Perl 6 mailing list summary for the weeks of August 1-19, 2006
Summary updates
This is the final installment of the older summaries. After this week, I
plan to post weekly summaries on Sundays.
Parrot Porters
[perl #39750] [EMAIL PROTECTED]: tru64 core dump: t/examples/japh_12.pasm
<http://xrl.us/qvh3>
This thread was previously completely summarized in July.
In [perl #39750] <http://xrl.us/qvh5>, Jarkko Hietaniemi posted about a
core dump. Chip Salzenberg fixed it.
[perl #39751] unbug - [EMAIL PROTECTED]: tru64 core dump:
t/dynoplibs/myops_4.pir
<http://xrl.us/qvh6>
This thread was previously completely summarized in July.
In [perl #39750] <http://xrl.us/qvh5>, Jarkko Hietaniemi posted about a
core dump. Will Coleda replied that this is how the test should behave, and
wondered if the test should be skipped or the opcode removed. Chip
Salzenberg wrote a haiku about the test, and Jarkko poetically requested
that the test harness be informed about Parrot's tendency to obey suicide
requests.
[perl #39752] [EMAIL PROTECTED]: tru64 core dump: t/op/lexicals_27.pir
<http://xrl.us/qvh7>
This thread was previously completely summarized in July.
Jarkko Hietaniemi had another bug to report in [perl #39752]
<http://xrl.us/qvh8>; Chip Salzenberg replied that it had been fixed.
[perl #39754] [EMAIL PROTECTED]: tru64 core dump:
t/pmc/resizablebooleanarray_20.pasm <http://xrl.us/qvh9>
This thread was previously completely summarized in July.
Jarkko Hietaniemi's issue in [perl #39754] <http://xrl.us/qvia> now waits
for a rewrite of `ResizeableBooleanArray`.
[perl #39756] [EMAIL PROTECTED]: tru64 core dump: t/examples/japh_10.pasm
<http://xrl.us/qvib>
This thread was previously completely summarized in July.
Jarkko Hietaniemi created another bug report, this time in ticket [perl
#39756] <http://xrl.us/qvic>. This made Chip Salzenberg thing the test
harness was ignoring segmentation faults, but Leopold Toetsch explained
that it was Test::* TODO.
Building a Fedora package <http://xrl.us/qvid>
This thread was previously completely summarized in July.
Steven Pritchard reported that he had been working on building Fedora
Extras packages for Parrot and Pugs and had a few issues on x86_64. Joshua
Hoblitt asked if perhaps this was x86_64 with multilib support, as the
current build system does not support multilib builds.
Attached to the same thread was ticket [perl #39860] <http://xrl.us/qvie>
where Joshua Hoblitt requested that Configure.pl should accept `--libdir`.
Later he attached a patch.
[perl #39776] [BUG] PGE core dump <http://xrl.us/qvif>
This thread was previously completely summarized in July.
Kevin Tew created ticket [perl #39776] <http://xrl.us/qvig> to report a PGE
core dump. Patrick R. Michaud identified a possible syntax error. Chip
Salzenberg wanted to know what the appropriate response was, a core dump or
an exception.
[perl #39778] Segfault when using a Namespace with an Iterator
<http://xrl.us/qvih>
This thread was previously completely summarized in July.
Matt Diephouse created ticket [perl #39778] <http://xrl.us/qvii> because
trying to use an Iterator with a NameSpace makes Parrot segfault. Chip
Salzenberg, Bob Rogers, Leopold Toetsch and Matt discussed how to fix it.
[perl #39856] TODO: Produce Single PBC from Multiple PIR Files with -o
<http://xrl.us/qvij>
This thread was previously completely summarized in July.
In ticket [perl #39856] <http://xrl.us/qvik>, chromatic requested that
Parrot be able to produce a single PBC file from a command like `parrot -o
all_files.pbc file1.pir file2.pir ...`. Joshua Hoblitt thought it would be
a good idea to define how symbol conflicts should be handled. Leopold
Toetsch believed that `pbc_merge` can handle this, but Allison Randall
explained that it doesn't quite address the request. Chip Salzenberg felt
that was a bug, although Allison preferred to call it a feature. Jerry Gay
clarified how it worked by citing documentation and including tests.
[perl #39905] [TODO] TGE - line number reporting. <http://xrl.us/qvim>
This thread was previously completely summarized in July.
Ticket [perl #39905] <http://xrl.us/qvin> was created by Will Coleda.
Errors in an embedded PIR should be reported against the original .tg file
rather than the generated PIR file. Patrick R. Michaud asked how this could
be addressed, and Will included documentation.
[perl #39986] Create Default PackFile for New Interpreter (Parrot C API)
<http://xrl.us/qvio>
This thread was previously completely summarized in July.
chromatic created ticket [perl #39986] <http://xrl.us/qvip> to report that
there is no default PackFile for the situation where he has created a new
interpreter with `Parrot_new()` and wants to compile and run code through
`Parrot_compile_string()` without having to read and load bytecode from
disk.
It was reported that this was fixed in r13613, but chromatic had further
questions. There was further discussion between chromatic, Leopold Toetsch
and Chip Salzenberg. Chip requested some cleaning on namespace pollution,
and Andy Lester replied that it is on his list.
Continuations and inferior runloops: Analysis and POC <http://xrl.us/qviq>
Leopold Toetsch presented an analysis of the the inferior runloops problem
- embedding foreign language runloops in the parrot runtime. The problems
encompass continuation support and garbage collection amongst other things.
Bob Rogers replied proposing that foreign data be explicitly tracked. Leo
mentioned that this has been previously discussed and dismissed.
[perl #39997] [PATCH] PGE P5 Test Cleanup <http://xrl.us/qvir>
This thread was previously completely summarized in July.
David Romano submitted a patch to clean up the regex tests taken from Perl
5. Jerry Gay raised the an issue with the data file being shared with Perl
5, after which David supplied an additional patch, moving todo/skip
information to an outside file. Jerry applied.
PGE/TGE and the future. <http://xrl.us/qvis>
This thread was previously completely summarized in July.
Kevin Tew wanted information about TGE design goals and plans. Allison
Randal replied that the plan is to have a single set of OST node classes
used by all languages, a task Vishal Soni is currently at work on. She
included a link to the first draft of the compiler tools plan, which
Patrick R. Michaud is at work on updating so that it can be converted to a
PDD.
[perl #40002] TGE Refactor / Compiler Tools Object <http://xrl.us/qvit>
This thread was previously completely summarized in July.
Kevin Tew created ticket [perl #40002] <http://xrl.us/qviu> to ask a number
of questions about the TGE refactor and cited several other tickets. There
were some replies in August.
Patrick R. Michaud replied with information about the plans for TGE.
Allison Randal and Patrick R. Michaud contributed summaries of a
conversation they had at OSCON. Jerry Gay also commented.
resizablepmcarray, assign. <http://xrl.us/qviv>
This thread was previously completely summarized in July.
Will Coleda included some code which was not working in Tcl. He thought the
error originated in the Parrot code base, and asked some questions. Bob
Rogers had difficulty replicating the problem; chromatic suggested using a
newer Test::Harness because prove is not included in the older version.
Leopold Toetsch felt this was in conflict with the Parrot README
prerequisites list.
Will said that the problem was fixed in r13696. He also reported that Tcl's
test suite is being converted from using Perl to being self-hosting in Tcl.
Bob reported some additional problems in r13715. Will noted that Tcl was
undergoing a lot of changes this week. There was some further discussion
with Matt Diephouse on the Tcl changes.
[perl #40030] [PATCH] compiler/imcc missing dependency <http://xrl.us/qviw>
This thread was previously completely summarized in July.
In [perl #40030] <http://xrl.us/qvix>, Fonseka reported some missing
dependencies which were preventing compilation and provided a patch. In the
end it turned out to be a problem with an old Parrot installation.
[perl #40032] [CAGE] no previous prototype <http://xrl.us/qviy>
Will Coleda created ticket [perl #40032] <http://xrl.us/qviz> for the Cage
Cleaners. The topic was a warning in `src/encodings/utf8.c`.
[perl #40033] [CAGE] argument type warning <http://xrl.us/qvi2>
In ticket [perl #40033] <http://xrl.us/qvi3>, Will Coleda reported a
warning in function `Parrot_Continuation_invoke`. This was fixed in r13722.
[Proposed PATCH lib/Parrot/Test/Embedded.pm] Use Embedded Interpreter for PIR
Tests <http://xrl.us/qvi4>
chromatic posted a patch which adds a new test module. He wanted feedback
on it. Leopold Toetsch had two questions about chromatic's methods, which
chromatic answered.
Questions <http://xrl.us/qvi5>
chromatic posted several questions that he needs answered before he can add
Parrot::Embed to the repository. Most of the questions involved where it
should be located and how it can be integrated with the rest of the build.
Joshua Hoblitt had some suggestions, which led to a discussion between
chromatic, Joshua and Chris Dolan.
[perl #40039] [BUG] r13762: compilers/bcg/ fails to build with msvc --
unresolved external symbol <http://xrl.us/qvi6>
Jerry Gay reported a bug with r13762 in ticket [perl #40039]
<http://xrl.us/qvi7>. Parrot was successfully built, but `compilers/bcg/`
failed.
[svn:parrot-pdd] r13740 - in trunk: . docs/pdds <http://xrl.us/qvi8>
Allison Randal modified the PDD21. The change was to make namespace opcodes
accept arrays to enable multidimensional namespaces, and to remove
redundant globals and get/set methods. Matt Diephouse asked how the removal
of get/set methods would affect namespaces with a sigil or other unusual
name. Allison replied that the methods were part of the untyped interface;
the typed interface uses `find_namespace` and `add_namespace`.
Initialization and Finalization hooks <http://xrl.us/qvi9>
Kevin Tew discussed having BEGIN and END blocks be supported by Parrot.
Leopold Toetsch replied that he thought END might be currently supported,
and suggested looking at the last few tests in `t/pmc/exception.t` between
revisions r13768 and r13770.
Shared Lib fro Dynamic PMC <http://xrl.us/qvja>
Vishal Soni wanted to know the best location to store a shared library for
the byte code generator. chromatic thought it belonged in either `blib/lib`
or `runtime/parrot/dynext` and wasn't sure which should be used in this
case.
[perl #40048] [CAGE] t/distro/file_metadata.t doesn't like svk
<http://xrl.us/qvjb>
Will Coleda created ticket [perl #40048] <http://xrl.us/qvjc> to report
that `t/distro/file_metadata.t` fails because every test assumes that you
are using svn, but you could be using svk. chromatic submitted a patch.
[perl #40053] Empty PGE rule causes error. rule x { } <http://xrl.us/qvjd>
In ticket [perl #40053] <http://xrl.us/qvje>, Kevin Tew requested that an
empty PGE rule should give a more informative error message.
random thoughts on making parrot compiler faster <http://xrl.us/qvjf>
Joshua Hoblitt noted that much of the build time is spent on
`tools/build/pmc2c.pl` and `tools/build/c2str.pl`. He suggested limiting
this process to maintainers only.
Parrot news: namespace ops good, (some) namespace methods bad; key-accepting
ops <http://xrl.us/qybz>
Chip Salzenberg notified the list that the namespace opcodes are now
preferred over the methods. Allison Randal replied that a similar change
needs to be made for lexicals. The naming convention for these routines was
discussed.
Note: this thread is partly missing from Google's archives and is thus
linked to gmane.
Fix symbol table namespace pollution <http://xrl.us/qvjh>
Chip Salzenberg wanted external functions and variables to have names that
begin with `Parrot`. The request was added to the Cage Cleaners todo list.
There was a brief discussion on storing Parrot tasks in the todo list
versus storing them in RT.
Fix non-symbol-table namespace pollution in public headers
<http://xrl.us/qvji>
Chip Salzenberg noted that public headers in the `include/parrot` directory
should not declare or define any symbol which isn't specific to Parrot.
Symbols should be prefixed with `Parrot` or `PARROT`, although the exact
syntax is up for debate. The request became ticket [perl #40060]
<http://xrl.us/qvjj>.
[perl #40058] Disambiguate usage of class PMCs from class name lookup (pdd15,
pdd06, pdd19) <http://xrl.us/qvjk>
In ticket [perl #40058] <http://xrl.us/qvjm>, Chip Salzenberg stated that
there is ambiguity between class names and class PMCs which should be
resolved in a general way for all class-accepting opcodes. Martin D Kealey
wanted to know why HLL namespaces should be treated differently from
classes. chromatic replied that this was to protect against future
additions to 'parrot' breaking code. Pheme was an example of a language
that would have problems with the situation Martin described.
[perl #40059] [CAGE] Fix symbol table namespace pollution
<http://xrl.us/qvjn>
Chip Salzenberg created ticket [perl #40059] <http://xrl.us/qvjo> to state
that external functions and variables must have names that begin with
'Parrot'.
[perl #40063] [PDD] document subroutine attributes (pdd19_pir.pod)
<http://xrl.us/qvjp>
Ticket [perl #40063] <http://xrl.us/qvjq> requests that subroutine
attributes `:main`, `:immediate`, etc, be documented.
[perl #40064] [TODO] shootout example testing <http://xrl.us/qvjr>
Leopold Toetsch wanted tests added to `examples/shootout/*.pir`. Ticket
[perl #40064] <http://xrl.us/qvjs> tracks this.
[perl #40065] STM first merge tracking ticket <http://xrl.us/qvjt>
Ticket [perl #40065] <http://xrl.us/qvju> was created by Chip Salzenberg to
track the first merge of STM, which should make basic threading work again.
Chip later closed this ticket when the merge was done.
[perl #40066] rewrite ResizeableBooleanArray <http://xrl.us/qvjv>
Ticket [perl #40066] <http://xrl.us/qvjw> states that
`ResizeableBooleanArray` should be rewritten. `FixedBooleanArray` could use
an update as well.
[perl #40069] [PGE] "value" can't be used as a rule name.
<http://xrl.us/qvjx>
Ticket [perl #40069] <http://xrl.us/qvjy> described how a rule or token
named 'value' does not work properly. Patrick R. Michaud knew why the
problem was happening, but was not sure how to resolve it.
[perl #40072] [CAGE] tools/dev/mk_manifest_and_skip.pl requires svn
<http://xrl.us/qvjz>
In ticket [perl #40072] <http://xrl.us/qvj2>, Will Coleda noted that
`tools/dev/mk_manifest_and_skip.pl` uses hardcoded references to svn.
[perl #40087] [BUG] Segfault in key.pmc <http://xrl.us/qvj3>
Will Coleda created ticket [perl #40087] <http://xrl.us/qvj4> to report a
segmentation fault in `key.pmc`.
[perl #40090] [TODO] Tcl - make 'tcl.pbc --pir' work. <http://xrl.us/qvj5>
In ticket [perl #40090] <http://xrl.us/qvj6>, Will Coleda explained how
`make tcl.pbc --pir` should work.
a premature optimization <http://xrl.us/qvj7>
Will Coleda has been converting Tcl's test suite to Tcl, but one of the
larger test files is much slower. Leopold Toetsch and Matt Diephouse
offered suggestions on how to improve the code.
[perl #40102] [BUG] Using splice with an out-of-range array index gives bad
error <http://xrl.us/qvj8>
Matt Diephouse reported in ticket [perl #40102] <http://xrl.us/qvj9> that
the error message for using splice with an out-of-range array is unhelpful.
Leopold Toetsch fixed it with r13890.
[perl #40106] [PATCH] 15 more tests work for PGE::P5Regex
<http://xrl.us/qvka>
Kay-Uwe Huell submitted, via ticket [perl #40106] <http://xrl.us/qvkb>,
fixes to PGE::P5Regex needed to make 15 more tests work.
Failing t//tools/pbc_merge.t <http://xrl.us/qvkc>
chromatic had a test failure. He made a patch to highlight the error, which
he included in the post.
Failing parrot test (t/op/stringu.t) <http://xrl.us/qvkd>
Paul Cochrane reported a test which has been failing across multiple
updates.
parrot compile error <http://xrl.us/qvke>
Hugues reported a parrot compilation issue. Several others reported similar
issues, and Shawn H. Corey tried to help them out. Eventually Leopold
Toetsch asked everyone to post some more specific information, but no
resolution was reached.
[perl #40120] [TODO] parrothist.pod <http://xrl.us/qvkf>
Chip Salzenberg suggested in ticket [perl #40120] <http://xrl.us/qvkg> that
a document describing Parrot evolution be written.
[perl #40121] [TODO] Tcl - implement [lreplace] <http://xrl.us/qvkh>
In ticket [perl #40121] <http://xrl.us/qvki>, Will Coleda requested the
implementation of `lreplace`.
#parrotsketch meeting august 08, 2006 <http://xrl.us/qvkj>
The transcript of the #parrotsketch meeting has been made available
<http://xrl.us/qvkk>.
[perl #40122] [TODO] list_splice should work on any pmc that 'does array'
<http://xrl.us/qvkm>
Will Coleda created ticket [perl #40122] <http://xrl.us/qvkn> to address
code near the `list_splice()` function which prevents using it across
multiple types of PMCs. Leopold Toetsch replied that the implementation is
optimized for Array, and a more general solution can be created by
implementing the splice vtable function in the TclList PMC. Will closed the
ticket and created [perl #40125] [TODO] implement splice for
ResizablePMCArray <http://xrl.us/qvkp>.
[perl #40123] [TODO] push_eh + .param for argument mismatch errors
<http://xrl.us/qvkq>
In ticket [perl #40123] <http://xrl.us/qvkr>, Matt Diephouse wrote that he
would like to be able to use `.param` to catch argument mismatch errors
inside of a subroutine.
[perl #40124] [TODO] Document HLL mappings <http://xrl.us/qvks>
Will Coleda penned ticket [perl #40124] <http://xrl.us/qvkt> because he
believes that the use of HLL mappings should be documented. This would help
HLL authors know when to override, and when a change would have no effect.
[perl #40125] [TODO] implement splice for ResizablePMCArray
<http://xrl.us/qvkp>
Will Coleda created ticket [perl #40125] <http://xrl.us/qvku> to schedule
implementing a generic splice method. This request originated in [perl
#40122] [TODO] list_splice should work on any pmc that 'does array'
<http://xrl.us/qvkm>.
[perl #40132] [TODO] Remove set_pmc vtable method <http://xrl.us/qvkv>
Matt Diephouse noted that `array.pmc` has contained a note indicating that
`set_pmc` is deprecated for over a year now. In ticket [perl #40132]
<http://xrl.us/qvkw>, he requested that it be removed soon.
Parrot 0.4.6 Released! {p2-only announcement} <http://xrl.us/qvkx>
Chip Salzenberg reported the release of Parrot 0.4.6.
[perl #40135] [BUG] Tcl - t_tcl/subst.t failure. <http://xrl.us/qvky>
In ticket [perl #40135] <http://xrl.us/qvkz>, Will Coleda reported a
failing test and explained why the problem happened.
[perl #40138] [TODO] Tcl - add tcl's tests into our repository
<http://xrl.us/qvk2>
Will Coleda created ticket [perl #40138] <http://xrl.us/qvk3>. In it, he
stated that the Tcl test files should be added to the repository to allow
anyone to run the tests.
[perl #40143] [TODO] Tcl - support $tcl_precision <http://xrl.us/qvk4>
In ticket [perl #40143] <http://xrl.us/qvk5>, Matt Diephouse described what
he wanted from Tcl. Will Coleda added some additional instructions.
[perl #40145] [PATCH] t/compilers/imcc/syn/macro.t segfault
<http://xrl.us/qvk6>
Audrey Tang requested a regeneration of a patch. Leopold Toetsch applied it
as r13984. Ticket [perl #40145] <http://xrl.us/qvk7> tracked the request.
http://arstechnica.com/news.ars/post/20060810-7462.html <http://xrl.us/qvk8>
Will Coleda quoted an article <http://xrl.us/qvk9> which spoke of
increasing interest in Parrot among dynamic language enthusiasts.
Review of current stm branch code <http://xrl.us/qvma>
Chip Salzenberg reviewed the progress of Charles Reiss' Google Summer of
Code project to add Software Transactional Memory to parrot. Chip commented
that the work so far has been very good, and that he'd like Charles to
continue contributing to Parrot.
There were many questions and answers provided by both parties, and some C
coding style issues were also raised. Due to the highly technical nature of
this thread, we will not cover it at length.
Grammar question <http://xrl.us/qvmb>
Alberto Sim??es reposted a question of rules from #parrot. Patrick R.
Michaud offered an answer.
PGE grammar test file <http://xrl.us/qvmc>
Nuno Carvalho has been working on tests to evaluate simple PGE grammars.
Tests and other patches were included in the post.
Why consting is good <http://xrl.us/qvmd>
Andy Lester wrote some documentation about why consting is good. He added
it to the Parrot repository as `cage/consting.pod`. Elizabeth Mattijsen
mentioned her module Data::Reuse, which was inspired by Andy's work and
Data::Alias.
The discussion continued in minority report [was RE: Why consting is good]
<http://xrl.us/qvme>.
minority report [was RE: Why consting is good] <http://xrl.us/qvme>
Responding to Why consting is good <http://xrl.us/qvmd>, Tom Horsley
explained why he is not as fond of `const` as Andy Lester is. Andy replied
to some of Tom's points.
Common/Standard POST Implementation <http://xrl.us/qvmf>
Vishal Soni would like to see a standard POST implementation. After looking
at several of the implementations, Vishal came up with something which
could be used by all languages. The patch is attached. Allison Randal and
Patrick R. Michaud were asked to examine it.
Improve Library Path Handling <http://xrl.us/qvmg>
chromatic noted that it is currently possible to change the paths Parrot
uses to load libraries. He included tests which show the problem, and a
patch to fix it. There is still a issue that the `enum_lib_paths` constants
in `include/parrot/library.h` are not available via `iglobals.pasm`.
rt problem <http://xrl.us/qvmi>
Amir E. Aharoni sent a patch but could not find it among the RT tickets.
Amir wondered if the problem was RT or a user error.
Register Allocator <http://xrl.us/qvmj>
Vishal Soni offered some comments on the #parrotsketch IRC log, on the
subject of register allocation. Vishal offered to write a simple register
allocation scheme which will not be optimized but which might be adequate
for the development phase. Comments were requested.
[PROPOSED PATCH] Add Parrot::Embed to Repository <http://xrl.us/qvmk>
chromatic proposed adding Parrot::Embed to the repository, and provided a
patch. Jerry Gay had a question about the patch format. Jerry and chromatic
resolved the problem, and Jerry was able to report that, as chromatic
expected, Windows did not like the patch. Francois Perrad also reported
difficulties. chromatic planned to fix the problem.
[perl #40153] [CAGE] TEST_PROG <http://xrl.us/qvmm>
Will Coleda wanted anything referring to `TEST_PROG` to rename it to
'parrot' or 'executable' depending on context. Ticket [perl #40153]
<http://xrl.us/qvmn> tracks the issue.
[perl #40162] [PATCH] parrot/trunk/docs/art/pp001-intro.pod - spelling
<http://xrl.us/qvmp>
In ticket [perl #40162] <http://xrl.us/qvmq>, Amir E . Aharoni sent a patch
to correct some spelling errors.
[perl #40171] [TODO] - Can't use an Iterator with a DynLexPad PMC
<http://xrl.us/qvmr>
In ticket [perl #40171] <http://xrl.us/qyb2>, Matt Diephouse noted that
iterator and DynLexPad don't work well together. He wondered how to make an
arbitrary hash-like PMC to work with iterator.
[perl #40178] None Must Die <http://xrl.us/qvms>
In ticket [perl #40178] <http://xrl.us/qyb3>, Chip Salzenberg passed a
death sentence on the None class. Leopold Toetsch reported that he was
working on removing it from the public interface of Hash.
Leopold also noted that `$S0 = hsh['no_such_key']` used to return an empty
STRING* but will now return a NULL STRING*. Patrick R. Michaud noted that
this would break many things in PGE and provided some examples. Chip had
some suggestions on how to reduce the pain. Jerry Gay also had some advice.
Later, Patrick reported that there were only five things that would break
with the change.
[perl #40182] [TODO] Convert PGE::CodeString to .pmc <http://xrl.us/qvmt>
Will Coleda created ticket [perl #40182] <http://xrl.us/qyb4> to submit a
request from Patrick R. Michaud to have CodeString written in C instead of
PIR.
[perl #40187] [PATCH] PGE simple grammar test file <http://xrl.us/qvmu>
Nuno Carvalho contributed a patch ([perl #40160] <http://xrl.us/qyb5>)
which includes a rewritten PGE test file, adding some simple grammars.
Patrick R. Michaud applied.
This bug was [reposted][perl #40160] [PATCH] PGE simple grammar test file
<http://xrl.us/qvmo> when the mail seemed to go missing.
[perl #40191] [PATCH] PGE simple grammar test file revisited
<http://xrl.us/qyb6>
In ticket [perl #40191] <http://xrl.us/qyb7>, Nuno Carvalho submitted a
patch to make some changes to `t/compilers/pge/06-grammar.t`, namely to
remove t spacing, reorganize comments, and add an author section.
[REPATCH] Parrot::Embed Take Two <http://xrl.us/qyb8>
chromatic submitted a patch for the build file to try to resolve pkg_config
issues with various platforms. He had some questions about installing from
outside the Parrot tree on Windows. Leopold Toetsch offered a suggestion.
Re: [perl #39868] [CAGE] convert C<BEGIN { plan tests => NN; }> to C<use
Parrot::Test tests => NN;> <http://xrl.us/qyb9>
Nuno Carvalho included a patch found in [perl #39868] <http://xrl.us/qyca>
and noted problems that it caused.
Mono, Bundles, and Ahead-Of-Time Compilation <http://xrl.us/qycb>
chromatic linked to a blog post <http://xrl.us/qycc> by one of the Mono
<http://www.mono-project.com/> developers, which discussed Ahead-Of-Time
Compilation, to speed up startup times, asking what Parrot could take from
this.
Leopold Toetsch replied that Parrot already supports the basic features
needed for this, but that are not currently being tested or used, and would
probably need to be updated.
[perl #40200] t/pmc/threads.t test 16 fails under JIT (parrot -j)
<http://xrl.us/qycd>
In ticket [perl #40200] <http://xrl.us/qyce>, Chip Salzenberg reported that
since the STM merge, all of the `t/pmc/threads.t` succeed, but one of the
tests fails under JIT. Leopold Toetsch clarified that the problem is not
linked to JIT, but happens because two threads change the same global.
Language
===, =:=, ~~, eq and == revisited (blame ajs!) <http://xrl.us/qvmv>
In the thread that wouldn't die, more questions and clarifications on the
semantics of === were raised, especially with respect to arrays, the use of
sigils during assignment, and cloning.
Patch for S05 <http://xrl.us/qvmw>
Agent Zhang sent a patch to S05 which corrects a number of typos, and
thanked people from #perl6 who had helped.
Legacy Dereferencing Syntax Used in S05 <http://xrl.us/qvmx>
Agent Zhang reported that S05 makes use of syntax that isn't found in other
synopses. The question was whether the syntax was legacy. Larry Wall agreed
that it was a leftover, and said he would fix it.
Array Definitions <http://xrl.us/qvmy>
Arthur Ramos Jr. wrote to ask if there would be a change to matrix coding.
Audrey Tang pointed him to S09 and recommended reading the synopses over
the apocalypses and exegesis.
underscores in the core lib <http://xrl.us/qvmz>
Yuval Kogman asked why the naming convention for the core Perl 6 library
disallows underscores. Many people contributed their own opinion,
especially their negative one on the naming of the `.valid` (value id)
method. Larry later clarified his opinion on underscores, and what
constitutes as "core".
Patch for S06 (and also S03) <http://xrl.us/qvm2>
Agent Zhang sent a patch to S06 and S03 with typo corrections but had two
lingering issues: S06 is overly repetitive in explaining the differences
between 'inner' and 'of' return values, and 'unpacking tree node
parameters' needs clarification. In addition, S06 referred to both 'is
assoc('chain')' and 'is assoc('chaining')', which were normalized to
'chain'.
Daniel Hulme had some patches to the patch; Trey Harris took responsibility
for the confusion. There was a discussion on 'pointy subs/blocks' which was
resolved with a reference to Audrey Tang's commit r10478.
weak roles <http://xrl.us/qvm3>
Yuval Kogman proposed the idea of "weak roles". Mark Overmeer and Luke
Palmer replied bringing up issues with the idea. The general consensus
seems to be that delegates are the right way to solve this.
Type annotation on expressions <http://xrl.us/qvm4>
Yuval Kogman asked for a syntax to provide type annotations to expressions,
not just containers, in order to disambiguate MMD, and to provide type
inference hints. Daniel Hulme asked if this is any different than coercion,
and Luke Palmer debated the necessity of such a construct.
naming of the Str type <http://xrl.us/qvm5>
Darren Duncan wrote because he felt that the name of the Str data type is
ambiguous given the traditional use of the word 'string'. Later he updated
his musings to accept that Str is a good enough name for the data type.
David Green felt Darren's reasons made sense, but he did not object to Str.
request addition of immutable Blob type <http://xrl.us/qvm6>
Following a #perl6 discussion on character strings, Darren Duncan proposed
the addition of an immutable and portable arbitrary length Blob type. Yuval
Kogman asked about str (now called 'buf'). Buf is a mutable type.
[svn:perl6-synopsis] r10804 - doc/trunk/design/syn <http://xrl.us/qvm7>
Larry Wall committed a doc patch with the MAIN sub semantics. Luke Palmer
and Audrey Tang discussed issues with the current semantics of MAIN
declared as a method.
Module/Class Authorities <http://xrl.us/qvm8>
Stevan Little wondered if there can be more than one authority for a
module/class. He thought S11 indicated no. Darren Duncan thought perhaps
versions would be used to clarify. Jonathan Scott Duff believed it was a
social issue, and explained why he thought single authorities were the way
to go.
Relative running order between traits blocks <http://xrl.us/qvna>
Agent Zhang asked for clarification because S04 does not explain the
running order of FIRST, ENTER, NEXT, LEAVE and LAST. Tests for Pugs, which
should be verified, were linked to.
Apply the unspace rule into embedded comments <http://xrl.us/qvnb>
Agent Zhang wondered if the unspace rule could be used with the embedded
comment syntax.
Curious corner cases concerning closure cloning <http://xrl.us/qvnc>
Luke Palmer included some code with closures and wanted to know what they
did. Yuval Kogman offered some answers based upon his intuition. Jonathan
Scott Duff also offered his opinion.
Outlaw the "ddd" notation <http://xrl.us/qvnd>
Agent Zhang wanted to state in S02 that Perl 6's `o` and `o[]` replace Perl
5's `ddd`.
do EXPR <http://xrl.us/qvne>
Agent Zhang submitted a patch for S04 which came from a snippet in the Pugs
test suite.
Dumb doc question... <http://xrl.us/qvnf>
Mark J. Reed wanted to know where he could find a pod2html converter that
understands the Perl 6 POD. Agent Zhang replied that pod2html can handle
the synopses, except for the '=encoding' directive. Thom Boyer suggested
looking at the online HTML version of the specifications
<http://xrl.us/qycf> but noted that S29 is not included except by reference
to the SVN repository.
NEXT and the general loop statement <http://xrl.us/qvng>
Joe Gottman asked if NEXT was called before or after the update portion of
a loop statement. Larry Wall replied that a three argument loop is
implicitly using NEXT, and the block exiting blocks run in reverse order.
This led to further discussion also involving Luke Palmer and Jonathan
Scott Duff.
typo fix: trinary -> ternary <http://xrl.us/qvnh>
Mark Stosberg suggested a typo fix in `Perl6/Spec/Operator.pod`.
Random grammar typo in S02 <http://xrl.us/qvni>
Mark Reed reported a typo in S02.
String length <http://xrl.us/qvnj>
Mark Reed wondered why there is no `.length` method on Str, when StrPos and
StrLen types are defined. He proposed a `.length` method which would return
a StrLen.
Numerification of Order:: constants <http://xrl.us/qvnk>
Mark Reed quoted a section of S03 and asked if Order::Increase should
numerify to +1 and Order::Decrease to -1. Mark A. Biggar explained that `$a
cmp $b` can be seen as a sequence `($a, $b)`--if the sequence is
increasing, `$a cmp $b` returns Order::Increase or -1.
David Green pondered the question. Larry Wall noted that `cmp` does not
mean `eq`. He also expressed his feeling that 'but True' and 'but False'
are workarounds, and not many core interfaces will use solutions involving
'but'.
Invocant name in S12 <http://xrl.us/qycg>
Mark J. Reed noted that S12 never states the default invocant name. He
wondered if the term 'self' was introduced somewhere.
[svn:perl6-synopsis] r11115 - doc/trunk/design/syn <http://xrl.us/qych>
In response to a commit by Larry Wall, Markus Laire asked if switch `-abc`
would mean `-a -b -c`. Larry discussed some of his concerns but thought it
should probably be possible. Markus Laire understood Larry's point, and
further elaborated on his proposal.
multi-line comments, C macros, & Pod abuse <http://xrl.us/qyci>
Joshua Hoblitt asked how the `#if 0` C preprocessor idiom could be
converted to Perl. He asked if there was going to be a better approach than
using POD hacks. Some of the suggested approaches were to use multi-line
block comments, constant folded if statements, and quote operations with
balanced, arbitrary delimiters.
Acknowlegements
Yuval Kogman once again contributed summaries for some of the threads.
This summary was prepared using Mail::Summary::Tools <http://xrl.us/qvno>,
now available on CPAN.
If you appreciate Perl, consider [contributing to the Perl
Foundation](http://donate.perlfoundation.org) to help support the
development of Perl.
Thank you to everyone who has pointed out mistakes and offered suggestions
for improving this series. Comments on this summary can be sent to Ann
Barcomb, <[EMAIL PROTECTED]>.
Distribution
This summary can be found in the following places:
* use.perl.org <http://use.perl.org/>
* The pugs blog <http://pugs.blogs.com/>
* The perl6-announce mailing list <http://xrl.us/qycj>
See Also
* Perl Foundation activities <http://blog.perlfoundation.org>
* Perl 6 Development <http://dev.perl.org/perl6>
* Planet Perl Six <http://planetsix.perl.org/perl6>
