Hi,
ext Jonathan Markevich wrote:
>> Talking again about open source & personal point of view, NOT about
>> commercial software or products (such as Nokia device...).
>
> If OSS doesn't view itself as professional as commercial stuff, then it's
> guaranteed to fail in the long run. Look at the success stories of OSS;
> Firefox, Apache, Linux, OOo, and so on.
They have commercial interests behind them and full time developers
(Google sponsors Firefox, Sun OOo, most of Linux developers are employed
by companies etc). And you should see what the Linux kernel developers
require from users reporting bugs[1]. :-)
My comments were more about hobby Open Source projects in general.
I think most of the maemo 3rd party applications are such.
[1] There was recently a story about bug which required user to
git-bisect & rebuild kernel to find the change that had broken
the kernel for his setup. That took 5 hours. After that the
developers could find & fix the issue.
> They aren't based around the bitterness of a developer in a basement.
>
> Submitting a bugzilla report shouldn't take more than 10 minutes
>> 10 minutes is a long time for something that may not help. Multiply that
>> by
>> say 8-10 open source applications you are interested in, and you see why
>> it's not worth the effort.
>
>
> 10*10 minutes is still <2 hours. And you've then participated in
>> helping 10 different projects to (potentially) improve!
>
> Who has 2 hours to waste on something as fun as bugzilla? Even this email
> exchange is way more rewarding. I've put in a handful for certain projects,
> then watch them sit around doing nothing, only to get an email that says
> "WONTFIX". That's user hostile.
Yes, I know that it can be frustrating (I've had my share of that too).
But it's not user hostile, just pragmatic. Developers are limited and
have limited time so they need to prioritize it. Filing better bug
reports (i.e. things that developer can immediately reproduce) is one
way to prioritize the issues.
Or if it's about a feature, the developers might have a different
vision of their software.
>> a difference). Have you ever got the "Report this error to Microsoft"
>>> dialog box? That's what I'm talking about.
>>>
>> It's fairly similar to Ubuntu apport I think? Ubuntu's lacking
>> MS problem & solution database. Novell/SUSE has something similar
>> to that though I think.
>>
>> However, those are distributions, not individual upstream projects
>> like's discussed here.
>>
>> (Hm... Maybe maemo, as a basis for distribution could offer something
>> here.)
>
> NOW you're talking! I'm trying to promote the "many eyeballs" thing
> and not the "genius with a microscope" thing.
>
> If you go to effort of reporting the bug and actually reply questions
>>> on how the developers might be able to reproduce it, so that they
>>> can start investigating how to fix it, that shows that you actually
>>> care about the issue and that it's real.
>>>
>>>
>> So you say we have thousands of users per developer. Great! The user
>> should be able to email or whatever saying: xyz is crashing when I view
>> the
>> records. It means almost nothing to the dev, but if he gets 999 others
>> that
>> say exactly the same thing, it means the view records routine is horribly
>> broken.
>>
>
> Usually it means that users are using too old version of the SW
>> (e.g. because distro hasn't upgraded to latest version) and therefore
>> wasting developers time.
>>
>> On the other hand, if he gets 5 others, and notices they're all
>>> from non-latin alphabet countries, the dev is in the best position to put
>>> those pieces together. The DEV can make a bugzilla record. Maybe in
>>>
>>> Instead what usually happens is the instant any individual report comes in
>>> the dev starts shouting about how the user should use bugzilla (yet
>>> another
>>> application, another big learning curve, and yet another registration on
>>> the
>>> net), expecting every user to be a developer or professional-grade QA
>>> tester
>>> too.
>>>
>>> That makes the triage for the user easy; 3) dump the program.
>>>
>> Regardless of how important you may feel yourself :-), most Open Source
>>
>
> This is not about my own perceived importance, but the developer's.
> Is he really more valuable than thousands or hundreds of users?
It's not about developer importance, but what he has time to do.
1000*users surely can get more done than a single developer in
regards to bug handling.
If the issue is not important enough to user so that he reports
a bug, that gives a pretty clear message to the developer about
the importance of the issue.
>> developers really aren't doing what they do to please you or get more
>> users, but to solve the issues they have themselves or otherwise find
>> interesting/fun to solve. Having more users is nice only if they help
>> in that, otherwise they are just a drag. As it's possible that users at
>> some later point become contributors, and it's nice to hear that your
>> efforts are appreciated by others, Open Source developers are usually
>> nice for the users (if they behave reasonably).
>>
>
> Again, this is a very unprofessional way to develop.
Well, I was talking about hobby Open Source projects.
> Self-defeating.
> Might as well keep it closed source at this point.
Why? The program might still be useful for somebody.
> And makes the platform unattractive to users and investors
> (I also mean investors of mindshare and
> time as well as money)
Yes, distro/platform proving some easy way to provide error information
to developers (without swamping them) would indeed be a good thing...
(in addition to providing common bug tracking system so that users
need to register there only once)
> However, Open Source is about a community of people who want to improve
>> things *together*. If you just want to profit from their work without
>> contributing yourself in someway (even to some other project), well,
>> they're not going to miss you.
>>
> Do you say the same thing to developers? "If you don't want to respond to
> users' input, take a hike, we're not going to miss you".
Sorry I lost you?
If (hobby project) developers don't have time to deal with user input,
some of the users can take that role. In many of the successful Open
Source projects bug triaging is done by non-developers and developer
can then take look at it after it's been reproduced, especially if
triage person tells that there have been lots of similar kind of
issues.
> The best experiences I've had with developers on maemo so far is on the ITT
> forums. It's simple to report, you can see results, and takes only
> seconds. One registration for many many applications, and it's something
> you might want to do in the long run anyway.
That sounds an excellent way for the users to organize&collect the
information required for a good (reproducible) bug report. It's
then enough that one of them reports the bug.
- Eero
_______________________________________________
maemo-users mailing list
maemo-users@xxxxxxxxx
https://lists.maemo.org/mailman/listinfo/maemo-users
