> "TA" == Thomas Anders <[EMAIL PROTECTED]> writes:
TA> If there's *really* no way to solve the circular dependencies
TA> between libnetsnmpagent, libnetsnmphelpers (and libnetsnmpmibs?),
TA> then I'd have certain sympathy for collapsing them into one,
TA> because I wouldn't see any benefit of
Wes Hardaker wrote:
> A possible and easy solution is to put the helper code into libagent
> instead and leave a libhelpers around with about nothing in it (but a
> bogus function).
If there's *really* no way to solve the circular dependencies between
libnetsnmpagent, libnetsnmphelpers (and libne
Wes Hardaker wrote:
> TA> None of the areas touched with this flag (SENSORLIBS, LIBCRYPTO) have
> TA> ever caused any problems for me, so I think you've turned something off
> TA> for no good reason.
>
> I believe they did as reported by someone on something I now forget.
Is "Someone reported som
> "TA" == Thomas Anders <[EMAIL PROTECTED]> writes:
TA> Wes Hardaker wrote:
>> I refuse to release code that may work in some cases and not others
>> and there is no clear way to know or warn a user in advance. To that
>> effect I've now put the as-needed support under a
>> must-turn-on-to-ge
Wes Hardaker wrote:
> I refuse to release code that may work in some cases and not others
> and there is no clear way to know or warn a user in advance. To that
> effect I've now put the as-needed support under a
> must-turn-on-to-get-it flag (--enable-as-needed).
None of the areas touched with
> "TA" == Thomas Anders <[EMAIL PROTECTED]> writes:
>> 8) "as-needed" linking
>> Again, I haven't checked the relevant bugs, but this feels as
>> if it might
>> be a more significant issue than many of the others.
>> How damaging would it be to ship things as they currently stand?
TA> OSX is
