On Tue, 28 Aug 2012, Christian Sternagel wrote:
Sometimes it is just hard to know whether some snag was not detected yet
or is already in the pipeline... so I report anything ;).
After so many years, Isabelle pilelines have become looong, so it is very
likely that whatever you name is in there already. Nonetheless, you
should report whatever you observe and find relevant. There might be new
things coming from that that are *not* yet in any pipeline system from the
Ural, and it always helps to figure out priorities what needs to be pulled
or pushed forward a bit more quickly than 5-10 years.
Maybe it would be a good idea to collect "feature-requests" in the wiki?
We are back to the tracker issue, which has already become a running gag
on this mailing list. A wiki is very bad technology to track feature
reports or bug requests. There are dedicated trackers for that, but most
of them are also quite bad. More recently some non-bad trackers have
emerged, so it might be worth revisiting the question with a clear mind at
some point.
Apart from that, technology alone never solves a problem. It needs to be
supported by people behind it, and this is exactly where most
tracker-based project that I've ever seen are lacking. E.g. when I put
something on one of the many jEdit trackers on Sourceforge (another bad
technology), it takes a very long time to get any feedback on it, even
just that it has been seen by someone.
Other projects like Cygwin define themselves as explicitly tracker-free.
When I posted something on their mailing list -- an actual "issue" -- I've
got feedback almost as quickly as normally on isabelle-dev.
(The official place of collection is currently the mailing list, but it
is not that easy to extract this kind of information.) I mean, in order
to have a place to point like: "first look there and only if you do not
find your 'feature' there, write to the mailing list".
The mailing list archive also helps. It has the advantage that it
provides an historical context to some issue and its discussion, unlike a
wiki that is always known to be outdated without explicit clues. (This is
also a reason why it is so important to quote changeset ids when pointing
to anything in the repository.)
Makarius
_______________________________________________
isabelle-dev mailing list
isabelle-...@in.tum.de
https://mailmanbroy.informatik.tu-muenchen.de/mailman/listinfo/isabelle-dev
