Commit | Line | Data |
---|---|---|
920dae64 AT |
1 | .\" Automatically generated by Pod::Man v1.37, Pod::Parser v1.32 |
2 | .\" | |
3 | .\" Standard preamble: | |
4 | .\" ======================================================================== | |
5 | .de Sh \" Subsection heading | |
6 | .br | |
7 | .if t .Sp | |
8 | .ne 5 | |
9 | .PP | |
10 | \fB\\$1\fR | |
11 | .PP | |
12 | .. | |
13 | .de Sp \" Vertical space (when we can't use .PP) | |
14 | .if t .sp .5v | |
15 | .if n .sp | |
16 | .. | |
17 | .de Vb \" Begin verbatim text | |
18 | .ft CW | |
19 | .nf | |
20 | .ne \\$1 | |
21 | .. | |
22 | .de Ve \" End verbatim text | |
23 | .ft R | |
24 | .fi | |
25 | .. | |
26 | .\" Set up some character translations and predefined strings. \*(-- will | |
27 | .\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left | |
28 | .\" double quote, and \*(R" will give a right double quote. | will give a | |
29 | .\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used to | |
30 | .\" do unbreakable dashes and therefore won't be available. \*(C` and \*(C' | |
31 | .\" expand to `' in nroff, nothing in troff, for use with C<>. | |
32 | .tr \(*W-|\(bv\*(Tr | |
33 | .ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p' | |
34 | .ie n \{\ | |
35 | . ds -- \(*W- | |
36 | . ds PI pi | |
37 | . if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch | |
38 | . if (\n(.H=4u)&(1m=20u) .ds -- \(*W\h'-12u'\(*W\h'-8u'-\" diablo 12 pitch | |
39 | . ds L" "" | |
40 | . ds R" "" | |
41 | . ds C` "" | |
42 | . ds C' "" | |
43 | 'br\} | |
44 | .el\{\ | |
45 | . ds -- \|\(em\| | |
46 | . ds PI \(*p | |
47 | . ds L" `` | |
48 | . ds R" '' | |
49 | 'br\} | |
50 | .\" | |
51 | .\" If the F register is turned on, we'll generate index entries on stderr for | |
52 | .\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index | |
53 | .\" entries marked with X<> in POD. Of course, you'll have to process the | |
54 | .\" output yourself in some meaningful fashion. | |
55 | .if \nF \{\ | |
56 | . de IX | |
57 | . tm Index:\\$1\t\\n%\t"\\$2" | |
58 | .. | |
59 | . nr % 0 | |
60 | . rr F | |
61 | .\} | |
62 | .\" | |
63 | .\" For nroff, turn off justification. Always turn off hyphenation; it makes | |
64 | .\" way too many mistakes in technical documents. | |
65 | .hy 0 | |
66 | .if n .na | |
67 | .\" | |
68 | .\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2). | |
69 | .\" Fear. Run. Save yourself. No user-serviceable parts. | |
70 | . \" fudge factors for nroff and troff | |
71 | .if n \{\ | |
72 | . ds #H 0 | |
73 | . ds #V .8m | |
74 | . ds #F .3m | |
75 | . ds #[ \f1 | |
76 | . ds #] \fP | |
77 | .\} | |
78 | .if t \{\ | |
79 | . ds #H ((1u-(\\\\n(.fu%2u))*.13m) | |
80 | . ds #V .6m | |
81 | . ds #F 0 | |
82 | . ds #[ \& | |
83 | . ds #] \& | |
84 | .\} | |
85 | . \" simple accents for nroff and troff | |
86 | .if n \{\ | |
87 | . ds ' \& | |
88 | . ds ` \& | |
89 | . ds ^ \& | |
90 | . ds , \& | |
91 | . ds ~ ~ | |
92 | . ds / | |
93 | .\} | |
94 | .if t \{\ | |
95 | . ds ' \\k:\h'-(\\n(.wu*8/10-\*(#H)'\'\h"|\\n:u" | |
96 | . ds ` \\k:\h'-(\\n(.wu*8/10-\*(#H)'\`\h'|\\n:u' | |
97 | . ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'^\h'|\\n:u' | |
98 | . ds , \\k:\h'-(\\n(.wu*8/10)',\h'|\\n:u' | |
99 | . ds ~ \\k:\h'-(\\n(.wu-\*(#H-.1m)'~\h'|\\n:u' | |
100 | . ds / \\k:\h'-(\\n(.wu*8/10-\*(#H)'\z\(sl\h'|\\n:u' | |
101 | .\} | |
102 | . \" troff and (daisy-wheel) nroff accents | |
103 | .ds : \\k:\h'-(\\n(.wu*8/10-\*(#H+.1m+\*(#F)'\v'-\*(#V'\z.\h'.2m+\*(#F'.\h'|\\n:u'\v'\*(#V' | |
104 | .ds 8 \h'\*(#H'\(*b\h'-\*(#H' | |
105 | .ds o \\k:\h'-(\\n(.wu+\w'\(de'u-\*(#H)/2u'\v'-.3n'\*(#[\z\(de\v'.3n'\h'|\\n:u'\*(#] | |
106 | .ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\(hy\fP\v'.25m'\h'-\*(#H' | |
107 | .ds D- D\\k:\h'-\w'D'u'\v'-.11m'\z\(hy\v'.11m'\h'|\\n:u' | |
108 | .ds th \*(#[\v'.3m'\s+1I\s-1\v'-.3m'\h'-(\w'I'u*2/3)'\s-1o\s+1\*(#] | |
109 | .ds Th \*(#[\s+2I\s-2\h'-\w'I'u*3/5'\v'-.3m'o\v'.3m'\*(#] | |
110 | .ds ae a\h'-(\w'a'u*4/10)'e | |
111 | .ds Ae A\h'-(\w'A'u*4/10)'E | |
112 | . \" corrections for vroff | |
113 | .if v .ds ~ \\k:\h'-(\\n(.wu*9/10-\*(#H)'\s-2\u~\d\s+2\h'|\\n:u' | |
114 | .if v .ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'\v'-.4m'^\v'.4m'\h'|\\n:u' | |
115 | . \" for low resolution devices (crt and lpr) | |
116 | .if \n(.H>23 .if \n(.V>19 \ | |
117 | \{\ | |
118 | . ds : e | |
119 | . ds 8 ss | |
120 | . ds o a | |
121 | . ds d- d\h'-1'\(ga | |
122 | . ds D- D\h'-1'\(hy | |
123 | . ds th \o'bp' | |
124 | . ds Th \o'LP' | |
125 | . ds ae ae | |
126 | . ds Ae AE | |
127 | .\} | |
128 | .rm #[ #] #H #V #F C | |
129 | .\" ======================================================================== | |
130 | .\" | |
131 | .IX Title "PERLTODO 1" | |
132 | .TH PERLTODO 1 "2006-01-07" "perl v5.8.8" "Perl Programmers Reference Guide" | |
133 | .SH "NAME" | |
134 | perltodo \- Perl TO\-DO List | |
135 | .SH "DESCRIPTION" | |
136 | .IX Header "DESCRIPTION" | |
137 | This is a list of wishes for Perl. The tasks we think are smaller or easier | |
138 | are listed first. Anyone is welcome to work on any of these, but it's a good | |
139 | idea to first contact \fIperl5\-porters@perl.org\fR to avoid duplication of | |
140 | effort. By all means contact a pumpking privately first if you prefer. | |
141 | .PP | |
142 | Whilst patches to make the list shorter are most welcome, ideas to add to | |
143 | the list are also encouraged. Check the perl5\-porters archives for past | |
144 | ideas, and any discussion about them. One set of archives may be found at: | |
145 | .PP | |
146 | .Vb 1 | |
147 | \& http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/ | |
148 | .Ve | |
149 | .PP | |
150 | What can we offer you in return? Fame, fortune, and everlasting glory? Maybe | |
151 | not, but if your patch is incorporated, then we'll add your name to the | |
152 | \&\fI\s-1AUTHORS\s0\fR file, which ships in the official distribution. How many other | |
153 | programming languages offer you 1 line of immortality? | |
154 | .SH "The roadmap to 5.10" | |
155 | .IX Header "The roadmap to 5.10" | |
156 | The roadmap to 5.10 envisages feature based releases, as various items in this | |
157 | \&\s-1TODO\s0 are completed. | |
158 | .Sh "Needed for a 5.9.4 release" | |
159 | .IX Subsection "Needed for a 5.9.4 release" | |
160 | .IP "\(bu" 4 | |
161 | Review assertions. Review syntax to combine assertions. Assertions could take | |
162 | advantage of the lexical pragmas work. \*(L"What hooks would assertions need?\*(R" | |
163 | .Sh "Needed for a 5.9.5 release" | |
164 | .IX Subsection "Needed for a 5.9.5 release" | |
165 | .ie n .IP "* Implement ""_ prototype character""" 4 | |
166 | .el .IP "* Implement ``_ prototype character''" 4 | |
167 | .IX Item "Implement _ prototype character" | |
168 | .PD 0 | |
169 | .ie n .IP "* Implement ""state variables""" 4 | |
170 | .el .IP "* Implement ``state variables''" 4 | |
171 | .IX Item "Implement state variables" | |
172 | .PD | |
173 | .Sh "Needed for a 5.9.6 release" | |
174 | .IX Subsection "Needed for a 5.9.6 release" | |
175 | Stabilisation. If all goes well, this will be the equivalent of a 5.10\-beta. | |
176 | .SH "Tasks that only need Perl knowledge" | |
177 | .IX Header "Tasks that only need Perl knowledge" | |
178 | .Sh "common test code for timed bail out" | |
179 | .IX Subsection "common test code for timed bail out" | |
180 | Write portable self destruct code for tests to stop them burning \s-1CPU\s0 in | |
181 | infinite loops. This needs to avoid using alarm, as some of the tests are | |
182 | testing alarm/sleep or timers. | |
183 | .Sh "\s-1POD\s0 \-> \s-1HTML\s0 conversion in the core still sucks" | |
184 | .IX Subsection "POD -> HTML conversion in the core still sucks" | |
185 | Which is crazy given just how simple \s-1POD\s0 purports to be, and how simple \s-1HTML\s0 | |
186 | can be. It's not actually \fIas\fR simple as it sounds, particularly with the | |
187 | flexibility \s-1POD\s0 allows for \f(CW\*(C`=item\*(C'\fR, but it would be good to improve the | |
188 | visual appeal of the \s-1HTML\s0 generated, and to avoid it having any validation | |
189 | errors. See also \*(L"make \s-1HTML\s0 install work\*(R", as the layout of installation tree | |
190 | is needed to improve the cross\-linking. | |
191 | .PP | |
192 | The addition of \f(CW\*(C`Pod::Simple\*(C'\fR and its related modules may make this task | |
193 | easier to complete. | |
194 | .Sh "Parallel testing" | |
195 | .IX Subsection "Parallel testing" | |
196 | The core regression test suite is getting ever more comprehensive, which has | |
197 | the side effect that it takes longer to run. This isn't so good. Investigate | |
198 | whether it would be feasible to give the harness script the \fBoption\fR of | |
199 | running sets of tests in parallel. This would be useful for tests in | |
200 | \&\fIt/op/*.t\fR and \fIt/uni/*.t\fR and maybe some sets of tests in \fIlib/\fR. | |
201 | .PP | |
202 | Questions to answer | |
203 | .IP "1" 4 | |
204 | .IX Item "1" | |
205 | How does screen layout work when you're running more than one test? | |
206 | .IP "2" 4 | |
207 | .IX Item "2" | |
208 | How does the caller of test specify how many tests to run in parallel? | |
209 | .IP "3" 4 | |
210 | .IX Item "3" | |
211 | How do setup/teardown tests identify themselves? | |
212 | .PP | |
213 | Pugs already does parallel testing \- can their approach be re\-used? | |
214 | .Sh "Make Schwern poorer" | |
215 | .IX Subsection "Make Schwern poorer" | |
216 | We should have for everything. When all the core's modules are tested, | |
217 | Schwern has promised to donate to \f(CW$500\fR to \s-1TPF\s0. We may need volunteers to | |
218 | hold him upside down and shake vigorously in order to actually extract the | |
219 | cash. | |
220 | .PP | |
221 | See \fIt/lib/1_compile.t\fR for the 3 remaining modules that need tests. | |
222 | .Sh "Improve the coverage of the core tests" | |
223 | .IX Subsection "Improve the coverage of the core tests" | |
224 | Use Devel::Cover to ascertain the core's test coverage, then add tests that | |
225 | are currently missing. | |
226 | .Sh "test B" | |
227 | .IX Subsection "test B" | |
228 | A full test suite for the B module would be nice. | |
229 | .Sh "A decent benchmark" | |
230 | .IX Subsection "A decent benchmark" | |
231 | \&\f(CW\*(C`perlbench\*(C'\fR seems impervious to any recent changes made to the perl core. It | |
232 | would be useful to have a reasonable general benchmarking suite that roughly | |
233 | represented what current perl programs do, and measurably reported whether | |
234 | tweaks to the core improve, degrade or don't really affect performance, to | |
235 | guide people attempting to optimise the guts of perl. Gisle would welcome | |
236 | new tests for perlbench. | |
237 | .Sh "fix tainting bugs" | |
238 | .IX Subsection "fix tainting bugs" | |
239 | Fix the bugs revealed by running the test suite with the \f(CW\*(C`\-t\*(C'\fR switch (via | |
240 | \&\f(CW\*(C`make test.taintwarn\*(C'\fR). | |
241 | .Sh "Dual life everything" | |
242 | .IX Subsection "Dual life everything" | |
243 | As part of the \*(L"dists\*(R" plan, anything that doesn't belong in the smallest perl | |
244 | distribution needs to be dual lifed. Anything else can be too. Figure out what | |
245 | changes would be needed to package that module and its tests up for \s-1CPAN\s0, and | |
246 | do so. Test it with older perl releases, and fix the problems you find. | |
247 | .ie n .Sh "Improving ""threads::shared""" | |
248 | .el .Sh "Improving \f(CWthreads::shared\fP" | |
249 | .IX Subsection "Improving threads::shared" | |
250 | Investigate whether \f(CW\*(C`threads::shared\*(C'\fR could share aggregates properly with | |
251 | only Perl level changes to shared.pm | |
252 | .Sh "\s-1POSIX\s0 memory footprint" | |
253 | .IX Subsection "POSIX memory footprint" | |
254 | Ilya observed that use \s-1POSIX\s0; eats memory like there's no tomorrow, and at | |
255 | various times worked to cut it down. There is probably still fat to cut out \- | |
256 | for example \s-1POSIX\s0 passes Exporter some very memory hungry data structures. | |
257 | .SH "Tasks that need a little sysadmin-type knowledge" | |
258 | .IX Header "Tasks that need a little sysadmin-type knowledge" | |
259 | Or if you prefer, tasks that you would learn from, and broaden your skills | |
260 | base... | |
261 | .Sh "Relocatable perl" | |
262 | .IX Subsection "Relocatable perl" | |
263 | The C level patches needed to create a relocatable perl binary are done, as | |
264 | is the work on \fIConfig.pm\fR. All that's left to do is the \f(CW\*(C`Configure\*(C'\fR tweaking | |
265 | to let people specify how they want to do the install. | |
266 | .Sh "make \s-1HTML\s0 install work" | |
267 | .IX Subsection "make HTML install work" | |
268 | There is an \f(CW\*(C`installhtml\*(C'\fR target in the Makefile. It's marked as | |
269 | \&\*(L"experimental\*(R". It would be good to get this tested, make it work reliably, and | |
270 | remove the \*(L"experimental\*(R" tag. This would include | |
271 | .IP "1" 4 | |
272 | .IX Item "1" | |
273 | Checking that cross linking between various parts of the documentation works. | |
274 | In particular that links work between the modules (files with \s-1POD\s0 in \fIlib/\fR) | |
275 | and the core documentation (files in \fIpod/\fR) | |
276 | .IP "2" 4 | |
277 | .IX Item "2" | |
278 | Work out how to split \f(CW\*(C`perlfunc\*(C'\fR into chunks, preferably one per function | |
279 | group, preferably with general case code that could be used elsewhere. | |
280 | Challenges here are correctly identifying the groups of functions that go | |
281 | together, and making the right named external cross-links point to the right | |
282 | page. Things to be aware of are \f(CW\*(C`\-X\*(C'\fR, groups such as \f(CW\*(C`getpwnam\*(C'\fR to | |
283 | \&\f(CW\*(C`endservent\*(C'\fR, two or more \f(CW\*(C`=items\*(C'\fR giving the different parameter lists, such | |
284 | as | |
285 | .Sp | |
286 | .Vb 1 | |
287 | \& =item substr EXPR,OFFSET,LENGTH,REPLACEMENT | |
288 | .Ve | |
289 | .Sp | |
290 | .Vb 1 | |
291 | \& =item substr EXPR,OFFSET,LENGTH | |
292 | .Ve | |
293 | .Sp | |
294 | .Vb 1 | |
295 | \& =item substr EXPR,OFFSET | |
296 | .Ve | |
297 | .Sp | |
298 | and different parameter lists having different meanings. (eg \f(CW\*(C`select\*(C'\fR) | |
299 | .Sh "compressed man pages" | |
300 | .IX Subsection "compressed man pages" | |
301 | Be able to install them. This would probably need a configure test to see how | |
302 | the system does compressed man pages (same directory/different directory? | |
303 | same filename/different filename), as well as tweaking the \fIinstallman\fR script | |
304 | to compress as necessary. | |
305 | .Sh "Add a code coverage target to the Makefile" | |
306 | .IX Subsection "Add a code coverage target to the Makefile" | |
307 | Make it easy for anyone to run Devel::Cover on the core's tests. The steps | |
308 | to do this manually are roughly | |
309 | .IP "\(bu" 4 | |
310 | do a normal \f(CW\*(C`Configure\*(C'\fR, but include Devel::Cover as a module to install | |
311 | (see \fI\s-1INSTALL\s0\fR for how to do this) | |
312 | .IP "\(bu" 4 | |
313 | .Vb 1 | |
314 | \& make perl | |
315 | .Ve | |
316 | .IP "\(bu" 4 | |
317 | .Vb 1 | |
318 | \& cd t; HARNESS_PERL_SWITCHES=-MDevel::Cover ./perl -I../lib harness | |
319 | .Ve | |
320 | .IP "\(bu" 4 | |
321 | Process the resulting Devel::Cover database | |
322 | .PP | |
323 | This just give you the coverage of the \fI.pm\fRs. To also get the C level | |
324 | coverage you need to | |
325 | .IP "\(bu" 4 | |
326 | Additionally tell \f(CW\*(C`Configure\*(C'\fR to use the appropriate C compiler flags for | |
327 | \&\f(CW\*(C`gcov\*(C'\fR | |
328 | .IP "\(bu" 4 | |
329 | .Vb 1 | |
330 | \& make perl.gcov | |
331 | .Ve | |
332 | .Sp | |
333 | (instead of \f(CW\*(C`make perl\*(C'\fR) | |
334 | .IP "\(bu" 4 | |
335 | After running the tests run \f(CW\*(C`gcov\*(C'\fR to generate all the \fI.gcov\fR files. | |
336 | (Including down in the subdirectories of \fIext/\fR | |
337 | .IP "\(bu" 4 | |
338 | (From the top level perl directory) run \f(CW\*(C`gcov2perl\*(C'\fR on all the \f(CW\*(C`.gcov\*(C'\fR files | |
339 | to get their stats into the cover_db directory. | |
340 | .IP "\(bu" 4 | |
341 | Then process the Devel::Cover database | |
342 | .PP | |
343 | It would be good to add a single switch to \f(CW\*(C`Configure\*(C'\fR to specify that you | |
344 | wanted to perform perl level coverage, and another to specify C level | |
345 | coverage, and have \f(CW\*(C`Configure\*(C'\fR and the \fIMakefile\fR do all the right things | |
346 | automatically. | |
347 | .Sh "Make Config.pm cope with differences between build and installed perl" | |
348 | .IX Subsection "Make Config.pm cope with differences between build and installed perl" | |
349 | Quite often vendors ship a perl binary compiled with their (pay\-for) | |
350 | compilers. People install a free compiler, such as gcc. To work out how to | |
351 | build extensions, Perl interrogates \f(CW%Config\fR, so in this situation | |
352 | \&\f(CW%Config\fR describes compilers that aren't there, and extension building | |
353 | fails. This forces people into choosing between re-compiling perl themselves | |
354 | using the compiler they have, or only using modules that the vendor ships. | |
355 | .PP | |
356 | It would be good to find a way teach \f(CW\*(C`Config.pm\*(C'\fR about the installation setup, | |
357 | possibly involving probing at install time or later, so that the \f(CW%Config\fR in | |
358 | a binary distribution better describes the installed machine, when the | |
359 | installed machine differs from the build machine in some significant way. | |
360 | .Sh "make parallel builds work" | |
361 | .IX Subsection "make parallel builds work" | |
362 | Currently parallel builds (such as \f(CW\*(C`make \-j3\*(C'\fR) don't work reliably. We believe | |
363 | that this is due to incomplete dependency specification in the \fIMakefile\fR. | |
364 | It would be good if someone were able to track down the causes of these | |
365 | problems, so that parallel builds worked properly. | |
366 | .Sh "linker specification files" | |
367 | .IX Subsection "linker specification files" | |
368 | Some platforms mandate that you provide a list of a shared library's external | |
369 | symbols to the linker, so the core already has the infrastructure in place to | |
370 | do this for generating shared perl libraries. My understanding is that the | |
371 | \&\s-1GNU\s0 toolchain can accept an optional linker specification file, and restrict | |
372 | visibility just to symbols declared in that file. It would be good to extend | |
373 | \&\fImakedef.pl\fR to support this format, and to provide a means within | |
374 | \&\f(CW\*(C`Configure\*(C'\fR to enable it. This would allow Unix users to test that the | |
375 | export list is correct, and to build a perl that does not pollute the global | |
376 | namespace with private symbols. | |
377 | .SH "Tasks that need a little C knowledge" | |
378 | .IX Header "Tasks that need a little C knowledge" | |
379 | These tasks would need a little C knowledge, but don't need any specific | |
380 | background or experience with \s-1XS\s0, or how the Perl interpreter works | |
381 | .Sh "Make it clear from \-v if this is the exact official release" | |
382 | .IX Subsection "Make it clear from -v if this is the exact official release" | |
383 | Currently perl from \f(CW\*(C`p4\*(C'\fR/\f(CW\*(C`rsync\*(C'\fR ships with a \fIpatchlevel.h\fR file that | |
384 | usually defines one local patch, of the form \*(L"\s-1MAINT12345\s0\*(R" or \*(L"\s-1RC1\s0\*(R". The output | |
385 | of perl \-v doesn't report that a perl isn't an official release, and this | |
386 | information can get lost in bugs reports. Because of this, the minor version | |
387 | isn't bumped up until \s-1RC\s0 time, to minimise the possibility of versions of perl | |
388 | escaping that believe themselves to be newer than they actually are. | |
389 | .PP | |
390 | It would be useful to find an elegant way to have the \*(L"this is an interim | |
391 | maintenance release\*(R" or \*(L"this is a release candidate\*(R" in the terse \-v output, | |
392 | and have it so that it's easy for the pumpking to remove this just as the | |
393 | release tarball is rolled up. This way the version pulled out of rsync would | |
394 | always say \*(L"I'm a development release\*(R" and it would be safe to bump the | |
395 | reported minor version as soon as a release ships, which would aid perl | |
396 | developers. | |
397 | .PP | |
398 | This task is really about thinking of an elegant way to arrange the C source | |
399 | such that it's trivial for the Pumpking to flag \*(L"this is an official release\*(R" | |
400 | when making a tarball, yet leave the default source saying \*(L"I'm not the | |
401 | official release\*(R". | |
402 | .Sh "Tidy up global variables" | |
403 | .IX Subsection "Tidy up global variables" | |
404 | There's a note in \fIintrpvar.h\fR | |
405 | .PP | |
406 | .Vb 4 | |
407 | \& /* These two variables are needed to preserve 5.8.x bincompat because | |
408 | \& we can't change function prototypes of two exported functions. | |
409 | \& Probably should be taken out of blead soon, and relevant prototypes | |
410 | \& changed. */ | |
411 | .Ve | |
412 | .PP | |
413 | So doing this, and removing any of the unused variables still present would | |
414 | be good. | |
415 | .ie n .Sh "Ordering of ""global"" variables." | |
416 | .el .Sh "Ordering of ``global'' variables." | |
417 | .IX Subsection "Ordering of global variables." | |
418 | \&\fIthrdvar.h\fR and \fIintrpvarh\fR define the \*(L"global\*(R" variables that need to be | |
419 | per-thread under ithreads, where the variables are actually elements in a | |
420 | structure. As C dictates, the variables must be laid out in order of | |
421 | declaration. There is a comment | |
422 | \&\f(CW\*(C`/* Important ones in the first cache line (if alignment is done right) */\*(C'\fR | |
423 | which implies that at some point in the past the ordering was carefully chosen | |
424 | (at least in part). However, it's clear that the ordering is less than perfect, | |
425 | as currently there are things such as 7 \f(CW\*(C`bool\*(C'\fRs in a row, then something | |
426 | typically requiring 4 byte alignment, and then an odd \f(CW\*(C`bool\*(C'\fR later on. | |
427 | (\f(CW\*(C`bool\*(C'\fRs are typically defined as \f(CW\*(C`char\*(C'\fRs). So it would be good for someone | |
428 | to review the ordering of the variables, to see how much alignment padding can | |
429 | be removed. | |
430 | .Sh "bincompat functions" | |
431 | .IX Subsection "bincompat functions" | |
432 | There are lots of functions which are retained for binary compatibility. | |
433 | Clean these up. Move them to mathom.c, and don't compile for blead? | |
434 | .Sh "am I hot or not?" | |
435 | .IX Subsection "am I hot or not?" | |
436 | The idea of \fIpp_hot.c\fR is that it contains the \fIhot\fR ops, the ops that are | |
437 | most commonly used. The idea is that by grouping them, their object code will | |
438 | be adjacent in the executable, so they have a greater chance of already being | |
439 | in the \s-1CPU\s0 cache (or swapped in) due to being near another op already in use. | |
440 | .PP | |
441 | Except that it's not clear if these really are the most commonly used ops. So | |
442 | anyone feeling like exercising their skill with coverage and profiling tools | |
443 | might want to determine what ops \fIreally\fR are the most commonly used. And in | |
444 | turn suggest evictions and promotions to achieve a better \fIpp_hot.c\fR. | |
445 | .Sh "emulate the per-thread memory pool on Unix" | |
446 | .IX Subsection "emulate the per-thread memory pool on Unix" | |
447 | For Windows, ithreads allocates memory for each thread from a separate pool, | |
448 | which it discards at thread exit. It also checks that memory is \fIfree()\fRd to | |
449 | the correct pool. Neither check is done on Unix, so code developed there won't | |
450 | be subject to such strictures, so can harbour bugs that only show up when the | |
451 | code reaches Windows. | |
452 | .PP | |
453 | It would be good to be able to optionally emulate the Window pool system on | |
454 | Unix, to let developers who only have access to Unix, or want to use | |
455 | Unix-specific debugging tools, check for these problems. To do this would | |
456 | involve figuring out how the \f(CW\*(C`PerlMem_*\*(C'\fR macros wrap \f(CW\*(C`malloc()\*(C'\fR access, and | |
457 | providing a layer that records/checks the identity of the thread making the | |
458 | call, and recording all the memory allocated by each thread via this \s-1API\s0 so | |
459 | that it can be summarily \fIfree()\fRd at thread exit. One implementation idea | |
460 | would be to increase the size of allocation, and store the \f(CW\*(C`my_perl\*(C'\fR pointer | |
461 | (to identify the thread) at the start, along with pointers to make a linked | |
462 | list of blocks for this thread. To avoid alignment problems it would be | |
463 | necessary to do something like | |
464 | .PP | |
465 | .Vb 7 | |
466 | \& union memory_header_padded { | |
467 | \& struct memory_header { | |
468 | \& void *thread_id; /* For my_perl */ | |
469 | \& void *next; /* Pointer to next block for this thread */ | |
470 | \& } data; | |
471 | \& long double padding; /* whatever type has maximal alignment constraint */ | |
472 | \& }; | |
473 | .Ve | |
474 | .PP | |
475 | although \f(CW\*(C`long double\*(C'\fR might not be the only type to add to the padding | |
476 | union. | |
477 | .Sh "reduce duplication in sv_setsv_flags" | |
478 | .IX Subsection "reduce duplication in sv_setsv_flags" | |
479 | \&\f(CW\*(C`Perl_sv_setsv_flags\*(C'\fR has a comment | |
480 | \&\f(CW\*(C`/* There's a lot of redundancy below but we're going for speed here */\*(C'\fR | |
481 | .PP | |
482 | Whilst this was true 10 years ago, the growing disparity between \s-1RAM\s0 and \s-1CPU\s0 | |
483 | speeds mean that the trade offs have changed. In addition, the duplicate code | |
484 | adds to the maintenance burden. It would be good to see how much of the | |
485 | redundancy can be pruned, particular in the less common paths. (Profiling | |
486 | tools at the ready...). For example, why does the test for | |
487 | \&\*(L"Can't redefine active sort subroutine\*(R" need to occur in two places? | |
488 | .SH "Tasks that need a knowledge of XS" | |
489 | .IX Header "Tasks that need a knowledge of XS" | |
490 | These tasks would need C knowledge, and roughly the level of knowledge of | |
491 | the perl \s-1API\s0 that comes from writing modules that use \s-1XS\s0 to interface to | |
492 | C. | |
493 | .Sh "IPv6" | |
494 | .IX Subsection "IPv6" | |
495 | Clean this up. Check everything in core works | |
496 | .ie n .Sh "shrink ""GV""\fPs, \f(CW""CV""s" | |
497 | .el .Sh "shrink \f(CWGV\fPs, \f(CWCV\fPs" | |
498 | .IX Subsection "shrink GVs, CVs" | |
499 | By removing unused elements and careful re\-ordering, the structures for \f(CW\*(C`AV\*(C'\fRs | |
500 | and \f(CW\*(C`HV\*(C'\fRs have recently been shrunk considerably. It's probable that the same | |
501 | approach would find savings in \f(CW\*(C`GV\*(C'\fRs and \f(CW\*(C`CV\*(C'\fRs, if not all the other | |
502 | larger\-than\-\f(CW\*(C`PVMG\*(C'\fR types. | |
503 | .Sh "merge Perl_sv_2[inpu]v" | |
504 | .IX Subsection "merge Perl_sv_2[inpu]v" | |
505 | There's a lot of code shared between \f(CW\*(C`Perl_sv_2iv_flags\*(C'\fR, | |
506 | \&\f(CW\*(C`Perl_sv_2uv_flags\*(C'\fR, \f(CW\*(C`Perl_sv_2nv\*(C'\fR, and \f(CW\*(C`Perl_sv_2pv_flags\*(C'\fR. It would be | |
507 | interesting to see if some of it can be merged into common shared static | |
508 | functions. In particular, \f(CW\*(C`Perl_sv_2uv_flags\*(C'\fR started out as a cut&paste | |
509 | from \f(CW\*(C`Perl_sv_2iv_flags\*(C'\fR around 5.005_50 time, and it may be possible to | |
510 | replace both with a single function that returns a value or union which is | |
511 | split out by the macros in \fIsv.h\fR | |
512 | .Sh "\s-1UTF8\s0 caching code" | |
513 | .IX Subsection "UTF8 caching code" | |
514 | The string position/offset cache is not optional. It should be. | |
515 | .Sh "Implicit Latin 1 => Unicode translation" | |
516 | .IX Subsection "Implicit Latin 1 => Unicode translation" | |
517 | Conversions from byte strings to \s-1UTF\-8\s0 currently map high bit characters | |
518 | to Unicode without translation (or, depending on how you look at it, by | |
519 | implicitly assuming that the byte strings are in Latin\-1). As perl assumes | |
520 | the C locale by default, upgrading a string to \s-1UTF\-8\s0 may change the | |
521 | meaning of its contents regarding character classes, case mapping, etc. | |
522 | This should probably emit a warning (at least). | |
523 | .PP | |
524 | This task is incremental \- even a little bit of work on it will help. | |
525 | .Sh "autovivification" | |
526 | .IX Subsection "autovivification" | |
527 | Make all autovivification consistent w.r.t \s-1LVALUE/RVALUE\s0 and strict/no strict; | |
528 | .PP | |
529 | This task is incremental \- even a little bit of work on it will help. | |
530 | .Sh "Unicode in Filenames" | |
531 | .IX Subsection "Unicode in Filenames" | |
532 | chdir, chmod, chown, chroot, exec, glob, link, lstat, mkdir, open, | |
533 | opendir, qx, readdir, readlink, rename, rmdir, stat, symlink, sysopen, | |
534 | system, truncate, unlink, utime, \-X. All these could potentially accept | |
535 | Unicode filenames either as input or output (and in the case of system | |
536 | and qx Unicode in general, as input or output to/from the shell). | |
537 | Whether a filesystem \- an operating system pair understands Unicode in | |
538 | filenames varies. | |
539 | .PP | |
540 | Known combinations that have some level of understanding include | |
541 | Microsoft \s-1NTFS\s0, Apple \s-1HFS+\s0 (In Mac \s-1OS\s0 9 and X) and Apple \s-1UFS\s0 (in Mac | |
542 | \&\s-1OS\s0 X), \s-1NFS\s0 v4 is rumored to be Unicode, and of course Plan 9. How to | |
543 | create Unicode filenames, what forms of Unicode are accepted and used | |
544 | (\s-1UCS\-2\s0, \s-1UTF\-16\s0, \s-1UTF\-8\s0), what (if any) is the normalization form used, | |
545 | and so on, varies. Finding the right level of interfacing to Perl | |
546 | requires some thought. Remember that an \s-1OS\s0 does not implicate a | |
547 | filesystem. | |
548 | .PP | |
549 | (The Windows \-C command flag \*(L"wide \s-1API\s0 support\*(R" has been at least | |
550 | temporarily retired in 5.8.1, and the \-C has been repurposed, see | |
551 | perlrun.) | |
552 | .ie n .Sh "Unicode in %ENV" | |
553 | .el .Sh "Unicode in \f(CW%ENV\fP" | |
554 | .IX Subsection "Unicode in %ENV" | |
555 | Currently the \f(CW%ENV\fR entries are always byte strings. | |
556 | .Sh "use less 'memory'" | |
557 | .IX Subsection "use less 'memory'" | |
558 | Investigate trade offs to switch out perl's choices on memory usage. | |
559 | Particularly perl should be able to give memory back. | |
560 | .PP | |
561 | This task is incremental \- even a little bit of work on it will help. | |
562 | .ie n .Sh "Re-implement "":unique"" in a way that is actually thread-safe" | |
563 | .el .Sh "Re-implement \f(CW:unique\fP in a way that is actually thread-safe" | |
564 | .IX Subsection "Re-implement :unique in a way that is actually thread-safe" | |
565 | The old implementation made bad assumptions on several levels. A good 90% | |
566 | solution might be just to make \f(CW\*(C`:unique\*(C'\fR work to share the string buffer | |
567 | of SvPVs. That way large constant strings can be shared between ithreads, | |
568 | such as the configuration information in \fIConfig\fR. | |
569 | .Sh "Make tainting consistent" | |
570 | .IX Subsection "Make tainting consistent" | |
571 | Tainting would be easier to use if it didn't take documented shortcuts and | |
572 | allow taint to \*(L"leak\*(R" everywhere within an expression. | |
573 | .Sh "readpipe(\s-1LIST\s0)" | |
574 | .IX Subsection "readpipe(LIST)" | |
575 | \&\fIsystem()\fR accepts a \s-1LIST\s0 syntax (and a \s-1PROGRAM\s0 \s-1LIST\s0 syntax) to avoid | |
576 | running a shell. \fIreadpipe()\fR (the function behind qx//) could be similarly | |
577 | extended. | |
578 | .SH "Tasks that need a knowledge of the interpreter" | |
579 | .IX Header "Tasks that need a knowledge of the interpreter" | |
580 | These tasks would need C knowledge, and knowledge of how the interpreter works, | |
581 | or a willingness to learn. | |
582 | .Sh "lexical pragmas" | |
583 | .IX Subsection "lexical pragmas" | |
584 | Document the new support for lexical pragmas in 5.9.3 and how %^H works. | |
585 | Maybe \f(CW\*(C`re\*(C'\fR, \f(CW\*(C`encoding\*(C'\fR, maybe other pragmas could be made lexical. | |
586 | .Sh "Attach/detach debugger from running program" | |
587 | .IX Subsection "Attach/detach debugger from running program" | |
588 | The old perltodo notes "With \f(CW\*(C`gdb\*(C'\fR, you can attach the debugger to a running | |
589 | program if you pass the process \s-1ID\s0. It would be good to do this with the Perl | |
590 | debugger on a running Perl program, although I'm not sure how it would be | |
591 | done." ssh and screen do this with named pipes in /tmp. Maybe we can too. | |
592 | .Sh "Constant folding" | |
593 | .IX Subsection "Constant folding" | |
594 | The peephole optimiser should trap errors during constant folding, and give | |
595 | up on the folding, rather than bailing out at compile time. It is quite | |
596 | possible that the unfoldable constant is in unreachable code, eg something | |
597 | akin to \f(CW\*(C`$a = 0/0 if 0;\*(C'\fR | |
598 | .Sh "\s-1LVALUE\s0 functions for lists" | |
599 | .IX Subsection "LVALUE functions for lists" | |
600 | The old perltodo notes that lvalue functions don't work for list or hash | |
601 | slices. This would be good to fix. | |
602 | .Sh "\s-1LVALUE\s0 functions in the debugger" | |
603 | .IX Subsection "LVALUE functions in the debugger" | |
604 | The old perltodo notes that lvalue functions don't work in the debugger. This | |
605 | would be good to fix. | |
606 | .Sh "_ prototype character" | |
607 | .IX Subsection "_ prototype character" | |
608 | Study the possibility of adding a new prototype character, \f(CW\*(C`_\*(C'\fR, meaning | |
609 | \&\*(L"this argument defaults to \f(CW$_\fR\*(R". | |
610 | .Sh "state variables" | |
611 | .IX Subsection "state variables" | |
612 | \&\f(CW\*(C`my $foo if 0;\*(C'\fR is deprecated, and should be replaced with | |
613 | \&\f(CW\*(C`state $x = "initial value\en";\*(C'\fR the syntax from Perl 6. | |
614 | .Sh "@INC source filter to Filter::Simple" | |
615 | .IX Subsection "@INC source filter to Filter::Simple" | |
616 | The second return value from a sub in \f(CW@INC\fR can be a source filter. This isn't | |
617 | documented. It should be changed to use Filter::Simple, tested and documented. | |
618 | .Sh "regexp optimiser optional" | |
619 | .IX Subsection "regexp optimiser optional" | |
620 | The regexp optimiser is not optional. It should configurable to be, to allow | |
621 | its performance to be measured, and its bugs to be easily demonstrated. | |
622 | .Sh "\s-1UNITCHECK\s0" | |
623 | .IX Subsection "UNITCHECK" | |
624 | Introduce a new special block, \s-1UNITCHECK\s0, which is run at the end of a | |
625 | compilation unit (module, file, eval(\s-1STRING\s0) block). This will correspond to | |
626 | the Perl 6 \s-1CHECK\s0. Perl 5's \s-1CHECK\s0 cannot be changed or removed because the | |
627 | O.pm/B.pm backend framework depends on it. | |
628 | .Sh "optional optimizer" | |
629 | .IX Subsection "optional optimizer" | |
630 | Make the peephole optimizer optional. Currently it performs two tasks as | |
631 | it walks the optree \- genuine peephole optimisations, and necessary fixups of | |
632 | ops. It would be good to find an efficient way to switch out the | |
633 | optimisations whilst keeping the fixups. | |
634 | .Sh "You \s-1WANT\s0 *how* many" | |
635 | .IX Subsection "You WANT *how* many" | |
636 | Currently contexts are void, scalar and list. split has a special mechanism in | |
637 | place to pass in the number of return values wanted. It would be useful to | |
638 | have a general mechanism for this, backwards compatible and little speed hit. | |
639 | This would allow proposals such as short circuiting sort to be implemented | |
640 | as a module on \s-1CPAN\s0. | |
641 | .Sh "lexical aliases" | |
642 | .IX Subsection "lexical aliases" | |
643 | Allow lexical aliases (maybe via the syntax \f(CW\*(C`my \e$alias = \e$foo\*(C'\fR. | |
644 | .Sh "entersub \s-1XS\s0 vs Perl" | |
645 | .IX Subsection "entersub XS vs Perl" | |
646 | At the moment pp_entersub is huge, and has code to deal with entering both | |
647 | perl and \s-1XS\s0 subroutines. Subroutine implementations rarely change between | |
648 | perl and \s-1XS\s0 at run time, so investigate using 2 ops to enter subs (one for | |
649 | \&\s-1XS\s0, one for perl) and swap between if a sub is redefined. | |
650 | .Sh "Self ties" | |
651 | .IX Subsection "Self ties" | |
652 | self ties are currently illegal because they caused too many segfaults. Maybe | |
653 | the causes of these could be tracked down and self-ties on all types re\- | |
654 | instated. | |
655 | .ie n .Sh "Optimize away @_" | |
656 | .el .Sh "Optimize away \f(CW@_\fP" | |
657 | .IX Subsection "Optimize away @_" | |
658 | The old perltodo notes \*(L"Look at the \*(R"reification" code in \f(CW\*(C`av.c\*(C'\fR". | |
659 | .Sh "What hooks would assertions need?" | |
660 | .IX Subsection "What hooks would assertions need?" | |
661 | Assertions are in the core, and work. However, assertions needed to be added | |
662 | as a core patch, rather than an \s-1XS\s0 module in ext, or a \s-1CPAN\s0 module, because | |
663 | the core has no hooks in the necessary places. It would be useful to | |
664 | investigate what hooks would need to be added to make it possible to provide | |
665 | the full assertion support from a \s-1CPAN\s0 module, so that we aren't constraining | |
666 | the imagination of future \s-1CPAN\s0 authors. | |
667 | .SH "Big projects" | |
668 | .IX Header "Big projects" | |
669 | Tasks that will get your name mentioned in the description of the \*(L"Highlights | |
670 | of 5.10\*(R" | |
671 | .Sh "make ithreads more robust" | |
672 | .IX Subsection "make ithreads more robust" | |
673 | Generally make ithreads more robust. See also \*(L"iCOW\*(R" | |
674 | .PP | |
675 | This task is incremental \- even a little bit of work on it will help, and | |
676 | will be greatly appreciated. | |
677 | .Sh "iCOW" | |
678 | .IX Subsection "iCOW" | |
679 | Sarathy and Arthur have a proposal for an improved Copy On Write which | |
680 | specifically will be able to \s-1COW\s0 new ithreads. If this can be implemented | |
681 | it would be a good thing. | |
682 | .Sh "(?{...}) closures in regexps" | |
683 | .IX Subsection "(?{...}) closures in regexps" | |
684 | Fix (or rewrite) the implementation of the \f(CW\*(C`/(?{...})/\*(C'\fR closures. | |
685 | .Sh "A re-entrant regexp engine" | |
686 | .IX Subsection "A re-entrant regexp engine" | |
687 | This will allow the use of a regex from inside (?{ }), (??{ }) and | |
688 | (?(?{ })|) constructs. |