October 2006 - Yhc - Haskell.org

Warning: scons fullclean recommended
by Neil Mitchell 15 Oct '06

15 Oct '06

Hi, Just a small warning, I would recommend people do a scons fullclean, as I just fixed a bug that will possibly result in corrupt hbc files (unlikely) but definately in corrupt ycr files (yhc core output). In wrapping the Id type, deriving Show was used, which outputs "Id n", instead of n, hence LAMBDA256 was output as LAMBDAId 256 - a valid string, but a bit weird. I'd recommend a scons fullclean to get rid of these oddities. Thanks Neil

2 2

Finfo flags
by Robert Dockins 15 Oct '06

15 Oct '06

Can anyone tell me what the Finfo flags field means and is used for? The C runtime defines this enum, which I assume is related: enum FInfoFlags { FFL_NONE = 0x00, FFL_INVIS = 0x01, FFL_LAMBDA = 0x02, FFL_PRIM_APPLY = 0x04 }; They were added in about May, and seem to relate to Hat, based on patch comments. -- Rob Dockins Talk softly and drive a Sherman tank. Laugh hard, it's a long way to the bank. -- TMBG

1 0

-package cpphs ? (build problem)
by Greg Buchholz 14 Oct '06

14 Oct '06

When I try the build process with scons, I get a failure when it tries to link the object files together to produce executables. The command... ghc src/compiler98/FixSyntax.o ...(etc)... -o inst/bin/yhc ...complains about... src/compiler98/Front.o(.text+0x155d): In function `s3My_1_alt': : undefined reference to `LanguageziPreprocessorziUnlit_unlit_closure' src/compiler98/Front.o(.text+0x1a56): In function `s3pb_info': : undefined reference to `LanguageziPreprocessorziCpphsziOptions_CpphsMacro_closure' src/compiler98/Front.o(.text+0x1ab1): In function `s3OV_1_alt': ...etc. If I try the command by hand, and add "-package cpphs", I can get a yhc executable. I'm wondering if there is something strange about my setup (Linux, ghc-6.4.2, scons v0.96.1.D001) which would cause this problem, or is there something I can change in the configuration files to get things to build properly? Thanks, Greg Buchholz

4 13

buildbot failure in gentoo_x86
by andrewjwilkinsonw＠gmail.com 13 Oct '06

13 Oct '06

The Buildbot has detected a failed build of gentoo_x86. Full details are available at: http://www.indiegigs.co.uk:8010/gentoo_x86/builds/186 Buildbot URL: http://www.indiegigs.co.uk:8010/ Build Reason: Build Source Stamp: HEAD Blamelist: Andrew Wilkinson <aw(a)cs.york.ac.uk> BUILD FAILED: failed darcs sincerely, -The Buildbot

1 0

Re: [Yhc] Re: Could not build yhc
by Andrew Wilkinson 13 Oct '06

13 Oct '06

On 10/13/06, Dimitry Golubovsky <golubovsky(a)gmail.com> wrote: > Thanks, I'll try that tonight. Is this documented anywhere (minimal > version of Python)? The only version requirement I saw was for scons, > and the latter is OK with even pre 1.6 No, I'm not sure it is documented. I hadn't even thought that some might be using an unsupported version of Python. 2.3 was released more than three years ago... I have added a test which will tell you if you have an old version of Python. I will also alter the wiki. Also how critical is to have gmp 3.3? I have 3.1 > I don't know about that. The build process doesn't require any particular version so it'll at least try to build. Andrew

1 0

Re: Could not build yhc
by Neil Mitchell 13 Oct '06

13 Oct '06

Hi Dimitry, No idea, but I'm sure Andrew does. I have sent this on to the Yhc mailing list where I am sure someone will no the answer. Thanks Neil (Please CC Dimitry in on the responses to this) On 10/13/06, Dimitry Golubovsky <golubovsky(a)gmail.com> wrote: > Neil, > > I just tried to build yhc using the instructions on Haskellwiki. > > I installed a newer Python (2.2.3), scons, and the latest Subversion. > > I got this error message (actually, this is the second attempt, the > first attempt failed after: > > bootstrap/packages/yhc-base/1.0/Data/Ratio.hi -> > src/packages/yhc-base-1.0/Data/Ratio.hi > bootstrap/packages/yhc-base/1.0/Data/Ratio.hbc -> > src/packages/yhc-base-1.0/Data/Ratio.hbc ) > > =============== > > bash$ scons > scons: Reading SConscript files ... > Checking for architecture... x86 > Checking for operating system... linux2 > Bootstrapping... > Done. > AttributeError: 'module' object has no attribute 'walk': > File "Sconstruct", line 70: > yhi_objs, yhcfiles, library, testerfiles = SConscript("src/SConscript") > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 581: > return apply(method, args, kw) > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 508: > return apply(_SConscript, [self.fs,] + files, {'exports' : exports}) > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 239: > exec _file_ in stack[-1].globals > File "src/SConscript", line 5: > yhc_main = SConscript("compiler98/SConscript") > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 581: > return apply(method, args, kw) > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 508: > return apply(_SConscript, [self.fs,] + files, {'exports' : exports}) > File "/previous/public/Python-2.2.3/lib/python2.2/site-packages/SCons/Script/SConscript.py", > line 239: > exec _file_ in stack[-1].globals > File "src/compiler98/SConscript", line 7: > for root, dir, files in list(os.walk(".")): > > ============== > > Do I need a newer Python than 2.2.3? > > -- > Dimitry Golubovsky > > Anywhere on the Web >

2 1

Version Numbers
by Neil Mitchell 13 Oct '06

13 Oct '06

Hi Everyone, We need to come up with a versioning scheme that can serve Yhc forevermore. This is partly needed because the Debian packagers need a version, but we probably need it anyway. Do yhc --version and you'll see what i mean... So what should we have as the version numbering scheme? I propose: <major>.<minor>.<date>.<number> major/minor being the normal release style version numbers. We might stick with 0.0 for a long time to come, but thats ok. I'd personally be happy with just one major/minor number, and increment it every time we feel like it - the major/minor distinction will have little meaning otherwise. date is just the date, i.e. 20060905 for the 5th of September 2006, so its in a nice sorted way. number starts at 0 each day, and goes up. This is required so each buildbot created version is a separate atomic version. I would very much like our build process to be tagging the darcs repo, collecting the builds that fall out and sending an email. By doing this every build becomes unique. Thoughts? Complaints? Ideas? Thanks Neil

1 0

Move to darcs.haskell.org complete
by Neil Mitchell 09 Oct '06

09 Oct '06

Hi, The official repo for Yhc is now at: http://darcs.haskell.org/yhc Please do not push patches to the York one, they will be thrown away. We will be abandoning the york one at some point soon. We have a new darcs web: http://darcs.haskell.org/darcsweb/darcsweb.cgi?r=yhc;a=summary If you do not have commit access, then everything should be as before. If you do (which now includes a few non-York people, and was the entire reason for the move!) "scons push" will do a push. Thanks Neil

1 0

darcs patch: Implement IdSupply for real (and 5 more)
by Samuel Bronson 04 Oct '06

04 Oct '06

Tue Oct 3 13:25:52 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Implement IdSupply for real Tue Oct 3 14:15:25 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Begin turning ImportState into a record type... Tue Oct 3 16:44:31 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Refactor ImportState and IExtract some Tue Oct 3 17:07:19 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Refactor IExtract.importData to use the pseudo-monad Tue Oct 3 18:52:05 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * More refactoring in IExtract Tue Oct 3 19:53:00 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Refactor IExtract.importField

1 1

darcs update
by Neil Mitchell 02 Oct '06

02 Oct '06

Hi, We are moving to a new darcs repo on haskell.org, so that non-York people can get ssh keys and access and have a chance to play in our repo. We now have a new repo: darcs.haskell.org/yhc Unfortunately I haven't tagged the repo yet, so --partial doesn't work. I am currently the only person with commit access, we're working on getting access for everyone else. Can no one commit to the old repo for the next day or so, until everyone has access to the new repo. Don't start pulling from there until I've got checkpointing set up, or it will take a while :) Thanks Neil

1 0