@mfikes: ^
@johanatan: Line 28 is commented out in master https://github.com/mfikes/planck/blob/master/planck-cljs/src/planck/repl.cljs#L28
do i need to pull?
I think so, yes
ok
@johanatan: The root cause of this is that I had Planck in dev set to use HEAD of the ClojureScript compiler, which changed.
ahh, interesting
It is now set to use a fixed version of the compiler
cool
so, are the cljs parts of planck-c zipped up and placed in bundle.c ?
Yes, essentially planck-cljs
produces a bunch of JavaScript which ends up being bundled in bundle.c
ahh, nice. so the cljs parts are constant between 1.0 and 2.0 ?
Yes. It hasn’t changed a bit 🙂 Only the native half.
ok 🙂
@tord: The crash you mentioned should now be fixed on master. https://github.com/mfikes/planck/issues/326
Cool, thanks, @mfikes!
@mfikes: unfortunately shell exec in plain C is more compliated than it needs to be due to popen not allowing stderr to be read as a pipe
probably going to have to end up with a solution like the last code sample from this page (which only returns stderr instead of both stderr and stdout): http://stackoverflow.com/questions/280571/how-to-control-popen-stdin-stdout-stderr-redirection
i.e., the one posted by kangear
wow
lol
it would've been like a few liner if I could've used system()
or if popen()
supported stderr
as it stands now, it looks like it will be roughly the same number of lines as the obj-c one (give or take)
@mfikes: just discovered a perhaps simpler way. use popen()
in 'r' mode to get stdout. also use freopen()
to redirect the parent process' stderr (which is where the child's stderr output is sent) to a tmp file. then read stderr from the tmpfile and freopen()
back to original stderr stream.
does that sound acceptable to you?
[this would be far less code but a bit more complex in that a tmp file is involved]