Re: [racket-users] raco distribute creating too many subdirectories

2019-02-16 Thread 'John Clements' via Racket Users
Good to know; I also use define-runtime-path with abandon, and was not aware of possible consequences of casting too wide a net. Thanks! John > On Feb 10, 2019, at 09:49, Philip McGrath wrote: > > I think this is probably a consequence of the quirk of `define-runtime-path` > with directories

Re: [racket-users] raco distribute creating too many subdirectories

2019-02-10 Thread Philip McGrath
I think this is probably a consequence of the quirk of `define-runtime-path` with directories that I reported at: https://github.com/racket/racket/issues/2336 I'm still not clear on whether the current behavior is intended and, if so, how one ought to do this sort of thing to work robustly with th

Re: [racket-users] raco distribute creating too many subdirectories

2019-02-10 Thread P. Baillet
Hello, After some analysis, I found out this is because I was using `(define-runtime-path` the wrong way. My code uses FFI and at some point, I thought it would be wise to use code I didn’t understand. The following snippet reproduces the issue: ``` #lang racket (require racket/runtime-path)

Re: [racket-users] raco distribute creating too many subdirectories

2019-02-08 Thread George Neuner
On 2/8/2019 5:25 PM, P. Baillet wrote: I’ve been playing with |raco distribute| to build small docker images with racket binaries in them and it seems that the command sometimes creates a very deep tree and crashes in the end (with Racket 7.2): On some platforms [those that use AUFS], a D

[racket-users] raco distribute creating too many subdirectories

2019-02-08 Thread P. Baillet
Hello list, I’ve been playing with `raco distribute` to build small docker images with racket binaries in them and it seems that the command sometimes creates a very deep tree and crashes in the end (with Racket 7.2): ``` root@a8ec24e4167d:/racket/pipo# raco distribute app tlf copy-file: cann