cvs commit: ports/devel/adime Makefile ports/x11-wm/icewm Makefile ports/graphics/scr2png Makefile ports/x11/xbindkeys Makefile

Sahil Tandon sahil+cvs-ports at tandon.net
Mon Feb 20 02:47:58 UTC 2012


On Mon, 2012-02-20 at 01:23:36 +0000, Alexey Dokuchaev wrote:

> On Sun, Feb 19, 2012 at 03:39:38PM -0500, Sahil Tandon wrote:
> > On Sun, 2012-02-19 at 19:11:06 +0000, Alexey Dokuchaev wrote:
> > > So far it seems that my assessments are not being counterargued with
> > > valid technical points and facts.
> > 
> > Let me point you, once again, to the thread on freebsd-ports in which
> > arguments have already been made against specifying library numbers in
> > LIB_DEPENDS.  For me, mi@ and dougb@'s arguments are more compelling
> > than yours.
> 
> "Wine argument" does not sound too technical to me.  

I have no clue what that means, but at this juncture it is irrelevant.

> I asked a question, how can one guarantee if some arbitrary port is
> ABI dependent or not.  I have not heard the answer so far.  I myself
> do not see it either.  Thus, the whole point that specifying exact ABI
> version might not be required is moot.  Thus, it's safer to always
> specify it.  At least it allows to catch versions inconsistencies
> earlier.

No, it is not always safer to specify it; and by the time I saw your
latest contribution to this thread, I see that dougb@ has done a
splendid job of trying to explain the opposing viewpoint to you rather
clearly.  I trust that has satiated your desire for people to repeat the
technical components of discussion that had already been archived on
freebsd-ports.

-- 
Sahil Tandon


More information about the cvs-ports mailing list