have the skills to do something, learn! It can cost a grand total of £0.00 to learn adequately enough how to program a new feature into an application. All you need is a bit of patience and commitment.

Kind regards,
Bodsda

mivo
November 2nd, 2009, 07:07 PM
What I see M. Shuttleworth do is try to organize, coordinate and streamline this process, which is probably benificial to both the developers and all distros.

When he is able to polish his own distro so that distro upgrade is actually usable and doesn't break installations, and when even hardcore Ubuntu users no longer feel they need to recommend reinstalls for new versions, perhaps then he will be in a position to lecture others on polishing and streamlining. There are other distros that are much more polished and streamlined. There is enough resentment towards Ubuntu among developers already (Ubuntu takes much more than it gives back), and his approach will not make that any better.

And marketing? Anyone with a few millions at their disposal could do that. And is it working well? Several of the mainstream vendors that used to ship machines with Ubuntu have stopped offering them. Shipit is no longer handing out free CDs to most people. Media coverage has declined other than at BBC (and then it is negative).

Ubuntu started great, but I sometimes wonder if some of the original ideals got lost somewhere on the way. I see more stagnation than progress.

BackwardsDown
November 2nd, 2009, 07:11 PM
There is enough resentment towards Ubuntu among developers already (Ubuntu takes much more than it gives back), and his approach will not make that any better.
It gives us the most popular distro at the moment. That's what I call 'giving back'. And there are ubuntu devs working on other stuff, like upstart.


Shipit is no longer handing out free CDs to most people
People that complain on this forum are not 'most people'. All the new people to Ubuntu can get a free cd just fine.

mivo
November 2nd, 2009, 07:17 PM
People that complain on this forum are not 'most people'. All the new people to Ubuntu can get a free cd just fine.

How do you know? A friend of mine who never posted here and never requested a CD before (because she never wanted to try Linux before) was declined. I'll send her one to the US. (Actually, a set, not just Ubuntu.)

koenn
November 2nd, 2009, 07:22 PM
You like to give your own twist to things don't you?
in stead of blindly going along with your twist ? Definitely.



Did you listen to the whole speech and the feedback from the room too?

Absolutely.






No one said anything about orders. go back and read my PS again... he tries to get across that they need to have the user in mind and not develop from a developers point of view... that was also what his story about design was... that developers don't know what a program should look like... (what the end user will like, etc).


PS: what he says doesn't have to do with a reward system, but it has to do with the reason I say a reward system is needed... to get the developers organised...

Shuttleworth talks about "usability". This is not entirely the same as "what the users want".

What talks about is getting the entire system of devs, projects, distros, the interaction between them, the timing and scheduling of releases, ... coordinated. Not entirely the same as "get the devs organized".

So much for 'your own twist'.

As for rewards :
In that same speech, Shuttleworth explains what the benefits would be of that whole effort to coordinate releases, devs, etc. For the devs / upstream projects, this would mean, amongst others, that they'd be able to predict/control which of their releases would be included in any given distro, allowing them to focus bug fixes or new features to that version of their software, in stead of having to maintain multiple production versions in parallel. Bug reports or other feedback from users would also become easier to handle, because it would be limited to 1 version of any given program, not a multitude of different versions in Ubuntu, Debian, Red Hat, Fedora, SLES, OpenSuse, ..
And obviously, less maintenance would free up time for new development.

So the reward would be in something they care about, and takes up the major part of their waking hours.
Somehow I think that would have a slightly bigger impact than some extra $$.

Keyper7
November 2nd, 2009, 07:43 PM
Two little bits of wisdom for this thread:

http://weblog.obso1337.org/2008/four-words-for-funpidgin/ (don't forget to read the update)
http://www.youtube.com/watch?v=bU4FktauK24 (watch from 1:15)

issih
November 2nd, 2009, 07:46 PM
Money should not decide what features get developed...whatever provides the best solution for the most users should.

Now to be slightly controversial - and please be aware that I say this as someone who develops software, the absolute WORST judge of any piece of software is the developer.

Partly because of pride, but mostly purely because of overfamiliarity. If I make a menu item called bob, and clicking on bob turns all the screen colours negative, copies all the unselected text and then puts the system into an infinite loop printing out "chickens are tasty", then when I click bob and all these ridiculous things happen, I think "excellent it works" not "WTF".

Developers know too much about the quirks of their own design, and often we stop once it works, not once its ready.

Quality control of the user experience is lacking in our ecosystem, and pretending it isn't will not help anyone.

koenn
November 2nd, 2009, 07:47 PM
When he is able to polish his own distro so that distro upgrade is actually usable and doesn't break installations, and when even hardcore Ubuntu users no longer feel they need to recommend reinstalls for new versions, perhaps then he will be in a position to lecture others on polishing and streamlining. There are other distros that are much more polished and streamlined.
I thought it was clear I was talking about organisation and (business) processes, and coordination and interaction between groups of people. I don't see how a technical issue such as this is relevant




There is enough resentment towards Ubuntu among developers already (Ubuntu takes much more than it gives back), and his approach will not make that any better.

What BackwardsDown said. And there are other examples: Gnome, X, ...
Besides that, Canonical's focus is clearly not on development, so they probably don't have all that much to give back in terms of lines of code.



And marketing? Anyone with a few millions at their disposal could do that. And is it working well? Several of the mainstream vendors that used to ship machines with Ubuntu have stopped offering them. Shipit is no longer handing out free CDs to most people. Media coverage has declined other than at BBC (and then it is negative).
Anyone could do it.
And 25 years after the first attempt to create a linux binary distribution, you know, the kind you don't have to compile from scratch, and after numerous attempts (RedHad, Lindows, Suse, Mandrake, Mandriva, Corel, ... ) Canonical is apparently the first one that might have a chance of pulling it of.

Have a look around on these forums and wade through all those questions of people who obviously only yesterday touched linux for the first time in their lives.

Looks to me Canonical knows what it's doing.

koenn
November 2nd, 2009, 08:03 PM
Two little bits of wisdom for this thread:

http://weblog.obso1337.org/2008/four-words-for-funpidgin/ (don't forget to read the update)
http://www.youtube.com/watch?v=bU4FktauK24 (watch from 1:15)

:)

Mr. Picklesworth
November 2nd, 2009, 08:03 PM
PS: what he says doesn't have to do with a reward system, but it has to do with the reason I say a reward system is needed... to get the developers organised...

This sounds, to me, like you view free software developers as employees. They aren't, you are not their employer, and for many of them this is a hobby, not a career.

When you throw in bounties and act like their boss, that changes.

For more on that note, there is a cautionary tale out there somewhere about the fallout from a bounty which was offered to the person (or people) to implement good colour management (or something along those lines - memory is hazy here) for GIMP. Interestingly, that bounty was offered by our own Mark S :)

On another note, I agree entirely with what issih says about quality control, but we don't need money to change that. We just need sane designers. I think GNOME is good at this with its usability mailing list that people comfortably post to, and Ubuntu has some nice movement in this way, too.

Pasdar
November 2nd, 2009, 08:37 PM
This sounds, to me, like you view free software developers as employees. They aren't, you are not their employer, and for many of them this is a hobby, not a career.

When you throw in bounties and act like their boss, that changes.

For more on that note, there is a cautionary tale out there somewhere about the fallout from a bounty which was offered to the person (or people) to implement good colour management (or something along those lines - memory is hazy here) for GIMP. Interestingly, that bounty was offered by our own Mark S :)

On another note, I agree entirely with what issih says about quality control, but we don't need money to change that. We just need sane designers. I think GNOME is good at this with its usability mailing list that people comfortably post to, and Ubuntu has some nice movement in this way, too.
While I am not a developer any more, I have considerable programming experience and know what I'm talking about. It's just that I chose a different path somewhere along my career.

I consider myself a normal user now since I haven't programmed (other than MATLAB, which I don't really consider programming) for many years. What I and many other non-developers would like to see is for the OS to wow us. Ubuntu does not wow you, it "works well enough" and this is what you see from many users on this forum.

What I see is a scene going in every direction imaginable at snail pace. I don't know what these programmers have learned, maybe they're self taught. However at university we also learnt the management and business side of programming and the way I see it be done is not what you learn at uni. Everything is done "at the coding level" as they say... nothing else seems to be considered. This will lead to a dead end each time, which is also what you see in real.

Note that I'm not talking about the major projects being worked on, like the kernel... but general application development in the scene.

23meg
November 2nd, 2009, 08:48 PM
You want to "steer" free software developers, or development, without being their boss? The time-proven way of doing that is to pay with your time, rather than your money (http://www2.bryceharrington.org:8080/drupal/node/52).

I'd rather work with people than attempt to boss them around. It produces superior results.

joebodo
November 2nd, 2009, 09:02 PM
Marketing is usually the driving force behind what goes into products. A set of features is requested, resources are assigned, and the work is done. The sales team sells the product by showing how these features provide benefit to potential customers.

In this model, any user requests from customers will only be implemented if they can be sold as new features in subsequent releases. There must be a cost benefit for allocating resources to do any type of work. Problems will only be fixed if there is the potential to lose existing customers,

In open source development, users can supply feedback on any issue or requested feature. It is up to the developer to decide if the work should be done. Open source has a much better potential to create customer focused software since usability requests can be implemented without having to justify the development expense.

Pasdar
November 2nd, 2009, 09:26 PM
You want to "steer" free software developers, or development, without being their boss? The time-proven way of doing that is to pay with your time, rather than your money (http://www2.bryceharrington.org:8080/drupal/node/52).

I'd rather work with people than attempt to boss them around. It produces superior results.
This is a good idea for some small project... e.g. nice program... here I made a good looking theme for it... or, here I made a filter for it.. but not on the large scale financing of good high potential programs I'm talking about.

Whether you like it or not, the best programs are made by paid developers, paid for their code... and not anything else and mostly they're closed source. the whole team behind GIMP can apparently not do what one guy is doing ON HIS OWN with pixel image editor for 10 operatung systems!!... i just hope the final gets released. Unfortunately, the greatest things always end up being closed-source. they know the value of their code, everyone else was somehow too incompetent to make it, so you don't want to lose your income by open-sourcing it... thats the taught behind these guys not open sourcing.. and I don't blame them... they want to earn money...

Unfortunately pixel image editor is an example of a developer that cant be trusted.. thats why I think it would be great to support open source developments like for example GIMP. the problem is... many of these project could move at a much faster pace, they could be greater, they could etc etc, but they not, because i think scene is not organised enough and the developers don't really have that much of an idea of what that specific segment of users want.

For example Im a professional photoshop user. However I hear these kids on the forum say GIMP is great, its good enough, etc. Yeah, maybe great for your school project, or some hobby work, having some fun, making a lame signature for on the forum... but this is NOT the group GIMP developers should be looking at. GIMP developers should look at those guys who have the highest of standard... expect professional features like found in photoshop...

The question is, do you want to create a replacement for MS paint or Adobe Photoshop, a replacement for MS Movie editor or Adobe premiere... do you get what I mean... there is too much, "good enough for normal use" in this scene, and too little... "wow, i don't need windows ever again"... just so you know, I don't even run windows... what I do is out of principle. while most people on this forum dual boot into windows and then hypocritically claim linux/ubuntu is all they need and its great and dont critisize.. thats ********... time to grow up

JillSwift
November 2nd, 2009, 09:50 PM
Pasdar, repeating an assertion doesn't add veracity to it. You're arguing your unfounded beliefs about development and providing no supporting evidence.

Pasdar
November 2nd, 2009, 09:53 PM
Pasdar, repeating an assertion doesn't add veracity to it. You're arguing your unfounded beliefs about development and providing no supporting evidence.
Look at the reasons most people dual boot (other than games). Its because they need a certain program. We have alternatives for everything, but all alternatives are "good enough" for normal use. e.g. some kid from school can use open office, but someone from university might find that ms equations does not function in open office.

What evidence would you like? Just read people's posts...

Screwdriver0815
November 2nd, 2009, 10:03 PM
Look at the reasons most people dual boot (other than games). Its because they need a certain program. We have alternatives for everything, but all alternatives are "good enough" for normal use. e.g. some kid from school can use open office, but someone from university might find that ms equations does not function in open office.

What evidence would you like? Just read people's posts...
as I read lots of peoples posts and I always searched for an exact list of features Gimp is lacking when compared to Photoshop...

it seems you are a professional in this area: so please tell me: which features does Photoshop have, which Gimp doesn't have? could you please list them? Thanks!

JillSwift
November 2nd, 2009, 10:04 PM
Look at the reasons most people dual boot (other than games). Its because they need a certain program. We have alternatives for everything, but all alternatives are "good enough" for normal use. e.g. some kid from school can use open office, but someone from university might find that ms equations does not function in open office.Yeah, they may want to use OOo Formula instead of MSOffice Equations for that.


What evidence would you like? Just read people's posts...
You mean where folks get annoyed that GIMP doesn't work like Photoshop? Does that make GIMP inferior to Photoshop, or does it just make it different? That's all very unquantifiable.

Or are you referring to the far less frequent posts from real pros who can cite what GIMP lacks compared to the the professional toolbox? If so, how does that translate into the claim that development methodology is at issue?

23meg
November 2nd, 2009, 10:09 PM
This is a good idea for some small project... e.g. nice program... here I made a good looking theme for it... or, here I made a filter for it.. but not on the large scale financing of good high potential programs I'm talking about.


It's the basic way of working through which free software has reached where it is today, which is remarkable.

That is not to say that where it is today is good enough and we should stick with the status quo and not experiment with new ways of doing things, but that it isn't necessarily a "small project" thing.

Pasdar
November 2nd, 2009, 10:21 PM
as I read lots of peoples posts and I always searched for an exact list of features Gimp is lacking when compared to Photoshop...

it seems you are a professional in this area: so please tell me: which features does Photoshop have, which Gimp doesn't have? could you please list them? Thanks!
I wont mention familiarity... because they cant be blamed for that...

the user interface and general way of usage is not made with the professional user in mind (e.g. simple example, right click brush resize).
limited color space
layer effects
limited filters and filter options
color adjustments, blending effects, text adding issues

there are countless other things... with ten thousand submissions Novell listed Photoshop as the number one most requested port for linux... just so you know... there really is a huge difference between gimp and photoshop for pro editors.

koenn
November 2nd, 2009, 10:48 PM
T.. its good enough, etc. Yeah, maybe great for your school project, or some hobby work, having some fun, making a lame signature for on the forum... but this is NOT the group GIMP developers should be looking at....
Maybe it is exactly the group of users GIMP should be lloking at.

Let's take this "good enough" first. What's wrong with "good enough" ? In my book, good enough means : does what i need it to do. I'm not a graphic designer. I don't work in pre-press. I need something to crop a picture. GIM lets me do that. People with a hobby in photography or creating graphic stuff may need a (quite) a bit more than that, but unless they're semi-pros ar pros, GIMP will be able to handle their needs.

Photoshop is a professonal tool. It's popular on windows because people wanted something a little more powerful than ms-paint, and they only alternative was to get a pirated copy of Photoshop - the Windows ports of Gimp came a bit too late, so now you end up with a userbase that is used to photoshop, and wont look at anything else.

And of course, the real pros will have Macs with an entire Adobe Designer suite and Quark, ... - that's hardly the market GIMP is aiming at.

zekopeko
November 2nd, 2009, 10:53 PM
Perhaps the OP should remember that developers aren't peons to do your (or anyones for that matter) bidding.
FLOSS developers that aren't being paid to do the work on projects are under no obligations do follow user demands.

And what the OP is talking about is usability which Ubuntu is pushing in the community. This is different to "what user(s) want". And it's ridiculous to think that "users that pay have a say" in commercial offering on every little detail.
Users want to do specific tasks and software vendors (once they see the market for a feature) do usability test to determine what exactly, and more importantly how, to implement "the want" in the next release of their offerings.

When you literally listen to end users you get million options and a messy code base.
But when you listen to what (the majority of) users want to do, preform usability tests and refine your design you get a "sane default" feature that actually preforms better and if done right makes the thing the user wanted easier or better for said user to use.

23meg
November 2nd, 2009, 11:04 PM
When you literally listen to end users you get million options and a messy code base.

The first rule of usability is "Don't listen to users" (http://www.useit.com/alertbox/20010805.html).

aysiu
November 2nd, 2009, 11:09 PM
The first rule of usability is "Don't listen to users" (http://www.useit.com/alertbox/20010805.html).
From that article:
Too frequently, I hear about companies basing their designs on user input obtained through misguided methods. A typical example? Create a few alternative designs, show them to a group of users, and ask which one they prefer. Wrong. If the users have not actually tried to use the designs, they'll base their comments on surface features. Such input often contrasts strongly with feedback based on real use. But most of us actually use Ubuntu and its default programs. We aren't just looking at demos and picking which demo we think looks best.

If users actually use an interface, their input is valuable, because it's based on real experience.

23meg
November 2nd, 2009, 11:14 PM
But most of us actually use Ubuntu and its default programs. We aren't just looking at demos and picking which demo we think looks best.

If users actually use an interface, their input is valuable, because it's based on real experience.

It's not that it's not valuable at all, but watching people use an interface reveals much more valuable insights than hearing people's reports of using it.

From the article:


However, when collecting preference data, you must take human nature into account. When talking about past behavior, users self-reported data is typically three steps removed from the truth:

In answering questions (particularly in a focus group), people bend the truth to be closer to what they think you want to hear or what's socially acceptable.

In telling you what they do, people are really telling you what they remember doing. Human memory is very fallible, especially regarding the small details that are crucial for interface design. Users cannot remember some details at all, such as interface elements that they didn't see.

In reporting what they do remember, people rationalize their behavior. Countless times I have heard statements like "I would have seen the button if it had been bigger." Maybe. All we know is that the user didn't see the button.

aysiu
November 2nd, 2009, 11:19 PM
It's not that it's not valuable at all, but watching people use an interface reveals much more valuable insights than hearing people's reports of using it.

From the article:
Well, if they can arrange to watch users, great.

But to be honest most of the feedback I've seen isn't based on fuzzy memories. People have legitimate interface design gripes... that doesn't mean they are necessarily right. The developer may have been thinking a certain way, and that certain way may have a certain logic to it, but two different ways can have their own merits.

Usually, suggestions do come out of real experience (How come I can't do something I used to be able to do? Why did this behavior (mouse movement, key combination) produce one outcome in a program and another outcome in a different program?).

On the other hand, I have yet to come across a developer's choice in Linux that didn't make some kind of sense, even if I disagreed with it. Usually, even if the developer agrees with you in theory, implementing the change isn't always easy or worth it for the small gains in consistency or familiarity. Sometimes certain regressions appear in the short-term for long-term eventual stability.

And sometimes developers just screw up (they are human, after all).

zekopeko
November 2nd, 2009, 11:23 PM
The first rule of usability is "Don't listen to users" (http://www.useit.com/alertbox/20010805.html).

Should re-phrase what I think.
Don't listen to how users want something to behave/look, but what they want to do.
More of a passive interaction then an active one.
The user doesn't want (pseudo GIMP example that might not reflect real GIMP problems) "GIMP to do brush resizing like Photoshop" but to make brush resizing simpler and more intuitive.
That's what I was aiming for with my statement.

zekopeko
November 2nd, 2009, 11:28 PM
Well, if they can arrange to watch users, great.

But to be honest most of the feedback I've seen isn't based on fuzzy memories. People have legitimate interface design gripes... that doesn't mean they are necessarily right. The developer may have been thinking a certain way, and that certain way may have a certain logic to it, but two different ways can have their own merits.

Usually, suggestions do come out of real experience (How come I can't do something I used to be able to do? Why did this behavior (mouse movement, key combination) produce one outcome in a program and another outcome in a different program?).

Office 2007 is one of the best interfaces that I have used. Yes people complained a lot but that was because they didn't use it.
That thing is so intuitive and easy. It informs me about every little option with nice "super tooltips" and every tool is placed in the proper category, and can be found just by some logical exploration.


On the other hand, I have yet to come across a developer's choice in Linux that didn't make some kind of sense, even if I disagreed with it. Usually, even if the developer agrees with you in theory, implementing the change isn't always easy or worth it for the small gains in consistency or familiarity. Sometimes certain regressions appear in the short-term for long-term eventual stability.

And sometimes developers just screw up (they are human, after all).

Well what makes sense to a developer might not make sense for the majority of users :)

23meg
November 2nd, 2009, 11:28 PM
Well, if they can arrange to watch users, great.

Ideally, separate people, usability testers and designers should, not developers. Developers should work on implementing design specifications that are produced based on qualitative usability testing.

Canonical currently has a User Experience and Design team that does precisely this.


But to be honest most of the feedback I've seen isn't based on fuzzy memories. People have legitimate interface design gripes... that doesn't mean they are necessarily right. The developer may have been thinking a certain way, and that certain way may have a certain logic to it, but two different ways can have their own merits.

Most of the good feedback I've seen in bug reports isn't, and it's actually valuable. It constitutes what I estimate as less than 1% of the design commentary (not feedback) that people produce online, mostly on web forums, blogs and comments.

cariboo907
November 3rd, 2009, 12:41 AM
Just to enforce what 23meg and zekopeko were saying, look at the sceenshots to see what happens when you let a developer design the program interface. The program in question was sold for many thousands of dollars to logging companies all over North America.

zekopeko
November 3rd, 2009, 12:51 AM
Just to enforce what 23meg and zekopeko were saying, look at the sceenshots to see what happens when you let a developer design the program interface. The program in question was sold for many thousands of dollars to logging companies all over North America.

My eyes!!!!! They BURN!!!!

issih
November 3rd, 2009, 12:54 AM
Whilst I somewhat started the mini thread explaining why developers are a bad source of user interface design, I have to fundamentally disagree with the idea of separating the developer from the process entirely and merely handing them a requirement sheet.

In fact I strongly advise those at canonical to rethink this decision immediately - it WILL NOT WORK.

Developers are not little atomic units that can be commanded to do the bidding of a grand system architect. They are people, and they produce their best work if they understand the big picture, the overall design philosophy and the reasoning behind it. If they know why something is needed they will come up with a design that is easier to work with, simpler to extend in ways that suit the requirements and far far better overall.

The management driven belief (sadly encouraged by poor OO programming philosophy) that you can separate everything out into little bits and put it together into a coherent whole at the end is absolutely wrong.

Good code needs as many people as possible to understand the point of it from the very start.

Change tack - now.

zekopeko
November 3rd, 2009, 01:06 AM
Whilst I somewhat started the mini thread explaining why developers are a bad source of user interface design, I have to fundamentally disagree with the idea of separating the developer from the process entirely and merely handing them a requirement sheet.

In fact I strongly advise those at canonical to rethink this decision immediately - it WILL NOT WORK.

Developers are not little atomic units that can be commanded to do the bidding of a grand system architect. They are people, and they produce their best work if they understand the big picture, the overall design philosophy and the reasoning behind it. If they know why something is needed they will come up with a design that is easier to work with, simpler to extend in ways that suit the requirements and far far better overall.

The management driven belief (sadly encouraged by poor OO programming philosophy) that you can separate everything out into little bits and put it together into a coherent whole at the end is absolutely wrong.

Good code needs as many people as possible to understand the point of it from the very start.

Change tack - now.

I think that you are missing the point. Coders aren't the best interface designers out there. They don't get formal training in that but in how to write nice, maintainable code.
Look at the SoftwareCenter spec in the wiki. mpt designed the user interactions but it was up to mvo to implement the thing the way he thought was best.
I think that coders should at least have some understanding of fundamentals of interface design (at least in FLOSS world).
Complex interactions should be designed by interface designers since this is their field of expertise.

issih
November 3rd, 2009, 01:16 AM
Nope, I absolutely get it, in fact I was the first person in this thread to state, categorically, that developers are the worst possible designers of user interaction.

My point however stands, and I stand by it. You give someone a set of interactions and they will code just that, no more no less, and any change away from that specific design will be difficult and intractable. If you include the developer in the design decisions, so that they follow the thinking behind it, then the whole system will be geared to produce that experience. It will be more easily adaptable, without damaging the under-ridding philosophy, and put quite simply - better.

Its equivalent to the difference between a program and a framework. Give a developer a list of needs and you get a program. Tell them what you want to achieve, what you want to achieve with it and why, and you get a framework...which in the end will be worth a million times more.

Mr. Picklesworth
November 3rd, 2009, 01:28 AM
Look at the reasons most people dual boot (other than games). Its because they need a certain program. We have alternatives for everything, but all alternatives are "good enough" for normal use. e.g. some kid from school can use open office, but someone from university might find that ms equations does not function in open office.

What evidence would you like? Just read people's posts...

I consider Blender (especially 2.5) more than “good enough.” In fact, I consider it best of class as a 3D modelling, rendering and graphics (+video) compositing tool, partly because it is so darn fast. It is one such “alternative.”

And on the topic of 2.5, previous (and current) versions of Blender have definitely had a very geeky design. 2.5 has a more refined design process you may want to take a look at. It getting there did not involve excluding users who couldn't pay for bounties.

FreeDesktop.org is another place you may want to look at. I think they are doing a good job leading this desktop in a cohesive direction. While the project is a bit shaky at times, it has definitely emerged as a single entity that GNOME, KDE and XFCE developers alike can look to (and point at) for guidance.



Here is that cautionary tale I mentioned:
http://dneary.free.fr/gimp_bounties.html


And, lastly, if you have a serious usability concern - for a GNOME project especially - and it is in context for the purpose of the software, I doubt that you will be ignored. Believe it or not, people who make free software and actually distribute it (instead of just sitting on it) really do care about their work.

wersdaluv
November 3rd, 2009, 02:10 AM
Interesting discussion going on here.

I think, it's best to have legitimate usability testing using the tools available today.

papangul
November 3rd, 2009, 03:42 AM
The first rule of usability is "Don't listen to users" (http://www.useit.com/alertbox/20010805.html).
Why are you treating that article as if it is a sermon of God written on stone? While that approach might be essential in some situations, other approaches or viewpoints can prove to be equally valid in other situations or both viewpoints might apply in a particular situation.

A no. of articles like that are published everyday and they almost always present an 'one side of the coin' viewpoint(mostly when the other side viewpoint becomes very dominant), if someone takes them very seriously - he will end up having a distorted one sided view of the world.

23meg
November 3rd, 2009, 03:59 AM
Why are you treating that article as if it is a sermon of God written on stone?

I'm not. I just agree with the gist of it.

Pasdar
November 3rd, 2009, 08:13 AM
I consider Blender (especially 2.5) more than “good enough.” In fact, I consider it best of class as a 3D modelling, rendering and graphics (+video) compositing tool, partly because it is so darn fast. It is one such “alternative.”

And on the topic of 2.5, previous (and current) versions of Blender have definitely had a very geeky design. 2.5 has a more refined design process you may want to take a look at. It getting there did not involve excluding users who couldn't pay for bounties.

FreeDesktop.org is another place you may want to look at. I think they are doing a good job leading this desktop in a cohesive direction. While the project is a bit shaky at times, it has definitely emerged as a single entity that GNOME, KDE and XFCE developers alike can look to (and point at) for guidance.



Here is that cautionary tale I mentioned:
http://dneary.free.fr/gimp_bounties.html


And, lastly, if you have a serious usability concern - for a GNOME project especially - and it is in context for the purpose of the software, I doubt that you will be ignored. Believe it or not, people who make free software and actually distribute it (instead of just sitting on it) really do care about their work.

One of the problems the Linux scene has are users who think that just because they're happy and content everyone else should be too. Just because they can run something, nothing is wrong with the software, etc, etc. It's very me-me-me rather than having anything to do with "community" as many proclaim.

Well there are some programs that are really good. The free video editing tools on Linux are way better than the free tools on Windows. Since I don't go that far with video editing, i'm content with that. However, that doesn't mean its where it should be. Many Adobe Premiere users are probably looking for a premiere alternative and the Linux software most likely will not deliver for them.

I read your link. Do you know what it makes me think? That just as I thought we are dealing here with a bunch of "guys" (the developers) who are self taught (or have studied pre-bachelor level, or maybe studied in ancient times) and making software/working on project as a hobby next to their work. I don't mean to be insulting, but this is just the way it seems to be sadly.

They couldn't deal with Mark's offer because they are basically a disorganized hobby club. This is a textbook example of a project going wrong because it doesn't follow the organisational rules. It seems like everyone is working on 'something' and 'it'll get done when its done'.

You know what will happen if you email an MS developer about something you want him to add for X amount of money? He'll tell you to talk to the manager, IF he takes you serious. A group can not function efficiently without a leader.

Crunchy the Headcrab
November 3rd, 2009, 08:29 AM
I don't think you are being fair to the developers. Most of the useful software that has been created is because some developer or group of developers has had a need for something so they made it themselves. They didn't make it for you. They made it because they wanted to. However, given the hacker culture that surrounds Linux and the GNU, they were more than willing to share their work and their code with the community. Furthermore, many of them graciously improve their projects to help more users every day. Yeah, some of them are paid to do it and those that are answer to the company that hired them, not you. The others are just being plain helpful, for free.

Pasdar
November 3rd, 2009, 08:33 AM
I don't think you are being fair to the developers. Most of the useful software that has been created is because some developer or group of developers has had a need for something so they made it themselves. They didn't make it for you. They made it because they wanted to. However, given the hacker culture that surrounds Linux and the GNU, they were more than willing to share their work and their code with the community. Furthermore, many of them graciously improve their projects to help more users every day. Yeah, some of them are paid to do it and those that are answer to the company that hired them, not you. The others are just being plain helpful, for free.
If we talk about hacker culture, then I think about slackware. Not about the business called Ubuntu, something you use. Canonical wants Ubuntu and everything with it to become a lean mean business machine... you're talking about a tweaking hobby club that everyone contributes to with code and talks about new hacks.... why do you use Ubuntu anyway?

Crunchy the Headcrab
November 3rd, 2009, 08:39 AM
If we talk about hacker culture, then I think about slackware. Not about the business called Ubuntu, something you use. Canonical wants Ubuntu and everything with it to become a lean mean business machine... you're talking about a tweaking hobby club that everyone contributes to with code and talks about new hacks.... why do you use Ubuntu anyway?
I use ubuntu mostly because I like the repositories full of easily installable software that mostly other people than Canonical have made. :) I have also used Fedora, OpenSuse, Debian, and Arch.

Pasdar
November 3rd, 2009, 11:02 AM
I use ubuntu mostly because I like the repositories full of easily installable software that mostly other people than Canonical have made. :) I have also used Fedora, OpenSuse, Debian, and Arch.

Correct, thats why I use it too. Canonical wants to deliver what the masses want (like Apple and MS), not what some tweakers want. If it wanted to do that it would never have grown beyond anything similar to slackware and die a slow death like slackware/gentoo/etc... it doesn't want to be a distro for hermits... we have enough of those... in fact we have 17000 distros and counting.


This scene needs to mature, learn to self criticize... do any of you think I or anyone else on this forum criticizing wouldn't be here if we didn't care?

One other thing I hate about this scene is that they criticize anything that aims to look like the competition. E.g. critics of the new single window system of GIMP, the look of abiword, etc, etc... especially anything that is similar to anything on any windows system... woe onto them, lol... they'll be but raped in the showers within seconds...

what these critics need to get through their head is that these big shot companies employ MANY people and some even psychologists to find out what exactly would be the best features, what interface would be best, most efficient, most ergonomic, etc, etc... they have different teams working on different things, communication is at a very high standard and everything is strictly organised and according to time tables.

Since 'we' don't have the money to do these things, we have to do it the guerrilla way.... look at the competition... product X is most used... people love the look and the usage... I (the developer) hate it... but since that's what seems to be what users like... let me adapt mine to have everything placed in the same way.... surely with the thousands of employees they have and the millions of buyers they know better....

However, for some reason... maybe pride... the above is ALMOST absent in the scene...

I'm really happy with the KDE developers team... i think they get this... just take a look at the wide range of programs they make and are working on... people criticize them... but you just wait until everything has matured... we'll see who was right... (in case you take this the wrong way.. im not talking about the flashy design perse, but the applications, like e.g. Krita... how they fasion them, what features they add, etc, etc)...

Keyper7
November 3rd, 2009, 01:28 PM
I'm really happy with the KDE developers team... i think they get this... just take a look at the wide range of programs they make and are working on... people criticize them... but you just wait until everything has matured... we'll see who was right... (in case you take this the wrong way.. im not talking about the flashy design perse, but the applications, like e.g. Krita... how they fasion them, what features they add, etc, etc)...

If you are happy about the work of a WORLDWIDE team of MASSIVE size that is responsible for HUNDREDS (if not thousands) of popular applications and maintains ONE OF THE TWO MOST USED desktop environments in the Linux scene, would you please be kind enough to stop saying that the Linux scene has development problems as a whole?

Over-generalization was the problem of your posts since the beginning of this thread. Generalizing never achieves anything: only results in solutions for the wrong problem.

blueturtl
November 3rd, 2009, 04:21 PM
The Linux scene is such that developers in the Linux scene are all over the place, going in every direction into infinity and this results in minimal progress in infinity directions. Furthermore, the products are so-so, because they're looking at it from a developer point of view and not from a user pov.
Linux developers most commonly are users of their own software. They create software so that it satisfies their needs. If others like how a particular developer envisions their program, the project grows. If you don't like a particular software piece, pick another one that better matches your tastes or make your own.


It is also difficult for them to make money from their written source. People can be hired to work on a software project be it OSS or closed source. How is this difficult?


In the Linux scene, the developer is at the forefront of development. He decides what is best for the user and works on it in his free time. If you tell the developer to do something (go in a certain direction), you'll most likely get the virtual middle finger from him (he will tell you to do it yourself). This is the very opposite of the paid closed-source market where the user is king. He gets what he wants because he paid for it and you better deliver... otherwise they will not pay anymore.
If I were a developer and someone came to tell me that they think a piece of software I made specifically for myself doesn't work the way they'd like it to work you'd probably be right about my reaction. Boohoo. But this being "the Linux scene" if there are others who want the software changed the project can easily be forked. It seems from how you write that you see this as a weakness but actually this is why in "the Linux scene" the user is king, and not in the closed source world like you say. In LinuxVille Everybody gets what they want (as long as they contribute). In a closed source world you can cry all you want to Bill Gates about having IE removed from Windows because it would make a better product. You can't do it yourself, and you can't pay for anyone else to do it either. It's not going to happen, period! Commercial interests dictate software in the closed world and more often than not those interest are in contrast to what the user wants.


The only way to change this and for us to become the people who are steering the developers in a certain direction, is by becoming their financiers. A donation here, and one there doesn't help. Donations should be regular and in return for something specific. It should be organized. For example (just an example):
Group A, X number of people want the pitvi developer(s) (a video editing tool) to become top-notch... so they together finance the developer(s) to work full time on it.

Groupd B, X number of people want GIMP to have every function available in Photoshop and they're willing to pay the developers to specifically work on achieving that.

What do you guys think? Can it be done, should it be done?

Frankly I think you have no clue about what makes open source software so great. I can agree on some general ideas that you present (back your favorite projects with monetary donations, coordination and user feedback is a good thing), but to steer developers? Steer users? Why steer when you can just differentiate. When you have the power to fork and better yet pay someone to fork for you. Why do we have to take something that others like and change it when we can create a new solution for us? Because then we'll have 10 000 distros and ten different editors and...

I fail to see what the problem is.

andrew.46
November 7th, 2009, 11:50 AM
Hi Pasdar,


If we talk about hacker culture, then I think about slackware. Not about the business called Ubuntu, something you use. Canonical wants Ubuntu and everything with it to become a lean mean business machine... you're talking about a tweaking hobby club that everyone contributes to with code and talks about new hacks.... why do you use Ubuntu anyway?

As both a committed Slackware user and some one who has worked with Ubuntu long and hard enough to become a Member I can tell you the common idea with both distros is that people use them to 'Have fun'. Newer ideas about 'business machines' and development cycles etc are certainly not why I personally use Ubuntu.

Andrew

alexk554
November 7th, 2009, 12:34 PM
If I were a developer and someone came to tell me that they think a piece of software I made specifically for myself doesn't work the way they'd like it to work you'd probably be right about my reaction. Boohoo. But this being "the Linux scene" if there are others who want the software changed the project can easily be forked.

I might get lambasted for this but whatever.

This is what really irks me about the Open Source community. Its quite selfish in that regard, very few people are interested in working towards something, each person just wants to make their own program, or "have fun" as has been mentioned a few times in this thread. If you dont like it, stuff off, write your own, etc.

Most hardcore Linux dudes dont care (or realise) the fact that Linux will never grow beyond what it is unless it changes. (and then they say things like "When Linux has dominant market share...." wtf?)

i.e., Ubuntu used to have Pidgin, now it has (damn, forgot what its called). These 2 apps do virtually the same job, what if the 2 developers (or teams) that made those apps worked together on making a single amazing app instead?

But no, FOSS is all horizontal growth. New apps that do similar things, new distros, new installers, etc. with not too much vertical growth.

koenn
November 7th, 2009, 01:15 PM
self-interest, or selfishness as you call it, is the driving force behind (biological) evolution, behind some of the worlds largest and (so far) most successful economies. It works, apparently.

Unique to FOSS is that it's not only selfish, it's also collaborative : they use open standards so programs can easily integrate witch each other, they have open source so developers can cooperate easily, they can even use their competitors' work to further their own.

So everyone is selfish, and everyone benefits from it. Pretty clever way of doing things, if you ask me.

Besides that, "Having Fun" is an excellent motivator.

As for ("vertical" vs "horizontal") growth - whatever happened to Soft Landing Linux ?
Never heard of it ? That's because "the market", the user base, will decide which projects grow and become successful, and which ones will quietly disappear.

mivo
November 7th, 2009, 01:29 PM
But no, FOSS is all horizontal growth. New apps that do similar things, new distros, new installers, etc. with not too much vertical growth.

This is true, and I agree that it is one of the factors that hinder Linux's growth, if growth means "bigger desktop market share". Many, perhaps most, desktop users today consider a computer a mere tool, like a phone, a toaster or a TV set. They do not want a lot of choice or freedom, they just want to get the task done.

The booklet that comes with Windows 7 talks about how things were made easier, with fewer obstacles, more automated configuration, fewer interactions with anything that's under the hat. They deliberately stress how the OS tries to stay out of the way as much as possible. Some here will say that it is a mindset that Microsoft and Apple foster, but I believe it is the other way around: they offer what people want. (I really like Windows 7 on my work machine, by the way, it is wonderfully done and it's very obvious to me why it is so successful).

Now, most Linux users are different. They like choice. They want options. They cherish freedom. If one distro doesn't work for you, or the devs refuse to tackle some issues (like a dist upgrade that has been broken for years now), you just pick another distro. Some things might be a bit different, but it's still Linux and you can use the same software. The same is true for plenty of applications or utilities. If one doesn't satisfy you, you can try a different one. Or even modify it.

Linux does have people who prefer less choice and just want their computer to work. Ubuntu probably has most of them, because it is a very friendly distro that offers you a complete system without requiring much input or Linux-specific knowledge.

But these people are not the ones who develop software, or who really contribute in any significant way (which is fine). They also don't pay for the OS or for the applications. A Windows user, who might contribute just as little, at least pays for the OS (often indirectly) or, sometimes, for software, and this money then goes into the industry, making it attractive to continue making software that is tailored for this audience. This isn't the case in the Linux scene.

People have to pay bills, food, rent. If they do something in their spare time, they do it for fun or recognition. They have every right to be selfish, because they do not get paid. And those who do get paid do what their employer wants.

blueturtl
November 7th, 2009, 01:35 PM
I might get lambasted for this but whatever.

This is what really irks me about the Open Source community. Its quite selfish in that regard, very few people are interested in working towards something, each person just wants to make their own program, or "have fun" as has been mentioned a few times in this thread. If you dont like it, stuff off, write your own, etc.

Most hardcore Linux dudes don't care (or realize) the fact that Linux will never grow beyond what it is unless it changes. (and then they say things like "When Linux has dominant market share...." wtf?)

i.e., Ubuntu used to have Pidgin, now it has (damn, forgot what its called). These 2 apps do virtually the same job, what if the 2 developers (or teams) that made those apps worked together on making a single amazing app instead?

But no, FOSS is all horizontal growth. New apps that do similar things, new distros, new installers, etc. with not too much vertical growth.

I can totally understand what you're saying, and I would agree with you if I thought that Linux needs bigger market share or vertical growth. That is something most posters for the "unified distro/community" seem to want. Bigger market share. I'm not saying there wouldn't be benefits to having a larger share of the market. There most certainly would.

But I also believe in freedom. Freedom to choose what suits you best isn't satisfied when you have "a single amazing app". It will be amazing for some and ok to terrible for others. Unification just leads to minorities being trampled (because obviously the "single amazing app" will have to appeal to the majority). The open source model might appear selfish on the surface, but really it is much better than democracy for example. It satisfies everyone.

That, I believe is much more important than market share in itself, or any benefits a larger segment of the market might bring us.

Mornedhel
November 7th, 2009, 01:45 PM
Most hardcore Linux dudes dont care (or realise) the fact that Linux will never grow beyond what it is unless it changes. (and then they say things like "When Linux has dominant market share...." wtf?)

See what you did there ? "Linux will never grow beyond what it is": you already assumed that growth was determined by market share. My own definition of growth for software is more centered around features. Someone else's might be ease of use.

As a Linux user, I do in fact not care about attracting users at all cost. Do not assume that because I use Ubuntu, I subscribe to the theory that bug #1 is, in fact, of higher priority than others. I use Ubuntu because it works fine for me.

Also, I have never encountered a "hardcore Linux dude" that didn't care about market share and *then* said "When Linux has dominant market share". Usually the people who say that do care about market share (and may or may not be hardcore Linux dudes).

hobo14
November 7th, 2009, 02:14 PM
@Pasdar
Apart from the OP, you've come across as very..."abrasive", to put it nicely.
You aren't likely to persuade people to agree with your points of view without at least a little politeness.

To answer your question bluntly: No, it won't work. The phrase "trying to herd cats" comes to mind.

geogur
November 7th, 2009, 04:07 PM
even hearding cats is possible if you use the right approach ! as far as linux developers they are doing a outstanding job . With out them we would be under the control of profit minded control freeks !!!!!!

saulgoode
November 7th, 2009, 10:02 PM
Originally Posted by blueturtl
If I were a developer and someone came to tell me that they think a piece of software I made specifically for myself doesn't work the way they'd like it to work you'd probably be right about my reaction. Boohoo. But this being "the Linux scene" if there are others who want the software changed the project can easily be forked.


I might get lambasted for this but whatever.

This is what really irks me about the Open Source community. Its quite selfish in that regard, very few people are interested in working towards something, each person just wants to make their own program, or "have fun" as has been mentioned a few times in this thread. If you dont like it, stuff off, write your own, etc.
People in the Free Software community ARE interested in working towards something; that is the reason why they contribute. Maybe YOU don't like what they are working towards, or that they are not working towards the same "something" that YOU want -- but that only means that YOU are the one with selfish expectations ("developers should be working on what I want, not what they want").