This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
inlining autoloaded constants is TODOne.
[perl5.git] / pod / perltodo.pod
CommitLineData
7711098a
GS
1=head1 NAME
2
3perltodo - Perl TO-DO List
4
5=head1 DESCRIPTION
e50bb9a1 6
0bdfc961
NC
7This is a list of wishes for Perl. The tasks we think are smaller or easier
8are listed first. Anyone is welcome to work on any of these, but it's a good
9idea to first contact I<perl5-porters@perl.org> to avoid duplication of
10effort. By all means contact a pumpking privately first if you prefer.
e50bb9a1 11
0bdfc961
NC
12Whilst patches to make the list shorter are most welcome, ideas to add to
13the list are also encouraged. Check the perl5-porters archives for past
14ideas, and any discussion about them. One set of archives may be found at:
e50bb9a1 15
0bdfc961 16 http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/
938c8732 17
617eabfa
NC
18What can we offer you in return? Fame, fortune, and everlasting glory? Maybe
19not, but if your patch is incorporated, then we'll add your name to the
20F<AUTHORS> file, which ships in the official distribution. How many other
21programming languages offer you 1 line of immortality?
938c8732 22
4e577f8b 23=head1 The roadmap to 5.10
938c8732 24
4e577f8b
NC
25The roadmap to 5.10 envisages feature based releases, as various items in this
26TODO are completed.
27
28=head2 Needed for a 5.9.3 release
29
30=over
31
32=item *
33Implement L</lexical pragmas>
34
35=back
36
37=head2 Needed for a 5.9.4 release
38
39=over
40
41=item *
42Review assertions. Review syntax to combine assertions. Can assertions take
43advantage of the lexical pragams work? L</What hooks would assertions need?>
44
45=back
46
47=head2 Needed for a 5.9.5 release
48
49=over
50
51=item *
52Implement L</_ prototype character>
53
54=item *
55Implement L</state variables>
56
57=back
58
59=head2 Needed for a 5.9.6 release
60
61Stabilisation. If all goes well, this will be the equivalent of a 5.10-beta.
e50bb9a1 62
0bdfc961 63=head1 Tasks that only need Perl knowledge
e50bb9a1 64
0bdfc961 65=head2 common test code for timed bail out
e50bb9a1 66
0bdfc961
NC
67Write portable self destruct code for tests to stop them burning CPU in
68infinite loops. This needs to avoid using alarm, as some of the tests are
69testing alarm/sleep or timers.
e50bb9a1 70
0bdfc961 71=head2 POD -> HTML conversion in the core still sucks
e50bb9a1 72
938c8732 73Which is crazy given just how simple POD purports to be, and how simple HTML
adebf063
NC
74can be. It's not actually I<as> simple as it sounds, particularly with the
75flexibility POD allows for C<=item>, but it would be good to improve the
76visual appeal of the HTML generated, and to avoid it having any validation
77errors. See also L</make HTML install work>, as the layout of installation tree
78is needed to improve the cross-linking.
938c8732 79
dc0fb092
SP
80The addition of C<Pod::Simple> and its related modules may make this task
81easier to complete.
82
aa237293
NC
83=head2 Parallel testing
84
85The core regression test suite is getting ever more comprehensive, which has
86the side effect that it takes longer to run. This isn't so good. Investigate
87whether it would be feasible to give the harness script the B<option> of
88running sets of tests in parallel. This would be useful for tests in
89F<t/op/*.t> and F<t/uni/*.t> and maybe some sets of tests in F<lib/>.
90
91Questions to answer
92
93=over 4
94
95=item 1
96
97How does screen layout work when you're running more than one test?
98
99=item 2
100
101How does the caller of test specify how many tests to run in parallel?
102
103=item 3
104
105How do setup/teardown tests identify themselves?
106
107=back
108
109Pugs already does parallel testing - can their approach be re-used?
110
0bdfc961 111=head2 Make Schwern poorer
e50bb9a1 112
0bdfc961
NC
113We should have for everything. When all the core's modules are tested,
114Schwern has promised to donate to $500 to TPF. We may need volunteers to
115hold him upside down and shake vigorously in order to actually extract the
116cash.
3958b146 117
0bdfc961 118See F<t/lib/1_compile.t> for the 3 remaining modules that need tests.
e50bb9a1 119
0bdfc961 120=head2 Improve the coverage of the core tests
e50bb9a1 121
0bdfc961
NC
122Use Devel::Cover to ascertain the core's test coverage, then add tests that
123are currently missing.
30222c0f 124
0bdfc961 125=head2 test B
e50bb9a1 126
0bdfc961 127A full test suite for the B module would be nice.
e50bb9a1 128
0bdfc961 129=head2 A decent benchmark
e50bb9a1 130
617eabfa 131C<perlbench> seems impervious to any recent changes made to the perl core. It
0bdfc961
NC
132would be useful to have a reasonable general benchmarking suite that roughly
133represented what current perl programs do, and measurably reported whether
134tweaks to the core improve, degrade or don't really affect performance, to
135guide people attempting to optimise the guts of perl. Gisle would welcome
136new tests for perlbench.
6168cf99 137
0bdfc961 138=head2 fix tainting bugs
6168cf99 139
0bdfc961
NC
140Fix the bugs revealed by running the test suite with the C<-t> switch (via
141C<make test.taintwarn>).
e50bb9a1 142
0bdfc961 143=head2 Dual life everything
e50bb9a1 144
0bdfc961
NC
145As part of the "dists" plan, anything that doesn't belong in the smallest perl
146distribution needs to be dual lifed. Anything else can be too. Figure out what
147changes would be needed to package that module and its tests up for CPAN, and
148do so. Test it with older perl releases, and fix the problems you find.
e50bb9a1 149
0bdfc961 150=head2 Improving C<threads::shared>
722d2a37 151
0bdfc961
NC
152Investigate whether C<threads::shared> could share aggregates properly with
153only Perl level changes to shared.pm
722d2a37 154
0bdfc961 155=head2 POSIX memory footprint
e50bb9a1 156
0bdfc961
NC
157Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
158various times worked to cut it down. There is probably still fat to cut out -
159for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 160
e50bb9a1 161
e50bb9a1 162
e50bb9a1 163
e50bb9a1 164
adebf063 165
adebf063 166
0bdfc961 167=head1 Tasks that need a little sysadmin-type knowledge
e50bb9a1 168
0bdfc961
NC
169Or if you prefer, tasks that you would learn from, and broaden your skills
170base...
e50bb9a1 171
617eabfa
NC
172=head2 Relocatable perl
173
174The C level patches needed to create a relocatable perl binary are done, as
175is the work on F<Config.pm>. All that's left to do is the C<Configure> tweaking
176to let people specify how they want to do the install.
177
cd793d32 178=head2 make HTML install work
e50bb9a1 179
adebf063
NC
180There is an C<installhtml> target in the Makefile. It's marked as
181"experimental". It would be good to get this tested, make it work reliably, and
182remove the "experimental" tag. This would include
183
184=over 4
185
186=item 1
187
188Checking that cross linking between various parts of the documentation works.
189In particular that links work between the modules (files with POD in F<lib/>)
190and the core documentation (files in F<pod/>)
191
192=item 2
193
617eabfa
NC
194Work out how to split C<perlfunc> into chunks, preferably one per function
195group, preferably with general case code that could be used elsewhere.
196Challenges here are correctly identifying the groups of functions that go
197together, and making the right named external cross-links point to the right
198page. Things to be aware of are C<-X>, groups such as C<getpwnam> to
199C<endservent>, two or more C<=items> giving the different parameter lists, such
200as
adebf063
NC
201
202 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
203
204 =item substr EXPR,OFFSET,LENGTH
205
206 =item substr EXPR,OFFSET
207
208and different parameter lists having different meanings. (eg C<select>)
209
210=back
3a89a73c 211
0bdfc961
NC
212=head2 compressed man pages
213
214Be able to install them. This would probably need a configure test to see how
215the system does compressed man pages (same directory/different directory?
216same filename/different filename), as well as tweaking the F<installman> script
217to compress as necessary.
218
30222c0f
NC
219=head2 Add a code coverage target to the Makefile
220
221Make it easy for anyone to run Devel::Cover on the core's tests. The steps
222to do this manually are roughly
223
224=over 4
225
226=item *
227
228do a normal C<Configure>, but include Devel::Cover as a module to install
229(see F<INSTALL> for how to do this)
230
231=item *
232
233 make perl
234
235=item *
236
237 cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness
238
239=item *
240
241Process the resulting Devel::Cover database
242
243=back
244
245This just give you the coverage of the F<.pm>s. To also get the C level
246coverage you need to
247
248=over 4
249
250=item *
251
252Additionally tell C<Configure> to use the appropriate C compiler flags for
253C<gcov>
254
255=item *
256
257 make perl.gcov
258
259(instead of C<make perl>)
260
261=item *
262
263After running the tests run C<gcov> to generate all the F<.gcov> files.
264(Including down in the subdirectories of F<ext/>
265
266=item *
267
268(From the top level perl directory) run C<gcov2perl> on all the C<.gcov> files
269to get their stats into the cover_db directory.
270
271=item *
272
273Then process the Devel::Cover database
274
275=back
276
277It would be good to add a single switch to C<Configure> to specify that you
278wanted to perform perl level coverage, and another to specify C level
279coverage, and have C<Configure> and the F<Makefile> do all the right things
280automatically.
281
0bdfc961
NC
282=head2 Make Config.pm cope with differences between build and installed perl
283
284Quite often vendors ship a perl binary compiled with their (pay-for)
285compilers. People install a free compiler, such as gcc. To work out how to
286build extensions, Perl interrogates C<%Config>, so in this situation
287C<%Config> describes compilers that aren't there, and extension building
288fails. This forces people into choosing between re-compiling perl themselves
289using the compiler they have, or only using modules that the vendor ships.
290
291It would be good to find a way teach C<Config.pm> about the installation setup,
292possibly involving probing at install time or later, so that the C<%Config> in
293a binary distribution better describes the installed machine, when the
294installed machine differs from the build machine in some significant way.
295
46925299 296=head2 make parallel builds work
0bdfc961 297
46925299
NC
298Currently parallel builds (such as C<make -j3>) don't work reliably. We believe
299that this is due to incomplete dependency specification in the F<Makefile>.
300It would be good if someone were able to track down the causes of these
301problems, so that parallel builds worked properly.
0bdfc961 302
728f4ecd
NC
303=head2 linker specification files
304
305Some platforms mandate that you provide a list of a shared library's external
306symbols to the linker, so the core already has the infrastructure in place to
307do this for generating shared perl libraries. My understanding is that the
308GNU toolchain can accept an optional linker specification file, and restrict
309visibility just to symbols declared in that file. It would be good to extend
310F<makedef.pl> to support this format, and to provide a means within
311C<Configure> to enable it. This would allow Unix users to test that the
312export list is correct, and to build a perl that does not pollute the global
313namespace with private symbols.
314
8523e164 315
0bdfc961
NC
316
317
318=head1 Tasks that need a little C knowledge
319
320These tasks would need a little C knowledge, but don't need any specific
321background or experience with XS, or how the Perl interpreter works
322
323=head2 Make it clear from -v if this is the exact official release
89007cb3 324
617eabfa
NC
325Currently perl from C<p4>/C<rsync> ships with a F<patchlevel.h> file that
326usually defines one local patch, of the form "MAINT12345" or "RC1". The output
327of perl -v doesn't report that a perl isn't an official release, and this
89007cb3 328information can get lost in bugs reports. Because of this, the minor version
fa11829f 329isn't bumped up until RC time, to minimise the possibility of versions of perl
89007cb3
NC
330escaping that believe themselves to be newer than they actually are.
331
332It would be useful to find an elegant way to have the "this is an interim
333maintenance release" or "this is a release candidate" in the terse -v output,
334and have it so that it's easy for the pumpking to remove this just as the
335release tarball is rolled up. This way the version pulled out of rsync would
336always say "I'm a development release" and it would be safe to bump the
337reported minor version as soon as a release ships, which would aid perl
338developers.
339
0bdfc961
NC
340This task is really about thinking of an elegant way to arrange the C source
341such that it's trivial for the Pumpking to flag "this is an official release"
342when making a tarball, yet leave the default source saying "I'm not the
343official release".
344
6d4cb3f4
NC
345=head2 Tidy up global variables
346
347There's a note in F<intrpvar.h>
348
349 /* These two variables are needed to preserve 5.8.x bincompat because
350 we can't change function prototypes of two exported functions.
351 Probably should be taken out of blead soon, and relevant prototypes
352 changed. */
353
354So doing this, and removing any of the unused variables still present would
355be good.
356
0f788cd2
NC
357=head2 Ordering of "global" variables.
358
359F<thrdvar.h> and F<intrpvarh> define the "global" variables that need to be
360per-thread under ithreads, where the variables are actually elements in a
361structure. As C dictates, the variables must be laid out in order of
362declaration. There is a comment
363C</* Important ones in the first cache line (if alignment is done right) */>
364which implies that at some point in the past the ordering was carefully chosen
365(at least in part). However, it's clear that the ordering is less than perfect,
366as currently there are things such as 7 C<bool>s in a row, then something
367typically requiring 4 byte alignment, and then an odd C<bool> later on.
368(C<bool>s are typically defined as C<char>s). So it would be good for someone
369to review the ordering of the variables, to see how much alignment padding can
370be removed.
371
0bdfc961
NC
372=head2 bincompat functions
373
374There are lots of functions which are retained for binary compatibility.
375Clean these up. Move them to mathom.c, and don't compile for blead?
376
62403a3c
NC
377=head2 am I hot or not?
378
379The idea of F<pp_hot.c> is that it contains the I<hot> ops, the ops that are
380most commonly used. The idea is that by grouping them, their object code will
381be adjacent in the executable, so they have a greater chance of already being
382in the CPU cache (or swapped in) due to being near another op already in use.
383
384Except that it's not clear if these really are the most commonly used ops. So
385anyone feeling like exercising their skill with coverage and profiling tools
386might want to determine what ops I<really> are the most commonly used. And in
387turn suggest evictions and promotions to achieve a better F<pp_hot.c>.
388
c99e3826
NC
389=head2 emulate the per-thread memory pool on Unix
390
391For Windows, ithreads allocates memory for each thread from a separate pool,
392which it discards at thread exit. It also checks that memory is free()d to
393the correct pool. Neither check is done on Unix, so code developed there won't
394be subject to such strictures, so can harbour bugs that only show up when the
395code reaches Windows.
396
397It would be good to be able to optionally emulate the Window pool system on
398Unix, to let developers who only have access to Unix, or want to use
399Unix-specific debugging tools, check for these problems. To do this would
400involve figuring out how the C<PerlMem_*> macros wrap C<malloc()> access, and
401providing a layer that records/checks the identity of the thread making the
402call, and recording all the memory allocated by each thread via this API so
403that it can be summarily free()d at thread exit. One implementation idea
404would be to increase the size of allocation, and store the C<my_perl> pointer
405(to identify the thread) at the start, along with pointers to make a linked
406list of blocks for this thread. To avoid alignment problems it would be
407necessary to do something like
408
409 union memory_header_padded {
410 struct memory_header {
411 void *thread_id; /* For my_perl */
412 void *next; /* Pointer to next block for this thread */
413 } data;
414 long double padding; /* whatever type has maximal alignment constraint */
415 };
416
417
418although C<long double> might not be the only type to add to the padding
419union.
62403a3c 420
077e3186
NC
421=head2 reduce duplication in sv_setsv_flags
422
423C<Perl_sv_setsv_flags> has a comment
424C</* There's a lot of redundancy below but we're going for speed here */>
425
426Whilst this was true 10 years ago, the growing disparity between RAM and CPU
427speeds mean that the trade offs have changed. In addition, the duplicate code
428adds to the maintenance burden. It would be good to see how much of the
429redundancy can be pruned, particular in the less common paths. (Profiling
430tools at the ready...). For example, why does the test for
431"Can't redefine active sort subroutine" need to occur in two places?
432
0bdfc961
NC
433
434
435
0bdfc961 436=head1 Tasks that need a knowledge of XS
e50bb9a1 437
0bdfc961
NC
438These tasks would need C knowledge, and roughly the level of knowledge of
439the perl API that comes from writing modules that use XS to interface to
440C.
441
442=head2 IPv6
443
444Clean this up. Check everything in core works
445
4a750395
NC
446=head2 shrink C<GV>s, C<CV>s
447
448By removing unused elements and careful re-ordering, the structures for C<AV>s
449and C<HV>s have recently been shrunk considerably. It's probable that the same
450approach would find savings in C<GV>s and C<CV>s, if not all the other
451larger-than-C<PVMG> types.
452
e593da1a
NC
453=head2 merge Perl_sv_2[inpu]v
454
455There's a lot of code shared between C<Perl_sv_2iv_flags>,
456C<Perl_sv_2uv_flags>, C<Perl_sv_2nv>, and C<Perl_sv_2pv_flags>. It would be
457interesting to see if some of it can be merged into common shared static
458functions. In particular, C<Perl_sv_2uv_flags> started out as a cut&paste
459from C<Perl_sv_2iv_flags> around 5.005_50 time, and it may be possible to
460replace both with a single function that returns a value or union which is
461split out by the macros in F<sv.h>
462
0bdfc961
NC
463=head2 UTF8 caching code
464
465The string position/offset cache is not optional. It should be.
466
467=head2 Implicit Latin 1 => Unicode translation
468
469Conversions from byte strings to UTF-8 currently map high bit characters
470to Unicode without translation (or, depending on how you look at it, by
471implicitly assuming that the byte strings are in Latin-1). As perl assumes
472the C locale by default, upgrading a string to UTF-8 may change the
473meaning of its contents regarding character classes, case mapping, etc.
474This should probably emit a warning (at least).
475
476This task is incremental - even a little bit of work on it will help.
e50bb9a1 477
cd793d32 478=head2 autovivification
e50bb9a1 479
cd793d32 480Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
e50bb9a1 481
0bdfc961 482This task is incremental - even a little bit of work on it will help.
e50bb9a1 483
0bdfc961 484=head2 Unicode in Filenames
e50bb9a1 485
0bdfc961
NC
486chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
487opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
488system, truncate, unlink, utime, -X. All these could potentially accept
489Unicode filenames either as input or output (and in the case of system
490and qx Unicode in general, as input or output to/from the shell).
491Whether a filesystem - an operating system pair understands Unicode in
492filenames varies.
e50bb9a1 493
0bdfc961
NC
494Known combinations that have some level of understanding include
495Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
496OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
497create Unicode filenames, what forms of Unicode are accepted and used
498(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
499and so on, varies. Finding the right level of interfacing to Perl
500requires some thought. Remember that an OS does not implicate a
501filesystem.
e50bb9a1 502
0bdfc961
NC
503(The Windows -C command flag "wide API support" has been at least
504temporarily retired in 5.8.1, and the -C has been repurposed, see
505L<perlrun>.)
969e704b 506
0bdfc961 507=head2 Unicode in %ENV
969e704b 508
0bdfc961 509Currently the %ENV entries are always byte strings.
e50bb9a1 510
0bdfc961 511=head2 use less 'memory'
e50bb9a1 512
0bdfc961
NC
513Investigate trade offs to switch out perl's choices on memory usage.
514Particularly perl should be able to give memory back.
e50bb9a1 515
0bdfc961 516This task is incremental - even a little bit of work on it will help.
0abe3f7c 517
0bdfc961 518=head2 Re-implement C<:unique> in a way that is actually thread-safe
0abe3f7c 519
0bdfc961
NC
520The old implementation made bad assumptions on several levels. A good 90%
521solution might be just to make C<:unique> work to share the string buffer
522of SvPVs. That way large constant strings can be shared between ithreads,
523such as the configuration information in F<Config>.
0abe3f7c 524
0bdfc961 525=head2 Make tainting consistent
0abe3f7c 526
0bdfc961
NC
527Tainting would be easier to use if it didn't take documented shortcuts and
528allow taint to "leak" everywhere within an expression.
0abe3f7c 529
0bdfc961 530=head2 readpipe(LIST)
0abe3f7c 531
0bdfc961
NC
532system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
533running a shell. readpipe() (the function behind qx//) could be similarly
534extended.
0abe3f7c 535
e50bb9a1 536
e50bb9a1 537
e50bb9a1 538
f86a8bc5 539
0bdfc961 540=head1 Tasks that need a knowledge of the interpreter
3298bd4d 541
0bdfc961
NC
542These tasks would need C knowledge, and knowledge of how the interpreter works,
543or a willingness to learn.
3298bd4d 544
0bdfc961
NC
545=head2 lexical pragmas
546
547Reimplement the mechanism of lexical pragmas to be more extensible. Fix
548current pragmas that don't work well (or at all) with lexical scopes or in
549run-time eval(STRING) (C<sort>, C<re>, C<encoding> for example). MJD has a
550preliminary patch that implements this.
0562c0e3 551
d10fc472 552=head2 Attach/detach debugger from running program
1626a787 553
cd793d32
NC
554The old perltodo notes "With C<gdb>, you can attach the debugger to a running
555program if you pass the process ID. It would be good to do this with the Perl
0bdfc961
NC
556debugger on a running Perl program, although I'm not sure how it would be
557done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
1626a787 558
0bdfc961 559=head2 Constant folding
80b46460 560
0bdfc961
NC
561The peephole optimiser should trap errors during constant folding, and give
562up on the folding, rather than bailing out at compile time. It is quite
563possible that the unfoldable constant is in unreachable code, eg something
564akin to C<$a = 0/0 if 0;>
565
566=head2 LVALUE functions for lists
567
568The old perltodo notes that lvalue functions don't work for list or hash
569slices. This would be good to fix.
570
571=head2 LVALUE functions in the debugger
572
573The old perltodo notes that lvalue functions don't work in the debugger. This
574would be good to fix.
575
576=head2 _ prototype character
577
578Study the possibility of adding a new prototype character, C<_>, meaning
579"this argument defaults to $_".
580
4e577f8b
NC
581=head2 state variables
582
583C<my $foo if 0;> is deprecated, and should be replaced with
584C<state $x = "initial value\n";> the syntax from Perl 6.
585
0bdfc961
NC
586=head2 @INC source filter to Filter::Simple
587
588The second return value from a sub in @INC can be a source filter. This isn't
589documented. It should be changed to use Filter::Simple, tested and documented.
590
591=head2 regexp optimiser optional
592
593The regexp optimiser is not optional. It should configurable to be, to allow
594its performance to be measured, and its bugs to be easily demonstrated.
595
596=head2 UNITCHECK
597
598Introduce a new special block, UNITCHECK, which is run at the end of a
599compilation unit (module, file, eval(STRING) block). This will correspond to
600the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the
601O.pm/B.pm backend framework depends on it.
602
603=head2 optional optimizer
604
605Make the peephole optimizer optional. Currently it performs two tasks as
606it walks the optree - genuine peephole optimisations, and necessary fixups of
607ops. It would be good to find an efficient way to switch out the
608optimisations whilst keeping the fixups.
609
610=head2 You WANT *how* many
611
612Currently contexts are void, scalar and list. split has a special mechanism in
613place to pass in the number of return values wanted. It would be useful to
614have a general mechanism for this, backwards compatible and little speed hit.
615This would allow proposals such as short circuiting sort to be implemented
616as a module on CPAN.
617
618=head2 lexical aliases
619
620Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
621
622=head2 entersub XS vs Perl
623
624At the moment pp_entersub is huge, and has code to deal with entering both
625perl and XS subroutines. Subroutine implementations rarely change between
626perl and XS at run time, so investigate using 2 ops to enter subs (one for
627XS, one for perl) and swap between if a sub is redefined.
2810d901
NC
628
629=head2 Self ties
630
631self ties are currently illegal because they caused too many segfaults. Maybe
632the causes of these could be tracked down and self-ties on all types re-
633instated.
0bdfc961
NC
634
635=head2 Optimize away @_
636
637The old perltodo notes "Look at the "reification" code in C<av.c>".
638
0bdfc961
NC
639=head2 What hooks would assertions need?
640
641Assertions are in the core, and work. However, assertions needed to be added
642as a core patch, rather than an XS module in ext, or a CPAN module, because
643the core has no hooks in the necessary places. It would be useful to
644investigate what hooks would need to be added to make it possible to provide
645the full assertion support from a CPAN module, so that we aren't constraining
646the imagination of future CPAN authors.
647
648
649
650
651
0bdfc961
NC
652=head1 Big projects
653
654Tasks that will get your name mentioned in the description of the "Highlights
655of 5.10"
656
657=head2 make ithreads more robust
658
4e577f8b 659Generally make ithreads more robust. See also L</iCOW>
0bdfc961
NC
660
661This task is incremental - even a little bit of work on it will help, and
662will be greatly appreciated.
663
664=head2 iCOW
665
666Sarathy and Arthur have a proposal for an improved Copy On Write which
667specifically will be able to COW new ithreads. If this can be implemented
668it would be a good thing.
669
670=head2 (?{...}) closures in regexps
671
672Fix (or rewrite) the implementation of the C</(?{...})/> closures.
673
674=head2 A re-entrant regexp engine
675
676This will allow the use of a regex from inside (?{ }), (??{ }) and
677(?(?{ })|) constructs.