Skip to content

Instantly share code, notes, and snippets.

@Whateverable

Whateverable/query

Created January 28, 2019 14:42
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save Whateverable/1b6ac36880bd7b186d0134c01f145354 to your computer and use it in GitHub Desktop.
Save Whateverable/1b6ac36880bd7b186d0134c01f145354 to your computer and use it in GitHub Desktop.
Download ZIP
quotable6
Raw
result-#perl6-dev.md

https://gist.github.com/ugexe/31e22b5924319f502f62970fca34e3a0 # why does this code act different? it looks the same, but it fails spectests (and additionally fixes Grammar::HTTP which was broke from uncurse branch merge in march-ish)
That came in as part of uncurse IIRC. Match is pretty low-level so some of the methods aren't Perl 6 routines, but NQP ones
so they were rewritten as part of uncurse then?
.ask TimToady can `uncurse` branch be deleted from Rakudo's repo now?
05:37Z <Zoffix> TimToady: can `uncurse` branch be deleted from Rakudo's repo now?
another uncurse curse?
AlexDaniel: now THAT looks like uncurse curse :)
Well, it's definitely uncurse. We already had a bug with capture markers due to it.
TimToady: I think https://rt.perl.org/Ticket/Display.html?id=131187 may be uncurse fallout, if you get chance to take a look
yes, if I recall correctly these 20 unbuildable commits are due to uncurse merge
jnthn: yes, that's almost certainly uncurse related; unifying $!to with $!pos was a slight overreach
¦ rakudo/nom: The unification of $!to and $!pos in the uncurse branch caused the )>
And I don't know if latest uncurse bug was found by tests
(and making *that* possible is why we got uncurse)
guesses that probably snuck in at some point due to hitting 'u' one too many times in vim, since it certainly wasn't part of the uncurse effort...
¦ nqp/master: 694725e31c | �Merge commit '0c53231dd9930d2cbaea6d03d2655d6d7b322476' into uncursej�
¦ nqp/master: 97af4c5d13 | �Merge commit '2a31c20393225f0c6ac1a21127f73ddae8ec0b5a' into uncursej�
¦ nqp/master: 9aa9c9e53f | �Merge branch 'uncursej'�
huh, something from the uncurse branch?
fwiw my very grammar involved compiler (https://github.com/spitsh/spitsh) is broken too after uncurse got merged.
then I'm working on an it-would-be-nice feature, mostly just to make sure uncurse didn't cause it
TimToady: how should I log the uncurse changes in the Changelog?
well, that's definitely uncurse territory
reminds: need to write down uncurse into ChangeLog; should should I write for it?
well, uncurse is about getting rid of the Cursor object by merging Match and Cursor together
¦ rakudo/nom: 9aa9c9e Merge branch 'uncursej'
¦ nqp/uncurse: 211b0f2ba3 | (Pawel Murias)++ | src/vm/moar/profiler/template.html
¦ nqp/uncurse: �Fix bitrot in moarvm profiler�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/211b0f2ba3�
¦ nqp/uncurse: 30a965887f | (Pawel Murias)++ | src/vm/js/Compiler.nqp
¦ nqp/uncurse: �[js] Stuff to fix nqp::p6initied�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/30a965887f�
¦ nqp/uncurse: 9063d0ec3f | TimToady++ | 2 files
¦ nqp/uncurse: �Merge branch 'master' into uncurse�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/9063d0ec3f�
¦ rakudo/uncurse: �6 commits �pushed by (Elizabeth Mattijsen)++, (Timo Paulssen)++, TimToady++
¦ rakudo/uncurse: c16cdb2ce7 | �Bump NQP version for Allocations tab in profile fix�
¦ rakudo/uncurse: 354a4dbbda | �Update ChangeLog for latest NQP bump�
¦ rakudo/uncurse: 17d34cdc91 | �allow for faster Enum ~~ Enum�
¦ rakudo/uncurse: 89457f8d57 | �make Numeric.ACCEPTS(Any) about 15% faster�
¦ rakudo/uncurse: e0e0800897 | �accidentally committed wrong version�
¦ rakudo/uncurse: b7f0a665f1 | �Merge branch 'nom' into uncurse�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/89fd91c1ea...b7f0a665f1�
Zoffix: lemme know when you're in a state that I can merge uncurse, cuz I'm now thinking it would better to get it into this release so we don't risk divergent nqp bootstraps (and to get the 1-2% performance boost)
has nqp-j been correctly rebootstrapped after the adding of constant support in the uncurse branch?
¦ nqp/uncurse: a44e6dedc5 | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �reduce old match object to a singleton�
¦ nqp/uncurse:
¦ nqp/uncurse: Still can't get it quite down to a boolean, but at least we can avoid
¦ nqp/uncurse: allocating every time MATCH is called.
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/a44e6dedc5�
¦ rakudo/uncurse: 63e76d7afe | TimToady++ | 2 files
¦ rakudo/uncurse: �reuse NQP's singleton "did match" type�
¦ rakudo/uncurse:
¦ rakudo/uncurse: (though for the moment we allocate our own singleton)
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/63e76d7afe�
¦ rakudo/uncurse: �13 commits �pushed by MasterDuke17++, (Zoffix Znet)++, TimToady++
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/63e76d7afe...fbf19d88da�
¦ rakudo/uncurse: d46e8df4cb | (Zoffix Znet)++ | src/core/IO/Pipe.pm
¦ rakudo/uncurse: �[io grant] Add IO::Pipe .path and .IO methods�
¦ rakudo/uncurse:
¦ rakudo/uncurse: That return an IO::Path type object. This fixes LTA error
¦ rakudo/uncurse: of .IO/.path crashing, since the inherited methods from IO::Handle
¦ rakudo/uncurse: try to instantiate an IO::Path from an empty string.
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/d46e8df4cb�
¦ rakudo/uncurse: c01ebea0a0 | (Zoffix Znet)++ | src/core/IO/Path.pm
¦ rakudo/uncurse: �[io grant] Make IO::Path.mkdir return invocant on success�
¦ rakudo/uncurse:
¦ rakudo/uncurse: Per discussion[^1], to ameliorate possible fallout from changes to
¦ rakudo/uncurse: &mkdir[^2], make the method return the invocant instead of True.
¦ rakudo/uncurse:
¦ rakudo/uncurse: Since all IO::Path:Ds are True, the method's success can still be
¦ rakudo/uncurse: judged in boolean context, and this way &mkdir returns its old
¦ rakudo/uncurse: �<…commit message has 5 more lines…>�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/c01ebea0a0�
¦ rakudo/uncurse: 762c63ce64 | TimToady++ | 2 files
¦ rakudo/uncurse: �Merge branch 'nom' into uncurse�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/762c63ce64�
Hm, I'm guessing uncurse didn't get merged yet?
But if uncurse ain't merged it must be something else :S
jnthn: and no, I don't think uncurse is merged
No, uncurse wasn't merged.
jnthn: I had not planned on merging uncurse till after the release, since the IO work is already a fairly big perturbation, but I could probably be argued into merging it for this release
¦ nqp/uncurse: �14 commits �pushed by (Pawel Murias)++, (Timo Paulssen)++, TimToady++
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/compare/a44e6dedc5...005a2b7a20�
¦ rakudo/uncurse: �9 commits �pushed by (Zoffix Znet)++, (Samantha McVey)++, (Timo Paulssen)++, TimToady++
¦ rakudo/uncurse: 1f689a94e1 | �[io grant] Fix up IO::Handle.Str�
¦ rakudo/uncurse: 490ffd1562 | �[io grant] Do not use self.Str in IO::Path errors�
¦ rakudo/uncurse: 40217edfb1 | �[io grant] Swap .child to .concat-with in all the guts�
¦ rakudo/uncurse: 0e36bb26ba | �[io grant] Make IO::Spec::Win32!canon-cat 2.3x faster�
¦ rakudo/uncurse: 822566f1ad | �Bump NQP for case insensitive regex 2x speed boost�
¦ rakudo/uncurse: 20af51fef9 | �bump nqp and moarvm for better multithreading performance�
¦ rakudo/uncurse: fd503f89ed | �Revert "[io grant] Remove `role IO` and its .umask method"�
¦ rakudo/uncurse: c95c4a75d1 | �[io grant] Make IO::Path/IO::Special do IO role�
¦ rakudo/uncurse: 89fd91c1ea | �Merge branch 'nom' into uncurse�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/762c63ce64...89fd91c1ea�
uncurse does seem to be 1-2% faster compiling the setting, and it does seem to be in a stable state, or at least a metastable state, and it does pass one TODO for Match.perl, so maybe it's worth merging earlier
¦ nqp/uncurse: 1708955631 | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �make false match stringify to ''�
¦ nqp/uncurse:
¦ nqp/uncurse: (rather than blowing up mysteriously with a negative length)
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/1708955631�
¦ nqp/uncurse: 98d2333913 | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �Match.Int should return, like, an int�
¦ nqp/uncurse:
¦ nqp/uncurse: The fact that it was returning a num was confusing rakudo.
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/98d2333913�
¦ rakudo/uncurse: 74a55548cc | TimToady++ | src/core/Match.pm
¦ rakudo/uncurse: �bulletproof Match more; Capture does own BUILD�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/74a55548cc�
¦ rakudo/uncurse: cbae90c6f9 | TimToady++ | src/core/Match.pm
¦ rakudo/uncurse: �submatches come from target, not orig�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/cbae90c6f9�
¦ rakudo/uncurse: 3487674be9 | TimToady++ | src/core/IO/Path.pm
¦ rakudo/uncurse: �enforce Bool type on $!is-absolute setting�
¦ rakudo/uncurse:
¦ rakudo/uncurse: Something's sneaking through as Int and giving the assignment heartburn.
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/3487674be9�
¦ nqp/uncurse: 341375da59 | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �submatches come from target, not orig�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/341375da59�
¦ nqp/uncurse: 223eeb0666 | (Jonathan Worthington)++ | src/how/NQPClassHOW.nqp
¦ nqp/uncurse: �Re-work --rxtrace to not trash the cache.�
¦ nqp/uncurse:
¦ nqp/uncurse: Which, aside from not breaking a bunch of things in the parse when
¦ nqp/uncurse: using uncurse, will also make it a load faster.
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/223eeb0666�
¦ nqp/uncurse: �10 commits �pushed by (Samantha McVey)++, (Zoffix Znet)++, MasterDuke17++, TimToady++
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/compare/223eeb0666...f038a43890�
¦ rakudo/uncurse: �12 commits �pushed by (Zoffix Znet)++, usev6++, (Samantha McVey)++, TimToady++
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/3487674be9...394c3c85fa�
¦ nqp/uncurse: 02c141f7f6 | TimToady++ | 7 files
¦ nqp/uncurse: �remove now-useless .CURSOR calls�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/02c141f7f6�
¦ rakudo/uncurse: �10 commits �pushed by (Zoffix Znet)++, usev6++, (Timo Paulssen)++, TimToady++
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/6d04acf7dc...331c51f85d�
¦ nqp/uncurse: fcff8e222e | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �allow !cursor_init to be used by a BUILD�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/fcff8e222e�
¦ rakudo/uncurse: 5069294b56 | TimToady++ | 2 files
¦ rakudo/uncurse: �Capture and Match shouldn't provide method new�
¦ rakudo/uncurse:
¦ rakudo/uncurse: (Especially when BUILD will work just fine.) Otherwise we force
¦ rakudo/uncurse: new grammars that have their own attributes to declare their own
¦ rakudo/uncurse: new, which should be unnecessary.
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/5069294b56�
¦ rakudo/uncurse: 18bb9be217 | TimToady++ | src/core/Match.pm
¦ rakudo/uncurse: �do not override nqp's make/made methods�
¦ rakudo/uncurse:
¦ rakudo/uncurse: NQPMatchRole from nqp now provides all our $!made service.
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/18bb9be217�
¦ nqp/uncurse: bd4f38a416 | TimToady++ | src/QRegex/Cursor.nqp
¦ nqp/uncurse: �PRECURSOR should not trim $!orig�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/commit/bd4f38a416�
progress report on uncurse, we're now down to blowing only 25 spectest .t files completely out of the water :)
anyhoo that's the direction I'm thinking once the uncurse branch settles down
¦ rakudo/uncurse: cdd625b68c | TimToady++ | 11 files
¦ rakudo/uncurse: �unify Match with Cursor�
¦ rakudo/uncurse:
¦ rakudo/uncurse: This just follows the same changes that nqp does for unification.
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/cdd625b68c�
¦ rakudo/uncurse: b123ab223b | TimToady++ | 2 files
¦ rakudo/uncurse: �move Match code from Cursor.pm to Match.pm�
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/b123ab223b�
progress report: the uncurse branch now successfully parses the setting, but blows up somewhere in the precompiler
¦ nqp/uncurse: �17 commits �pushed by (Samantha McVey)++, (Pawel Murias)++, TimToady++
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/compare/e32ad16b0e...6439b08a94�
¦ rakudo/uncurse: �30 commits �pushed by (Zoffix Znet)++, (Julien Simonet)++, (Elizabeth Mattijsen)++, (Samantha McVey)++, (Daniel Green)++, (David Warring)++, lizmat++, TimToady++
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/compare/b123ab223b...3dbd05631d�
¦ rakudo/uncurse: 6d04acf7dc | TimToady++ | src/core/Attribute.pm
¦ rakudo/uncurse: �get rid of hacky read of .HOW�
¦ rakudo/uncurse:
¦ rakudo/uncurse: (It was a mystery why we ever needed it in the first place, but for
¦ rakudo/uncurse: a while the following method call to .generate_accessor would not
¦ rakudo/uncurse: work without it.)
¦ rakudo/uncurse: �review: https://github.com/rakudo/rakudo/commit/6d04acf7dc�
¦ nqp/uncurse: �4 commits �pushed by TimToady++
¦ nqp/uncurse: 2b73538d8d | �rebootstrap to get constant declarator�
¦ nqp/uncurse: 0c53231dd9 | �start to unify NQPMatch with NQPCursor�
¦ nqp/uncurse: 2a31c20393 | �rebootstrap with temp shim to find lost NQPCursor�
¦ nqp/uncurse: e32ad16b0e | �finish s/NQPCursor/NQPMatch/ tranformation�
¦ nqp/uncurse: �review: https://github.com/perl6/nqp/compare/0000000000...e32ad16b0e�
TimToady: is the nqp/uncurse branch supposed to have any immediate effects for something like rakudo's parsing?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment