Debian Codegraph @

README

raw
1 NAME
2 common::sense - save a tree AND a kitten, use common::sense!
3
4 SYNOPSIS
5 use common::sense;
6
7 # Supposed to be mostly the same, with much lower memory usage, as:
8
9 # use utf8;
10 # use strict qw(vars subs);
11 # use feature qw(say state switch);
12 # use feature qw(unicode_strings unicode_eval current_sub fc evalbytes);
13 # no feature qw(array_base);
14 # no warnings;
15 # use warnings qw(FATAL closed threads internal debugging pack
16 # prototype inplace io pipe unpack malloc glob
17 # digit printf layer reserved taint closure semicolon);
18 # no warnings qw(exec newline unopened);
19
20 DESCRIPTION
21 “Nothing is more fairly distributed than common sense: no one thinks
22 he needs more of it than he already has.”
23
24 – René Descartes
25
26 This module implements some sane defaults for Perl programs, as defined
27 by two typical (or not so typical - use your common sense) specimens of
28 Perl coders. In fact, after working out details on which warnings and
29 strict modes to enable and make fatal, we found that we (and our code
30 written so far, and others) fully agree on every option, even though we
31 never used warnings before, so it seems this module indeed reflects a
32 "common" sense among some long-time Perl coders.
33
34 The basic philosophy behind the choices made in common::sense can be
35 summarised as: "enforcing strict policies to catch as many bugs as
36 possible, while at the same time, not limiting the expressive power
37 available to the programmer".
38
39 Two typical examples of how this philosophy is applied in practise is
40 the handling of uninitialised and malloc warnings:
41
42 *uninitialised*
43 "undef" is a well-defined feature of perl, and enabling warnings for
44 using it rarely catches any bugs, but considerably limits you in
45 what you can do, so uninitialised warnings are disabled.
46
47 *malloc*
48 Freeing something twice on the C level is a serious bug, usually
49 causing memory corruption. It often leads to side effects much later
50 in the program and there are no advantages to not reporting this, so
51 malloc warnings are fatal by default.
52
53 Unfortunately, there is no fine-grained warning control in perl, so
54 often whole groups of useful warnings had to be excluded because of a
55 single useless warning (for example, perl puts an arbitrary limit on the
56 length of text you can match with some regexes before emitting a
57 warning, making the whole "regexp" category useless).
58
59 What follows is a more thorough discussion of what this module does, and
60 why it does it, and what the advantages (and disadvantages) of this
61 approach are.
62
63 RATIONALE
64 use utf8
65 While it's not common sense to write your programs in UTF-8, it's
66 quickly becoming the most common encoding, is the designated future
67 default encoding for perl sources, and the most convenient encoding
68 available (you can do really nice quoting tricks...). Experience has
69 shown that our programs were either all pure ascii or utf-8, both of
70 which will stay the same.
71
72 There are few drawbacks to enabling UTF-8 source code by default
73 (mainly some speed hits due to bugs in older versions of perl), so
74 this module enables UTF-8 source code encoding by default.
75
76 use strict qw(subs vars)
77 Using "use strict" is definitely common sense, but "use strict
78 'refs'" definitely overshoots its usefulness. After almost two
79 decades of Perl hacking, we decided that it does more harm than
80 being useful. Specifically, constructs like these:
81
82 @{ $var->[0] }
83
84 Must be written like this (or similarly), when "use strict 'refs'"
85 is in scope, and $var can legally be "undef":
86
87 @{ $var->[0] || [] }
88
89 This is annoying, and doesn't shield against obvious mistakes such
90 as using "", so one would even have to write (at least for the time
91 being):
92
93 @{ defined $var->[0] ? $var->[0] : [] }
94
95 ... which nobody with a bit of common sense would consider writing:
96 clear code is clearly something else.
97
98 Curiously enough, sometimes perl is not so strict, as this works
99 even with "use strict" in scope:
100
101 for (@{ $var->[0] }) { ...
102
103 If that isn't hypocrisy! And all that from a mere program!
104
105 use feature qw(say state given ...)
106 We found it annoying that we always have to enable extra features.
107 If something breaks because it didn't anticipate future changes, so
108 be it. 5.10 broke almost all our XS modules and nobody cared either
109 (or at least I know of nobody who really complained about gratuitous
110 changes - as opposed to bugs).
111
112 Few modules that are not actively maintained work with newer
113 versions of Perl, regardless of use feature or not, so a new major
114 perl release means changes to many modules - new keywords are just
115 the tip of the iceberg.
116
117 If your code isn't alive, it's dead, Jim - be an active maintainer.
118
119 But nobody forces you to use those extra features in modules meant
120 for older versions of perl - common::sense of course works there as
121 well. There is also an important other mode where having additional
122 features by default is useful: commandline hacks and internal use
123 scripts: See "much reduced typing", below.
124
125 There is one notable exception: "unicode_eval" is not enabled by
126 default. In our opinion, "use feature" had one main effect - newer
127 perl versions don't value backwards compatibility and the ability to
128 write modules for multiple perl versions much, after all, you can
129 use feature.
130
131 "unicode_eval" doesn't add a new feature, it breaks an existing
132 function.
133
134 no warnings, but a lot of new errors
135 Ah, the dreaded warnings. Even worse, the horribly dreaded "-w"
136 switch: Even though we don't care if other people use warnings (and
137 certainly there are useful ones), a lot of warnings simply go
138 against the spirit of Perl.
139
140 Most prominently, the warnings related to "undef". There is nothing
141 wrong with "undef": it has well-defined semantics, it is useful, and
142 spitting out warnings you never asked for is just evil.
143
144 The result was that every one of our modules did "no warnings" in
145 the past, to avoid somebody accidentally using and forcing his bad
146 standards on our code. Of course, this switched off all warnings,
147 even the useful ones. Not a good situation. Really, the "-w" switch
148 should only enable warnings for the main program only.
149
150 Funnily enough, perllexwarn explicitly mentions "-w" (and not in a
151 favourable way, calling it outright "wrong"), but standard
152 utilities, such as prove, or MakeMaker when running "make test",
153 still enable them blindly.
154
155 For version 2 of common::sense, we finally sat down a few hours and
156 went through *every single warning message*, identifying - according
157 to common sense - all the useful ones.
158
159 This resulted in the rather impressive list in the SYNOPSIS. When we
160 weren't sure, we didn't include the warning, so the list might grow
161 in the future (we might have made a mistake, too, so the list might
162 shrink as well).
163
164 Note the presence of "FATAL" in the list: we do not think that the
165 conditions caught by these warnings are worthy of a warning, we
166 *insist* that they are worthy of *stopping* your program,
167 *instantly*. They are *bugs*!
168
169 Therefore we consider "common::sense" to be much stricter than "use
170 warnings", which is good if you are into strict things (we are not,
171 actually, but these things tend to be subjective).
172
173 After deciding on the list, we ran the module against all of our
174 code that uses "common::sense" (that is almost all of our code), and
175 found only one occurrence where one of them caused a problem: one of
176 elmex's (unreleased) modules contained:
177
178 $fmt =~ s/([^\s\[]*)\[( [^\]]* )\]/\x0$1\x1$2\x0/xgo;
179
180 We quickly agreed that indeed the code should be changed, even
181 though it happened to do the right thing when the warning was
182 switched off.
183
184 much reduced typing
185 Especially with version 2.0 of common::sense, the amount of
186 boilerplate code you need to add to get *this* policy is daunting.
187 Nobody would write this out in throwaway scripts, commandline hacks
188 or in quick internal-use scripts.
189
190 By using common::sense you get a defined set of policies (ours, but
191 maybe yours, too, if you accept them), and they are easy to apply to
192 your scripts: typing "use common::sense;" is even shorter than "use
193 warnings; use strict; use feature ...".
194
195 And you can immediately use the features of your installed perl,
196 which is more difficult in code you release, but not usually an
197 issue for internal-use code (downgrades of your production perl
198 should be rare, right?).
199
200 mucho reduced memory usage
201 Just using all those pragmas mentioned in the SYNOPSIS together
202 wastes <blink>*776 kilobytes*</blink> of precious memory in my perl,
203 for *every single perl process using our code*, which on our
204 machines, is a lot. In comparison, this module only uses *four*
205 kilobytes (I even had to write it out so it looks like more) of
206 memory on the same platform.
207
208 The money/time/effort/electricity invested in these gigabytes
209 (probably petabytes globally!) of wasted memory could easily save 42
210 trees, and a kitten!
211
212 Unfortunately, until everybody applies more common sense, there will
213 still often be modules that pull in the monster pragmas. But one can
214 hope...
215
216 THERE IS NO 'no common::sense'!!!! !!!! !!
217 This module doesn't offer an unimport. First of all, it wastes even more
218 memory, second, and more importantly, who with even a bit of common
219 sense would want no common sense?
220
221 STABILITY AND FUTURE VERSIONS
222 Future versions might change just about everything in this module. We
223 might test our modules and upload new ones working with newer versions
224 of this module, and leave you standing in the rain because we didn't
225 tell you. In fact, we did so when switching from 1.0 to 2.0, which
226 enabled gobs of warnings, and made them FATAL on top.
227
228 Maybe we will load some nifty modules that try to emulate "say" or so
229 with perls older than 5.10 (this module, of course, should work with
230 older perl versions - supporting 5.8 for example is just common sense at
231 this time. Maybe not in the future, but of course you can trust our
232 common sense to be consistent with, uhm, our opinion).
233
234 WHAT OTHER PEOPLE HAD TO SAY ABOUT THIS MODULE
235 apeiron
236
237 "... wow"
238 "I hope common::sense is a joke."
239
240 crab
241
242 "i wonder how it would be if joerg schilling wrote perl modules."
243
244 Adam Kennedy
245
246 "Very interesting, efficient, and potentially something I'd use all the time."
247 [...]
248 "So no common::sense for me, alas."
249
250 H.Merijn Brand
251
252 "Just one more reason to drop JSON::XS from my distribution list"
253
254 Pista Palo
255
256 "Something in short supply these days..."
257
258 Steffen Schwigon
259
260 "This module is quite for sure *not* just a repetition of all the other
261 'use strict, use warnings'-approaches, and it's also not the opposite.
262 [...] And for its chosen middle-way it's also not the worst name ever.
263 And everything is documented."
264
265 BKB
266
267 "[Deleted - thanks to Steffen Schwigon for pointing out this review was
268 in error.]"
269
270 Somni
271
272 "the arrogance of the guy"
273 "I swear he tacked somenoe else's name onto the module
274 just so he could use the royal 'we' in the documentation"
275
276 Anonymous Monk
277
278 "You just gotta love this thing, its got META.json!!!"
279
280 dngor
281
282 "Heh. '"<elmex at ta-sa.org>"' The quotes are semantic
283 distancing from that e-mail address."
284
285 Jerad Pierce
286
287 "Awful name (not a proper pragma), and the SYNOPSIS doesn't tell you
288 anything either. Nor is it clear what features have to do with "common
289 sense" or discipline."
290
291 acme
292
293 "THERE IS NO 'no common::sense'!!!! !!!! !!"
294
295 apeiron (meta-comment about us commenting^Wquoting his comment)
296
297 "How about quoting this: get a clue, you fucktarded amoeba."
298
299 quanth
300
301 "common sense is beautiful, json::xs is fast, Anyevent, EV are fast and
302 furious. I love mlehmannware ;)"
303
304 apeiron
305
306 "... it's mlehmann's view of what common sense is. His view of common
307 sense is certainly uncommon, insofar as anyone with a clue disagrees
308 with him."
309
310 apeiron (another meta-comment)
311
312 "apeiron wonders if his little informant is here to steal more quotes"
313
314 ew73
315
316 "... I never got past the SYNOPSIS before calling it shit."
317 [...]
318 How come no one ever quotes me. :("
319
320 chip (not willing to explain his cryptic questions about links in
321 Changes files)
322
323 "I'm willing to ask the question I've asked. I'm not willing to go
324 through the whole dance you apparently have choreographed. Either
325 answer the completely obvious question, or tell me to fuck off again."
326
327 FREQUENTLY ASKED QUESTIONS
328 Or frequently-come-up confusions.
329
330 Is this module meant to be serious?
331 Yes, we would have put it under the "Acme::" namespace otherwise.
332
333 But the manpage is written in a funny/stupid/... way?
334 This was meant to make it clear that our common sense is a
335 subjective thing and other people can use their own notions, taking
336 the steam out of anybody who might be offended (as some people are
337 always offended no matter what you do).
338
339 This was a failure.
340
341 But we hope the manpage still is somewhat entertaining even though
342 it explains boring rationale.
343
344 Why do you impose your conventions on my code?
345 For some reason people keep thinking that "common::sense" imposes
346 process-wide limits, even though the SYNOPSIS makes it clear that it
347 works like other similar modules - i.e. only within the scope that
348 "use"s them.
349
350 So, no, we don't - nobody is forced to use this module, and using a
351 module that relies on common::sense does not impose anything on you.
352
353 Why do you think only your notion of common::sense is valid?
354 Well, we don't, and have clearly written this in the documentation
355 to every single release. We were just faster than anybody else
356 w.r.t. to grabbing the namespace.
357
358 But everybody knows that you have to use strict and use warnings, why do
359 you disable them?
360 Well, we don't do this either - we selectively disagree with the
361 usefulness of some warnings over others. This module is aimed at
362 experienced Perl programmers, not people migrating from other
363 languages who might be surprised about stuff such as "undef". On the
364 other hand, this does not exclude the usefulness of this module for
365 total newbies, due to its strictness in enforcing policy, while at
366 the same time not limiting the expressive power of perl.
367
368 This module is considerably *more* strict than the canonical "use
369 strict; use warnings", as it makes all its warnings fatal in nature,
370 so you can not get away with as many things as with the canonical
371 approach.
372
373 This was not implemented in version 1.0 because of the daunting
374 number of warning categories and the difficulty in getting exactly
375 the set of warnings you wish (i.e. look at the SYNOPSIS in how
376 complicated it is to get a specific set of warnings - it is not
377 reasonable to put this into every module, the maintenance effort
378 would be enormous).
379
380 But many modules "use strict" or "use warnings", so the memory savings
381 do not apply?
382 I suddenly feel sad...
383
384 But yes, that's true. Fortunately "common::sense" still uses only a
385 miniscule amount of RAM.
386
387 But it adds another dependency to your modules!
388 It's a fact, yeah. But it's trivial to install, most popular modules
389 have many more dependencies. And we consider dependencies a good
390 thing - it leads to better APIs, more thought about interworking of
391 modules and so on.
392
393 Why do you use JSON and not YAML for your META.yml?
394 This is not true - YAML supports a large subset of JSON, and this
395 subset is what META.yml is written in, so it would be correct to say
396 "the META.yml is written in a common subset of YAML and JSON".
397
398 The META.yml follows the YAML, JSON and META.yml specifications, and
399 is correctly parsed by CPAN, so if you have trouble with it, the
400 problem is likely on your side.
401
402 But! But!
403 Yeah, we know.
404
405 AUTHOR
406 Marc Lehmann <[email protected]>
407 http://home.schmorp.de/
408
409 Robin Redeker, "<elmex at ta-sa.org>".
410