- User Since
- Jan 13 2018, 6:43 AM (79 w, 19 h)
Fri, Jul 19
Tue, Jul 16
Fri, Jul 5
Thu, Jun 27
Apr 2 2019
Oct 28 2018
Oct 26 2018
Aug 22 2018
Aug 21 2018
@bgamari done. Check that the chengelog/release-notes updated reasonably.
- Update T104121 reference output to reflect current Unicode-related definitions.
Aug 20 2018
@Azel WCsubst.c is generated by ubconfc indeed, but I don't think this is done on rebuilding. Git history says WCsubst.c was updated 4 years ago last time. And the manual update using the current Unicode spec (v.12) fixes several bugs (Trac #5518, Trac #15525), which is the aim of this revision. We just need to decide what to do with T10412. /cc @bgamari
Aug 17 2018
- update unicode002 test according to Unicode v12
Aug 16 2018
The failure on OS X doesn't seem to have anything to do with the changes. @bgamari, since CI finished, do you want to look further at this diff?
Aug 14 2018
Second test, unicode002, I still can't wrap my mind around it. Basically, it generates a huge table with the result of 7 character tests for first 6554 codepoints. The way this test is mainted, I guess, is just replace the whole reference table with every Unicode standard update (like this was done, e.g., in this commit).
Two tests are failing.
"state transformer" → "state monad" / "ST" (whichever is meant), fix Trac #15189
Update Unicode tables to v. 12 of the standard
Aug 4 2018
@bgamari can you hint what does Lease a working copy for Linux/amd64 builds mean: CI seems to have felt on this step.
Aug 2 2018
Thanks, Ben! My wording is certainly not too accurate. I'm not sure how to improve it nicely, though. Feel free to edit the summary as you wish!
@alpmestan thank you again! This does seem to work! Sorry for the mess… Now it is clear how to fight with this in the future.
- fighting Cabal submodule
It seems to help for terminfo, but Cabal is still dirty.
- update submodules
I'm lost with git submodules. If only anyone could tell me the magic spell to rollback the change in libraries/Cabal and libraries/terminfo… Or is it fine now?
Jul 28 2018
Fix a typo
May 15 2018
Apr 30 2018
@simonpj here is the summary:
@simonpj it is done now.
Make updNameCacheTc use updNameCacheIO
Apr 28 2018
@simonpj do you think it is good enough after last revision?
Apr 22 2018
The task of forcing mod and occ before doing updNameCache moved
from the callers to callee.
Apr 18 2018
@simonpj do you mean adding mod and occ to just updNameCache or to updNameCacheIO also? The latter seems inviting, but I'm not sure it is possible for NameCacheUpdater business to keep up. Otherwise, a bunch of seq before calls to updNameCacheIO will have to remain.
Wow, I missed the half of calls (for a silly reason). Then probably inlining liftIO lookupOrigIO everywhere is not that good. I could easily change its definition for liftIO, but if @simonpj not sure, I'd better hold on.
Apr 17 2018
This has been fixed by Simon himself.
Apr 14 2018
Apr 6 2018
Do we actually need lookupOrig, or lookupOrigIO alone is enough? I see just a couple of usages of the former.
Apr 2 2018
- Rename updateNameCacheCallback in lookupNameCache.
- Move a comment about seqing parameters from lookupOrig to lookupOrigIO.
Mar 31 2018
- Erase initTcForLookup from tests also.
In fact, lookupOrig is also IO-heavy and I can easily do a lookupOrigIO variant.
Mar 30 2018
After some fiddling with Tc/IO-changes I think this is clear besides one thing: dataConInfoPtrToName does lookupOrig at the very end, and I'm not sure how to handle it inside IO. Before, in simplCore setting we used CoreM to do this. But what could we do inside plain IO?..
Yes, I think, I can do that.
Mar 29 2018
I've done what Simon asked for.
- inline small auxiliary function
It seems the CI fails because of test T4891. I should investigate it.
- fix unused import
- remove initTcForLookup which is not used anymore
Mar 28 2018
- Move conicalizeModuleIfHome to more appropriate place.
- Liberate thNameToGhcName from typechecker API.
Mar 25 2018
- factor out canonicalizeModuleIfHome from initiTc and lookupGlobal
Mar 20 2018
Thanks! I also dig into thNameToGhcName a bit more, and I think there could be something out there allowing removal of initTcForLookup. Not sure yet.
Mar 17 2018
I've done one of two things discussed above (2): added lookupGlobal_maybe. For the other one (1), expurging initTcForLookup, I described my troubles on Trac, let me repeat that here.
- Add lookupGlobal_maybe for clients willing to recover from lookup failure
Mar 15 2018
- remove old, commented out version of lookupGlobal
Will try 1) expurge initTcForLookup, 2) make lookupGlobal_maybe — later this week. Thanks for the comments!