[vorbis] Ogg bitstream spec question

[Graham Mitchell]

As a detached observer who's seen the last few emails, perhaps I can provide 
some neutral perspective.  Not exactly an email I expected I'd be writing 
when I got up early to get to church, but I'm always happy to be of service 
when I can.  Note that I am not Michael Smith, who responded to your first 
email.

Your first email *was* unclear.  It mentions nothing about tunnelling by that 
name, though now that I know that's what you meant, I can go back and find 
that subject in there somewhere.  What it appears you are doing (though, 
again, now with additional clarification on your part I see this was 
misinterpretation) is saying something to the effect of:
"It seems your spec allows X, though I don't understand why.  This makes it 
harder to write certain pieces of the decoder.  If you would clarify your 
spec in ways I find convenient, it would allow me to take some shortcuts."

This is not what you said in your first email, but this is how it probably 
read to most of the people on this list, me (and apparently Michael) 
included.  On this list we get a lot of people attempting to implement the 
spec who are upset about some of the difficulties inherent in Ogg's flexible 
design, so this is an understandable misunderstanding.

Hence, Michael's statement of:

> >We're not going to change the bitstream format so that your implementation
> > can be slightly more convenient. The format is fixed (though future
> > versions may differ, of course).

This is reasonable, IMO.  Of course, I've been on this list for quite a while 
in vorbis terms, so I've seen a lot of things like this come by.

You apparently, don't have the same perspective on this as I do, and have 
flown off the handle.  As a person who does not, in any way, speak for 
xiph.org, allow me to answer some of your questions.

> Woah!  What happened to open source?  Community effort?  Listening to
> your users?

"Listening to your users" does not extend to "implementing their pet requests 
for them", which is what it sounded like you were asking.  The core 
developers are way busy just coding more basic functionality and don't have 
time to hardly even weigh in on more academic side projects whose value is 
not immediately apparent.  Yes, this is a bad state of affairs, I know.

> You sound like you personally own the Ogg project.  Do you
> speak for everyone at Xiph.Org?

No, that would be Monty.  :)  But Michael is a repected developer who's 
contributed a lot, and probably most of the other developers would misread 
your email the same way.  In particular, Michael devotes more of his week to 
dealing with mailing list issues than any of the other developers.

> Such a rude rebuff of someone that is
> making a feature suggestion reminds me of the kind of service I get from
> the help desks of closed source products - not what I have become used
> to from the (large number of) other open source projects with which I
> have been involved.  If you want the Ogg project to live beyond your own
> personal input, you might want to reconsider that point of view.

Let me just say, "You catch more flies with honey than with vinegar."  That's 
a saying in the States.  Comparing xiph with a closed-source product 1) isn't 
going to help your argument in the slightest, 2) hurts your argument, and 3) 
shows that you don't really understand how to conduct a mature, reasonable 
argument on the Internet.

> I am more than 70% through the implementation of the Ogg bitstream spec...
        [snip]
> FYI - I am not suggesting modifying any existing spec.  Nor am I
> suggesting anything specific to any implementation.  I am /asking/
> whether or not it is considered useful to tunnel Ogg within Ogg and
> /suggesting/ that an additional spec that defines a standard mechanism
> for tunnelling Ogg within Ogg (similar to the way in which the Vorbis
> spec specifies Vorbis-in-Ogg) would be useful if that were the case.

If these had been in the first email you sent, we wouldn't be having this 
discussion right now because there would have been no misunderstanding.

As someone with absolutely no authority, may I congratulate you on your 
efforts to implement the ogg bitstream spec.  Additional "clean" 
implementations help the xiph.org guys find and clear up any mistakes (not 
many of those left now) or sections which are unclear in the spec documents.

Also, please feel free to define your own spec for "tunnelling Ogg within 
Ogg".  When you're done, we'd love to hear about it.  I'm sure once it's 
posted here it will generate some useful discussion.  I can predict, however, 
that none of the core developers will offer much in the way of suggestions 
*before* you start.  Add much as they'd like to be able to, they simply don't 
have the time.

Of course, some targeted questions to the mailing list (probably vorbis-dev) 
along the way will probably be answered.

Hope this helps clear up the misunderstanding and doesn't cause you to pack up 
your coding skills and leave.  The vorbis cause needs as many talented coders 
as is can get, and just because you have less-than-optimal mailing list 
communication skills doesn't mean they don't want your input.

Now I'm off to get ready for church.  It's before 7a.m. where I am.  By the 
way, I spent over half an hour composing this email.  Consider crafting your 
mails a little more carefully, too.


-- 
Graham Mitchell - computer science teacher, Leander High School
Axioms of RISC design:
	1. Simplicity favors regularity.
	2. Smaller is faster.
	3. Good design demands compromise.
	4. Make the common case fast.

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.