« first day (2100 days earlier)      last day (851 days later) » 

att
att
02:19
> sometimes schizophrenia
 
4 hours later…
06:03
On my bike this morning, I suddenly realised that ⊢⍤⌿⍥, (when the arguments have the same rank) can be written as ⍸⍛⊇
If argument ranks differ, then ⍸⍛⊇ is actually superior, as it works along leading axes, conceptually collapsing leading axes of the arguments until the left argument is a vector.
@Adám Oh dang! We might actually get Before?! Were will go on the keyboard, BTW?
I'm proposing either f or F.
Hah. A nice conservative option.
On that note, does Dyalog have an official release schedule?
No.
Haha. So much communicated in 3 characters. Cheers.
06:28
@B.Wilson Well, f would be progressive, because that's already _
Hrm. F is also _ on my setxkeyboard apl -variant dyalog layout, so I have 3 separate ways to type underbar.
That's a mistake, really.
Mistake in xkeyboard-config's definition, you mean?
Yes.
See Linux layouts here. Clearly, _ is only on f, not on F.
@Adám Ah! Took me a bit to see that. Very nice. It's a trifle sad that bottom horizontal bars all fail to line up pairwise with whatever font is getting used here.
Is From also potentially slated for 19?
06:41
Yes, and Depth.
@Adám Hrm. I might have to track down if this is just a me error. Will file PR if not.
Could be that they have consistently equated shifted state to non-shifted state where there's no assigned character for the shifted state. Check out U; does it give ?
That's exactly what's happening. ↑↓⊃∩∪ are also mapped to the shifted key as well.
Using f would make the best symmetry with on j (note that f and j are the "knobbed" keys), but it also makes sense to have and _ together, and I dislike removing traditional bindings.
@B.Wilson APL ligatures?
07:15
Yeeeah. That feels so good :P
 
1 hour later…
08:34
@Adám Ah, I hadn't seen the j-f symmetry. I tend to have biased preferences for beauty, symmetry and the like, but working in industry has corrupted me with another preference for not breaking the status quo.
Nice, at least one person can appreciate Dyalog's commitment to maintaining backwards compatibility.
09:33
Just out of curiosity, and if you can share, what does Dyalog's internal dev process look like? Vis-à-vis Hyrum's Law and Dyalog's age, I would imagine you need a large CI pipeline or something to ensure solid backwards compatibility.
10:19
@B.Wilson Not sure if it answers your question, but we run nightly QAs.
 
2 hours later…
12:22
@Adám Somewhat. Am interested in the details, actually, as much as can be shared at least.
The majority of my work is in devops and dev infrastructure development, so I'm really curious about the kinds of policies and tech in place on such a long-running project. I imagine it has gone through many evolutions.
 
2 hours later…
14:14
@B.Wilson We don't really have secrets at Dyalog, other than user and customer information. However, I'm neither in operations, nor management, so my knowledge is limited.
@B.Wilson We use SVN and Git to track changes, Mantis and GitHub issues for tracking issues. Any changes to a primitives and a currently released version must go through full peer review. Internal knowledge sharing and project management is on an internal MediaWiki. New developments (other than bug fixes) have project plans with sign-offs from multiple parties: CTO, Architect, User Representative, Development Manager, Documentation Manager, QA Manager, and Software Security Group.
14:29
watched the podcast and was wondering. Was all that just because jeremy wanted to add an empty row in a matrix? Not sure if i understood the problem.
@Jeremygee Well, it was all due to Jeremy's tweet, but what he actually wanted to do (if anything), is obvious.
@Adám or U (Under(score)) which is also free and close to ∘, ⍤, ○ and ⍥?
That's a good one too. I'll keep that in mind when we get to keyboarding for new primitives.
14:52
Will have to look into broadcasting. It was an interesting discussion nonetheless.
15:26
ok, so im kinda stumped on this one. ive got a vector of vectors of vectors of scalars. I wish to unnest the innermost vector. how would i do that?
Can you give an example of input and desired output?
never mind, a step i used earlier to generate that vector was faulty. ive got what i wanted
on another note, is there any way i can force stop printing in the repl?
ive printed a huge vector by accident
Press Ctrl+C
tried already
And for next time, do ]rows on -fold=3
It'll truncate any vertically overflowing output.
15:33
mhm thanks
i think ill just restart the thing :P
nvm it stopped finally
Which OS and interface?
E.g. Ctrl+Break should do it on Windows.
windows
and whatever thing thing is
That's the normal Windows IDE.
Action>Interrupt should do it as well.
ah ty
ah wait its giving me those nested vectors again, argh
You can also click the Dyalog app logo in the Windows systray to issue interrupts.
15:39
ah wait i just realized strings are char vectors
then its gving me the output i want
 
6 hours later…
21:26
@user20316045 Hi Erick Vick. If you want to participate here, please email [email protected]

« first day (2100 days earlier)      last day (851 days later) »