@JosephWright I just saw in the CTAN announcement of greek-fontenc "\lccode/\uccode corrections for Unicode". greek-euenc.def now contains a long list of lccode/uccode settings.
@JosephWright OK. I simply wondered if greek-euenc.def is the correct place. I checked a few and they seemed to set the same values as unicode-letters.def.
I see the L3 team added \exp:w for full expansion. This, however, is different from \romannumeral-`\Q (or \romannumeral-0, I don't know what exactly is it, I never used it explicitly). Why that difference? Why now \romannumeral is “enough”?
I defined some time ago something in this lines \let \exp:w \romannumeral \def \now:w { -`\Q } and then it had to be called like \exp:w\now:w. But now just\exp:w is enough?
@Johannes_B -- i have had some chance, but have not yet come to a firm conclusion. i am still struggling with the (unfortunately) widespread "belief" that document class = template. i can see some reasons behind that opinion, but strongly feel it is the wrong way to go; my struggle is at least partly with myself -- am i simply being stubborn? anyhow, tomorrow is a holiday here, and (as soon as i finish editing the tugboat proceedings issue) contemplation of templates is next on my list.
@barbarabeeton The question of what a template is is unclear yet and needs to be specified. In the post i used my point of view of what i think templates are. This isn't fixed and open for discussion. I just didn't knew how to express that, that's mainly why i asked for your opinion. The question of how to get the text out there and how to actually discuss needs some attention as well. But TUGboat first.
@Manuel \exp:w = \romannumeral but with a defined expansion purpose. There's also now \exp_end_continue_f:w` = `^^@ ~ -`\Q
@Manuel You can\t avoid needing \exp:w\exp_end_continue_f:w for cases where the first token is hit by \exp_after:wN and must force expansion in one shot
@MartinScharrer I'll resend
@Manuel We are (once again) fixing inconsistencies in expl3: here we'd been using \tex_romannumeral:D rather too much for comfort, and a name reflecting the use was needed. Expect changes in the case changing area for the same reason.
@JosephWright No there wasn't really any discussion, it just got added on the back of a comment I made while failing to fix \Uchar's surrogate pair handling.
@PauloCereda Yes. Quite crowded place, actually: it was a splendid day. But I did the narrow road when almost everybody was having lunch, so not much traffic. :)