Armin K. wrote these words on 01/25/13 06:22 CST: > On 01/25/2013 01:09 PM, Randy McMurchy wrote: >> On 1/24/2013 7:44 PM, Armin K. wrote: >>> Just curious. I was always using template/template.xml for that case. >> I made that template years ago. If I recall correctly, there was an example >> parameter and an example option. They were there so there would be one of >> each >> as a reminder to other editors. >> >> > > Yeah, that makes sense now. Could you correct current template? I think > that Andy modified it.
Sure. I originally tried to create a template that you could use and perform a sed command or two to make it into something easy to use for new packages. That was the idea. > Sure, that's okay with me. Now I noticed, as I said in the other mail, > that --enable-cups is required to enable CUPS backend. And it does not > use the CUPS at build time, but backend is practicaly useless without > CUPS at runtime ... I agree. It is about the only package that builds something for another package without it being installed. It is strange. But if you look at the book, I think the instructions I made for it make sense. Don't you? >>> This is strange. All tests passed on my side. You need to run them from >>> a GUI though. >> I did. Here is what I got: >> > > http://www.linuxfromscratch.org/~krejzi/check.txt > > This is mine. I don't know why are you getting that. I build from a very, very old version of Hummingbird X-Windows server on an old Windows XP machine using ssh to the host machine. It works for almost everything. However, it is not GL enabled. So for some tests, I actually use X on my host machine where it is GL enabled. I will try that for these tests. This is why you see "you must perform the tests from a GL enabled host or some tests will fail" in some package instructions. This is how I find that out. >>> I've always wanted to ask for this one. >>> >>> Can't we just stop explaining --whateverdir=/path/to/whatever switches? >>> I mean, it should be obvious that it is setting installation paths for, >>> lets say, docs, executable, whatever. >> I agree But if they are already in the book, what is the harm in keeping >> them? >> > > There is no harm, but I guess we can stop doing that in the future (for > new packages, etc), if you agree. You bet. docdir, libexecdir, sysconfdir, all of these types of things really do not need to be explained unless it is some bizarre location being set. But that doesn't happen very often. -- Randy rmlscsi: [bogomips 1003.23] [GNU ld version 2.16.1] [gcc (GCC) 4.0.3] [GNU C Library stable release version 2.3.6] [Linux 2.6.14.3 i686] 06:26:00 up 50 days, 16:25, 1 user, load average: 0.51, 0.35, 0.14 -- http://linuxfromscratch.org/mailman/listinfo/blfs-book FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
