RS> I just strip all but the minimal sequences from ANSI file descriptions.
RS> I could do something similar for ANSI in message bodies too, but I
RS> suppose someone might be wanting to post animated ANSIs and want them to
RS> display as intended.

That is a concern for sure. My take on that at the time was that there have
probably been maybe 10 animated ANSIs worth viewing in the history of ANSI art
and that modern buffering on TCP/IP and terminals could also make ANSImation
inconsistant. I felt the benefits far outweighed the ANSImation issue.

I do have an "undocumented" command that will do a raw message dump to the
screen without any pre-processing which would allow ANSImation in those rare
cases, but so far no one has ever brought this up.

RS> Do you translate the ANSI clearing and cursor movement sequence to a
RS> different set of ANSI sequences or are you actually using the ANSI
RS> region definition sequences to define this "message viewport" and
RS> letting the terminal handle that?

I am adjusting it. So a 1;1H would translate to the top of the message view
and not actually the first row and first column of the user's terminal, and
then any cursor movements are offset like that as well.

It doesn't use any terminal-based boundaries/scroll regions but the end result
is about the same as if it did.

RS> I'm curious how absolute cursor positioning works when the coordinates
RS> are outside of this "message viewport". Sounds like a lot of work and
RS> opportunity for wonkiness. :-)

It was a lot of work. At the time I was driven by some cooler ideas like the
art collaboration through message bases. None of that was ever used though, so
at the end of the day it is probably most noticed when someone posts an ANSI
with a clear screen lol.

I don't think the effort to build something like that is worth it if its just
for that clear screen "use-case" (if you can call it that).

.... Evil triumphs when good people do nothing. - Einstein

Re: Re: ANSI Ad
By: g00r00 to Rob Swindell on Mon Mar 28 2022 11:17 am

> RS> I just strip all but the minimal sequences from ANSI file descriptions.
> RS> I could do something similar for ANSI in message bodies too, but I
> RS> suppose someone might be wanting to post animated ANSIs and want them
> RS> to display as intended.
>
> That is a concern for sure. My take on that at the time was that there have
> probably been maybe 10 animated ANSIs worth viewing in the history of ANSI
> art and that modern buffering on TCP/IP and terminals could also make
> ANSImation inconsistant. I felt the benefits far outweighed the ANSImation
> issue.

Oh, I've seen some awesome animated ANSIs. I assume you've seen the Star Wars
one (wasn't that ANSI?), but yeah, posting them to message bases has always
been pretty uncommon. And with today's transfer rates, the result is usually
not what was intended (especially those written back in the 2400-9600bps days).

> I do have an "undocumented" command that will do a raw message dump to the
> screen without any pre-processing which would allow ANSImation in those rare
> cases, but so far no one has ever brought this up.
>
> RS> Do you translate the ANSI clearing and cursor movement sequence to a
> RS> different set of ANSI sequences or are you actually using the ANSI
> RS> region definition sequences to define this "message viewport" and
> RS> letting the terminal handle that?
>
> I am adjusting it. So a 1;1H would translate to the top of the message view
> and not actually the first row and first column of the user's terminal, and
> then any cursor movements are offset like that as well.
>
> It doesn't use any terminal-based boundaries/scroll regions but the end
> result is about the same as if it did.

Okay, I'm doing the same/similar in my msglist module. I just render the ANSI
to a virtual CGA-style screen buffer and then send the relevant portions of
that buffer to the user as they scroll the message body. So if there's any
overwriting or clearing in the ANSI, they only get/see the final result. This
discussion inspired that enhancement, so thanks to Joe!

> RS> I'm curious how absolute cursor positioning works when the coordinates
> RS> are outside of this "message viewport". Sounds like a lot of work and
> RS> opportunity for wonkiness. :-)
>
> It was a lot of work. At the time I was driven by some cooler ideas like
> the art collaboration through message bases. None of that was ever used
> though, so at the end of the day it is probably most noticed when someone
> posts an ANSI with a clear screen lol.
>
> I don't think the effort to build something like that is worth it if its
> just for that clear screen "use-case" (if you can call it that).

I see ANSIs BBS ads posted that use a lot of vertical cursor movement too. Not
just screen clears.

Filtering out the old "ANSI bombs" (keyboard remapping via ANSI.SYS) is a good
idea too. Remember those?
--
digital man (rob)

This Is Spinal Tap quote #44:
It really, it does disturb me, but i'll rise above it; I'm a professional.
Norco, CA WX: 58.9�F, 70.0% humidity, 8 mph ENE wind, 0.00 inches rain/24hrs