
Hi Folks, To give you a rough idea of how GHC activity is ramping up, take a look at the number of tickets assigned to recent milestones: 6.4.2: 27 tickets 6.6: 48 tickets 6.6.1: 107 tickets (61 open) 6.8: 134 tickets (117 open) We're focussed on getting 6.6.1 out of the door at the moment. We still have 61 open tickets though, so this message is a plea for help: if you have some time to spare, and would like to help out, please get involved! Here is a query that gives you the list of bugs milestoned for 6.6.1 with no current owner: http://hackage.haskell.org/trac/ghc/query?status=new&status=assigned&status=reopened&group=milestone&milestone=6.6.1&owner=&order=priority I've also listed the bugs at the end of this message. Some of the bugs we've marked as low priority (blue background), those are the ones we consider would be "nice to fix" but not essential for 6.6.1 (some of these are quite easy, incedentally). The protocol is, if you start working on a bug then claim it by setting yourself as the owner. Other guidelines for working on GHC can be found here: http://hackage.haskell.org/trac/ghc/wiki/WorkingConventions in particular see the section at the end of that page about use of the bug tracker. If you need permission to modify tickets, please ask either myself or Ian. Cheers, Simon open tickets for 6.6.1 with no owner: 742 Graphics.SOE runs very slowly under win32. 804 Signal handlers always installed, evem in a DLL 933 Separate compilation fails with existential records 955 more object-code blow-up in ghc-6.6 vs. ghc-6.4.2 (both with optimization) 957 No way to use -lgmp from a non-standard location 970 GHCi crashes under Windows Millenium 976 GHCi crashes on Windows 98 987 X11: foreign declarations use Haskell types instead of C ones 1010 ghci crashes when running out of heap. 1014 Mangled module name in __stginit_Module symbol 1025 -ddump-minimal-imports works wrongly 1035 Win32 package doesn't appear in the online library docs 1044 library docs on web have broken link 1075 ghci unhandled exception on startup (memory access violation) 1076 runhaskell crash with .hi/.o files in the directory 1082 Documentation missing from MacOS PPC binary 1093 Windows: haddock-html fields are wrong in package.conf 1109 lockFile: fd out of range 1110 Setting PATH needed in Windows Vista 1128 The impossible happened, code commented 604 Windows installer: use WiX 609 Useful optimisation for set-cost-centre 831 GHCi user interface bug 839 Generate documentation for built-in types and primitve operations 850 threaded RTS uses SIGALRM 956 improving error messages #1 998 Tab-completion of filenames does not work in GHCi 6.6 1077 documentation error and omission 1095 make boot under includes/ doesn't run make depend 1096 More make boot / make depend problems 1098 Broken link in the User's Guide 1117 [2,4..10] is not a good list producer 1119 openBinaryFile: does not exist (No such file or directory) i386-unknown-linux 629 IO library locking doesn't count readers

| -----Original Message----- | From: Simon Marlow | Sent: 02 February 2007 09:54 | | We're focussed on getting 6.6.1 out of the door at the moment. We still have 61 | open tickets though, so this message is a plea for help: if you have some time | to spare, and would like to help out, please get involved! Just to reinforce what Simon's message said.... We think that the reason GHC 6.6.1 has such a lot of open bugs is not because quality has fallen, but rather because more and more people are using GHC, in more and more interesting ways. That's good news. And hopefully it also means that there are more and more people who are well-qualified to help find and fix these corner cases. GHC is not our compiler, it's *your* compiler. The more you join in and help us, the better GHC will become. This is quite important, because what may seem like a corner case to us can be a show-stopper for the person who trips over it; and if that happens too much, they just switch languages. 61 open bugs is 61 too many. So do please lend us a hand. Even characterising a bug, narrowing down just what the problem is, can be a huge help. Simon and I and Ian, and others, are ready to help with specific information, the Commentary is an increasingly useful resource. (Please to add material there too, as you find it out.) Thanks; and a big thank-you to those who have already contributed. Simon PS: even if you are looking at a 6.6.1 bug, the usual place to start is with the HEAD; fix and patch the HEAD, and Ian will merge the patch to the branch. Only if the bug is branch-specific should you work on the branch. See http://hackage.haskell.org/trac/ghc/wiki/WorkingConventions
participants (2)
-
Simon Marlow
-
Simon Peyton-Jones