Date: Sun, 21 Mar 2004 10:12:37 +0100 From: Oliver Eikemeier <eikemeier@fillmore-labs.com> To: Doug Barton <DougB@FreeBSD.org> Cc: Sergey Matveychuk <sem@ciam.ru> Subject: Re: mozilla and bind Message-ID: <405D5C85.10706@fillmore-labs.com> In-Reply-To: <405D0268.4070901@FreeBSD.org> References: <405B9799.5000008@ciam.ru> <20040319184525.E85086@qbhto.arg> <405BB3D7.5090905@fillmore-labs.com> <405D0268.4070901@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Doug Barton wrote:
> Oliver Eikemeier wrote:
>
>>Since this may still overwrite the include files without changing the
>>PKGNAME,
>>it seems like a good idea to keep the build conflicts with bind8-base
>>ports.
>>
>>Moreover we can run into problems if the port is installed with
>>PORT_REPLACES_BASE_BIND8_INCLUDES, without PORT_REPLACES_BASE_BIND8 and
>>/usr/local/include is in the path.
>
> The version I committed this morning included defining
> PORT_REPLACES_BASE_BIND8 if PORT_REPLACES_BASE_BIND8_INCLUDES was
> defined. For last night I wanted to get the user something to test, but
> I do appreciate you bringing up this concern.
Currently the problem the problem only shifted from people installing the
port with PORT_REPLACES_BASE_BIND8 to PORT_REPLACES_BASE_BIND8_INCLUDES,
which will be admittedly less users.
The commit does not really solve the problems. May I suggest one of the
following:
- issue a big, fat warning what problems will occur if the port is
installed with PORT_REPLACES_BASE_BIND8_INCLUDES, including a way
to restore the previous state in case of problems
- add another PKGNAMESUFFIX, so that ports known to conflict with this
installation could test for it
Thanks
Oliver
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?405D5C85.10706>