On 12/12/13 10:15, Nikolay Orlyuk wrote:
Agree about *necessary* when you just want your package to work.
But in some cases reason for me to have options is to check different
configurations of package, to reveal some bugs and give feedback to upstream.
If upstream will decide that this is very obscure case they may drop support for
such configuration at all. And I'd prefer to try to force upstream to drop
IPv4-only configurations if those packages have problems. Exherbo forces
upstreams busy with fixing/patching their code and dropping legacy code.
This is exactly one of the cases, where you should go the other route, as
testing specific configurations (especially when something goes wrong) involves
modifications of the exheres quite often.
"Being able to test" is not a valid case for adding an option.
Apart from that, you can also use econfs EXTRA_ECONF to temporarily set a
configure parameter. Overall, I'd prefer a modified exheres.
But for existing packages for which someone already spend some effort to make it
configurable why would we want to drop that option?
That's not the question, the question is: Why would we want that option?
_______________________________________________
Exherbo-dev mailing list
[email protected]
http://lists.exherbo.org/mailman/listinfo/exherbo-dev
