@jeapostrophe pkgd appears to be down
@stchang: ping
@jwmeert has joined the channel
hi @jwmeert
@toblux has joined the channel
,doc fails with (in a plain repl in OSX’s terminal.app): 0:131: execution error: “file:///MyApplications/dev/lisp/Racket/doc/reference/strings.html#(def._((lib._racket%2Fstring..rkt)._string-trim))” doesn’t understand the “open location” message. (–1708)
running open $ABOVE_URL
opens just fine
(except that it ignores the anchor, but that’s separate)
any hints? or suggestions where to poke?
@zenspider: what version of racket is this on?
sorry. 6.9 and latest osx too
stable, right? not a snapshot?
I ask because I know that code recently changed, but I don’t think it made it into the release
correct. I’m not leet :stuck_out_tongue:
maybe try a snapshot and see if you get the same error? the answer would be informative, at least.
@zenspider I see you switched here from #racket
on this so I will, too. :slightly_smiling_face: Found the thing this reminded me of: https://groups.google.com/d/msg/racket-users/Jw3jMB2viCk/ggJbXHu5BQAJ
hrm… left side of nightlies (6.9.0.4) or right side (current)?
and wow… can I check this with minimal? 90% smaller and I’m in a cafe
@zenspider: you can probably check it with minimal, but you’ll want to use racket/help
instead of xrepl
, since the latter is not shipped with minimal racket, and it ends up pulling in a lot of dependencies
also, 6.9.0.4 and current are the same version, the latter is just a stable alias for things like CI scripts
is there a way to isolate this? in ruby we can set GEM_HOME to create a temporary repository that you can later throw away
moot point:
10058 % ./bin/raco pkg install --dry-run racket-help
Resolving "racket-help" via <https://www.cs.utah.edu/plt/snapshots/20170527-7f1ab90/catalog/>
tcp-connect: connection failed
detail: host not found
address: <http://www.cs.utah.edu\|www.cs.utah.edu>
port number: 443
step: 1
system error: nodename nor servname provided, or not known; errno=8
I mean the module racket/help
. it’s in the base
package; you won’t have to install it.
also, the snapshot should maintain its own set of packages.
I can’t figure this out… ¯_(ツ)_/¯
> (help string-trim)
standard-module-name-resolver: collection not found
for module path: help/help-utils
oh, right. apparently that requires the racket-doc
package to be installed, which admittedly makes sense.
(since it’s opening the docs)
yeah. as far as I can tell, minimal is too minimal for this :slightly_smiling_face:
but it’s racket-doc
that will bring in the whole universe of dependencies. :stuck_out_tongue:
since racket-doc
currently links to other docs, including the web server docs… which means you’ll end up installing the web server (and a whole bunch of other packages) if you “just” install the docs.
yeah… I’ll just download the minimal+gui+docs package. it’s 80mb (instead of 9!)… it’ll just be slower on this cafe network
that’s probably the safest bet, yes
maybe it should have a more descriptive name than “gui–6.9.0.4-x86_64-macosx.dmg” … maybe “racket” should be in there somewhere. :stuck_out_tongue:
compared to: min-racket–6.9.0.4-x86_64-macosx.dmg
ok. confirmed that on 6.9.0.4 that ,doc
is fixed. I wonder if I should steal the patch and try to compile it into my current version
@zenspider: Make sure you try running the snapshot from outside of your home directory (that is, the snapshot distribution itself should be placed outside your home directory). IIUC, the issues around that feature stem from very odd macOS behavior around opening file URLs in/outside of your home directory.
If you had the snapshot in your home directory and the release outside of it, you might get false positives.
I ran it straight out of the disk image
that’s probably safe then
is there a bar-chart plot type somewhere? I’m trying to hack it into discrete-histogram and it’s screwing up w/ my x-axis.
ah. it ignores negative values even if you set y-min
considering the irc bridge is down, I’ll ask here
how can I custom print objects? the reference says printable<%> should be implemented, and that it seems to be a wrapper around prop:custom-write
but the docs for prop:custom-write say it’s deprecated in favor of gen:custom-write
@andreiformiga: gen:custom-write
is for structs; it still uses prop:custom-write
under the hood. if you’re using racket/class
, I’m pretty sure printable<%>
is the right thing to implement.
the docs are confusing then
my guess is that the printable<%>
docs weren’t updated when gen:custom-write
was added, since the latter is (I believe) a relatively recent addition.
yes, I see printable<%>
is still defined in terms of prop:custom-write
as it probably ought to be. I somewhat doubt prop:custom-write
itself is really deprecated; I think using it directly with struct
via a #:property
clause probably is.
@mrm has joined the channel