IRC channel logs

<lechner>Hi, I have a long association list of symbols. At what point should I use a hash-table instead, please?

<old>lechner: I saw a benchmark on that recently. It will depends on what you do with it and how it's allocated.

<old>I think it was in the ~40 that you might want to make to hashtable

<old>I might be wrong

<old>s/make/move

<old>I think for example, if your alist is fixed in size and is allocated with literals, you can probably scale better than if you dynamically generare it

<old>Probably the allocation for literals will make it so that you hit less cache line for general access. But even with that, I would say that whenever you start going in 10^2 values, you are better off with hash-table

<mwette>lechner: If this is for foo-symtol-tab you can add (define bar-htab (alist->hashq foo-symbol-tab)) to your .nyacc (aka .ffi) file.

<mwette>nyacc-3.02.0 released today; next effort: revisit C pre-processor

<adhoc>mwette: is that all written in guile ?

<mwette>adhoc: yes

<ekaitz>mwette: yay!

<ekaitz>janneke: ^ nyacc release!

<old>mwette: I don't why, but your mail render some NO-BREAK SPACE (codepoint 160, #o240, #xa0) in some place. Is this intended?

<mwette>old: sorry, my email client is thunderbird. I should send unformatted, but I get lazy. (I surrendered to the populist email clients a while ago; I used RMAIL for a long time before that.)

<old>I see, it's okay

<old>whish Emacs could just not print it as a red underline

<rlb>old: offhand, I'd guess that may be configurable, but I don't know the context all that well.

<rlb>I know it's configurable in some places/ways: https://www.gnu.org/software/emacs/manual/html_node/emacs/Useless-Whitespace.html

<lechner>old / rlb / thanks! my alist has 5442 elements.

<lechner>mwette / thanks for that!

<janneke>how do i get more than 11 stack frames on a backtrace? the essential information seems to be chopped-off..

<mwette>janneke: does ,bt accept a `#:full #t' argument

<janneke>mwette: ah that could be, i'm running from a compilation buffer so, hmm

<mwette>(debug-set! depth 40)

<janneke>mwette: thanks!

<mwette>yw

<old>where is this documented?

<janneke>old: https://www.gnu.org/software/guile/manual/html_node/Debug-Options.html

<janneke>(when you know what to search for, it's kinda easy :)

<old>oh right, that's kind of nested in the documentation

<janneke>if i had found it, i wouldn't have asked it here...

<old>I've been reading this manual for many years now, I keep finding stuff I've overlook. like this

<janneke>iwbn if "6.11.7 Returning and Accepting Multiple Values" would (also) suggest srfi-71 instead of/next to srfi-8, /me thinks

<identity>ITYM SRFI 11

<daviid>fwwi, imo, next to, not instead, receive is by far the best (to me) ...

IRC channel logs

2026-01-15.log