This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Re-order the TODO tasks based on the skills they need, putting the
[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
938c8732 18
938c8732 19
938c8732 20
e50bb9a1 21
0bdfc961 22=head1 Tasks that only need Perl knowledge
e50bb9a1 23
0bdfc961 24=head2 common test code for timed bail out
e50bb9a1 25
0bdfc961
NC
26Write portable self destruct code for tests to stop them burning CPU in
27infinite loops. This needs to avoid using alarm, as some of the tests are
28testing alarm/sleep or timers.
e50bb9a1 29
0bdfc961 30=head2 POD -> HTML conversion in the core still sucks
e50bb9a1 31
938c8732 32Which is crazy given just how simple POD purports to be, and how simple HTML
adebf063
NC
33can be. It's not actually I<as> simple as it sounds, particularly with the
34flexibility POD allows for C<=item>, but it would be good to improve the
35visual appeal of the HTML generated, and to avoid it having any validation
36errors. See also L</make HTML install work>, as the layout of installation tree
37is needed to improve the cross-linking.
938c8732 38
0bdfc961 39=head2 Make Schwern poorer
e50bb9a1 40
0bdfc961
NC
41We should have for everything. When all the core's modules are tested,
42Schwern has promised to donate to $500 to TPF. We may need volunteers to
43hold him upside down and shake vigorously in order to actually extract the
44cash.
3958b146 45
0bdfc961 46See F<t/lib/1_compile.t> for the 3 remaining modules that need tests.
e50bb9a1 47
0bdfc961 48=head2 Improve the coverage of the core tests
e50bb9a1 49
0bdfc961
NC
50Use Devel::Cover to ascertain the core's test coverage, then add tests that
51are currently missing.
52=head2 test B
e50bb9a1 53
0bdfc961 54A full test suite for the B module would be nice.
e50bb9a1 55
0bdfc961 56=head2 A decent benchmark
e50bb9a1 57
0bdfc961
NC
58perlbench seems impervious to any recent changes made to the perl core. It
59would be useful to have a reasonable general benchmarking suite that roughly
60represented what current perl programs do, and measurably reported whether
61tweaks to the core improve, degrade or don't really affect performance, to
62guide people attempting to optimise the guts of perl. Gisle would welcome
63new tests for perlbench.
6168cf99 64
0bdfc961 65=head2 fix tainting bugs
6168cf99 66
0bdfc961
NC
67Fix the bugs revealed by running the test suite with the C<-t> switch (via
68C<make test.taintwarn>).
e50bb9a1 69
0bdfc961 70=head2 Dual life everything
e50bb9a1 71
0bdfc961
NC
72As part of the "dists" plan, anything that doesn't belong in the smallest perl
73distribution needs to be dual lifed. Anything else can be too. Figure out what
74changes would be needed to package that module and its tests up for CPAN, and
75do so. Test it with older perl releases, and fix the problems you find.
e50bb9a1 76
0bdfc961 77=head2 Improving C<threads::shared>
722d2a37 78
0bdfc961
NC
79Investigate whether C<threads::shared> could share aggregates properly with
80only Perl level changes to shared.pm
722d2a37 81
0bdfc961 82=head2 POSIX memory footprint
e50bb9a1 83
0bdfc961
NC
84Ilya observed that use POSIX; eats memory like there's no tomorrow, and at
85various times worked to cut it down. There is probably still fat to cut out -
86for example POSIX passes Exporter some very memory hungry data structures.
e50bb9a1 87
e50bb9a1 88
e50bb9a1 89
e50bb9a1 90
e50bb9a1 91
adebf063 92
adebf063 93
0bdfc961 94=head1 Tasks that need a little sysadmin-type knowledge
e50bb9a1 95
0bdfc961
NC
96Or if you prefer, tasks that you would learn from, and broaden your skills
97base...
e50bb9a1 98
cd793d32 99=head2 make HTML install work
e50bb9a1 100
adebf063
NC
101There is an C<installhtml> target in the Makefile. It's marked as
102"experimental". It would be good to get this tested, make it work reliably, and
103remove the "experimental" tag. This would include
104
105=over 4
106
107=item 1
108
109Checking that cross linking between various parts of the documentation works.
110In particular that links work between the modules (files with POD in F<lib/>)
111and the core documentation (files in F<pod/>)
112
113=item 2
114
115Work out how to split perlfunc into chunks, preferably one per function group,
116preferably with general case code that could be used elsewhere. Challenges
117here are correctly identifying the groups of functions that go together, and
118making the right named external cross-links point to the right page. Things to
119be aware of are C<-X>, groups such as C<getpwnam> to C<endservent>, two or
120more C<=items> giving the different parameter lists, such as
121
122 =item substr EXPR,OFFSET,LENGTH,REPLACEMENT
123
124 =item substr EXPR,OFFSET,LENGTH
125
126 =item substr EXPR,OFFSET
127
128and different parameter lists having different meanings. (eg C<select>)
129
130=back
3a89a73c 131
0bdfc961
NC
132=head2 compressed man pages
133
134Be able to install them. This would probably need a configure test to see how
135the system does compressed man pages (same directory/different directory?
136same filename/different filename), as well as tweaking the F<installman> script
137to compress as necessary.
138
139=head2 Make Config.pm cope with differences between build and installed perl
140
141Quite often vendors ship a perl binary compiled with their (pay-for)
142compilers. People install a free compiler, such as gcc. To work out how to
143build extensions, Perl interrogates C<%Config>, so in this situation
144C<%Config> describes compilers that aren't there, and extension building
145fails. This forces people into choosing between re-compiling perl themselves
146using the compiler they have, or only using modules that the vendor ships.
147
148It would be good to find a way teach C<Config.pm> about the installation setup,
149possibly involving probing at install time or later, so that the C<%Config> in
150a binary distribution better describes the installed machine, when the
151installed machine differs from the build machine in some significant way.
152
153=head2 Relocatable perl
154
155The C level patches needed to create a relocatable perl binary are done, as
156is the work on Config.pm. All that's left to do is the C<Configure> tweaking
157to let people specify how they want to do the install.
158
159
160
161
162
163=head1 Tasks that need a little C knowledge
164
165These tasks would need a little C knowledge, but don't need any specific
166background or experience with XS, or how the Perl interpreter works
167
168=head2 Make it clear from -v if this is the exact official release
89007cb3
NC
169
170Currently perl from p4/rsync ships with a patchlevel.h file that usually
171defines one local patch, of the form "MAINT12345" or "RC1". The output of
172perl -v doesn't report that a perl isn't an official release, and this
173information can get lost in bugs reports. Because of this, the minor version
fa11829f 174isn't bumped up until RC time, to minimise the possibility of versions of perl
89007cb3
NC
175escaping that believe themselves to be newer than they actually are.
176
177It would be useful to find an elegant way to have the "this is an interim
178maintenance release" or "this is a release candidate" in the terse -v output,
179and have it so that it's easy for the pumpking to remove this just as the
180release tarball is rolled up. This way the version pulled out of rsync would
181always say "I'm a development release" and it would be safe to bump the
182reported minor version as soon as a release ships, which would aid perl
183developers.
184
0bdfc961
NC
185This task is really about thinking of an elegant way to arrange the C source
186such that it's trivial for the Pumpking to flag "this is an official release"
187when making a tarball, yet leave the default source saying "I'm not the
188official release".
189
190=head2 bincompat functions
191
192There are lots of functions which are retained for binary compatibility.
193Clean these up. Move them to mathom.c, and don't compile for blead?
194
195
196
197
198
199=head1 Tasks that need a knowledge of XS
e50bb9a1 200
0bdfc961
NC
201These tasks would need C knowledge, and roughly the level of knowledge of
202the perl API that comes from writing modules that use XS to interface to
203C.
204
205=head2 IPv6
206
207Clean this up. Check everything in core works
208
209=head2 UTF8 caching code
210
211The string position/offset cache is not optional. It should be.
212
213=head2 Implicit Latin 1 => Unicode translation
214
215Conversions from byte strings to UTF-8 currently map high bit characters
216to Unicode without translation (or, depending on how you look at it, by
217implicitly assuming that the byte strings are in Latin-1). As perl assumes
218the C locale by default, upgrading a string to UTF-8 may change the
219meaning of its contents regarding character classes, case mapping, etc.
220This should probably emit a warning (at least).
221
222This task is incremental - even a little bit of work on it will help.
e50bb9a1 223
cd793d32 224=head2 autovivification
e50bb9a1 225
cd793d32 226Make all autovivification consistent w.r.t LVALUE/RVALUE and strict/no strict;
e50bb9a1 227
0bdfc961 228This task is incremental - even a little bit of work on it will help.
e50bb9a1 229
0bdfc961 230=head2 Unicode in Filenames
e50bb9a1 231
0bdfc961
NC
232chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open,
233opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen,
234system, truncate, unlink, utime, -X. All these could potentially accept
235Unicode filenames either as input or output (and in the case of system
236and qx Unicode in general, as input or output to/from the shell).
237Whether a filesystem - an operating system pair understands Unicode in
238filenames varies.
e50bb9a1 239
0bdfc961
NC
240Known combinations that have some level of understanding include
241Microsoft NTFS, Apple HFS+ (In Mac OS 9 and X) and Apple UFS (in Mac
242OS X), NFS v4 is rumored to be Unicode, and of course Plan 9. How to
243create Unicode filenames, what forms of Unicode are accepted and used
244(UCS-2, UTF-16, UTF-8), what (if any) is the normalization form used,
245and so on, varies. Finding the right level of interfacing to Perl
246requires some thought. Remember that an OS does not implicate a
247filesystem.
e50bb9a1 248
0bdfc961
NC
249(The Windows -C command flag "wide API support" has been at least
250temporarily retired in 5.8.1, and the -C has been repurposed, see
251L<perlrun>.)
969e704b 252
0bdfc961 253=head2 Unicode in %ENV
969e704b 254
0bdfc961 255Currently the %ENV entries are always byte strings.
e50bb9a1 256
0bdfc961 257=head2 use less 'memory'
e50bb9a1 258
0bdfc961
NC
259Investigate trade offs to switch out perl's choices on memory usage.
260Particularly perl should be able to give memory back.
e50bb9a1 261
0bdfc961 262This task is incremental - even a little bit of work on it will help.
0abe3f7c 263
0bdfc961 264=head2 Re-implement C<:unique> in a way that is actually thread-safe
0abe3f7c 265
0bdfc961
NC
266The old implementation made bad assumptions on several levels. A good 90%
267solution might be just to make C<:unique> work to share the string buffer
268of SvPVs. That way large constant strings can be shared between ithreads,
269such as the configuration information in F<Config>.
0abe3f7c 270
0bdfc961 271=head2 Make tainting consistent
0abe3f7c 272
0bdfc961
NC
273Tainting would be easier to use if it didn't take documented shortcuts and
274allow taint to "leak" everywhere within an expression.
0abe3f7c 275
0bdfc961 276=head2 readpipe(LIST)
0abe3f7c 277
0bdfc961
NC
278system() accepts a LIST syntax (and a PROGRAM LIST syntax) to avoid
279running a shell. readpipe() (the function behind qx//) could be similarly
280extended.
0abe3f7c 281
e50bb9a1 282
e50bb9a1 283
e50bb9a1 284
f86a8bc5 285
0bdfc961 286=head1 Tasks that need a knowledge of the interpreter
3298bd4d 287
0bdfc961
NC
288These tasks would need C knowledge, and knowledge of how the interpreter works,
289or a willingness to learn.
3298bd4d 290
0bdfc961
NC
291=head2 lexical pragmas
292
293Reimplement the mechanism of lexical pragmas to be more extensible. Fix
294current pragmas that don't work well (or at all) with lexical scopes or in
295run-time eval(STRING) (C<sort>, C<re>, C<encoding> for example). MJD has a
296preliminary patch that implements this.
0562c0e3 297
d10fc472 298=head2 Attach/detach debugger from running program
1626a787 299
cd793d32
NC
300The old perltodo notes "With C<gdb>, you can attach the debugger to a running
301program if you pass the process ID. It would be good to do this with the Perl
0bdfc961
NC
302debugger on a running Perl program, although I'm not sure how it would be
303done." ssh and screen do this with named pipes in /tmp. Maybe we can too.
1626a787 304
0bdfc961 305=head2 inlining autoloaded constants
d10fc472 306
0bdfc961
NC
307Currently the optimiser can inline constants when expressed as subroutines
308with prototype ($) that return a constant. Likewise, many packages wrapping
309C libraries export lots of constants as subroutines which are AUTOLOADed on
310demand. However, these have no prototypes, so can't be seen as constants by
311the optimiser. Some way of cheaply (low syntax, low memory overhead) to the
312perl compiler that a name is a constant would be great, so that it knows to
313call the AUTOLOAD routine at compile time, and then inline the constant.
80b46460 314
0bdfc961 315=head2 Constant folding
80b46460 316
0bdfc961
NC
317The peephole optimiser should trap errors during constant folding, and give
318up on the folding, rather than bailing out at compile time. It is quite
319possible that the unfoldable constant is in unreachable code, eg something
320akin to C<$a = 0/0 if 0;>
321
322=head2 LVALUE functions for lists
323
324The old perltodo notes that lvalue functions don't work for list or hash
325slices. This would be good to fix.
326
327=head2 LVALUE functions in the debugger
328
329The old perltodo notes that lvalue functions don't work in the debugger. This
330would be good to fix.
331
332=head2 _ prototype character
333
334Study the possibility of adding a new prototype character, C<_>, meaning
335"this argument defaults to $_".
336
337=head2 @INC source filter to Filter::Simple
338
339The second return value from a sub in @INC can be a source filter. This isn't
340documented. It should be changed to use Filter::Simple, tested and documented.
341
342=head2 regexp optimiser optional
343
344The regexp optimiser is not optional. It should configurable to be, to allow
345its performance to be measured, and its bugs to be easily demonstrated.
346
347=head2 UNITCHECK
348
349Introduce a new special block, UNITCHECK, which is run at the end of a
350compilation unit (module, file, eval(STRING) block). This will correspond to
351the Perl 6 CHECK. Perl 5's CHECK cannot be changed or removed because the
352O.pm/B.pm backend framework depends on it.
353
354=head2 optional optimizer
355
356Make the peephole optimizer optional. Currently it performs two tasks as
357it walks the optree - genuine peephole optimisations, and necessary fixups of
358ops. It would be good to find an efficient way to switch out the
359optimisations whilst keeping the fixups.
360
361=head2 You WANT *how* many
362
363Currently contexts are void, scalar and list. split has a special mechanism in
364place to pass in the number of return values wanted. It would be useful to
365have a general mechanism for this, backwards compatible and little speed hit.
366This would allow proposals such as short circuiting sort to be implemented
367as a module on CPAN.
368
369=head2 lexical aliases
370
371Allow lexical aliases (maybe via the syntax C<my \$alias = \$foo>.
372
373=head2 entersub XS vs Perl
374
375At the moment pp_entersub is huge, and has code to deal with entering both
376perl and XS subroutines. Subroutine implementations rarely change between
377perl and XS at run time, so investigate using 2 ops to enter subs (one for
378XS, one for perl) and swap between if a sub is redefined.
2810d901
NC
379
380=head2 Self ties
381
382self ties are currently illegal because they caused too many segfaults. Maybe
383the causes of these could be tracked down and self-ties on all types re-
384instated.
0bdfc961
NC
385
386=head2 Optimize away @_
387
388The old perltodo notes "Look at the "reification" code in C<av.c>".
389
390=head2 switch ops
391
392The old perltodo notes "Although we have C<Switch.pm> in core, Larry points to
393the dormant C<nswitch> and C<cswitch> ops in F<pp.c>; using these opcodes would
394be much faster."
395
396=head2 What hooks would assertions need?
397
398Assertions are in the core, and work. However, assertions needed to be added
399as a core patch, rather than an XS module in ext, or a CPAN module, because
400the core has no hooks in the necessary places. It would be useful to
401investigate what hooks would need to be added to make it possible to provide
402the full assertion support from a CPAN module, so that we aren't constraining
403the imagination of future CPAN authors.
404
405
406
407
408
409
410
411=head1 Big projects
412
413Tasks that will get your name mentioned in the description of the "Highlights
414of 5.10"
415
416=head2 make ithreads more robust
417
418Generally make ithreads more robust. See also L<iCOW>
419
420This task is incremental - even a little bit of work on it will help, and
421will be greatly appreciated.
422
423=head2 iCOW
424
425Sarathy and Arthur have a proposal for an improved Copy On Write which
426specifically will be able to COW new ithreads. If this can be implemented
427it would be a good thing.
428
429=head2 (?{...}) closures in regexps
430
431Fix (or rewrite) the implementation of the C</(?{...})/> closures.
432
433=head2 A re-entrant regexp engine
434
435This will allow the use of a regex from inside (?{ }), (??{ }) and
436(?(?{ })|) constructs.