Ulrich Weigand wrote:
> Two points:
>
> - Please don't reformat the code originating from Bochs; IMO it's
> much more important to keep clean diffs between the FMW and the
> Bochs versions so as to simplify back-and-forth porting of bug
> fixes and the like, than it is to enforce a common style.
Okay. Though I'd prefer the BOCHS code to be clearer itself ;)
About drivers, I had a look today at the DOSEMU sources,
which also have a pretty complete set of hardware. I don't
know whether there are advantages in using these drivers over
using BOCHS drivers, but it can never hurt to have a look (for
instance, the video code looks a lot better than BOCHS because
direct graphics access in full-screen mode is available).
> - If you want to reformat the FMW code itself, please don't mix up
> format changes with real modifications or new code. Keeping useful
> CVS diffs is very important; while a diff clearly marked as 'reformat
> only' can be completely ignored, a diff that contains two lines of
> crucial semantical change amidst thousands of lines of whitespace-only
> diffs is evil ;-)
>
> [ In fact, there's probably no point for you to send me a patch
> containing only the results of an 'indent' run; I'd suggest we
> agree on a set of indent options and I run it myself on my main
> tree ... ]
That sounds like a very good idea. Do you like my indent options ? ;)
PS I may as well do it myself, I do have CVS access... no need to
include it in a patch. It also doesn't have to be recorded in the
ChangeLog. Of course, it is probably handier to have one person
changing CVS only, even for something trivial like this...
-- Ramon
