version 2, including all changes.
.
Rev |
Author |
# |
Line |
1 |
perry |
1 |
PERLTODO |
|
|
2 |
!!!PERLTODO |
|
|
3 |
NAME |
|
|
4 |
DESCRIPTION |
|
|
5 |
Infrastructure |
|
|
6 |
Configure |
|
|
7 |
Perl Language |
|
|
8 |
Perl Internals |
|
|
9 |
Documentation |
|
|
10 |
Modules |
|
|
11 |
Tom's Wishes |
|
|
12 |
Win32 Stuff |
|
|
13 |
Would be nice to have |
|
|
14 |
Possible pragmas |
|
|
15 |
Optimizations |
|
|
16 |
Vague possibilities |
|
|
17 |
To Do Or Not To Do |
|
|
18 |
Threading |
|
|
19 |
Compiler |
|
|
20 |
Recently Finished Tasks |
|
|
21 |
---- |
|
|
22 |
!!NAME |
|
|
23 |
|
|
|
24 |
|
|
|
25 |
perltodo - Perl TO-DO List |
|
|
26 |
!!DESCRIPTION |
|
|
27 |
|
|
|
28 |
|
|
|
29 |
This is a list of wishes for Perl. It is maintained by |
|
|
30 |
Nathan Torkington for the Perl porters. Send updates to |
|
|
31 |
''perl5-porters@perl.org''. If you want to work on any of |
|
|
32 |
these projects, be sure to check the perl5-porters archives |
|
|
33 |
for past ideas, flames, and propaganda. This will save you |
|
|
34 |
time and also prevent you from implementing something that |
|
|
35 |
Larry has already vetoed. One set of archives may be found |
|
|
36 |
at: |
|
|
37 |
|
|
|
38 |
|
|
|
39 |
http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/ |
|
|
40 |
!!Infrastructure |
|
|
41 |
|
|
|
42 |
|
|
|
43 |
__Mailing list archives__ |
|
|
44 |
|
|
|
45 |
|
|
|
46 |
Chaim suggests contacting egroup and asking them to archive |
|
|
47 |
the other perl.org mailing lists. Probably not advocacy, but |
|
|
48 |
definitely perl6-porters, etc. |
|
|
49 |
|
|
|
50 |
|
|
|
51 |
__Bug tracking system__ |
|
|
52 |
|
|
|
53 |
|
|
|
54 |
Richard Foley ''richard@perl.org'' is writing one. We |
|
|
55 |
looked at several, like gnats and the Debian system, but at |
|
|
56 |
the time we investigated them, none met our needs. Since |
|
|
57 |
then, Jitterbug has matured, and may be worth |
|
|
58 |
reinvestigation. |
|
|
59 |
|
|
|
60 |
|
|
|
61 |
The system we've developed is the recipient of perlbug mail, |
|
|
62 |
and any followups it generates from perl5-porters. New bugs |
|
|
63 |
are entered into a mysql database, and sent on to |
|
|
64 |
perl5-porters with the subject line rewritten to include a |
|
|
65 |
``ticket number'' (unique ID for the new |
|
|
66 |
bug). If the incoming message already had a ticket number in |
|
|
67 |
the subject line, then the message is logged against that |
|
|
68 |
bug. There is a separate email interface (not forwarding to |
|
|
69 |
p5p) that permits porters to claim, categorize, and close |
|
|
70 |
tickets. |
|
|
71 |
|
|
|
72 |
|
|
|
73 |
There is also a web interface to the system at |
|
|
74 |
http://bugs.perl.org. |
|
|
75 |
|
|
|
76 |
|
|
|
77 |
The current delay in implementation is caused by perl.org |
|
|
78 |
lockups. One suspect is the mail handling system, possibly |
|
|
79 |
going into loops. |
|
|
80 |
|
|
|
81 |
|
|
|
82 |
We still desperately need a bugmaster, someone who will look |
|
|
83 |
at every new ``bug'' and kill those that we already know |
|
|
84 |
about, those that are not bugs at all, etc. |
|
|
85 |
|
|
|
86 |
|
|
|
87 |
__Regression Tests__ |
|
|
88 |
|
|
|
89 |
|
|
|
90 |
The test suite for Perl serves two needs: ensuring features |
|
|
91 |
work, and ensuring old bugs have not been reintroduced. Both |
|
|
92 |
need work. |
|
|
93 |
|
|
|
94 |
|
2 |
perry |
95 |
Brent !LaVelle (lavelle@metronet.com) has stepped forward to |
1 |
perry |
96 |
work on performance tests and improving the size of the test |
|
|
97 |
suite. |
|
|
98 |
|
|
|
99 |
|
|
|
100 |
Coverage |
|
|
101 |
|
|
|
102 |
|
|
|
103 |
Do the tests that come with Perl exercise every line (or |
|
|
104 |
every block, or ...) of the Perl interpreter, and if not |
|
|
105 |
then how can we make them do so? |
|
|
106 |
|
|
|
107 |
|
|
|
108 |
Regression |
|
|
109 |
|
|
|
110 |
|
|
|
111 |
No bug fixes should be made without a corresponding |
|
|
112 |
testsuite addition. This needs a dedicated enforcer, as the |
|
|
113 |
current pumpking is either too lazy or too stupid or both |
|
|
114 |
and lets enforcement wander all over the map. |
|
|
115 |
:-) |
|
|
116 |
|
|
|
117 |
|
|
|
118 |
__DIE__ |
|
|
119 |
|
|
|
120 |
|
|
|
121 |
Tests that fail need to be of a form that can be readily |
|
|
122 |
mailed to perlbug and diagnosed with minimal |
|
|
123 |
back-and-forth's to determine which test failed, due to what |
|
|
124 |
cause, etc. |
|
|
125 |
|
|
|
126 |
|
|
|
127 |
suidperl |
|
|
128 |
|
|
|
129 |
|
|
|
130 |
We need regression/sanity tests for suidperl |
|
|
131 |
|
|
|
132 |
|
|
|
133 |
The 25% slowdown from perl4 to perl5 |
|
|
134 |
|
|
|
135 |
|
|
|
136 |
This value may or may not be accurate, but it certainly is |
|
|
137 |
eye-catching. For some things perl5 is faster than perl4, |
|
|
138 |
but often the reliability and extensibility have come at a |
|
|
139 |
cost of speed. The benchmark suite that Gisle released |
|
|
140 |
earlier has been hailed as both a fantastic solution and as |
|
|
141 |
a source of entirely meaningless figures. Do we need to test |
|
|
142 |
``real applications''? Can you do so? Anyone have machines |
|
|
143 |
to dedicate to the task? Identify the things that have grown |
|
|
144 |
slower, and see if there's a way to make them |
|
|
145 |
faster. |
|
|
146 |
!!Configure |
|
|
147 |
|
|
|
148 |
|
|
|
149 |
Andy Dougherty maintain(eds) a list of ``todo'' items for |
|
|
150 |
the configure that comes with Perl. See Porting/pumpkin.pod |
|
|
151 |
in the latest source release. |
|
|
152 |
|
|
|
153 |
|
|
|
154 |
__Install HTML__ |
|
|
155 |
|
|
|
156 |
|
|
|
157 |
Have ``make install'' give you the option to install |
|
|
158 |
HTML as well. This would be part of |
|
|
159 |
Configure. Andy Wardley (certified Perl studmuffin) will |
|
|
160 |
look into the current problems of HTML |
|
|
161 |
installation--is 'installhtml' preventing this from |
|
|
162 |
happening cleanly, or is pod2html the problem? If the |
|
|
163 |
latter, Brad Appleton's pod work may fix the problem for |
|
|
164 |
free. |
|
|
165 |
!!Perl Language |
|
|
166 |
|
|
|
167 |
|
|
|
168 |
__64-bit Perl__ |
|
|
169 |
|
|
|
170 |
|
|
|
171 |
Verify complete 64 bit support so that the value of sysseek, |
|
|
172 |
or -s, or ''stat()'', or tell can fit into a |
|
|
173 |
perl number without losing precision. Work with the |
|
|
174 |
perl-64bit mailing list on perl.org. |
|
|
175 |
|
|
|
176 |
|
|
|
177 |
__Prototypes__ |
|
|
178 |
|
|
|
179 |
|
|
|
180 |
Named prototypes |
|
|
181 |
|
|
|
182 |
|
|
|
183 |
Add proper named prototypes that actually work |
|
|
184 |
usefully. |
|
|
185 |
|
|
|
186 |
|
|
|
187 |
Indirect objects |
|
|
188 |
|
|
|
189 |
|
|
|
190 |
Fix prototype bug that forgets indirect |
|
|
191 |
objects. |
|
|
192 |
|
|
|
193 |
|
|
|
194 |
Method calls |
|
|
195 |
|
|
|
196 |
|
|
|
197 |
Prototypes for method calls. |
|
|
198 |
|
|
|
199 |
|
|
|
200 |
Context |
|
|
201 |
|
|
|
202 |
|
|
|
203 |
Return context prototype declarations. |
|
|
204 |
|
|
|
205 |
|
|
|
206 |
Scoped subs |
|
|
207 |
|
|
|
208 |
|
|
|
209 |
lexically-scoped subs, e.g. my sub |
|
|
210 |
!!Perl Internals |
|
|
211 |
|
|
|
212 |
|
|
|
213 |
__magic_setisa__ |
|
|
214 |
|
|
|
215 |
|
|
|
216 |
magic_setisa should be made to update |
|
|
217 |
%FIELDS [[???] |
|
|
218 |
|
|
|
219 |
|
|
|
220 |
__Garbage Collection__ |
|
|
221 |
|
|
|
222 |
|
|
|
223 |
There was talk of a mark-and-sweep garbage collector at |
|
|
224 |
TPC2 , but the (to users) unpredictable |
|
|
225 |
nature of its behaviour put some off. Sarathy, I believe, |
|
|
226 |
did the work. Here's what he has to say: |
|
|
227 |
|
|
|
228 |
|
|
|
229 |
Yeah, I hope to implement it someday too. The points that |
|
|
230 |
were raised in TPC2 were all to do with |
|
|
231 |
calling ''DESTROY ()'' methods, but I |
|
|
232 |
think we can accommodate that by extending ''bless()'' to |
|
|
233 |
stash extra information for objects so we track their |
|
|
234 |
lifetime accurately for those that want their |
|
|
235 |
''DESTROY ()'' to be predictable (this |
|
|
236 |
will be a speed hit, naturally, and will therefore be |
|
|
237 |
optional, naturally. :) |
|
|
238 |
|
|
|
239 |
|
|
|
240 |
[[N.B. Don't even ask me about this now! When I have the time |
|
|
241 |
to write a cogent summary, I'll post it.] |
|
|
242 |
|
|
|
243 |
|
|
|
244 |
__Reliable signals__ |
|
|
245 |
|
|
|
246 |
|
|
|
247 |
Sarathy and Dan Sugalski are working on this. Chip posted a |
|
|
248 |
patch earlier, but it was not accepted into 5.005. The issue |
|
|
249 |
is tricky, because it has the potential to greatly slow down |
|
|
250 |
the core. |
|
|
251 |
|
|
|
252 |
|
|
|
253 |
There are at least three things to consider: |
|
|
254 |
|
|
|
255 |
|
|
|
256 |
Alternate ''runops()'' for signal despatch |
|
|
257 |
|
|
|
258 |
|
|
|
259 |
Sarathy and Dan are discussed this on |
|
|
260 |
perl5-porters. |
|
|
261 |
|
|
|
262 |
|
|
|
263 |
Figure out how to ''die()'' in delayed |
|
|
264 |
sighandler |
|
|
265 |
|
|
|
266 |
|
|
|
267 |
Add tests for Thread::Signal |
|
|
268 |
|
|
|
269 |
|
|
|
270 |
Automatic tests against CPAN |
|
|
271 |
|
|
|
272 |
|
|
|
273 |
Is there some way to automatically build all/most of |
|
|
274 |
CPAN with the new Perl and check that the |
|
|
275 |
modules there pass all the tests? |
|
|
276 |
|
|
|
277 |
|
|
|
278 |
__Interpolated regex performance bugs__ |
|
|
279 |
|
|
|
280 |
|
|
|
281 |
while ( |
|
|
282 |
The qr// syntax added in 5.005 has solved this problem, but it needs more thorough documentation. |
|
|
283 |
|
|
|
284 |
|
|
|
285 |
__Memory leaks from failed eval/regcomp__ |
|
|
286 |
|
|
|
287 |
|
|
|
288 |
The only known memory leaks in Perl are in failed code or |
|
|
289 |
regexp compilation. Fix this. Hugo Van Der Sanden will |
|
|
290 |
attempt this but won't have tuits until January |
|
|
291 |
1999. |
|
|
292 |
|
|
|
293 |
|
|
|
294 |
__Make XS easier to use__ |
|
|
295 |
|
|
|
296 |
|
|
|
297 |
There was interest in SWIG from porters, but |
|
|
298 |
nothing has happened lately. |
|
|
299 |
|
|
|
300 |
|
|
|
301 |
__Make embedded Perl easier to use__ |
|
|
302 |
|
|
|
303 |
|
|
|
304 |
This is probably difficult for the same reasons that `` |
|
|
305 |
XS For Dummies'' will be |
|
|
306 |
difficult. |
|
|
307 |
|
|
|
308 |
|
|
|
309 |
__Namespace cleanup__ |
|
|
310 |
|
|
|
311 |
|
|
|
312 |
CPP-space: restrict CPP symbols exported from headers |
|
|
313 |
header-space: move into CORE/perl/ |
|
|
314 |
API-space: begin list of things that constitute public api |
|
|
315 |
env-space: Configure should use PERL_CONFIG instead of CONFIG etc. |
|
|
316 |
|
|
|
317 |
|
|
|
318 |
__MULTIPLICITY__ |
|
|
319 |
|
|
|
320 |
|
|
|
321 |
Complete work on safe recursive interpreters |
|
|
322 |
Perl-. Sarathy says that a reference |
|
|
323 |
implementation exists. |
|
|
324 |
|
|
|
325 |
|
2 |
perry |
326 |
__!MacPerl__ |
1 |
perry |
327 |
|
|
|
328 |
|
|
|
329 |
Chris Nandor and Matthias Neeracher are working on better |
2 |
perry |
330 |
integrating !MacPerl into the Perl distribution. |
1 |
perry |
331 |
!!Documentation |
|
|
332 |
|
|
|
333 |
|
|
|
334 |
There's a lot of documentation that comes with Perl. The |
|
|
335 |
quantity of documentation makes it difficult for users to |
|
|
336 |
know which section of which manpage to read in order to |
|
|
337 |
solve their problem. Tom Christiansen has done much of the |
|
|
338 |
documentation work in the past. |
|
|
339 |
|
|
|
340 |
|
|
|
341 |
__A clear division into tutorial and |
|
|
342 |
reference__ |
|
|
343 |
|
|
|
344 |
|
|
|
345 |
Some manpages (e.g., perltoot and perlreftut) clearly set |
|
|
346 |
out to educate the reader about a subject. Other manpages |
|
|
347 |
(e.g., perlsub) are references for which there is no |
|
|
348 |
tutorial, or are references with a slight tutorial bent. If |
|
|
349 |
things are either tutorial or reference, then the reader |
|
|
350 |
knows which manpage to read to learn about a subject, and |
|
|
351 |
which manpage to read to learn all about an aspect of that |
|
|
352 |
subject. Part of the solution to this is: |
|
|
353 |
|
|
|
354 |
|
|
|
355 |
__Remove the artificial distinction between operators and |
|
|
356 |
functions__ |
|
|
357 |
|
|
|
358 |
|
|
|
359 |
History shows us that users, and often porters, aren't clear |
|
|
360 |
on the operator-function distinction. The present split in |
|
|
361 |
reference material between perlfunc and perlop hinders user |
|
|
362 |
navigation. Given that perlfunc is by far the larger of the |
|
|
363 |
two, move operator reference into perlfunc. |
|
|
364 |
|
|
|
365 |
|
|
|
366 |
__More tutorials__ |
|
|
367 |
|
|
|
368 |
|
|
|
369 |
More documents of a tutorial nature could help. Here are |
|
|
370 |
some candidates: |
|
|
371 |
|
|
|
372 |
|
|
|
373 |
Regular expressions |
|
|
374 |
|
|
|
375 |
|
|
|
376 |
Robin Berjon (r.berjon@ltconsulting.net) has |
|
|
377 |
volunteered. |
|
|
378 |
|
|
|
379 |
|
|
|
380 |
I/O |
|
|
381 |
|
|
|
382 |
|
|
|
383 |
Mark-Jason Dominus (mjd@plover.com) has an outline for |
|
|
384 |
perliotut. |
|
|
385 |
|
|
|
386 |
|
|
|
387 |
pack/unpack |
|
|
388 |
|
|
|
389 |
|
|
|
390 |
This is badly needed. There has been some discussion on the |
|
|
391 |
subject on perl5-porters. |
|
|
392 |
|
|
|
393 |
|
|
|
394 |
Debugging |
|
|
395 |
|
|
|
396 |
|
|
|
397 |
Ronald Kimball (rjk@linguist.dartmouth.edu) has |
|
|
398 |
volunteered. |
|
|
399 |
|
|
|
400 |
|
|
|
401 |
__Include a search tool__ |
|
|
402 |
|
|
|
403 |
|
|
|
404 |
perldoc should be able to 'grep' fulltext indices of |
|
|
405 |
installed POD files. This would let people |
|
|
406 |
say: |
|
|
407 |
|
|
|
408 |
|
|
|
409 |
perldoc -find printing numbers with commas |
|
|
410 |
and get back the perlfaq entry on 'commify'. |
|
|
411 |
|
|
|
412 |
|
|
|
413 |
This solution, however, requires documentation to contain |
|
|
414 |
the keywords the user is searching for. Even when the users |
|
|
415 |
know what they're looking for, often they can't spell |
|
|
416 |
it. |
|
|
417 |
|
|
|
418 |
|
|
|
419 |
__Include a locate tool__ |
|
|
420 |
|
|
|
421 |
|
|
|
422 |
perldoc should be able to help people find the manpages on a |
|
|
423 |
particular high-level subject: |
|
|
424 |
|
|
|
425 |
|
|
|
426 |
perldoc -find web |
|
|
427 |
would tell them manpages, web pages, and books with material on web programming. Similarly perldoc -find databases, perldoc -find references and so on. |
|
|
428 |
|
|
|
429 |
|
|
|
430 |
We need something in the vicinity of: |
|
|
431 |
|
|
|
432 |
|
|
|
433 |
% perl -help random stuff |
|
|
434 |
No documentation for perl function `random stuff' found |
|
|
435 |
The following entry in perlfunc.pod matches /random/a: |
|
|
436 |
=item rand EXPR |
|
|
437 |
=item rand |
|
|
438 |
Returns a random fractional number greater than or equal to C |
|
|
439 |
(Note: If your rand function consistently returns numbers that are too |
|
|
440 |
large or too small, then your version of Perl was probably compiled |
|
|
441 |
with the wrong number of RANDBITS.) |
|
|
442 |
The following pod pages seem to have /stuff/a: |
|
|
443 |
perlfunc.pod (7 hits) |
|
|
444 |
perlfaq7.pod (6 hits) |
|
|
445 |
perlmod.pod (4 hits) |
|
|
446 |
perlsyn.pod (3 hits) |
|
|
447 |
perlfaq8.pod (2 hits) |
|
|
448 |
perlipc.pod (2 hits) |
|
|
449 |
perl5004delta.pod (1 hit) |
|
|
450 |
perl5005delta.pod (1 hit) |
|
|
451 |
perlcall.pod (1 hit) |
|
|
452 |
perldelta.pod (1 hit) |
|
|
453 |
perlfaq3.pod (1 hit) |
|
|
454 |
perlfaq5.pod (1 hit) |
|
|
455 |
perlhist.pod (1 hit) |
|
|
456 |
perlref.pod (1 hit) |
|
|
457 |
perltoc.pod (1 hit) |
|
|
458 |
perltrap.pod (1 hit) |
|
|
459 |
Proceed to open perlfunc.pod? [[y] n |
|
|
460 |
Do you want to speak perl interactively? [[y] n |
|
|
461 |
Should I dial 911? [[y] n |
|
|
462 |
Do you need psychiatric help? [[y] y |
|
|
463 |
|
|
|
464 |
|
|
|
465 |
__Separate function manpages by default__ |
|
|
466 |
|
|
|
467 |
|
|
|
468 |
Perl should install 'manpages' for every function/operator |
|
|
469 |
into the 3pl or 3p manual section. By default. The splitman |
|
|
470 |
program in the Perl source distribution does the work of |
|
|
471 |
turning big perlfunc into little 3p pages. |
|
|
472 |
|
|
|
473 |
|
|
|
474 |
__Users can't find the manpages__ |
|
|
475 |
|
|
|
476 |
|
|
|
477 |
Make perldoc tell users what they need to add to |
|
|
478 |
their .login or .cshrc to set their MANPATH |
|
|
479 |
correctly. |
|
|
480 |
|
|
|
481 |
|
|
|
482 |
__Install ALL Documentation__ |
|
|
483 |
|
|
|
484 |
|
|
|
485 |
Make the standard documentation kit include the |
|
|
486 |
VMS , OS/2 , Win32, Threads, |
|
|
487 |
etc information. installperl and pod/Makefile should know |
|
|
488 |
enough to copy README .foo to perlfoo.pod |
|
|
489 |
before building everything, when appropriate. |
|
|
490 |
|
|
|
491 |
|
|
|
492 |
__Outstanding issues to be documented__ |
|
|
493 |
|
|
|
494 |
|
|
|
495 |
Tom has a list of 5.005_5* features or changes that require |
|
|
496 |
documentation. |
|
|
497 |
|
|
|
498 |
|
|
|
499 |
Create one document that coherently explains the delta |
|
|
500 |
between the last camel release and the current release. |
|
|
501 |
perldelta was supposed to be that, but no longer. The things |
|
|
502 |
in perldelta never seemed to get placed in the right places |
|
|
503 |
in the real manpages, either. This needs work. |
|
|
504 |
|
|
|
505 |
|
|
|
506 |
__Adapt www.linuxhq.com for Perl__ |
|
|
507 |
|
|
|
508 |
|
|
|
509 |
This should help glorify documentation and get more people |
|
|
510 |
involved in perl development. |
|
|
511 |
|
|
|
512 |
|
|
|
513 |
__Replace man with a perl program__ |
|
|
514 |
|
|
|
515 |
|
|
|
516 |
Can we reimplement man in Perl? Tom has a start. I believe |
|
|
517 |
some of the Linux systems distribute a manalike. |
|
|
518 |
Alternatively, build on perldoc to remove the unfeatures |
|
|
519 |
like ``is slow'' and ``has no apropos''. |
|
|
520 |
|
|
|
521 |
|
|
|
522 |
__Unicode tutorial__ |
|
|
523 |
|
|
|
524 |
|
|
|
525 |
We could use more work on helping people understand Perl's |
|
|
526 |
new Unicode support that Larry has created. |
|
|
527 |
!!Modules |
|
|
528 |
|
|
|
529 |
|
|
|
530 |
__Update the POSIX extension to conform with |
|
|
531 |
the POSIX 1003.1 Edition 2__ |
|
|
532 |
|
|
|
533 |
|
|
|
534 |
The current state of the POSIX extension is |
|
|
535 |
as of Edition 1, 1991, whereas the Edition 2 came out in |
|
|
536 |
1996. ISO/IEC 9945:1-1996(E), |
|
|
537 |
ANSI/IEEE Std 1003.1, 1996 Edition. |
|
|
538 |
ISBN 1-55937-573-6. The updates were legion: |
|
|
539 |
threads, IPC , and real time |
|
|
540 |
extensions. |
|
|
541 |
|
|
|
542 |
|
|
|
543 |
__Module versions__ |
|
|
544 |
|
|
|
545 |
|
|
|
546 |
Automate the checking of versions in the standard |
|
|
547 |
distribution so it's easy for a pumpking to check whether |
|
|
548 |
CPAN has a newer version that we should be |
|
|
549 |
including? |
|
|
550 |
|
|
|
551 |
|
|
|
552 |
__New modules__ |
|
|
553 |
|
|
|
554 |
|
|
|
555 |
Which modules should be added to the standard distribution? |
|
|
556 |
This ties in with the SDK discussed on the |
|
|
557 |
perl-sdk list at perl.org. |
|
|
558 |
|
|
|
559 |
|
|
|
560 |
__Profiler__ |
|
|
561 |
|
|
|
562 |
|
|
|
563 |
Make the profiler (Devel::DProf) part of the standard |
|
|
564 |
release, and document it well. |
|
|
565 |
|
|
|
566 |
|
|
|
567 |
__Tie Modules__ |
|
|
568 |
|
|
|
569 |
|
2 |
perry |
570 |
!VecArray |
1 |
perry |
571 |
|
|
|
572 |
|
|
|
573 |
Implement array using ''vec()''. Nathan Torkington has |
|
|
574 |
working code to do this. |
|
|
575 |
|
|
|
576 |
|
2 |
perry |
577 |
!SubstrArray |
1 |
perry |
578 |
|
|
|
579 |
|
|
|
580 |
Implement array using ''substr()'' |
|
|
581 |
|
|
|
582 |
|
2 |
perry |
583 |
!VirtualArray |
1 |
perry |
584 |
|
|
|
585 |
|
|
|
586 |
Implement array using a file |
|
|
587 |
|
|
|
588 |
|
2 |
perry |
589 |
!ShiftSplice |
1 |
perry |
590 |
|
|
|
591 |
|
|
|
592 |
Defines shift et al in terms of splice method |
|
|
593 |
|
|
|
594 |
|
|
|
595 |
__Procedural options__ |
|
|
596 |
|
|
|
597 |
|
|
|
598 |
Support procedural interfaces for the common cases of Perl's |
|
|
599 |
gratuitously OOO modules. Tom objects to |
|
|
600 |
``use IO::File'' reading many thousands of lines of |
|
|
601 |
code. |
|
|
602 |
|
|
|
603 |
|
|
|
604 |
__RPC__ |
|
|
605 |
|
|
|
606 |
|
|
|
607 |
Write a module for transparent, portable remote procedure |
|
|
608 |
calls. (Not core). This touches on the CORBA |
|
|
609 |
and ILU work. |
|
|
610 |
|
|
|
611 |
|
|
|
612 |
__y2k localtime/gmtime__ |
|
|
613 |
|
|
|
614 |
|
|
|
615 |
Write a module, Y2k::Catch, which overloads localtime and |
|
|
616 |
gmtime's returned year value and catches ``bad'' attempts to |
|
|
617 |
use it. |
|
|
618 |
|
|
|
619 |
|
|
|
620 |
__Export File::Find variables__ |
|
|
621 |
|
|
|
622 |
|
|
|
623 |
Make File::Find export $name etc manually, at least |
|
|
624 |
if asked to. |
|
|
625 |
|
|
|
626 |
|
|
|
627 |
__Ioctl__ |
|
|
628 |
|
|
|
629 |
|
|
|
630 |
Finish a proper Ioctl module. |
|
|
631 |
|
|
|
632 |
|
|
|
633 |
__Debugger attach/detach__ |
|
|
634 |
|
|
|
635 |
|
|
|
636 |
Permit a user to debug an already-running |
|
|
637 |
program. |
|
|
638 |
|
|
|
639 |
|
|
|
640 |
__Regular Expression debugger__ |
|
|
641 |
|
|
|
642 |
|
|
|
643 |
Create a visual profiler/debugger tool that stepped you |
|
|
644 |
through the execution of a regular expression point by |
|
|
645 |
point. Ilya has a module to color-code and display regular |
|
|
646 |
expression parses and executions. There's something at |
|
|
647 |
http://tkworld.org/ that might be a good start, it's a |
|
|
648 |
Tk/Tcl RE wizard, that builds regexen of many |
|
|
649 |
flavours. |
|
|
650 |
|
|
|
651 |
|
|
|
652 |
__Alternative RE Syntax__ |
|
|
653 |
|
|
|
654 |
|
|
|
655 |
Make an alternative regular expression syntax that is |
|
|
656 |
accessed through a module. For instance, |
|
|
657 |
|
|
|
658 |
|
|
|
659 |
use RE; |
|
|
660 |
$re = start_of_line() |
|
|
661 |
- |
|
|
662 |
if (/$re/) { |
|
|
663 |
print |
|
|
664 |
Newbies to regular expressions typically only use a subset of the full language. Perhaps you wouldn't have to implement the full feature set. |
|
|
665 |
|
|
|
666 |
|
|
|
667 |
__Bundled modules__ |
|
|
668 |
|
|
|
669 |
|
|
|
670 |
Nicholas Clark (nick@flirble.org) had a patch for storing |
|
|
671 |
modules in zipped format. This needs exploring and |
|
|
672 |
concluding. |
|
|
673 |
|
|
|
674 |
|
|
|
675 |
__Expect__ |
|
|
676 |
|
|
|
677 |
|
|
|
678 |
Adopt IO::Tty, make it as portable as Don Libes' ``expect'' |
|
|
679 |
(can we link against expect code?), and perfect a Perl |
|
|
680 |
version of expect. IO::Tty and expect could then be |
|
|
681 |
distributed as part of the core distribution, replacing |
|
|
682 |
Comm.pl and other hacks. |
|
|
683 |
|
|
|
684 |
|
|
|
685 |
__GUI::Native__ |
|
|
686 |
|
|
|
687 |
|
|
|
688 |
A simple-to-use interface to native graphical abilities |
|
|
689 |
would be welcomed. Oh, Perl's access Tk is nice enough, and |
|
|
690 |
reasonably portable, but it's not particularly as fast as |
|
|
691 |
one would like. Simple access to the mouse's cut buffer or |
|
|
692 |
mouse-presses shouldn't required loading a few terabytes of |
|
|
693 |
Tk code. |
|
|
694 |
|
|
|
695 |
|
|
|
696 |
__Update semibroken auxiliary tools; h2ph, a2p, |
|
|
697 |
etc.__ |
|
|
698 |
|
|
|
699 |
|
|
|
700 |
Kurt Starsinic is working on h2ph. mjd has fixed bugs in a2p |
|
|
701 |
in the past. a2p apparently doesn't work on nawk and gawk |
|
|
702 |
extensions. Graham Barr has an Include module that does h2ph |
|
|
703 |
work at runtime. |
|
|
704 |
|
|
|
705 |
|
|
|
706 |
__pod2html__ |
|
|
707 |
|
|
|
708 |
|
|
|
709 |
A short-term fix: pod2html generates absolute |
|
|
710 |
HTML links. Make it generate relative |
|
|
711 |
links. |
|
|
712 |
|
|
|
713 |
|
|
|
714 |
__Podchecker__ |
|
|
715 |
|
|
|
716 |
|
|
|
717 |
Something like lint for Pod would be good. Something that |
|
|
718 |
catches common errors as well as gross ones. Brad Appleton |
2 |
perry |
719 |
is putting together something as part of his !PodParser |
1 |
perry |
720 |
work. |
|
|
721 |
!!Tom's Wishes |
|
|
722 |
|
|
|
723 |
|
|
|
724 |
__Webperl__ |
|
|
725 |
|
|
|
726 |
|
|
|
727 |
Design a webperl environment that's as tightly integrated |
|
|
728 |
and as easy-to-use as Perl's current command-line |
|
|
729 |
environment. |
|
|
730 |
|
|
|
731 |
|
|
|
732 |
__Mobile agents__ |
|
|
733 |
|
|
|
734 |
|
|
|
735 |
More work on a safe and secure execution environment for |
|
|
736 |
mobile agents would be neat; the Safe.pm module is a start, |
|
|
737 |
but there's a still a lot to be done in that area. Adopt |
|
|
738 |
Penguin? |
|
|
739 |
|
|
|
740 |
|
|
|
741 |
__POSIX on non-POSIX__ |
|
|
742 |
|
|
|
743 |
|
|
|
744 |
Standard programming constructs for non-POSIX systems would |
|
|
745 |
help a lot of programmers stuck on primitive, legacy |
|
|
746 |
systems. For example, Microsoft still hasn't made a usable |
|
|
747 |
POSIX interface on their clunky systems, |
|
|
748 |
which means that standard operations such as ''alarm()'' |
|
|
749 |
and ''fork()'', both critical for sophisticated |
|
|
750 |
client-server programming, must both be kludged |
|
|
751 |
around. |
|
|
752 |
|
|
|
753 |
|
|
|
754 |
I'm unsure whether Tom means to emulate alarm( )and |
|
|
755 |
''fork()'', or merely to provide a document like |
|
|
756 |
perlport.pod to say which features are portable and which |
|
|
757 |
are not. |
|
|
758 |
|
|
|
759 |
|
|
|
760 |
__Portable installations__ |
|
|
761 |
|
|
|
762 |
|
|
|
763 |
Figure out a portable semi-gelled installation, that is, one |
|
|
764 |
without full paths. Larry has said that he's thinking about |
|
|
765 |
this. Ilya pointed out that ''perllib_mangle()'' is good |
|
|
766 |
for this. |
|
|
767 |
!!Win32 Stuff |
|
|
768 |
|
|
|
769 |
|
|
|
770 |
__Rename new headers to be consistent with the |
|
|
771 |
rest__ |
|
|
772 |
|
|
|
773 |
|
|
|
774 |
__Sort out the__ ''spawnvp()'' |
|
|
775 |
__mess__ |
|
|
776 |
|
|
|
777 |
|
|
|
778 |
__Work out DLL versioning__ |
|
|
779 |
|
|
|
780 |
|
|
|
781 |
__Style-check__ |
|
|
782 |
!!Would be nice to have |
|
|
783 |
|
|
|
784 |
|
|
|
785 |
pack ``(stuff)*'' |
|
|
786 |
|
|
|
787 |
|
|
|
788 |
Contiguous bitfields in pack/unpack |
|
|
789 |
|
|
|
790 |
|
|
|
791 |
lexperl |
|
|
792 |
|
|
|
793 |
|
|
|
794 |
Bundled perl preprocessor |
|
|
795 |
|
|
|
796 |
|
|
|
797 |
Use posix calls internally where possible |
|
|
798 |
|
|
|
799 |
|
|
|
800 |
format BOTTOM |
|
|
801 |
|
|
|
802 |
|
|
|
803 |
-i rename file only when successfully changed |
|
|
804 |
|
|
|
805 |
|
|
|
806 |
All ARGV input should act like |
|
|
807 |
|
|
|
808 |
|
|
|
809 |
report HANDLE [[formats]. |
|
|
810 |
|
|
|
811 |
|
|
|
812 |
support in perlmain to rerun debugger |
|
|
813 |
|
|
|
814 |
|
|
|
815 |
lvalue functions |
|
|
816 |
|
|
|
817 |
|
|
|
818 |
Tuomas Lukka, on behalf of the PDL project, |
|
|
819 |
greatly desires this and Ilya has a patch for it (probably |
|
|
820 |
against an older version of Perl). Tuomas points out that |
|
|
821 |
what PDL really wants is lvalue |
|
|
822 |
''methods'', not just subs. |
|
|
823 |
!!Possible pragmas |
|
|
824 |
|
|
|
825 |
|
|
|
826 |
(use less memory, CPU ) |
|
|
827 |
!!Optimizations |
|
|
828 |
|
|
|
829 |
|
|
|
830 |
__constant function cache__ |
|
|
831 |
|
|
|
832 |
|
|
|
833 |
__foreach(reverse...)__ |
|
|
834 |
|
|
|
835 |
|
|
|
836 |
__Cache eval tree__ |
|
|
837 |
|
|
|
838 |
|
|
|
839 |
Unless lexical outer scope used (mark in |
|
|
840 |
|
|
|
841 |
|
|
|
842 |
__rcatmaybe__ |
|
|
843 |
|
|
|
844 |
|
|
|
845 |
__Shrink opcode tables__ |
|
|
846 |
|
|
|
847 |
|
|
|
848 |
Via multiple implementations selected in peep. |
|
|
849 |
|
|
|
850 |
|
|
|
851 |
__Cache hash value__ |
|
|
852 |
|
|
|
853 |
|
|
|
854 |
Not a win, according to Guido. |
|
|
855 |
|
|
|
856 |
|
|
|
857 |
__Optimize away__ @_ __where |
|
|
858 |
possible__ |
|
|
859 |
|
|
|
860 |
|
|
|
861 |
__Optimize sort by {__ $a ____ |
|
|
862 |
$b __}__ |
|
|
863 |
|
|
|
864 |
|
|
|
865 |
Greg Bacon added several more sort optimizations. These have |
|
|
866 |
made it into 5.005_55, thanks to Hans Mulder. |
|
|
867 |
|
|
|
868 |
|
|
|
869 |
__Rewrite regexp parser for better integrated |
|
|
870 |
optimization__ |
|
|
871 |
|
|
|
872 |
|
|
|
873 |
The regexp parser was rewritten for 5.005. Ilya's the regexp |
|
|
874 |
guru. |
|
|
875 |
!!Vague possibilities |
|
|
876 |
|
|
|
877 |
|
|
|
878 |
ref function in list context |
|
|
879 |
|
|
|
880 |
|
|
|
881 |
This seems impossible to do without substantially breaking |
|
|
882 |
code. |
|
|
883 |
|
|
|
884 |
|
|
|
885 |
make tr/// return histogram in list context? |
|
|
886 |
|
|
|
887 |
|
|
|
888 |
Loop control on do{} et al |
|
|
889 |
|
|
|
890 |
|
|
|
891 |
Explicit switch statements |
|
|
892 |
|
|
|
893 |
|
|
|
894 |
Nobody has yet managed to come up with a switch syntax that |
|
|
895 |
would allow for mixed hash, constant, regexp checks. Submit |
|
|
896 |
implementation with syntax, please. |
|
|
897 |
|
|
|
898 |
|
|
|
899 |
compile to real threaded code |
|
|
900 |
|
|
|
901 |
|
|
|
902 |
structured types |
|
|
903 |
|
|
|
904 |
|
|
|
905 |
Modifiable $1 et al |
|
|
906 |
|
|
|
907 |
|
|
|
908 |
The intent is for this to be a means of editing the matched |
|
|
909 |
portions of the target string. |
|
|
910 |
!!To Do Or Not To Do |
|
|
911 |
|
|
|
912 |
|
|
|
913 |
These are things that have been discussed in the past and |
|
|
914 |
roundly criticized for being of questionable |
|
|
915 |
value. |
|
|
916 |
|
|
|
917 |
|
|
|
918 |
__Making__ ''my()'' __work on ``package'' |
|
|
919 |
variables__ |
|
|
920 |
|
|
|
921 |
|
|
|
922 |
Being able to say my($Foo::Bar), something that sounds |
|
|
923 |
ludicrous and the 5.6 pumpking has mocked. |
|
|
924 |
|
|
|
925 |
|
|
|
926 |
__``or'' testing defined not truth__ |
|
|
927 |
|
|
|
928 |
|
|
|
929 |
We tell people that can be used to give a default value to a |
|
|
930 |
variable: |
|
|
931 |
|
|
|
932 |
|
|
|
933 |
$children = shift 5; # default is 5 children |
|
|
934 |
which is almost (but not): |
|
|
935 |
|
|
|
936 |
|
|
|
937 |
$children = shift; |
|
|
938 |
$children = 5 unless $children; |
|
|
939 |
but if the first argument was given and is ``0'', then it will be considered false by and 5 used instead. Really we want an -like operator that behaves like: |
|
|
940 |
|
|
|
941 |
|
|
|
942 |
$children = shift; |
|
|
943 |
$children = 5 unless defined $children; |
|
|
944 |
Namely, a that tests defined-ness rather than truth. One was discussed, and a patch submitted, but the objections were many. While there were objections, many still feel the need. At least it was decided that ?? is the best name for the operator. |
|
|
945 |
|
|
|
946 |
|
|
|
947 |
__``dynamic'' lexicals__ |
|
|
948 |
|
|
|
949 |
|
|
|
950 |
my $x; |
|
|
951 |
sub foo { |
|
|
952 |
local $x; |
|
|
953 |
} |
|
|
954 |
Localizing, as Tim Bunce points out, is a separate concept from whether the variable is global or lexical. Chip Salzenberg had an implementation once, but Larry thought it had potential to confuse. |
|
|
955 |
|
|
|
956 |
|
|
|
957 |
__``class''-based, rather than package-based |
|
|
958 |
``lexicals''__ |
|
|
959 |
|
|
|
960 |
|
|
|
961 |
This is like what the Alias module provides, but the |
|
|
962 |
variables would be lexicals reserved by perl at |
|
|
963 |
compile-time, which really are indices pointing into the |
|
|
964 |
pseudo-hash object visible inside every method so |
|
|
965 |
declared. |
|
|
966 |
!!Threading |
|
|
967 |
|
|
|
968 |
|
|
|
969 |
__Modules__ |
|
|
970 |
|
|
|
971 |
|
|
|
972 |
Which of the standard modules are thread-safe? Which |
|
|
973 |
CPAN modules? How easy is it to fix those |
|
|
974 |
non-safe modules? |
|
|
975 |
|
|
|
976 |
|
|
|
977 |
__Testing__ |
|
|
978 |
|
|
|
979 |
|
|
|
980 |
Threading is still experimental. Every reproducible bug |
|
|
981 |
identifies something else for us to fix. Find and submit |
|
|
982 |
more of these problems. |
|
|
983 |
|
|
|
984 |
|
|
|
985 |
__$AUTOLOAD__ |
|
|
986 |
|
|
|
987 |
|
|
|
988 |
__exit/die__ |
|
|
989 |
|
|
|
990 |
|
|
|
991 |
Consistent semantics for exit/die in threads. |
|
|
992 |
|
|
|
993 |
|
|
|
994 |
__External threads__ |
|
|
995 |
|
|
|
996 |
|
|
|
997 |
Better support for externally created threads. |
|
|
998 |
|
|
|
999 |
|
|
|
1000 |
__Thread::Pool__ |
|
|
1001 |
|
|
|
1002 |
|
|
|
1003 |
__thread-safety__ |
|
|
1004 |
|
|
|
1005 |
|
|
|
1006 |
Spot-check globals like statcache and global GVs for |
|
|
1007 |
thread-safety. Part done__ |
|
|
1008 |
__ |
|
|
1009 |
|
|
|
1010 |
|
|
|
1011 |
__Per-thread GVs__ |
|
|
1012 |
|
|
|
1013 |
|
|
|
1014 |
According to Sarathy, this would make @_ be the |
|
|
1015 |
same in threaded and non-threaded, as well as helping solve |
|
|
1016 |
problems like filehandles (the same filehandle currently |
|
|
1017 |
cannot be used in two threads). |
|
|
1018 |
!!Compiler |
|
|
1019 |
|
|
|
1020 |
|
|
|
1021 |
__Optimization__ |
|
|
1022 |
|
|
|
1023 |
|
|
|
1024 |
The compiler's back-end code-generators for creating |
|
|
1025 |
bytecode or compilable C code could use optimization |
|
|
1026 |
work. |
|
|
1027 |
|
|
|
1028 |
|
|
|
1029 |
__Byteperl__ |
|
|
1030 |
|
|
|
1031 |
|
|
|
1032 |
Figure out how and where byteperl will be built for the |
|
|
1033 |
various platforms. |
|
|
1034 |
|
|
|
1035 |
|
|
|
1036 |
__Precompiled modules__ |
|
|
1037 |
|
|
|
1038 |
|
|
|
1039 |
Save byte-compiled modules on disk. |
|
|
1040 |
|
|
|
1041 |
|
|
|
1042 |
__Executables__ |
|
|
1043 |
|
|
|
1044 |
|
|
|
1045 |
Auto-produce executable. |
|
|
1046 |
|
|
|
1047 |
|
|
|
1048 |
__Typed lexicals__ |
|
|
1049 |
|
|
|
1050 |
|
|
|
1051 |
Typed lexicals should affect B::CC::load_pad. |
|
|
1052 |
|
|
|
1053 |
|
|
|
1054 |
__Win32__ |
|
|
1055 |
|
|
|
1056 |
|
|
|
1057 |
Workarounds to help Win32 dynamic loading. |
|
|
1058 |
|
|
|
1059 |
|
|
|
1060 |
__END blocks__ |
|
|
1061 |
|
|
|
1062 |
|
|
|
1063 |
END blocks need saving in compiled output, |
|
|
1064 |
now that CHECK blocks are |
|
|
1065 |
available. |
|
|
1066 |
|
|
|
1067 |
|
|
|
1068 |
___AUTOLOAD__ |
|
|
1069 |
|
|
|
1070 |
|
|
|
1071 |
_AUTOLOAD prodding. |
|
|
1072 |
|
|
|
1073 |
|
|
|
1074 |
__comppadlist__ |
|
|
1075 |
|
|
|
1076 |
|
|
|
1077 |
Fix comppadlist (names in comppad_name can have fake SvCUR |
|
|
1078 |
from where newASSIGNOP steals the field). |
|
|
1079 |
|
|
|
1080 |
|
|
|
1081 |
__Cached compilation__ |
|
|
1082 |
|
|
|
1083 |
|
|
|
1084 |
Can we install modules as bytecode? |
|
|
1085 |
!!Recently Finished Tasks |
|
|
1086 |
|
|
|
1087 |
|
|
|
1088 |
__Figure a way out of $^(capital letter)__ |
|
|
1089 |
|
|
|
1090 |
|
|
|
1091 |
Figure out a clean way to extend $^(capital letter) beyond |
|
|
1092 |
the 26 alphabets. (${^WORD} maybe?) |
|
|
1093 |
|
|
|
1094 |
|
|
|
1095 |
Mark-Jason Dominus sent a patch which went into |
|
|
1096 |
5.005_56. |
|
|
1097 |
|
|
|
1098 |
|
|
|
1099 |
__Filenames__ |
|
|
1100 |
|
|
|
1101 |
|
|
|
1102 |
Keep filenames in the distribution and in the standard |
|
|
1103 |
module set be 8.3 friendly where feasible. Good luck |
|
|
1104 |
changing the standard modules, though. |
|
|
1105 |
|
|
|
1106 |
|
|
|
1107 |
__Foreign lines__ |
|
|
1108 |
|
|
|
1109 |
|
|
|
1110 |
Perl should be more generous in accepting foreign line |
|
|
1111 |
terminations. Mostly __done__ in 5.005. |
|
|
1112 |
|
|
|
1113 |
|
|
|
1114 |
__Namespace cleanup__ |
|
|
1115 |
|
|
|
1116 |
|
|
|
1117 |
symbol-space: |
|
|
1118 |
CPP-space: stop malloc()/free() pollution unless asked |
|
|
1119 |
|
|
|
1120 |
|
|
|
1121 |
__ISA .pm__ |
|
|
1122 |
|
|
|
1123 |
|
|
|
1124 |
Rename and alter ISA .pm. __Done__. It is |
|
|
1125 |
now base.pm. |
|
|
1126 |
|
|
|
1127 |
|
|
|
1128 |
__gettimeofday__ |
|
|
1129 |
|
|
|
1130 |
|
2 |
perry |
1131 |
See Time::!HiRes. |
1 |
perry |
1132 |
|
|
|
1133 |
|
|
|
1134 |
__autocroak?__ |
|
|
1135 |
|
|
|
1136 |
|
|
|
1137 |
This is the Fatal.pm module, so any builtin that does not |
|
|
1138 |
return success automatically ''die()''s. If you're |
|
|
1139 |
feeling brave, tie this in with the unified exceptions |
|
|
1140 |
scheme. |
|
|
1141 |
---- |