luis changed the topic of #slime to: SLIME, the Superior Lisp Interaction Mode for Emacs | https://common-lisp.net/project/slime | https://irclog.tymoon.eu/freenode/%23slime | https://irclog.whitequark.org/slime
<cgay> luis: sbcl
<cgay> I've noticed in the past that sometimes M-. is off by many lines even when I haven't edited the file.
<cgay> I'm not as familiar with our local Lisp environment as I could be so it's possible it's due to, um, something non-sbcl.
<luis> cgay: if you can cook up a test case, that'd be great.
prite has joined #slime
scymtym has quit [Ping timeout: 240 seconds]
aerique has joined #slime
scymtym has joined #slime
<fiddlerwoaroof> luis: a random, small commit for something that's been irritating me
<jackdaniel> my biggest contribution to slime was adding stolen phrase to the welcome message: "Are we consing yet?" :-)
<luis> jackdaniel: good job!
<jackdaniel> I know, I wouldn't brag about it otherwise ;p
<jackdaniel> (and that's folks how you get an op on freenode channel ,)
<jackdaniel> but more seriously: I have issue with navigating to source code
<jackdaniel> not sure if ecl fault or slime, I see "error in process filter: slime-dcase failed: nil"
<jackdaniel> hum, otoh I have outdated slime, I'll try with a new one
<luis> fiddlerwoaroof: I think I have a simpler fix. Replied to the PR.
<luis> jackdaniel: did updating SLIME fix it?
<jackdaniel> no, but it is ecl to blame
<jackdaniel> (possibly in swank's ecl backend, not sure yet)
<luis> jackdaniel: I'm adding a testing suite for swank, and I'm precisely starting with source location tests, so you might find that useful.
<luis> I got side-tracked with picking/writing the test framework. Classical mistake. I've avoided this mistake in the past, not sure what happened this time around. :-)
<jackdaniel> thank you. maybe you'll find this test useful: ECL can't tell the source location of methods (but it works just fine with generic functions per se and "normal" functions)
<luis> ACK
<jackdaniel> I'm not sure if it is a bug to return NIL or maybe should ECL signal an error
<jackdaniel> (or ecl's swank backend)
<luis> jackdaniel: swank/backend:converting-errors-to-error-location suggests you should always return some sort of location object
<luis> Hmm, actually MAKE-ERROR-LOCATION returns an s-exp, not a location object
<luis> jackdaniel: anyway, use one of those (or both). Seems to be what the SBCL and ACL backends do.
<jackdaniel> so I should return something in `(:error "Not implemented yet") spirit? (+/- how it should be implemented)
<luis> jackdaniel: something like this, maybe: https://github.com/slime/slime/blob/master/swank/allegro.lisp#L664
scymtym has quit [Ping timeout: 268 seconds]
pritambaral has joined #slime
prite has quit [Read error: Connection reset by peer]
aerique has quit [Ping timeout: 244 seconds]
<fiddlerwoaroof> luis: your changes are definitely better
makomo has joined #slime
makomo has quit [Quit: WeeChat 2.2]
<luis> fiddlerwoaroof: I'll go ahead and commit the simpler fix then
cgay has quit [Remote host closed the connection]