[Ardour-Users] [64studio-users] Just a suggestion about how to handle bug reports

Wed Dec 17 09:43:44 PST 2008

>> Testing of RCs often is unwanted. I tried to make a howto and needed 
>> help and that was unwanted. And people don't report bugs to the right
>>  places, 1. because they don't know the right places and 2. because
>> they know, that they will get bashed.
>
> If I don't do my homework, then I expect to find a less than comforting
> response.  I do not get coddled anywhere else in life, I do not see why
> this should be any different.

And here we have the problem, I think. Software users actually do get coddled in almost every commercial app support environment. They aren't really expected to do all this research before they make a call to Ableton's support line about the playback marker not keeping up, for example.  There's a perception in the vast majority of users (I was guilty of this myself at one point) that reporting bugs in Open Source projects is as simple as writing the list and telling them what went wrong. But, after reading this email thread, we can see that's just not the case in at least some circumstances.

I've found the best way to get support in open source projects is to research and investigate the bug to the point where you're almost looking at the code and screaming "Look, here's the problem!"  The less investigative work the developers have to do, the more quickly your bug will get resolved.  Of course, the caveat to that is, you really have to know your shit, and it certainly doesn't hurt to know how to read C code.  I'm not implying that the developers are lazy, just that this is their volunteer work, their free time, and no developer (that I know of ) enjoys tracking down bugs.

I applaud the efforts of some projects that work to make bug reporting easier (like having a crash dialog with a button that says "Send bug report"), but I really think this perception needs to be balanced against the oft-advertised "Help support open source; report bugs!" mantra.

Michael Bechard

----- Original Message ----
From: Gustin Johnson <gustin at echostar.ca>
To: Ralf Mardorf <ralf.mardorf at alice-dsl.net>
Cc: lmms-devel at lists.sf.net; Ardor Users Mailing List <ardour-users at lists.ardour.org>; 64studio-users mailing list <64studio-users at 64studio.com>
Sent: Tuesday, December 16, 2008 8:43:36 PM
Subject: Re: [Ardour-Users] [64studio-users] Just a suggestion about how to handle bug reports

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ralf Mardorf wrote:
> Hi Gustin :)
> 
> it's easy to get named a troll. What should a user do, that has
> troubles with JACK?
> 
It depends, this is where research comes in handy.

> I like the 64 Studio community, unfortunately the version for the 64 
> Studio 2.1 stable version is 0.103.
> 
> It's silly if a user reports a bug about that version, while the JACK
>  developers needs statements about 0.116.

So, if you look in to this, the Debian way is to contact package
maintainer, who then contacts upstream as necessary.
> 
> There are distros like Suse, that supports 0.116 by community
> packages, e.g. for Suse it's the Packman repository.
> 
> I had contact to the packager by the Linux Club.
> 
> People reported that bug e.g. to the Ardour forums, a long time
> before I had this bug. There they say, it's a JACK bug, but this
> isn't true, it was a packman bug. http://ardour.org/node/2271

You did ask the question in the wrong place.  The person who gave the
correct answer actually did not do you any favours.
> 
> I don't want to talk again about this, because I'm personal involved
> in this, maybe someone of those who replied off-list can give other 
> examples, I won't quote private mails.
> 
Doesn't matter to me.  I see nothing wrong with that thread.

> Testing of RCs often is unwanted. I tried to make a howto and needed 
> help and that was unwanted. And people don't report bugs to the right
>  places, 1. because they don't know the right places and 2. because
> they know, that they will get bashed.

If I don't do my homework, then I expect to find a less than comforting
response.  I do not get coddled anywhere else in life, I do not see why
this should be any different.
> 
> When I reported the JACK package bug for Packman I get sarcasms.  I
> was a troll, who made everything wrong, because I have no knowledge.
> Nobody but me should have this bug too, written by the packer, who
> asserted that he makes music with the packages from Packman. This
> must be a lie, because this package is broken for everyone.
> 
I know I have suspected you to be a troll, in fact I am still not
entirely convinced that you are not.  It may be a language issue, but
you do seem to find a very inflammatory way of communicating in English.
Subject lines like "Linux has no future for multimedia" are going to
get a hostile response.  I am pretty sure you are smart enough to know
that.

> 1. If you like to help and write a howto, but you need some help
> because you e.g. don't know that compilers differ for several Linux,
> you need to add headers for some applications, than you do something
> that is unwanted. 2. If you report that something don't work, you
> need help, than you are a troll, because it's not true.
> 
> I'm not the only one. Why can you find bugs in the web, that are not 
> reported to the right people?

It is easier to post your problems to a blog rather to figure out who
you need to talk to.  I am guilty of this as well.
> 
> I got mails off-list, with statements similar to "I recently got
> raked over the coals by ...".
> 
I can tell you it has happened to me on more than one occasion.  It is
even worse when it happens in real life by someone like Aaron Seigo in
front of your friends.  I may have considered the responses to be harsh,
but in every single case I learned something and came away better for
it.  There is an old saying "if you can't handle the heat, stay out of
the kitchen".

> Which distribution can I use, that is wanted by the developers? How
> do I

Ask them.  Most of the ones I have dealt with don't care.

> have to report bugs? When I say "Hello, there's a bug" I don't want
> the answer, that I'm a noob, nobody has this bug, while the whole
> community has got the same bug.
> 
> I'm not the "Girl, interrupted" when I reported a bug and I'm not a 
> individual case. I get crazed, when people are unconvincable, they
> don't need to excuse, while people like me get banned.
> 
If you have been banned more than once, then I would look at your
procedure.  Just once can be an accident, more than that indicates a
problem with your communication methodologies.

> The only thing users can do, is to report bugs anonymous, in forums 
> somewhere in the web, instead of talking to the right people.

That helps no one.  No, what users should do is participate.  The
information is there, the tools are there, there is really no excuse
other than "I didn't want to put forth the effort so I didn't".  This is
valid, but one should not expect a response.
> 
> Misspelling for a user who reports something should be allowed.

Depends on the community running that particular forum.  I am not going
to dictate terms to people who do work for me for 0$.

> Posting a bug that was posted before, should be allowed.

I disagree with this.  If a bug has been posted before, and you try to
post it again, it usually means you did not do your homework.

> Posting a bug of a front or back end to the back or front end people 
> should be allowed.

Depends on the community running that particular forum.  I tend to obey
the rules of whatever community I am a guest in, virtual or physical.

> To be stupid should be allowed.
> 
> But it isn't.
> 
You really should read this link or find a quality German translation.
http://www.catb.org/~esr/faqs/smart-questions.html
Actually anyone who has emailed you off list should read this link too.

> PS: I have a bad conscience because I wrote this troll like stuff. I

So, you know that you write troll like stuff, and are still surprised by
the hostile response?

> also get off-list mails that this will break the work. But I guess
> it's a topic, if Linux is also for Users. I never developed anything
> for Linux, but I was a developer too. I don't understand the problem
> with

I am not sure what to say here.  Hundreds of thousands of volunteer
developers, thousands of different projects all makes for a complicated
situation.  Seems pretty straight forward to me.

> bug reports. At the moment it looks like I'm not banned in much
> mailing lists, as it was written.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFJSGdYwRXgH3rKGfMRAv6HAJ9prWJ+7BZC2SULij7CEPDyPK9X5ACbBbJ+
KA34P/sdXPsJJwY5EEDTMFA=
=UnRj
-----END PGP SIGNATURE-----
_______________________________________________
Ardour-Users mailing list
ardour-users at lists.ardour.org
http://lists.ardour.org/listinfo.cgi/ardour-users-ardour.org