On Fri, Oct 30 2009, Tobi wrote: > Manoj Srivastava schrieb: > >> 1. SPECIAL_VDR_SUFFIX=devel make -f debian/rules build >> 2. make -f debian/rules SPECIAL_VDR_SUFFIX=devel build >> 3. SPECIAL_VDR_SUFFIX=devel ./debian/rules build >> 4. ./debian/rules SPECIAL_VDR_SUFFIX=devel build >> Giving you differing results is confusing enough to anyone >> building the packages manually (You know, as free software folks, the >> buildds are not the sole focus of our packaging) that I think it is >> good that the policy is specific enough to block these. > > I just don't think this is a problem at all. If you use > SPECIAL_VDR_SUFFIX=devel, you do this for a reason. It's very specific > for this set of packages and without reading the documentation, you > wouldn't even consider setting SPECIAL_VDR_SUFFIX=devel. And if you > read the documentation, you know exactly, how to use > SPECIAL_VDR_SUFFIX.
Oh, I am doing this for the reasons specified. But I have been assured by policy that it is a makefile, so I think any of the 4 commands will give me the super-duper devel version, right? Wrong! And that is why this is a horrendous idea. > In general: IMHO the policy shouldn't force implementation details, it > should just enforce the interface. The interface definition behind this is: calling make -f debian/rules and ./debian/rules should result in identical behaviour, all other things being equal. Either invocation should deal identically with the environment, and with variables set on the command line. There. manoj -- Revolution, n.: A form of government abroad. Manoj Srivastava <sriva...@debian.org> <http://www.debian.org/~srivasta/> 1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org