IRC channel logs

<rlb>ACTION notes that debugging rebases across the utf-32 to uint-8 .go representation is a bit tedous (requires a full bootstrap in order to pass all the tests).

<rlb>wingo: (or anyone) correct to say that string= doesn't "do" normalization?

<RhodiumToad>is string= supposed to be testing for actual equality, or collation equivalence?

<RhodiumToad>looking at the manual, string=? is explicitly documented as not doing normalization

<rlb>RhodiumToad: thanks

<rlb>That's also what the previous code appeared to do, just wanted to be sure.

<wingo>rlb: have you seen https://hsivonen.fi/string-length/

<RhodiumToad>I think that got linked here before?

<RhodiumToad>ACTION has certainly seen it recently, and here seems the most likely place

<lloda>should repeated #:kwarg be an error

<wingo>sometimes you end up extending a kwargs procedure with (lambda args (apply proc (append args #:kw-to-override val)))

<wingo>not so nice but i think it is a thing

<wingo>probably should be a warning if you repeat #:kwarg in any given function application tho

<wingo>like a diagnostic -W1 warning or so

<lloda>makes sense

<sarna>does guile have something like type hints, contracts, or similar?

<hwpplayer1> https://termbin.com/5434 what is the year that started for the FSF's Copyright year is it 1995 ?

<hwpplayer1>What is before ?

<hwpplayer1>Who has the Copyright before I mean

<rlb>wingo: hmm, don't think so - and thanks. (I'd assumed we wouldn't be required to do anything expensive in string=, just wanted to be sure I'd understood things right.)

<rlb>wingo: one thing I found out the hard way yesterday -- switching to utf-8 may reveal corruption more often/sooner :)

<rlb>Hit trouble because an early free had mangled a string in a way that very quickly crashed the utf-8 routines.

<wingo>good evening

<RhodiumToad>evening

IRC channel logs

2023-06-20.log