Re: [Lame-dev] Released lame 3.90.1

[Alexander L. <Alexander@Leidinger.net>]

On 26 Dez, Darin Morrison wrote:

>> Is there someone who gives you money if you "get things done"? I assume
>> not, so why not wait a day or two more, but perhaps get some inspiration
>> from some comments and improve the existing work you already did? Or
>> better, first talk here (on lame-dev, the "forum" where the developers
>> (should) discuss about the development of LAME), present your idea, get
>> perhaps inspired by others and improve the design, and then implement
>> this design (-> less work to do and the code may be better).
> 
> Why wait a day or two to fix a bug when it can be fixed now?  Waiting a day

You fixed it, and commited it. But making releases isn't your part in
the project (and it isn't my part). If you want to make releases talk
with Mark about the open position as a release engineer.

> or two means more people encode with a buggy version and produce non-optimal
> files.  Meanwhile, at least one capable person knows about and can fix the
> bug, but you are saying they should wait for no good reason?  Hrmm.  Sorry

You already fixed the bug. I just complain about 3.90.2. See below.

> but this kind of attitude is leads to slow development times, missed release
> dates, buggy versions being in use longer than they should be, and other
> things like that which the LAME project has already had enough of.
> 
> As for discussing things on lame-dev, I don't do that much for a few
> reasons.  One, this list isn't very interesting.  Sorry to say, but the
> traffic is low and there is not much discussion about quality, which is what
> I am primarily interested in.  When there is discussion, it isn't

I hadn't encoding quality in mind. You talked in the past about an
userfriendly interface, and you are going to work on it with
--alt-preset. That's fine, but we (you, J.D. and me) already talked
about it. I think I was able to give you the impression that I'm
interested in a userfriendly interface too, wasn't I? So why not just
tell us "Hey, I want to do ... (e.g. the "fast" attribute to the preset)
and I plan to do it this way. What do you think?"?

> synchronized with what is already known and being discussed on web based
> forums.  Most comments from users related to quality mentioned on this list
> do not get addressed.  At the very least, none of mine ever really were by

Perhaps because most of the people here don't have much time or think
that there are people out there, which can comment on those questions
more competently (e.g. I haven't done as much listening tests as you
did, so I think I better let you comment on such questions than
answering them on my own).

> anyone, and that's the main reason I actually started coding myself.  I came

That's the main reason for most contributions in an open source project.
If you want it to be done right, you have to do it yourself. Sometimes
there's another one with the same interest, and you can load balance the
work.

> from the web based forums, so thats where I continue to work.  In addition,
> I dislike mailing lists.  Web based forums manage data much better IMO and
> it is immensly easier to keep track of multiple in depth discussions.  I
> would suggest that if the LAME developers actually want to push LAME forward
> in quality and usability in the future, that they pay more attention to
> these forums, because that's where the action is at whether anyone likes it
> or not.

You are a LAME developer and you pay attention to it. And you improve
lame. We can't make every change of your work the default, we are
limited by some contraints (e.g. "-V x"), but some of the work can be
defaulted (and I already suggested to you to talk with Mark about some
of the changes).

>> Most people perhaps check freshmeat or slashdot. I don't know who feels
>> responsible to tell them about the new release, but maybe he (or a
>> slashdot editor) reads it and waits until 3.91.
> 
> The majority of the people using the --alt-presets are going to check
> Hydrogenaudio, because that is all of the development and testing for this
> has gone on in the past.  In addition, many of the people using LAME who are

And those people aren't the audiency my proposed newsitem is aimed at.
The people at hydrogenaudio mostly know about it, I'm talking about the
"I want a program which generates MP3's." people which don't look at HQ
encoding sites.

> interested in high quality in general, know to check this place or at least
> r3mix.net.  I think that if you are going to put up a notice about this
> issue, and do release a 3.90.2 (still unlikely I guess?), you should at
> least add a notice that I have released a compile that fixed this issue (a
> few days ago by now).

When I release 3.91 we didn't had a 3.90.2, so I didn't have to talk
about it. People which downloaded 3.90.2 from hydrogenaudio know it is a
inofficial bugfix (at least you said it's marked as an inofficial bugfix
release). So they should know they didn't have to download 3.91.

>> Does it also state, that this isn't a official released version and
>> bugreports have to be made to hydrogenaudio.org?
> 
> It states in the file_id.diz that it is an unofficial version, and the
> header for the .exe is changed.  However, I find it a little bit amusing
> that people would be hung up on a trivial issue as the amount of "bug
> reports" increasing.  The minute the LAME dev starts to get increased bug

It's just one issue. Making indistinguishable binaries - one with a bug,
one with a bugfix - isn't good habbit because of more than only this
issue. An user can't know which version he has only by looking at it.

Bye,
Alexander.

-- 
   "One world, one web, one program"  -- Microsoft promotional ad
         "Ein Volk, ein Reich, ein Fuehrer"  -- Adolf Hitler

http://www.Leidinger.net                       Alexander @ Leidinger.net
  GPG fingerprint = C518 BC70 E67F 143F BE91  3365 79E2 9C60 B006 3FE7