Stuart Henderson <[email protected]> writes: > On 2020/01/03 21:03, Stuart Henderson wrote: >> On 2020/01/03 11:33, Jeremy Evans wrote: >> > On 01/03 01:49, Kurt Mosiejczuk wrote: >> > > On Fri, Jan 03, 2020 at 09:23:24AM -0800, Jeremy Evans wrote: >> > > > On 01/03 12:12, Xiyue Deng wrote: >> > > > > Friendly ping. Also adding Jeremy to CC as he last worked on this >> > > > > port. >> > > >> > > > Unfortunately, I can't be of much help. If your interest is in getting >> > > > passenger to work on longsoon purely for the challenge, I wish you the >> > > > best. If you actually want to run a Ruby app on longsoon, use Puma or >> > > > Nginx/Unicorn, both of which are in ports, should build on longsoon, >> > > > and >> > > > are much more suitable. >> > > >> > > At least on sparc64, the lack of www/ruby-passenger keeps any version >> > > of nginx from building. >> > >> > I wasn't aware of that. We should probably modify the nginx port to >> > only enable the use of www/ruby-passenger on amd64 and i386 (and other >> > arches it builds correctly on, if any). >> >> Done. We already had no_passenger PSEUDO_FLAVOR scaffolding so I have >> set ONLY_FOR_ARCHS-passenger in www/nginx. >> > > BTW: If the real problem was indeed "I can't build nginx *because* > ruby-passenger doesn't build" then saying that in the first place would > have got a fix made much more quickly :-)
Funny thing is, initially it was the reason for my first build failure report[1]. But as time went by it gradually became how far I could get it to work. Unfortunately despite the itch to make it work I realize I don't have the expertise and should give up on the linker issue and came to the conclusion of "don't touch rake" which doesn't honor any environmental variables and does everything in adhoc ways. But thanks to you guys at least now it's also not blocking nginx on mips64el anymore :) [1] https://marc.info/?l=openbsd-ports&m=152359916007397&w=2
signature.asc
Description: PGP signature
