Head stability [Was Re: [Fwd: Re: [PATCH] Fix Name mangling in HEAD]]
Christopher R. Hertel
crh at nts.umn.edu
Tue Mar 26 14:48:01 GMT 2002
In order to manage this sort of environment, we really do need to move
toward a more modular design. We should be able, for example, to come up
with a an API for name mangling and isolate the mangling code. It was
*close* to this a while ago (but I understand that retro-fitting Unicode
is likely to cause trouble).
Point is, if we could build solid boundaries between modules then it would
be possible to swap entire subsections (eg., name mangling) without
impacting the rest of the system. (I'm not talking about run-time
loadable modules. You'd need to recompile but that should be it.)
Experimental name mangling code could be swapped in for testing, etc.
Yes, I know that it is difficult to get people to test, but if we could
swap things in and out like this then we could put experimental code into
HEAD, and back it out in a hurry if there were identified problems. Once
those problems were fixed the new code could go back into head *without*
losing the old code.
Well, okay, that's my 2cents. :P
> On Wed, Mar 27, 2002 at 07:46:55AM +1100, Andrew Bartlett wrote:
> > I'm not proposing this at this time. I am also not proposing that HEAD
> > somehow lose its experimental edge. What I am proposing is that failed
> > experiments are backed out, and that new code be added is on a 'no more
> > broken than before' basis (that is that the originator knows of no
> > reason why this code is any worse than the previous case). This is
> > certainly the unwritten basis I've worked on in my time on the team.
> Ah. Yes, ok :-). I'm a big fan of the "no more broken than before"
> strategy :-). Was the new stuff that bad ?
> > I've seen the amount of work this requires, and its not something I have
> > the time to manage at the moment.
> A wise man :-)
> Jeremy (who is in merge hell at the moment :-).
Christopher R. Hertel -)----- University of Minnesota
crh at nts.umn.edu Networking and Telecommunications Services
Ideals are like stars; you will not succeed in touching them
with your hands...you choose them as your guides, and following
them you will reach your destiny. --Carl Schultz
More information about the samba-technical