11 Feb
2013
11 Feb
'13
5:13 p.m.
Paul Davis:
if you're not willing to participate in the managed handling of a
bug, i'm
not interested in your bug.
of course, someone else might be, in which case you get lucky.
Please don't write things like that in a way that could make people
believe it's a general attitude.
If there is a bug in my software, I want people to report it any way
they like
(as long as it's electronically). Bug reports should always be
appreciated.
I think/hope this goes for many other developers.
On Sunday 10
February 2013 11:08:02 Ralf Mardorf wrote:
> When subscribed to the user and/or devel mailing list, I report
the bug
> to that list and most of the times I'm
not willing to subscribe
to a
bugtracker,
11 Feb
11 Feb
5:25 p.m.
On Mon, 2013-02-11 at 17:13 +0100, Kjetil Matheussen wrote:
Please don't write things like that in a way that
could make people
believe it's a general attitude.
If there is a bug in my software, I want people to report it any way
they like (as long as it's electronically). Bug reports should always
be appreciated.
I think/hope this goes for many other developers.
For Linux I experienced that bugs reported by email and feature requests
by email very often were done within a short time, by several coders.
5:46 p.m.
On Mon, 2013-02-11 at 17:25 +0100, Ralf Mardorf wrote:
On Mon, 2013-02-11 at 17:13 +0100, Kjetil Matheussen
wrote:
^^^^^
Linux audio
Please don't write things like that in a way
that could make people
believe it's a general attitude.
If there is a bug in my software, I want people to report it any way
they like (as long as it's electronically). Bug reports should always
be appreciated.
I think/hope this goes for many other developers.
For Linux I experienced that bugs reported by email and feature requests by email very often were done within a short time, by
several coders.
5:49 p.m.
On Mon, Feb 11, 2013 at 11:13 AM, Kjetil Matheussen <
k.s.matheussen(a)notam02.no> wrote:
Paul Davis:
if you're not willing to participate in the managed handling of a bug, i'm
not interested in your bug.
of course, someone else might be, in which case you get lucky.
Please don't write things like that in a way that could make people
believe it's a general attitude.
OK, let me very clear. That is MY attitude. For ARDOUR.
There are almost 1500 bugs still open at tracker.ardour.org. Anyone who
expects me (or anyone) to manage that number of bugs via email is someone
who has expectations of me that will not be met. Moreover, many bug fixes
involve totally asynchronous communication the bug reporter and/or other
people interested in the bug. I see no sensible way to work in this way
without the coordination that a bug tracker brings.
On Sunday
10 February 2013 11:08:02 Ralf Mardorf wrote:
> When subscribed to the user and/or devel mailing list, I report the
bug
to that
list and most of the times I'm not willing to subscribe to a
bugtracker,
6:07 p.m.
On 11.02.2013 17:49, Paul Davis wrote:
On Mon, Feb 11, 2013 at 11:13 AM, Kjetil Matheussen
wrote:
Thank you. :-)
Paul Davis:
Please don't write things like that in a way that could make people
believe it's a general attitude.
OK, let me very clear. That is MY attitude. For ARDOUR.
On Sunday
10 February 2013 11:08:02 Ralf Mardorf wrote:
> When subscribed to the user and/or devel mailing list, I
report the bug
> to that list and most of the times I'm
not willing to
subscribe to a
> bugtracker,
if you're not willing to participate in the managed handling of a
bug, i'm
not interested in your bug.
of course, someone else might be, in which case you get lucky. There are almost 1500 bugs still open at
tracker.ardour.org [2].
Anyone who expects me (or anyone) to manage that number of bugs via
email is someone who has expectations of me that will not be met.
Moreover, many bug fixes involve totally asynchronous communication
the bug reporter and/or other people interested in the bug. I see no
sensible way to work in this way without the coordination that a bug
tracker brings.
That's understandable.
6:43 p.m.
On Mon, Feb 11, 2013 at 8:49 PM, Paul Davis wrote:
There are almost 1500 bugs still open at
tracker.ardour.org.
That sounds very nearly unmanageable.
I've just taken a brief look.
222 new, confirmed and acknowledged for SVN/2.0-ongoing and 2.8.x
except latest 2.8.14. IMO, it would make sense to look through things
after releasing and "upgrade" bugs to newer software version or drop
them as needed.
You also have 182 waiting for feedback, some of them -- since 2009.
These could be shaved by introducing a policy of dropping odd reports
that haven't been followed up longer that a year.
Then there are obsolete reports that are still marked as "new", e.g.
http://tracker.ardour.org/view.php?id=2344
And some bugs have "new" status while their duplicates are closed as
resolved, e.g. http://tracker.ardour.org/view.php?id=1760
IMO, it would makes sense to stuff development and spend a boring week
ransacking the tracker. I have a gut feeling that you could end up
with 1K of actually useful reports. Or less.
Alexandre Prokoudine
http://libregraphicsworld.org
12 Feb
12 Feb
12:53 a.m.
New subject: [Ardour-Dev] So what do you think sucks about Linux audio ?
On Mon, Feb 11, 2013 at 12:43 PM, Alexandre Prokoudine <
alexandre.prokoudine(a)gmail.com> wrote:
That sounds very nearly unmanageable.
I've just taken a brief look.
222 new, confirmed and acknowledged for SVN/2.0-ongoing and 2.8.x
except latest 2.8.14. IMO, it would make sense to look through things
after releasing and "upgrade" bugs to newer software version or drop
them as needed.
There are some of us that have helped out in exactly this respect with
Ardour over the years. Do not underestimate how much time even doing this,
updating bugs, etc. takes to be honest.
Seablade
11 Feb
11 Feb
10:18 p.m.
On Mon, Feb 11, 2013 at 05:13:02PM +0100, Kjetil Matheussen wrote:
Paul Davis:
>if you're not willing to participate in the managed handling of a
>bug, i'm not interested in your bug.
'Your bug' is a bit missing the point. It's the author's bug, not
the user's.
If there is a bug in my software, I want people to
report it any way
they like (as long as it's electronically). Bug reports should always
be appreciated. I think/hope this goes for many other developers.
It's certainly like that here.
A bugtracker is a tool used by a developer or a team much in
the same way as an electronic or paper agenda is. If I want
a meeting with some person I'll call him/her or maybe his/her
secretary. I don't write myself into his/her agenda or todo-
list.
If some project gets a bug report via email or a mailing list
it's a simple matter for a developer to enter it into the bug
tracker of his/her choice. Or write it down on a post-it (TM).
It probably takes less time than reading, evaluating and
assigning it, which has to be done anyway.
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
10:45 p.m.
On Mon, Feb 11, 2013 at 4:18 PM, Fons Adriaensen <fons(a)linuxaudio.org>wrote;wrote:
this is bullshit. i make myself available to "some people" via IRC and
email and web forums for up to 15 hours a day, typically 6 or sometimes 7
days a week. the fact that i am not willing to add "oh sure, i'll just
paste everything you've just told me into the bug tracker" is a reflection
of the fact that both i and the bug reporter recognize that as a silly
waste of whatever time i have left.
On Mon, Feb 11, 2013 at 05:13:02PM +0100, Kjetil
Matheussen wrote:
it may be the author's bug, but the author may have zero reason to care
about it right now, or this month, or this year. it may, nevertheless, be
extremely useful to have it filed away so that when the author(s) do care
for one reason or another, attention will be drawn to it. this frequently
happens with ardour, just for the record. lots of bugs get filed in an area
that i'm not currently working on. weeks/months and occasionally years
later, i can go back and find them all when i do start focusing on that
again. in the meantime, calling it "the author's bug" is just a
convenience, even if it is also the truth. it doesn't actually help the
right thing happen at all.
Paul Davis:
>if you're not willing to participate in the managed handling of a
>bug, i'm not interested in your bug.
'Your bug' is a bit missing the point. It's the author's bug, not
the user's.
If there is a bug in my software, I want people
to report it any way
they like (as long as it's electronically). Bug reports should always
be appreciated. I think/hope this goes for many other developers.
It's certainly like that here.
A bugtracker is a tool used by a developer or a team much in
the same way as an electronic or paper agenda is. If I want
a meeting with some person I'll call him/her or maybe his/her
secretary. I don't write myself into his/her agenda or todo-
list.
If some project gets a bug report via email or a mailing list
it's a simple matter for a developer to enter it into the bug
tracker of his/her choice. Or write it down on a post-it (TM).
It probably takes less time than reading, evaluating and
assigning it, which has to be done anyway.
clearly, you're not working on any projects with the kind of bug flow that
i do, let alone ones i've worked on in the past.
at what point do you believe that this "the developer should take care of
it" scales itself out of existence? 5 bugs a day? 10 bugs a day? and what
happens when there are >1 developers? who "writes it down on a post-it(TM)"
or enters it in the tracker? whose bug is it then anyway?
and how does this cover feature requests, which to the user are often
indistinguishable from bug reports? if there are 10 people this week who
have an idea for how ardour should do something, does that i get to spend
10xN minutes entering these ideas in the only place that i can sensibly
review in the future?
11:15 p.m.
On Mon, Feb 11, 2013 at 04:45:09PM -0500, Paul Davis wrote:
this is bullshit. i make myself available to
"some people" via IRC and
email and web forums for up to 15 hours a day, typically 6 or sometimes 7
days a week.
Fine, but what purpose does this serve if no trace of all that
activity is kept on your side ?
clearly, you're not working on any projects with
the kind of bug flow that
i do, let alone ones i've worked on in the past.
I do get less bug reports, that's clear.
at what point do you believe that this "the
developer should take care of
it" scales itself out of existence? 5 bugs a day? 10 bugs a day? and what
happens when there are >1 developers? who "writes it down on a post-it(TM)"
or enters it in the tracker? whose bug is it then anyway?
Something doesn't add up here. Entering a bug in a tracker, if you use
one, is the first of *many* actions in the process of fixing it. It's
not the most time consuming one. If this tiny part of the process is
already overwhelming, what about the rest ?
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
12 Feb
12 Feb
10:10 a.m.
On 02/11/2013 11:15 PM, Fons Adriaensen wrote:
Something doesn't add up here. Entering a bug in a
tracker, if you use
one, is the first of*many* actions in the process of fixing it. It's
not the most time consuming one. If this tiny part of the process is
already overwhelming, what about the rest ?
Writing a good bug report can easily take 10 minutes or more. Plus every
developer should know the cost of a context switch.
Chances of a "report" happening via IRC or email already being well
structured and complete are lower, as the tracker interface will include
certain hints and brings a certain context/perspective to mind.
A developer of a long running, highly complex project will usually have
a virtually unlimited amount of work to do. Filing tickets that users
could file is a very inefficient use of valuable time.
If a user want a bug fixed, it's in their interest to file it properly
and to not waste a developers time. There are of course cases, where
talking to a developer first is better, say when it's not clear if it's
a bug or a feature, or when hunting a heisenbug. I don't know any
developer who would be more approachable, helpful and patient than Paul
Davis in such cases (as mad as he made me in design matters, sometimes ;)).
--
Thorsten Wilms
thorwil's design for free software:
http://thorwil.wordpress.com/
4226
days inactive
4227
days old
linux-audio-user@lists.linuxaudio.org
10 comments
7 participants
participants (7)
-
Alexandre Prokoudine -
Fons Adriaensen -
Kjetil Matheussen -
Paul Davis -
Ralf Mardorf -
Thomas Vecchione -
Thorsten Wilms