svn commit: r358704 - head/Mk

[Eitan Adler]

On 21 June 2014 09:17, Adam Weinberger <adamw at adamw.org> wrote:
> On 21 Jun, 2014, at 11:57, Eitan Adler <eadler at freebsd.org> wrote:
>
>> On 21 June 2014 07:23, Adam Weinberger <adamw at freebsd.org> wrote:
>>> -GHOSTSCRIPT_DESC?=     Ghostscript PDF support
>>> +GHOSTSCRIPT_DESC?=     Ghostscript support
>>
>> This description is now meaningless.  Can you propose a better alternative?
>
> It’s probably better to be vague and meaningless than wrong.

Just to note: it is perfectly fine to have a specific description
which is correct for most cases but wrong for some.  The description
can be overridden by ports locally.

> Someone will think of a more accurate descriptor, but in the meantime it’s better not to have an inaccurate one.

Alright.


>>> -LDAP_DESC?=            LDAP authentication support
>>> +LDAP_DESC?=            LDAP protocol support
>>
>> What functionality might i gain or lose if I turn this on/off?  LDAP
>> should likely not be a shared description at all.
>
> Sure it should. Tons of ports have LDAP support.

The question is not "do many ports have LDAP support" but "do many
ports provide the same functionality when enabling LDAP" or "Can most
ports describe LDAP support in a similar way that goes beyond just
'LDAP support'"?

> This feels a bit like pedantry, as “Kerberos support” and “Gopher protocol support” and “Unicode support” are no more or less useful, though people who need them will know to enable them.

I didn't like "Gopher protocol support" either but couldn't think of
something better.

“Unicode support” is very bad and it provides absolutely no context as
to what functionality will be gained when turning it on.




-- 
Eitan Adler
Source, Ports, Doc committer
Bugmeister, Ports Security teams