[haiku-commits] Re: haiku: hrev52055 - src/kits/media headers/os/media
- From: Dario Casalinuovo <b.vitruvio@xxxxxxxxx>
- To: haiku-commits@xxxxxxxxxxxxx
- Date: Fri, 27 Jul 2018 09:39:55 +0200
On Thu, Jul 26, 2018 at 7:06 PM Stephan Aßmus <superstippi@xxxxxx> wrote:
Hi,
Am 26.07.2018 um 17:52 schrieb Dario Casalinuovo:
However, I got enough for today, and sincerely my final answer is
"whatever". Do what you want, don't expect me to work on this or to
comment anymore.
OK. My initial mail in this thread wasn't directed at you, but I wanted
to draw attention to it from the person having done the commit. I have
no time to work on it either.
Did you consider, it is two weeks I try to draw attention to this commit?
I spent a great amount of time, more than you can imagine, digging in
the media_kit and fixed various long standing issues.
Just out of curiosity, are you aware that I worked on MediaKit before
you? I could say the exact same thing about myself, but why is it even
relevant here? I am thinking, when you suddenly tell people how much you
have worked on something, it isn't necessarily an argument in favor of
whatever your opinion is in some random discussion. It just looks like
"I don't want to waste time actually convincing someone with real
arguments." Sometimes, it is useful to get done quicker, but here? What
are you even trying to say?
I will tell the thing here: I think we should take the general design of
the media_kit but rewrite it from scratch.
You each time come in and say that media_kit is perfect, BMediaEventLooper
is the best and so on.
Now there are only two possibilities:
a) I am completely wrong.
b) You are completely wrong.
I sincerely thing you are not really thinking of the issues I pointed out
over the years, and I think you do so because you consider your opinion
better than mine. That is what I perceive from discussing with you. In past
you pointed out some interesting things in my discussion, for example I
remember the discussion about callbacks naming.
I am not perfect,
but I have a very precise idea of the major media_kit flaws, if you
don't want to listen me: I will not care.
Yes sure, I have an opinion on MediaKit, too, from working on it so
much. Again, why is this relevant?
It is relevant because, some flaws are just handbook ones. I can't believe
you don't understand those.
For example, do you remember your discussion with JBQ telling you that
ports have the bad problem of locking under certain cirmustances?
He is completely right, and this discussion was 20 years ago, if you didn't
listen to a BeOS developer, what's the chance you are going to listen to a
relatively-newbie like me?
Again, I don't pretend to be right, but I invite another time to question
yourself if you are taking my opinions, and in general the opinion of
everyone that goes against your convinctions.
Now, I joined this project as dev in this situation:
* Media kit is fantastic has clean API
* Media kit has far superior latency capabilities
* Media kit is super easy to use compared to other kits
At the begin I had lots of attacks about that, also from you. Now, people
seems to a bit have changed. I begin to see people mentioning that we have
to rewrite the API, and so on.
Even this discussion, about media_format is obviously related to the fact
we have to introduce BMediaFormat based on a BMessage-like class using
key,value primitives. Don't you want to consider my concerns? Let it be.
But I listen at you always as being a respectable programmer, but don't
come here and tell me that continuing to use a C function like memset in
place of C++ constructs (whatever construct, not necessarily my solution
which Ingo indeed pointed out as sub-efficient) is any good for the code
quality of the project.
Have fun.
--
Saluti,
Dario
Other related posts:
