- User Since
- Jan 8 2015, 11:07 PM (128 w, 13 h)
Wed, Jun 21
It turned out to be more involved than I thought to tweak the GHC pretty-printer to be aware of zero-argument lambdas. I opted for the simpler solution for just converting TH LamEs with zero patterns to their underly body Exp instead, which avoids that problem entirely.
- Actually check for -ddump-splices output
- Just convert zero-arg lambdas to their bodies
This isn't quite right yet - the GHC pretty-printer still prints out an arrow (->) needlessly. Also, the test isn't checking for this.
Sun, Jun 18
This ought to be merged into 8.2.
Tue, Jun 13
- Add Note, per SPJ's suggestion
Mon, Jun 12
Sat, Jun 10
Thu, Jun 8
- Accurate numbers
Awaiting a Harbormaster run to tell me how many bytes to use (since my machine gives unreliable numbers).
Mon, Jun 5
Trac #11785 seems to encompass the idea of merging Type and Kind TH functionality together, so I mentioned this business about repLTy and repLKind on that ticket. I also updated the comments to refer to Trac #11785.
- Mention Trac Trac #11785 in the comments
Sun, Jun 4
Fri, Jun 2
Thu, Jun 1
Wed, May 31
I'll put this on Phab in case this has any effect on the performance tests.
Tue, May 30
Sat, May 27
Fri, May 26
May 24 2017
Thank you for the detailed explanation. But I'm not clearer than before on my second question: why are the file/process opesrtions you just added not subject to the same issues that qRunIO has? After all, they're also IO operations. Where specifically in the GHC codebase does a cross compiler make the decision to, e.g., look up executables on the build machine instead of the host machine when running qFindExecutables?
My apologies for not being familiar with the previous work on this front, but I don't see why this is necessary. Why does adding a bunch of hardcoded IO operations to Quasi resolve this issue, but qRunIO alone does not? Where in the code are cross compilers making the distinction between files on the build machine and files on the host machine when running, e.g., qFindExecutables?
May 14 2017
- Include test output
May 12 2017
May 11 2017
May 6 2017
May 4 2017
May 3 2017
Apr 27 2017
- Remove errant runIO
Apr 26 2017
Apr 25 2017
Thanks for doing this, @adamse! I've left some inline remarks.
- Accept new test output
We should add the program in Trac #13609 as a test case, no?
Apr 24 2017
Oops, grabbed too many commits
Rebase (and cross fingers for a green Harbormaster build)
Apr 23 2017
ghci063 (as currently written) does not pass on Windows. Here is the error that I get:
Apr 20 2017
- simonpj's suggestions
Apr 19 2017
Apr 6 2017
Apr 4 2017
That was some quite impressive turnaround time.
Ought the title of this Diff be "Mark unfold as deprecated"?
We definitely need to point out this change in libraries/base/changelog.md. Note that the deadline for base-22.214.171.124 (which will be shipped with GHC 8.2 soon) has already passed, so this would be incorporated into the base-126.96.36.199 release. At the moment, however, libraries/base/changelog.md doesn't have a section for base-188.8.131.52—do you mind making this change yourself? Just put a new section at the top like:
Apr 1 2017
Looks good overall. Make sure to mention this in the 8.4 release notes!
Mar 31 2017
Mar 30 2017
It's not terribly clear to me how this works. This needs
Fantastic work, @dfeuer!
Mar 29 2017