[Fresco-devel] Re: automake for fresco
Stefan Seefeld <[email protected]>
| Newsgroups | gmane.comp.video.fresco.devel |
|---|---|
| Message-ID | <[email protected]> |
wturkal wrote: >>what advantages do you expect from using automake instead >>of our current build system ? > > > 1. Ability to build in and out of the source tree without > issue (think --disable-sdl and trying to build within the > source tree). hmm, given that the problem stems from a shortcoming in autoconf, which is used by automake, too, how does automake solve this issue ? > 2. Makefile.am is smaller than the Makefiles and usually > easier to understand. For instance, I have a top level > Makefile.am that only has like 20 lines. This includes the > logic for conditionally building certain subdirectories via > the enable/disable config commands for sdl, ggi, c++ demos, > python-runtime, and python-demos. right, Makefile.am files are more compact than Makefile.in files. That is not unexpected, given that the whole idea behind automake was to restrict the usage of 'make' to 'normal', i.e. the most frequent cases. But it is exactly this inflexibility that let me refrain from using it, i.e. I wanted to do things that simply weren't possible with automake. But may be you can prove me wrong. I'm curious to see your Makefile.am files, and to test whether they support all we want to be able to do. > 3. Reuse of the knowledge of the automake people in terms of > portability. In how far is an automake based Makefile (Makefile.am) more portable than our current Makefiles ? > 4. Automatic dependency checking from automake. dependency of what, from what ? > 5. Parallel build ability without as much work. I seriously doubt that automake can help us. After all, it's a matter to clearly indicating dependencies, nothing more. > 6. Less work when new packages/binaries/libraries are added. How so ? Stefan