Re: [Chicken-users] Need to do an install, any changes in the pipe line I should wait for?

[Peter Bex]

On Fri, Nov 08, 2013 at 10:45:05AM -0700, Matt Welland wrote:
> The primary purpose is to enable an installation of Megatest. With a little
> luck I hope to get Megatest, logpro and associated utilities complete and
> stable so I can install them centrally before the end of the year.

Oh, that's really cool!

I think if you're just running Megatest and you won't be installing any
"new" Scheme code it's enough to just test Megatest and if it works,
stick with it.  Usually, as new bugs get found and fixed these are when
people run into unexplored corners of CHICKEN and getting in trouble,
which you'll only run into when you're writing new code to run on that
old CHICKEN, which just happens to expose the bug.

There are always security issues which get discovered and may apply even
to unchanging systems, but that's a constant with ANY system you set up
and expect to keep running for a long time.

> A little grepping gives me the list of eggs below. I'll be doing some
> testing of 4.8.05 over the next week and wanted to hear from the dev's if
> there were patches in the queue I should wait for.

Just ensure you have the latest released versions of all these eggs.
Right now I'm intending to get 4.9.0 released soon, but this may take
on the order of several weeks or even months.  The roadmap in Trac is
a (very rough) indication of where we're at: https://bugs.call-cc.org/roadmap
As you can see, we've closed a whole slew of bugs and there are only
a handful remaining.  The critical/major ones need to be fixed before
the release, the others might get lifted over to be fixed for 4.10.0,
or in the "stability" cycle of 4.9.0.

Also, there are some known bugs in the irregex version used in 4.8.0
which are (AFAIK) fundamental to the way the submatch extraction is
implemented for the DFA-based matcher.  The irregex in 4.9.0 uses a
different implementation which does not have this problem.  See for
example https://bugs.call-cc.org/ticket/1047
This might be a problem for you if you intend to run user-supplied
regular expressions with submatches.

We're also slowly clearing up the small bugs found through the paranoia
results of DEBUGBUILD.  This may result in a few small fixes in existing
eggs.

Cheers,
Peter
-- 
http://www.more-magic.net