On 17 May 2017 at 16:19, Mathias Kresin <d...@kresin.me> wrote: > Hey Roman., > > independent of the work done by you, I worked on exactly the same > topic. But I've pushed it a bit further by switching all targets to > the generic boardname function and a few targets to the generic board > detection in basefiles/preinit. > > Furthermore I've converted some targets to get the boardname from the > device tree compatible string as suggested by John. Till now I wasn't > comfortable to push it somewhere, but with my latest changes it should > work in theory. You can find the changes at > https://git.lede-project.org/?p=lede/mkresin/staging.git;a=shortlog;h=refs/heads/boarddetection. > What is missing: testing! > > 2017-05-17 13:49 GMT+02:00 Roman Yeryomin <leroi.li...@gmail.com>: >> Hi John, >> >> the reasoning is that most scripts which need board_name, don't need >> anything else, and those which do need more, they need only a fraction >> of functions.sh > > Due to the fact that I had to touch all scripts using the board name, > I feel confident enough to reply here. I don't see any benefit in > moving the board_name function into an extra script. Most of the > scripts include (and require) functions.sh anyway.
I already explained that. functions.sh inclusion in many cases is a mistake and in other 90% of cases only one or two functions are needed other 10% need 3-5 functions so all that is very far from optimal both from maintenance/development and resource usage perspective > I would be happy I've you can review and runtime test the boardname > branch from my staging tree. > > Mathias _______________________________________________ Lede-dev mailing list Lede-dev@lists.infradead.org http://lists.infradead.org/mailman/listinfo/lede-dev
