> On Sat, 5 Jul 2014, Ulrich Mueller wrote:
> I don't see any different opinions. The patch included below renames
> all variables whose name starts with two underscores.
Committed to CVS.
pgpiyHzs2sctw.pgp
Description: PGP signature
> On Fri, 4 Jul 2014, Michał Górny wrote:
>> Since names like _USER or _PAM look a little dangerous, I suggest
>> __ECLASS for the variable's name.
>>
>> I have no strong preference for the variable's value. Is testing
>> for a non-empty value sufficient? Then we should assign something
>> si
Dnia 2014-07-04, o godz. 11:26:01
Ulrich Mueller napisał(a):
> Since names like _USER or _PAM look a little dangerous, I suggest
> __ECLASS for the variable's name.
>
> I have no strong preference for the variable's value. Is testing for
> a non-empty value sufficient? Then we should assign some
Hi all,
Several of our eclasses use a mechanism that prevents them from being
inherited more than once. Some eclasses test for a special value [1]:
___ECLASS_ONCE_EUTILS="recur -_+^+_- spank"
while others simply assign 1 and test for a non-empty variable [2]:
_GIT_R3=1
The problem with th
