<rlb>wingo: nevermind - the #nil handling is still broken in main, I had unwittingly left my patch applied to the working tree. The make-struct/no-tail issue appears to be new (after 3.0.7). ***Noisytoot is now known as N7t
***N7t is now known as Noisytoot
<sneek>I've been running for 2 days <sneek>This system has been up 2 days, 15 hours, 48 minutes ***zap1 is now known as zap
<leoprikler>given a SCM UTF-8 string, is there an "easy" way to determine the byte offset (and possibly count) of the char at a given (char) offset? <rlb>On the C side or scheme side? <leoprikler>For context, I plan to (indirectly) use a library, that uses C-style strings and byte offsets, whereas any Scheme API would use char offsets to be useful <rlb>Though iirc I think I discovered that u8_next() (also in the null terminated section), might not actually care about nulls, and is a bit easier to use than the corresponding non-null-term function. Might not be something you'd want to rely on, though. <rlb>But if you're writing this as an extension (outside guile), I suppose you might or might not want to depend on libunicode. <leoprikler>perhaps not, but within the context of a guile extension it might still make sense to use libunicode <leoprikler>do I understand this correctly, that I'd basically loop over u8_next() and then subtract new_ptr - old_ptr? <leoprikler>it'd still be nice if I could instead determine the width by… say code point <rlb>Right, or you could keep two counts, or... i.e. you just use the iterators to skip ahead the right amount for each char, assuming it's valid. Though iirc u8_next may not have comparable error handling... ***iskarian is now known as Guest7348
<rlb>wingo: tracked down the commit introducing the "Unbound variable: make-struct/no-tail" lokke compilation failure and filed a bug: https://debbugs.gnu.org/49404 I don't understand the situation well enough to pursue it further myself atm, but happy to assist if I can.