On 05-Apr-98 Chris G. Demetriou wrote:

>> I think it's generally significant if there are programs that needed revision
>> to work right with a revised system call.
>
> Except, those changes are really considered part of the syscall
> change.
>
> In general:
>
>       Make major change, leave things that would break broken.
>
> should be the exception to the rule.  "Make major change" implies (or
> should imply) "and fix anything that needed fixing because of it."

Hmm, in the past I've seen (at least some cases where) the implied was mentioned
in the major change.  That's why I was surprised to hear there was more.

But this is really a niggling point.  It would have been work to dig it all up
anyway, as the related fixes would always be scattered through time.

Besides, that's why I didn't know:  I tend to drop in and out of this mailing
list, and I haven't been following the changes over time.  I just barged in
recently, squinty-eyed and gasping for breath.

But given that I drop in and out of this list, one can only guess how long I'd
last on source-changes.  Which only helps the point about keeping CHANGES more
selective.


--Jim Howard  <jiho@mail.c-zone.net>


----------------------------------
E-Mail: jiho@mail.c-zone.net
Date: 04-Apr-98
Time: 21:39:58

This message was sent by XFMail
----------------------------------