[Freeciv-Dev] Re: AI cleanup - first wave

[vze2zq63@xxxxxxxxxxx]

Raimar Falke wrote:

> On Fri, Nov 30, 2001 at 08:58:55PM +0100, Petr Baudis wrote:
> I think a requirement should be that someone else go through the
> patches

All at once, or as they are submitted?

> How are these changes related to the ongoing style
> discussion?

I imagine they're much more conformant than the current code, so that 
shouldn't be much of an issue.
> I also noticed that you made changes which is the job of
> indent. Like this
>
> -    }
> -    else {
> +    } else {
>
> Another time for the "feed all code through indent" discussion?!

Ugh.  I'm in favor of such a move (though I can certainly see its 
drawbacks), but such discussions always result in much time spent and no 
changes being made.  The question, really, is: is more time going to be 
wasted figuring out code that indent mangles and working around the 
annotate problems, or fixing code that indent could easily handle (like 
in this case)?

> As for the style: we have to add another question:
>
>  A
>   if (is_barbarian(pplayer)) {
>      ....
>   /* If not barbarian, ... */
>   } else {
>
>  B
>   if (is_barbarian(pplayer)) {
>      ....
>   } else { /* If not barbarian, ... */
>
>  C
>   if (is_barbarian(pplayer)) {
>      ....
>   } else {
>     /* If not barbarian, ... */

I have in many places seen Thue's comments go like this:

D
  if (is_isometric) {
    ...
  } else { /* if (is_isometric) */
    ...
  }

I find these very hard to read.  I also find A very hard to read.  B 
would be my choice, but C is good as well.

> As for the question you have asked: we should try one function at a
> time and see how this works.

The problem with this is that if each patch has a one-week lag time 
before being commited it'll take way too long.  If you can handle a 
one-day lagtime, then it should be reasonable.  (How many functions are 
we talking about here, anyway?)
jason