Izumi Tsutsui wrote:
> garrett_damore@tadpole.com wrote:
>
>   
>> but everyone else said "just implement it and get some real-world
>> results".  So I'm doing that.
>>     
>
> Well, in that case, you don't have to change whole attachment
> before waiting comments. Changes to com.c, comvar.h, and one
> example of attachments are enough to review real code.
>
>   
>> If you (or anyone else) comes up with a much better way to do this, then
>> you  (or whoever else) is more than free to either change what I've done
>> or ask core@ to reject my proposal with an alternate proposal.  I don't
>> really care either way.
>>     
>
> >From my experience on NetBSD, rushing to commit without enough
> period often caused unpleasure flame. That is my concern.
> Actuall I won't object your COM_INIT_REGS so strongly if
> some other people say it's acceptable. Fortunately I have
> spare time to see/respond your mail in these two days,
> but I don't think all people have time in such short period.
>   

Getting it in a *branch* ready for use is meaningful.  The loudest
detractors (other than yourself), have been folks who've objected on the
basis of performance.  So having the ability to actually test it makes
sense.

By the way, this was discussed and I received other feedback than just
yours.  Including discussion from Charles, Christos, Jason, etc.

So this isn't happening entirely in a vacuum.

> How do you think the way yamt@ did on his many changes recently?
> It's quite reasonable and causes few unfortunate results, I think.
>   

Honestly, I've not followed yamt's changes, but I've sort of assumed
that folks who are more in tune with this have been following it.

    -- Garrett
> ---
> Izumi Tsutsui
>   


-- 
Garrett D'Amore, Principal Software Engineer
Tadpole Computer / Computing Technologies Division,
General Dynamics C4 Systems
http://www.tadpolecomputer.com/
Phone: 951 325-2134  Fax: 951 325-2191