Re: [Gnu-arch-users] details, details

[Thomas Lord]

On Wed, 2005-09-07 at 17:38 +0200, Karel Gardas wrote:

> That's interesting information. But let say I also do have some experience 
> with egcs/gcc these days at least from C++ user point of view and I can 
> only add that I was happy like a small child trying egcs1.0.1 since it 
> provided me with the C++ compiler which consumed _only_ about 30-40MB of 
> RAM for compiling our project in comparison with gcc of these days which 
> spent twice this amount. Note: 32MB RAM was quite a lot in 1996/7.
> 
> So thanks to egcs hackers I've been able to complete my school projects 
> with much less pain.
> 
> Also the second note: current GCC release manager is doing good job IMHO 
> provided as a commercial service by his small company to the big IT 
> sponsor. I think this is also needed to note since every coin has two 
> sides...
> 


Without contradicting myself, I second you on both points and will
add a third element of praise.

The current GCC project is arguably the most professionally run example
of a public free software project there is.   I don't mean that as 
faint praise, even though the bar of comparison to other projects is
pretty low -- their bar is very high.

The current GCC project is very respectful of the volunteer community --
mostly by greeting eager-beaver newbies with a kind of benign neglect
while being open and helpful to clearly competent and self-interested
contributors.   Day to day, it is largely an inter-corporate effort with
a few academics in the mix: good job.

I've praised them in those ways in the past and stand by that praise.

Could we have and should we have wound up with a more facile collection
of free software compiler technology?   I believe so.

Was it necessary to run roughshod over the GNU project?  I believe not.

Was a pattern demonstrated, and now repeated, for how to take over
projects and communities?  Certainly so.

Is GCC in a bit of a tough spot because of its intractability relative
to the demands of the day?  You betcha.   And I say that that's a
foreseeable consequence of the dog-pile approach that forced the initial
split and takeover.

Does "every coin have two sides"?   I'm not seeing any upside to the
trashing of the Arch project.


-t