[rt-devel] hacking on rt.

[Joshua Johnson]

--On Friday, February 08, 2002 12:29:16 PM -0600 Matt Knopp 
<mhat at netlag.com> wrote:

>
> I get the impression from their website that they don't even have a
> working win32 client.  Perhaps I'm missing something obvious, but under
> the NT section they describe the normal Win32 case as being:
>
>   "Most sites using Aegis and Windows NT together do so by running Aegis
>    on the Unix systems, but building and testing on the NT systems. The
>    work areas and repository are accessed via Samba or NFS."
>
> Not that it probably matters for 97% of RT/DBIx hackers.  Its just
> something I noticed and was somewhat surprised about.

	Yep, aegis runs suid. (needed to maintain security)  AFAIK that's not easy 
on Windows.

>
> Also, and this is just my own commentary.  Does RT really need a new
> version control system?  It seems like there is still a lot of things
> that could be added to RT, and futzing with the version control system
> is not (traditionally) a trivial task.  I dono, maybe a new vc system
> will help spurn rt development to new heights. . .
>
	I can't speak of the need for new version control for RT.

	But I personally note a difference between version control and 
process/configuration management.

	Aegis, for better or worse helps control the "how" of development and as 
such can add another layer of defense against stupid mistakes.  If 
configured it can prevent someone submitting a patch that doesn't build. It 
can require that a test case be submitted with the patch to prove the patch 
fixes the issue (must fail before patch and pass after patch). Things like 
this can easily be handled by the expertise of the developer for a while, 
but it does always seem to become increasingly difficult as things grow. 
And it can do other things like building a regression set, collecting 
metrics, etc. etc.

	There have been quite a few announcements of what I call "CVS 
second-generation" tools lately:

	BitKeeper
	Arch
	Subversion

	But these are all addressed to take on CVS's shortcomings, none of them 
tackle some of the issues that aegis does.  They all look really cool, 
don't get me wrong.  But they aren't an apples for apples comparison with 
aegis.

	Again, it all depends on the goals of your development project. I don't 
know what Jesse has in mind.


> Also, I'm inclined to believe that subversion will be the-right-thing
> when it becomes available, so I'm not sure I see the value in changing
> to something that may or may not be good enough and then possibly having
> to switch yet again in the future when subversion becomes ready.
>
	I'll only say that there seem to be many candidates for the "CVS 
second-generation" above.  The competition for the spot should help deliver 
a great product.


> Then again, that green grass over the fence looks a whole lot greener
> then the grass on this side.
>
> -Matt
>
>
>
>
> _______________________________________________
> rt-devel mailing list
> rt-devel at lists.fsck.com
> http://lists.fsck.com/mailman/listinfo/rt-devel