Re: [Bug-apl] Back to underline

[Juergen Sauermann]

HI, the old APL machines had underlined characters in their character set (and the use of underlined characters as, for example variable names) comes from that time. But no lowercase characters. Also all IBM EBCDIC machines have it because that character set has no lowercase letters but instead underlined uppercase characters. In my opinion that mapping (lowercase ←→ underlined uppercase) avoids all the problems mentioned so far. The problem with ⍴'A̲' is that it looks like one character but is actually two. I guess if I would implement combining underline today, then this would come as a trouble report tomorrow at latest. I haven't seen a machine with lowercase and underlined characters myself, but it seems to exist (and is it a Unicode machine?). IBM APL2 for PC seems not to have underlined characters. /// Jürgen On 08/15/2015 06:38 PM, Elias Mårtenson wrote: On 15 August 2015 at 21:31, Juergen Sauermann <juergen.sauerm...@t-online.de> wrote: Hi Fred , fine. Maybe the way forward is if you publish a patch in our Bits_and_Pieces directory. I would not like it to be a ./configure option because that would give us many #ifdefs all over the place and I (instead of Fred) would have the burden of maintaining it. It is actually far more more than only changing Avec.def. For example the built-in line editor (i.e. every line entered interactively) currently assumes that every character moves the cursor by one position. So the cursor will be positioned one character off for every underlined one. Also, there is a difference between APLs that have underlined chars in their character set and those that do not. Consider ⍴'A̲' (not sure if this displays properly everywhere - read: RHO QUOTE A COMBINING_UNDERLINE QUOTE) On a system with underlined characters this will return 1 while on Unicode with combining underline machine it gives 2. What kind of machine "with underline characters" is that? As far as I know, GNU APL depends on Unicode, under what circumstances would that return 1? Regards, Elias