[PATCH] diff portability fixes

[Stefan Seyfried]

On Tue, 9 Feb 2010 20:09:26 -0600
Rob Landley <rob at landley.net> wrote:

> I was with you up until the word "useful".
> 
> That's a coding style issue.  (We're not the only project to consider those a 
> pain and avoid them.  The compiler's already moving them up to the start of 
> the block and doing lifetime analysis on them anyway, we might as well be 
> explicit about it.)

A lot of Dan's "fixes" are actually fixing coding style as a side
effect ;) For example, I consider

	for (int i = 0; i < j; i++)

to be a "mixed declaration and code" issue. Dan fixed that. :)
I also consider

-	FILE_and_pos_t ft[2] = { { fp[0] }, { fp[1] } };
+	FILE_and_pos_t ft[2];
+	ft[0].ft_fp = fp[0];
+	ft[0].ft_pos = 0;
+	ft[1].ft_fp = fp[1];
+	ft[1].ft_pos = 0;
+

to be an improvement in readability, but maybe that's just me.

So he is doing cleanup of the code - and as a reward, he gets the
occasional compatibility fix for his weird compiler in. I think that's a
fair deal.

Besides him, nobody will ever check patches for issues besides the
current coding style guidelines, so yes, he has to keep an endless
churn of patches coming, but that's up to him. The day he stops being
interested in that compiler, probably nobody will care anymore, and the
patches will stop.

Personally, I found Dan's patches to not be very intrusive, he is doing
useful work, and I think he deserves the reward.

Just my 2 cents, of course
-- 
Stefan Seyfried

"Any ideas, John?"
"Well, surrounding them's out."