[Asterisk-Dev] 'tonezone' in chan_zap.c

alex at pilosoft.com alex at pilosoft.com
Tue Mar 1 16:35:43 MST 2005


On Tue, 1 Mar 2005, Brian Capouch wrote:

> Yep.  And then once the grepper pulls up the files, he can use the
> voluminous comments in the code itself then, to penetrate the logic of
> the half-dozen people who have separately hacked the code over time.
"If you could read it, it wouldn't have been called 'code'"

> I am doing a class about Asterisk this semester at the little college
> where I teach, and one of the first things I told my students was,
> "Don't post to the Asterisk lists!!  Ever!!"
Good. The very last thing we need on -dev are people who ask us for help
with their [literally] homework. This list is for *developers to talk to
other developers*. If you are not a developer, you do not qualify. If you 
use asterisk in your enterprise, you do not qualify. If you sell 
asterisk-based solutions and need developer's help, you do not qualify. 

> The truth of the matter is that the guru types on the Asterisk lists are
> far quicker to act like dicks to those who they deem beneath them than
> on any other OS list I subscribe to.  Whenever someone calls them out,
> they immediately cop this, "We're not being dicks; we're just too busy
> to screw with newbies" answer.
> 
> On the -users list I have come to appreciate Steven's helpfulness, even
> though it comes with the price of his occasional snippishness.
> 
> But on -dev, the rule of the day is that questions *about* the code are
> discouraged, even though this product is arguably the
> least-well-documented code I have ever tried to plow through.
Good. We need more people who *write* code, not people who have question
on how to read code. Big difference. 

> In my own case, I've been reading code for weeks, trying to figure out
> just how to use (and extend) the Realtime engine.  The Wiki has some
> documentation, but it is cursory, self-contradictory, and in general of
> very limited use.
So, why aren't you contributing to Wiki and adding necessary documentation
as you figure things out?

> The code?  Well, there's not a shred of documentation anywhere
> internally, and to make the chase interesting, there are data structures
> sprinkled about that define structures of pointers to functions,
> indirect calls to those functions buried uncommented inside other
> functions, and then for good measure a whole function defined as a macro
> (REALTIME_COMMON).  To that add fragments of functionality scattered
> throughout the pbx code, the configuration parser, the various channels,
> and the resources directory, and the thought of someone just reading it
> and coming to a good understanding becomes vanishingly small.
It is free, it works, and you can improve on it. What else can you ask
for?

> That sort of thing is withstandable when you can fire off a mail to the
> developer and ask for some quick direction.  Oftentimes a couple of
> sentences from the person who wrote the code can be enough to jump-start
> what is otherwise a futile exercise in reading someone else's mind.
It's free. You are not paying the developer in question. In fact, high 
chance that you are *competing* with developer in question as far as 
services go. You are very welcome to figure it out on your own.

> But not this bunch.  We are kept at bay by watching as the important
> people flame anyone on the outside who dares to ask for a bit of
> direction.
> 
> I am not so much angry about this as disheartened.  IMO, many people
> could be made into contributing members of the developer
> community--something I hope that everyone agrees is needed--but for
> their lack of enough of a masochistic bent to actually dare to ask for
> some background on this list and then be flamed to a crisp.
People aren't flamed here to a crisp for *asking a question*. People are 
flamed here for *NOT DOING THEIR HOMEWORK PRIOR TO ASKING A QUESTION*.

Original poster *should* have did basic grep *prior to asking a question* 
and then, posted results of grep with "uh, it looks like this code does 
foo, can you confirm this?".

> I'll shoulder on, and I'm sure the other posters here today will too,
> because the project is bigger and more important than any of our
> respective egos.
We don't need more posters, we need developers who write code. If you 
can't figure out your way around code, you probably won't be much of a 
help to us either.

> But dealing with all of this makes me pine for the sort of spirit I
> commonly see on other lists I'm on, where the developers are quick to
> help, and encouraging to those who are at a lesser stage of experience
> with the code than they are.  They know in the end their patience will
> help make the product better.
Show me the code.

-alex




More information about the asterisk-dev mailing list