Re: Re: cvs: php-src(PHP_5_3) / NEWS /ext/sockets sockets.c

1. This should be applied to PHP_5_2 (not big change and in no way can break anything) 2. You can always simply put a short comment like "Fixed in CVS, needs documenting" and reclassify the bug report as docu issue. There's no need for some fancy shortcuts or such. Just tune your process.. --Jani Arnaud Le Blanc wrote:

On Thursday 23 October 2008 08:54:19 Hannes Magnusson wrote:

On Wed, Oct 22, 2008 at 20:59, Arnaud Le Blanc <lbarnaud@php.net> wrote:

lbarnaud                Wed Oct 22 18:59:34 2008 UTC

 Modified files:              (Branch: PHP_5_3)
   /php-src    NEWS
   /php-src/ext/sockets        sockets.c

Log: MFH: Fixed bug #46360 (TCP_NODELAY constant for socket_{get,set}_option) http://cvs.php.net/viewvc.cgi/php-

src/NEWS?r1=1.2027.2.547.2.965.2.349&r2=1.2027.2.547.2.965.2.350&diff_format=u

Index: php-src/NEWS diff -u php-src/NEWS:1.2027.2.547.2.965.2.349 php-

src/NEWS:1.2027.2.547.2.965.2.350

--- php-src/NEWS:1.2027.2.547.2.965.2.349       Tue Oct 21 23:22:00 2008
+++ php-src/NEWS        Wed Oct 22 18:59:33 2008

@@ -18,6 +18,8 @@ - Fixed bug causing the algorithm parameter of mhash() to be modified.

(Scott)

+- Fixed bug #46360 (TCP_NODELAY constant for socket_{get,set}_option). + (bugs at trick dot vanstaveren dot us) - Fixed bug #46238 (Segmentation fault on static call with empty string

method).

(Felipe) - Fixed bug #46205 (Closure - Memory leaks when ReflectionException is

thrown).

http://cvs.php.net/viewvc.cgi/php-

src/ext/sockets/sockets.c?r1=1.171.2.9.2.14.2.7&r2=1.171.2.9.2.14.2.8&diff_format=u

Index: php-src/ext/sockets/sockets.c diff -u php-src/ext/sockets/sockets.c:1.171.2.9.2.14.2.7 php-

src/ext/sockets/sockets.c:1.171.2.9.2.14.2.8

--- php-src/ext/sockets/sockets.c:1.171.2.9.2.14.2.7    Tue Oct 21 

23:39:15 2008

+++ php-src/ext/sockets/sockets.c Wed Oct 22 18:59:33 2008 @@ -19,7 +19,7 @@

+----------------------------------------------------------------------+

*/ -/* $Id: sockets.c,v 1.171.2.9.2.14.2.7 2008/10/21 23:39:15 lbarnaud Exp $

*/

+/* $Id: sockets.c,v 1.171.2.9.2.14.2.8 2008/10/22 18:59:33 lbarnaud Exp $

*/

#ifdef HAVE_CONFIG_H #include "config.h" @@ -661,6 +661,9 @@

       REGISTER_LONG_CONSTANT("SO_ERROR",              SO_ERROR,               

CONST_CS | CONST_PERSISTENT);

REGISTER_LONG_CONSTANT("SOL_SOCKET", SOL_SOCKET,

CONST_CS | CONST_PERSISTENT);

REGISTER_LONG_CONSTANT("SOMAXCONN", SOMAXCONN,

CONST_CS | CONST_PERSISTENT);

+#ifdef TCP_NODELAY

+       REGISTER_LONG_CONSTANT("TCP_NODELAY",   TCP_NODELAY,    CONST_CS | 

CONST_PERSISTENT);

+#endif

I have cried so many times. Does anyone have any idea how we can help docwriters find commits that need to be documented? Personally I would like _all_ changes in functionality, new stuff and removed stuff to be attached to an bug report which then gets turned into a documentation bug report referencing the commit and preferably with a quick explanation from the developer on what changed including an example of usage. That would mean whenever you close a bug report you would get an extra step, asking if "the fix needs documenting" which would require answer and filling in the comment field before the report could be closed. That at least forces you to think, which the [DOC] tagging did not (and is probably the reason why it failed) and makes it 100% times easier for doc guys to do their work. -Hannes

Hi, Sorry, I did not think to add a [DOC] tag in the commit message.

That would mean whenever you close a bug report you would get an extra step, asking if "the fix needs documenting" which would require answer and filling in the comment field before the report could be closed.

+1 :) Or at least something like "Convert to Documentation bug" close the the "Submit" button on the bug form. Also, CC-ing the phpdoc mailing list while creating a "Documentation" bug, or while closing a "Feature request" bug. Regards, Arnaud