freenode/#clim - IRC Chatlog
Search
8:03:47
Krystof
For what it's worth, sbcl would like to discourage specializations of the stream argument of print-object
8:04:26
Krystof
the reason is to do with being able to guarantee printing conditions and restarts even in cases of heap or stack exhaustion
8:05:07
Krystof
we build a special discriminating function for print-object which always has the effective methods for printing conditions, restarts and one or two other things precomputed
8:05:27
Krystof
but our precomputation assumes that there aren't specializations of the stream argument; if there are, we give up and you're on your own :-)
9:25:55
dim
hi! how can I find out which of my ASDF dependencies or their dependencies require a given system, here CFFI (or actually abcl-cffi)?
9:41:21
jackdaniel
maybe because dim wants (format-graph-from-roots "cffi" #'print (lambda (s) (asdf:system-depends-on (asdf:find-system s)))) ;? :)
9:42:04
dim
I have too many problems with just being able to load pgloader in any CL implementation today to even consider doing McCLIM on macOS this week, you know
9:43:19
dim
SBCL 2.1.3 is broken (sigabort), CCL 1.12 is broken (sigabort), ECL is broken (too many open files), ABCL seems to work but I need to avoid loading some CFFI systems (that I will replace with JDBC components or Java native things anyway)
10:05:32
lukego
annoying that CLIM and ALEXANDRIA conflict on SIMPLE-PARSE-ERROR when you USE-PACKAGE both
10:10:29
scymtym
there is no choice for McCLIM since SIMPLE-PARSE-ERROR is in the CLIM specification
10:11:06
lukego
and actually the annoying part is mostly just my ignorance about how easy this is to resolve with :shadowing-import-from. kicks the can down the road a bit anyway...
10:12:00
lukego
hm it doesn't seem to be mentioned in the alexandria documentation so maybe it could even be changed there potentially
10:34:52
lukego
Since SIMPLE-PARSE-ERROR is not referenced in Alexandria's reference manual it could potentially be renamed to avoid a conflict. Or at the least, I can shadow it without worrying about confusing myself with the Alexandria docs in the future
10:35:58
lukego
but I'll probably tease my code into more structured packages later. first I need to celebrate the milestone that I've started testing and refactoring more than just sketching and rewriting.
10:45:45
jackdaniel
many people will disagree, but I find too many packages in a single project very taxing on maintanance
10:48:14
lukego
ACTION nods, more than one way to do it, glad to be able to just do what feels right