Re: Patch: One version number to rule them all

[Mark Mitchell <mark at codesourcery dot com>]

Rainer Orth wrote:
> Mark Mitchell writes:
>
>
> > I think it's a misallocation of resources for us to even pretend to 
> > maintain a script that Dan says is used once or twice a month, with 
> > relatively little success.  I don't doubt that the Coding Standards says 
> > we need to accept bugs by email, but I suspect that's been there since 
> > long before the web became so prevalent.
>
>
> But why remove what works now?  This whole discussion (and the previous
> ones) take up much more time than keeping the current gccbug and conversion
> scripts.

Every time I do a release or make a branch, I have to update the silly 
little mail script, which requires editing a file locally, cvs diff, cvs 
checkin, post the patch, etc.  Annoying, and rather pointless, if the 
script is getting used by two people a month.

> And (as I've stated several times before) I consider this web-centricity a
> nightmare for serious contributions.  With the same arguments, you could
> just as well remove gcc-patches and require patch submission via the web ;-( 

There's a key difference: people actually *use* gcc-patches.  If we got 
100 good bug reports a month from gccbug, I'd be singing a different tune.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304