[olsr-dev] code style issues

Hannes Gredler (spam-protected)
Mon Mar 26 14:05:58 CEST 2007


On Mon, Mar 26, 2007 at 01:13:16PM +0200, Bernd Petrovitsch wrote:
| On Mon, 2007-03-26 at 12:48 +0200, Hannes Gredler wrote:
| > On Mon, Mar 26, 2007 at 12:26:50PM +0200, Bernd Petrovitsch wrote:
| > | On Mon, 2007-03-26 at 11:04 +0200, Hannes Gredler wrote:
| [...]
| > | >   - little/no use of typedefs.
| > | 
| > | Which is IMHO actually a good thing. No need for another layer of
| > | (syntactic) obfuscation and appending "_type" is IMHO the same effort as
| > | prepending with
| > | "struct " .....
| > 
| > well a foo_t (aka POSIX style e.g. u_int8_t) is shorter than a "struct foo";
| 
| Gotcha;-) The "_t" suffix is reserved for the libc (similar to all
| __.*__ preprocessor #define's). More reserved stuff can be found on e.g.
| http://www.gnu.org/software/libc/manual/html_node/Reserved-Names.html
| Propose another one.

well if you proper prefix (another concern of the current tree)
with somehing like olsr_ the practical implication of a e.g. fictional
type like olsr_spf_result_t is IMO next to nothing.

anyway i hear you - so lets keep with struct foo then.

/hannes
 
| > and i find a e.g. avl_node_t pretty self speaking. i have no hard feelings
| > here either (its just what i am used to do over here).
| 
| Yes, it is also (if not mainly) a question of what people are used to.




More information about the Olsr-dev mailing list