On Wed, 2011-04-13 at 15:06 -0400, Allen S. Rout wrote: > > > > R2spec remains a project designed to make the life of Fedora's R > > packager's easier, so it has to remain guideline compliant. That is > > a reason why putting the source0 as URL also is not *in my opinion* > > a good approach (while I actually do like the --URL option). > > Well, like I said, from my perspective it's aesthetic, so I'd be > delighted with it. I'll try to push this on the git by the end of the week-end then :) > > There was a time when I thought I would try to get an "automatic > > cran builder" working but atm I have other priorities. > > Are you willing to record your preferences on this, especially with > respect to how my previous work didn't fit your picture? If you did, > then if I get around to it I could work in ways aligned with your > plans. Well at the moment I see more R2rpm/R2spec has a tool to generate rpm or spec individually as one would have to do to incorporate them into Fedora. It is a spec generator tool which push the work up to building the rpm (which it needs to fill the %file section anyway). Specs needs manual inspection and curation sometime, one needs to run rpmlint and so on. My thoughts were that there was space for a tool to build the whole dependency tree and run R2spec/R2rpm in the right order. In my attempts to build CRAN I used your R script for that. I could this way create 7 files and then I was just running these 7 files on after the other. My problem was more or less: - Takes time to build (~1 week for the whole run) - Lack of resources (although I improved on these but not enough) - I did it the dirty way by doing : build all R packages in fileX in mock - Install all rpm - build all R packages in fileX+1 in mock - install all rpm.... - restart for another arch/release I actually wonder, if one wants to give a shot at such project in a long term, if looking into koji would not be a nice alternative: build all the R packages from fileX for all arch/release - populate the repo as they are build - build all the R packages from fileX+1 for all arch/release... And this way at the end you have a repo which you know works. I always had the wish to support at least i686 and x86_64 on Fedora n and n-1 RHEL 5 and 6, so 6 to 8 (if I added rawhide) build for each packages. For this I though koji would make sense, but I never got around to try to install it and configure vm builders. - Not all the packages build, sometime because of the lack of outside dependencies. For this reason all the packages that fails have to be inspected by hand. - I have a script to update existing spec file but it does not check for changes in the dependency tree and cran as well as bioconductor are so good at changing the dependency tree at every releases... So for every release of bioconductor, we have to update all the spec files for most of their elements but the changelog basically (sounds easy when I say it this way ^^). Basically I would keep R2spec at a single package level and build on the top a cran2rpm or so. All this makes me think that I still have few thousands spec files hanging somewhere and that I should actually make them accessible somewhere (with a warning that a significant part of it is probably outdated). These were my thoughts when I was looking into it. Hope this help, Pierre _______________________________________________ r-devel mailing list r-devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/r-devel
