1 |
=head1 NAME |
2 |
|
3 |
common::sense - save a tree AND a kitten, use common::sense! |
4 |
|
5 |
=head1 SYNOPSIS |
6 |
|
7 |
use common::sense; |
8 |
|
9 |
# supposed to be the same, with much lower memory usage, as: |
10 |
# |
11 |
# use strict qw(vars subs); |
12 |
# use feature qw(say state switch); |
13 |
# no warnings; |
14 |
# use warnings qw(FATAL closed threads internal debugging pack substr malloc |
15 |
# unopened portable prototype inplace io pipe unpack regexp |
16 |
# deprecated exiting glob digit printf utf8 layer |
17 |
# reserved parenthesis taint closure semicolon); |
18 |
# no warnings qw(exec newline); |
19 |
|
20 |
=head1 DESCRIPTION |
21 |
|
22 |
This module implements some sane defaults for Perl programs, as defined by |
23 |
two typical (or not so typical - use your common sense) specimens of Perl |
24 |
coders. In fact, after working out details on which warnings and strict |
25 |
modes to enable and make fatal, we found that we (and our code written so |
26 |
far, and others) fully agree on every option, even though we never used |
27 |
warnings before, so it seems this module indeed reflects a "common" sense |
28 |
among some long-time Perl coders. |
29 |
|
30 |
The basic philosophy behind the choices made in common::sense can be |
31 |
summarised as: "enforcing strict policies to catch as many bugs as |
32 |
possible, while at the same time, not limiting the expressive power |
33 |
available to the programmer". |
34 |
|
35 |
Two typical examples of this philosophy are uninitialised and malloc |
36 |
warnings: |
37 |
|
38 |
C<undef> is a well-defined feature of perl, and enabling warnings for |
39 |
using it rarely catches any bugs, but considerably limits you in what you |
40 |
can do, so uninitialised warnings are disabled. |
41 |
|
42 |
Freeing something twice on the C level is a serious bug, usually causing |
43 |
memory corruption. It often leads to side effects much later in the |
44 |
program and there are no advantages to not reporting this, so malloc |
45 |
warnings are fatal by default. |
46 |
|
47 |
What follows is a more thorough discussion of what this module does, |
48 |
and why it does it, and what the advantages (and disadvantages) of this |
49 |
approach are. |
50 |
|
51 |
=over 4 |
52 |
|
53 |
=item use strict qw(subs vars) |
54 |
|
55 |
Using C<use strict> is definitely common sense, but C<use strict |
56 |
'refs'> definitely overshoots its usefulness. After almost two |
57 |
decades of Perl hacking, we decided that it does more harm than being |
58 |
useful. Specifically, constructs like these: |
59 |
|
60 |
@{ $var->[0] } |
61 |
|
62 |
Must be written like this (or similarly), when C<use strict 'refs'> is in |
63 |
scope, and C<$var> can legally be C<undef>: |
64 |
|
65 |
@{ $var->[0] || [] } |
66 |
|
67 |
This is annoying, and doesn't shield against obvious mistakes such as |
68 |
using C<"">, so one would even have to write (at least for the time |
69 |
being): |
70 |
|
71 |
@{ defined $var->[0] ? $var->[0] : [] } |
72 |
|
73 |
... which nobody with a bit of common sense would consider |
74 |
writing: clear code is clearly something else. |
75 |
|
76 |
Curiously enough, sometimes perl is not so strict, as this works even with |
77 |
C<use strict> in scope: |
78 |
|
79 |
for (@{ $var->[0] }) { ... |
80 |
|
81 |
If that isn't hypocrisy! And all that from a mere program! |
82 |
|
83 |
|
84 |
=item use feature qw(say state given) |
85 |
|
86 |
We found it annoying that we always have to enable extra features. If |
87 |
something breaks because it didn't anticipate future changes, so be |
88 |
it. 5.10 broke almost all our XS modules and nobody cared either (or at |
89 |
least I know of nobody who really complained about gratuitous changes - |
90 |
as opposed to bugs). |
91 |
|
92 |
Few modules that are not actively maintained work with newer versions of |
93 |
Perl, regardless of use feature or not, so a new major perl release means |
94 |
changes to many modules - new keywords are just the tip of the iceberg. |
95 |
|
96 |
If your code isn't alive, it's dead, Jim - be an active maintainer. |
97 |
|
98 |
But nobody forces you to use those extra features in modules meant for |
99 |
older versions of perl - common::sense of course works there as well. |
100 |
There is also an important other mode where having additional features by |
101 |
default is useful: commandline hacks and internal use scripts: See "much |
102 |
reduced typing", below. |
103 |
|
104 |
|
105 |
=item no warnings, but a lot of new errors |
106 |
|
107 |
Ah, the dreaded warnings. Even worse, the horribly dreaded C<-w> |
108 |
switch: Even though we don't care if other people use warnings (and |
109 |
certainly there are useful ones), a lot of warnings simply go against the |
110 |
spirit of Perl. |
111 |
|
112 |
Most prominently, the warnings related to C<undef>. There is nothing wrong |
113 |
with C<undef>: it has well-defined semantics, it is useful, and spitting |
114 |
out warnings you never asked for is just evil. |
115 |
|
116 |
The result was that every one of our modules did C<no warnings> in the |
117 |
past, to avoid somebody accidentally using and forcing his bad standards |
118 |
on our code. Of course, this switched off all warnings, even the useful |
119 |
ones. Not a good situation. Really, the C<-w> switch should only enable |
120 |
warnings for the main program only. |
121 |
|
122 |
Funnily enough, L<perllexwarn> explicitly mentions C<-w> (and not in a |
123 |
favourable way, calling it outright "wrong"), but standard utilities, such |
124 |
as L<prove>, or MakeMaker when running C<make test>, still enable them |
125 |
blindly. |
126 |
|
127 |
For version 2 of common::sense, we finally sat down a few hours and went |
128 |
through I<every single warning message>, identifiying - according to |
129 |
common sense - all the useful ones. |
130 |
|
131 |
This resulted in the rather impressive list in the SYNOPSIS. When we |
132 |
weren't sure, we didn't include the warning, so the list might grow in |
133 |
the future (we might have made a mistake, too, so the list might shrink |
134 |
as well). |
135 |
|
136 |
Note the presence of C<FATAL> in the list: we do not think that the |
137 |
conditions caught by these warnings are worthy of a warning, we I<insist> |
138 |
that they are worthy of I<stopping> your program, I<instantly>. They are |
139 |
I<bugs>! |
140 |
|
141 |
Therefore we consider C<common::sense> to be much stricter than C<use |
142 |
warnings>, which is good if you are into strict things (we are not, |
143 |
actually, but these things tend to be subjective). |
144 |
|
145 |
After deciding on the list, we ran the module against all of our code that |
146 |
uses C<common::sense> (that is almost all of our code), and found only one |
147 |
occurence where one of them caused a problem: one of elmex's (unreleased) |
148 |
modules contained: |
149 |
|
150 |
$fmt =~ s/([^\s\[]*)\[( [^\]]* )\]/\x0$1\x1$2\x0/xgo; |
151 |
|
152 |
We quickly agreed that indeed the code should be changed, even though it |
153 |
happened to do the right thing when the warning was switched off. |
154 |
|
155 |
|
156 |
=item much reduced typing |
157 |
|
158 |
Especially with version 2.0 of common::sense, the amount of boilerplate |
159 |
code you need to add to gte I<this> policy is daunting. Nobody would write |
160 |
this out in throwaway scripts, commandline hacks or in quick internal-use |
161 |
scripts. |
162 |
|
163 |
By using common::sense you get a defined set of policies (ours, but maybe |
164 |
yours, too, if you accept them), and they are easy to apply to your |
165 |
scripts: typing C<use common::sense;> is even shorter than C<use warnings; |
166 |
use strict; use feature ...>. |
167 |
|
168 |
And you can immediately use the features of your installed perl, which |
169 |
is more difficult in code you release, but not usually an issue for |
170 |
internal-use code (downgrades of your production perl should be rare, |
171 |
right?). |
172 |
|
173 |
|
174 |
=item mucho reduced memory usage |
175 |
|
176 |
Just using all those pragmas mentioned in the SYNOPSIS together wastes |
177 |
<blink>I<< B<776> kilobytes >></blink> of precious memory in my perl, for |
178 |
I<every single perl process using our code>, which on our machines, is a |
179 |
lot. In comparison, this module only uses I<< B<four> >> kilobytes (I even |
180 |
had to write it out so it looks like more) of memory on the same platform. |
181 |
|
182 |
The money/time/effort/electricity invested in these gigabytes (probably |
183 |
petabytes globally!) of wasted memory could easily save 42 trees, and a |
184 |
kitten! |
185 |
|
186 |
Unfortunately, until everybods applies more common sense, there will still |
187 |
often be modules that pull in the monster pragmas. But one can hope... |
188 |
|
189 |
=cut |
190 |
|
191 |
package common::sense; |
192 |
|
193 |
our $VERSION = '2.02'; |
194 |
|
195 |
# paste this into perl to find bitmask |
196 |
|
197 |
# no warnings; |
198 |
# use warnings qw(FATAL closed threads internal debugging pack substr malloc unopened portable prototype |
199 |
# inplace io pipe unpack regexp deprecated exiting glob digit printf |
200 |
# utf8 layer reserved parenthesis taint closure semicolon); |
201 |
# no warnings qw(exec newline); |
202 |
# BEGIN { warn join "", map "\\x$_", unpack "(H2)*", ${^WARNING_BITS}; exit 0 }; |
203 |
|
204 |
# overload should be included |
205 |
|
206 |
sub import { |
207 |
# verified with perl 5.8.0, 5.10.0 |
208 |
${^WARNING_BITS} ^= ${^WARNING_BITS} ^ "\xfc\x3f\xf3\x00\x0f\xf3\xcf\xc0\xf3\xfc\x33\x03"; |
209 |
|
210 |
# use strict vars subs |
211 |
$^H |= 0x00000600; |
212 |
|
213 |
# use feature |
214 |
$^H{feature_switch} = |
215 |
$^H{feature_say} = |
216 |
$^H{feature_state} = 1; |
217 |
} |
218 |
|
219 |
1; |
220 |
|
221 |
=back |
222 |
|
223 |
=head1 THERE IS NO 'no common::sense'!!!! !!!! !! |
224 |
|
225 |
This module doesn't offer an unimport. First of all, it wastes even more |
226 |
memory, second, and more importantly, who with even a bit of common sense |
227 |
would want no common sense? |
228 |
|
229 |
=head1 STABILITY AND FUTURE VERSIONS |
230 |
|
231 |
Future versions might change just about everything in this module. We |
232 |
might test our modules and upload new ones working with newer versions of |
233 |
this module, and leave you standing in the rain because we didn't tell |
234 |
you. In fact, we did so when switching from 1.0 to 2.0, which enabled gobs |
235 |
of warnings, and made them FATAL on top. |
236 |
|
237 |
Maybe we will load some nifty modules that try to emulate C<say> or so |
238 |
with perls older than 5.10 (this module, of course, should work with older |
239 |
perl versions - supporting 5.8 for example is just common sense at this |
240 |
time. Maybe not in the future, but of course you can trust our common |
241 |
sense to be consistent with, uhm, our opinion). |
242 |
|
243 |
=head1 WHAT OTHER PEOPLE HAD TO SAY ABOUT THIS MODULE |
244 |
|
245 |
apeiron |
246 |
|
247 |
"... wow" |
248 |
"I hope common::sense is a joke." |
249 |
|
250 |
crab |
251 |
|
252 |
"i wonder how it would be if joerg schilling wrote perl modules." |
253 |
|
254 |
Adam Kennedy |
255 |
|
256 |
"Very interesting, efficient, and potentially something I'd use all the time." |
257 |
[...] |
258 |
"So no common::sense for me, alas." |
259 |
|
260 |
H.Merijn Brand |
261 |
|
262 |
"Just one more reason to drop JSON::XS from my distribution list" |
263 |
|
264 |
Pista Palo |
265 |
|
266 |
"Something in short supply these days..." |
267 |
|
268 |
Steffen Schwigon |
269 |
|
270 |
"This module is quite for sure *not* just a repetition of all the other |
271 |
'use strict, use warnings'-approaches, and it's also not the opposite. |
272 |
[...] And for its chosen middle-way it's also not the worst name ever. |
273 |
And everything is documented." |
274 |
|
275 |
BKB |
276 |
|
277 |
"[Deleted - thanks to Steffen Schwigon for pointing out this review was |
278 |
in error.]" |
279 |
|
280 |
Somni |
281 |
|
282 |
"the arrogance of the guy" |
283 |
"I swear he tacked somenoe else's name onto the module |
284 |
just so he could use the royal 'we' in the documentation" |
285 |
|
286 |
Anonymous Monk |
287 |
|
288 |
"You just gotta love this thing, its got META.json!!!" |
289 |
|
290 |
dngor |
291 |
|
292 |
"Heh. '"<elmex at ta-sa.org>"' The quotes are semantic |
293 |
distancing from that e-mail address." |
294 |
|
295 |
Jerad Pierce |
296 |
|
297 |
"Awful name (not a proper pragma), and the SYNOPSIS doesn't tell you |
298 |
anything either. Nor is it clear what features have to do with "common |
299 |
sense" or discipline." |
300 |
|
301 |
acme |
302 |
|
303 |
"THERE IS NO 'no common::sense'!!!! !!!! !!" |
304 |
|
305 |
apeiron (meta-comment about us commenting^Wquoting his comment) |
306 |
|
307 |
"How about quoting this: get a clue, you fucktarded amoeba." |
308 |
|
309 |
quanth |
310 |
|
311 |
"common sense is beautiful, json::xs is fast, Anyevent, EV are fast and |
312 |
furious. I love mlehmannware ;)" |
313 |
|
314 |
=head1 FREQUQNTLY ASKED QUESTIONS |
315 |
|
316 |
Or frequently-come-up confusions. |
317 |
|
318 |
=over 4 |
319 |
|
320 |
=item Is this module meant to be serious? |
321 |
|
322 |
Yes, we would have put it under the C<Acme::> namespace otherwise. |
323 |
|
324 |
=item But the manpage is written in a funny/stupid/... way? |
325 |
|
326 |
This was meant to make it clear that our common sense is a subjective |
327 |
thing and other people can use their own notions, taking the steam out |
328 |
of anybody who might be offended (as some people are always offended no |
329 |
matter what you do). |
330 |
|
331 |
This was a failure. |
332 |
|
333 |
But we hope the manpage still is somewhat entertaining even though it |
334 |
explains boring rationale. |
335 |
|
336 |
=item Why do you impose your conventions on my code? |
337 |
|
338 |
For some reason people keep thinking that C<common::sense> imposes |
339 |
process-wide limits, even though the SYNOPSIS makes it clear that it works |
340 |
like other similar modules - only on the scope that uses them. |
341 |
|
342 |
So, no, we don't - nobody is forced to use this module, and using a module |
343 |
that relies on common::sense does not impose anything on you. |
344 |
|
345 |
=item Why do you think only your notion of common::sense is valid? |
346 |
|
347 |
Well, we don't, and have clearly written this in the documentation to |
348 |
every single release. We were just faster than anybody else w.r.t. to |
349 |
grabbing the namespace. |
350 |
|
351 |
=item But everybody knows that you have to use strict and use warnings, |
352 |
why do you disable them? |
353 |
|
354 |
Well, we don't do this either - we selectively disagree with the |
355 |
usefulness of some warnings over others. This module is aimed at |
356 |
experienced Perl programmers, not people migrating from other languages |
357 |
who might be surprised about stuff such as C<undef>. On the other hand, |
358 |
this does not exclude the usefulness of this module for total newbies, due |
359 |
to its strictness in enforcing policy, while at the same time not limiting |
360 |
the expresive power of perl. |
361 |
|
362 |
This module is considerably I<more> strict than the canonical C<use |
363 |
strict; use warnings>, as it makes all its warnings fatal in nature, so |
364 |
you can not get away with as many things as with the canonical approach. |
365 |
|
366 |
This was not implemented in version 1.0 because of the daunting number |
367 |
of warning categories and the difficulty in getting exactly the set of |
368 |
warnings you wish (i.e. look at the SYNOPSIS in how complicated it is to |
369 |
get a specific set of warnings - it is not reasonable to put this into |
370 |
every module, the maintainance effort would be enourmous). |
371 |
|
372 |
=item But many modules C<use strict> or C<use warnings>, so the memory |
373 |
savings do not apply? |
374 |
|
375 |
I am suddenly so sad. |
376 |
|
377 |
But yes, that's true. Fortunately C<common::sense> still uses only a |
378 |
miniscule amount of RAM. |
379 |
|
380 |
=item But it adds another dependency to your modules! |
381 |
|
382 |
It's a fact, yeah. But it's trivial to install, most popular modules have |
383 |
many more dependencies and we consider dependencies a good thing - it |
384 |
leads to better APIs, more thought about interworking of modules and so |
385 |
on. |
386 |
|
387 |
=item Why do you use JSON and not YAML for your META.yml? |
388 |
|
389 |
This is not true - YAML supports a large subset of JSON, and this subset |
390 |
is what META.yml is written in, so it would be correct to say "the |
391 |
META.yml is written in a common subset of YAML and JSON". |
392 |
|
393 |
The META.yml follows the YAML, JSON and META.yml specifications, and is |
394 |
correctly parsed by CPAN, so if you have trouble with it, the problem is |
395 |
likely on your side. |
396 |
|
397 |
=item But! But! |
398 |
|
399 |
Yeah, we know. |
400 |
|
401 |
=back |
402 |
|
403 |
=head1 AUTHOR |
404 |
|
405 |
Marc Lehmann <schmorp@schmorp.de> |
406 |
http://home.schmorp.de/ |
407 |
|
408 |
Robin Redeker, "<elmex at ta-sa.org>". |
409 |
|
410 |
=cut |
411 |
|