Rockbox.org home
release
dev builds
extras
themes manual
wiki
device status forums
mailing lists
IRC bugs
patches
dev guide



Rockbox mail archive

Subject: Re: Random numbers et al
From: Blue Chip (cs_bluechip_at_webtribe.net)
Date: 2003-04-08


> > There are zero docs on how the firmware works or how it "hangs together"
> > (its basic structure) and code comments are almost unheard of. So I am
> > struggling to work out what is going on. This is *NOT* intended as a
> > criticism, it is merely an honest observation.
>
>That is a correct observation. So maybe you could help us out to document
>things as you proceed and find out?

Believe me, I thought about it.
Personally, I /*hate*/ 'C' commenting. This is THE reason that I did not
offer to comment all the code up.
I realise that I have just left myself wiide-oopen for being told I am
"different" (once again), but I will not ask you to change either.

>This fact has, believe it or not, not been so an important factor for the ~50
>persons that have provided non-trivial source code contributions so far.

Clearly I'm just stupid then <shrugs>

>I think in general we appreciate help with the areas we believe we have
>problems with or where we still lack features (reading the feature requests,
>they are plenty). We don't have a performance problem (AFAIK) so rewriting
>stuff to assembler is _very_ low prio, and in most cases simply counter
>productive.

Quite frankly I am not buggering about with code I don't understand - it
causes far more aggro than it's worth.

>Why are you different or better than everyone else?

Oh Yeah, very nice. Is that straight out of a Dale Karnegie book?

>I guess that was a suggestion (haven't checked the log), and you're free to
>pick a task yourself. I don't understand how writing stuff we don't need are
>the only things you seem to be able to do.

I THOUGHT you needed the ROM space! I THOUGHT I was being helpful!

> > I have *NOT* "complained" about your rand(), what I said was nothing more
> > than MY opinon based on MY experience and observations of the code - I was
> > offering you an alternative which I honestly thought would benefit the
> > Rockbox effort - but again, I have only succeeded in offending :(
>
>... and I expressed that if replacing the current one with yours is a
>significant improvement, then I'm all for it.

No you did not! You said my improvement was "stuff we don't need" and
refused to tell me HOW the MT_rand() improved the play-list shuffle.

Perhaps I could run the playlist shuffle on a list of tracks called 1..100
and do it, say, ten times with each algorithm. Then I could post the
results and you could say CLEARLY and DECISIVELY which was produced with
which algorithm?? Would that be proof enough?

Maybe you could direct me to the "user complaints" about the old rand()
system - then I can follow the conversation thread and work out WHY this MT
solution offers any improvement.

>Much of the development of Rockbox is making patches (and often separate
>binaries) available, so that people can try out new concepts or new code
>before it goes into the main sources. You can do that too.

Yes I know I can - GPL kind of ensures that really!

>I don't want to be rude, to offend you or to be annoying.

Then perhaps you asked "Why are you different or better than everyone
else?" in error?

>I'm trying to express that I'm sure we all can cooperate fine. We just
>need to be allowed
>to argue about the priorities. We are many with opinions here.

I'm not offering anything else. I have nothing else to offer. So there is
NO "priority."

I'm offering "improvements that I *CAN* make" in exchange for "knowledge as
to how the rest of it works"

It seems stupid to offer to make changes I am unable to make through lack
of understanding! When I understand the threading system and interrupts
(etc) THEN I will bugger about with the code that uses them. Not until!

Bc



Page was last modified "Jan 10 2012" The Rockbox Crew
aaa