On Sat, Aug 02, 2025 at 01:20:32AM -0500, Parzival Wolfram wrote: > (My email client managed to only reply to one address in the email despite > hitting Reply All - Thunderbird, why do you betray me?)
I mostly use mutt on my local machines for replies... > I do have a functioning Solaris 7 machine, as stated, that can build and run > ncurses-6.5, and I've gone through the trouble of putting together patches > to make it largely C98 compatible. (Support is incomplete, as I'm not able > to get the leaked kernel sources to build, but pre-processors and gnulib can > handle a lot of the missing functionality.) If you have anything that you > need to run to test it, point out a tarball and I'll see what I can do. for checking an existing terminal description, I'd use tack. It should build with either ncurses or Solaris curses, the ncurses configuration would provide for extensions which you wouldn't (initially) use: https://invisible-island.net/ncurses/tack.html https://invisible-island.net/ncurses/tack.html#download https://github.com/ThomasDickey/tack-snapshots I've built it with Solaris 10 (using the SVr4 libraries -- the xpg4 stuff as I recall it isn't very useful in this instance). So (unless you've installed ncurses in some conflicting manner), you should be able to just ./configure && make ./tack and step through the menus ("n" is easy), to see what discrepancies pop up. > > Additionally, after scouring the aforementioned leaked kernel sources, this > is actually NOT the "Intel AT/386 for color card with monochrome display" > entry as listed, but instead an "AT&T 386 color console". > > On 8/1/25 7:33 PM, Thomas Dickey wrote: > > On Fri, Aug 01, 2025 at 06:29:43PM -0500, Parzival Wolfram wrote: > > > Are terminfo definitions still being added for/from ancient Unix flavors? > > > I'm running Solaris 7 i86pc (i486-pc-solaris2.7) and the "AT386" terminfo > > > file, defining the textmode shell on the physical machine, is not present > > > as > > > of ncurses-6.5 (along with several other files.) Would these be > > > appropriate > > > to add to ncurses, either directly from Solaris or remade based on the > > > infocmp output? infocmp and toe do seem to parse them, both on the Solaris > > > box and on a modern Linux machine. > > hmm - I hadn't exactly overlooked that, so much as thinking it was out of > > scope. Here's what I think you're talking about, from a 1996 file from SCO: > > > > -------------------------------------------------------------------------------- > > # Intel AT/386 for color card with monochrome display > > # > > AT386-M|at386-m|386AT-M|386at-m|at/386 console, ... > > # > > # AT&T 386 color console > > # > > AT386|at386|386AT|386at|at/386 console, ...for instance, this one. > > use=at386-m, ...but, you can see that it uses the indn/rin from the monochrome. I have copies of various terminfo databases, but they're also of various quality. > > -------------------------------------------------------------------------------- > > > > Mostly I'm interested in things that can be validated. > > > > https://invisible-island.net/ncurses/ncurses.faq.html#dbupdate > > > > It's possible that you have found an overlooked terminal description > > which is > > (a) supported by a (working) terminal emulator and > > (b) you can find a manual to cross-check the behavior > > > > But I see some problems to investigate, e.g., indn/rin are using "%P1" for > > one ("%p1" is expected), and omitting the parameter in the other. A later > > file from Solaris fixes the "%P1" typo, but still doesn't provide a > > plausible > > value for rin. > > > > I'd actually expect rin to be as in ecma+index, but might be persuaded > > otherwise, given a detailed manual: > > > > # ECMA-48 does not include the VT100 indexing and scroll-margins. It has > > its > > # own variation. > > ecma+index|ECMA-48 scroll up/down, > > indn=\E[%p1%dS, rin=\E[%p1%dT, > > > > ...bitsavers doesn't appear to have any useful AT&T terminal manuals. > > Testing the machine's an option (though getting tack to run on Solaris 7 > > might be challenging). The manpages with your system may describe the terminal in detail. > > > > -------------------------------------------------------------------------------- > > tic6: sorting by terminfo name order > > tic6: width = 60, tversion = 0, outform = 0 > > "/tmp/at386.ti", line 3, terminal 'AT386-M': indn uses 0 parameters, > > expected 1 > > "/tmp/at386.ti", line 3, terminal 'AT386-M': rin uses 0 parameters, > > expected 1 > > "/tmp/at386.ti", line 3, terminal 'AT386-M': checking sgr(none) > > "\e[10m\e[0m" > > compare to "\e[0;10m" > > unmatched "10m" > > "/tmp/at386.ti", line 3, terminal 'AT386-M': exit_attribute_mode differs > > from sgr(0) > > exit_attribute_mode="\e[0;10m" > > sgr(0)="\e[10m\e[0m" > > "/tmp/at386.ti", line 3, terminal 'AT386-M': exit_standout_mode matches > > exit_attribute_mode > > "/tmp/at386.ti", line 3, terminal 'AT386-M': exit_underline_mode matches > > exit_attribute_mode > > "/tmp/at386.ti", line 27, terminal 'AT386': indn uses 0 parameters, > > expected 1 > > "/tmp/at386.ti", line 27, terminal 'AT386': rin uses 0 parameters, expected > > 1 > > "/tmp/at386.ti", line 27, terminal 'AT386': checking sgr(none) "\e[10m\e[0m" > > compare to "\e[0;10m" > > unmatched "10m" > > "/tmp/at386.ti", line 27, terminal 'AT386': exit_attribute_mode differs > > from sgr(0) > > exit_attribute_mode="\e[0;10m" > > sgr(0)="\e[10m\e[0m" > > "/tmp/at386.ti", line 27, terminal 'AT386': exit_standout_mode matches > > exit_attribute_mode > > "/tmp/at386.ti", line 27, terminal 'AT386': exit_underline_mode matches > > exit_attribute_mode > > > -- Thomas E. Dickey <[email protected]> https://invisible-island.net
signature.asc
Description: PGP signature
