Tested it on #Windows now. Ok, it works on some Windows-10 machine, the terminal nowadays can do both #utf8 output and interpret some #ansi sequences ...

But: You have to enable both explicitly in your code using some proprietary Console APIs 🤯
https://github.com/Zirias/dos2ansi/commit/5a85965d96d4456d1e739427122f66e89fb358b6

New pre-release of dos2ansi: v0.2

* Works on #Windows, win32 binary (cross-compiled on #FreeBSD) attached
* Selectable input #codepage (so far only #cp437, #cp850 and #cp858)
* Selectable output format, #utf8, #utf16 or #utf16le, with or without #BOM

Still a few things to add, e.g. use #termcap/#terminfo or Windows Console API for "color output" when applicable ... we will see 😎

https://github.com/Zirias/dos2ansi/releases/tag/v0.2

Just released dos2ansi v0.4, with lots of #DOS #codepage s supported and a testmode to display them.

The next nice feature would be to use the actual terminal capabilities if output goes there. Very simple on *nix-like systems (#Linux, #FreeBSD, ...), just link #curses and use the termcap functions.

Thinking about #Windows again, either I keep relying on #UTF8 support (since #win7 IIRC? and still a bit buggy) and #ANSI sequences support (since #win10) .... OR I attempt to use the native #Console #API there (using special functions to write in #UTF16 and other special functions to set colors, which would require a major refactoring first 🙄)

https://github.com/Zirias/dos2ansi/releases/tag/v0.4

dos2ansi v0.6 released 🥳

This version brings specialized writers for direct terminal output, using #terminfo (via #curses) on non-windows and configuring the #ANSI writer for 256 colors on #Windows >= 10, otherwise using a writer for the legacy Windows Console API.

It *should* work fine on any Windows and any Unix-like system offering terminfo/curses, but I could only test on #FreeBSD and Windows Server 2022 myself, so it would be nice if someone could test on older Windows as well as other *nix systems like #Linux 😉

https://github.com/Zirias/dos2ansi/releases/tag/v0.6

Released v0.6.1 just fixing one stupid little regression from refactoring: The terminfo writer must never be used when output does *not* go to stdout 🙈

So far so good, I just thought there was yet another bug, testing redirected output on #Windows. I *always* got a file encoded as #UTF16LE, no matter which format I chose (and with #UTF8 chosen, the output was just broken). Meditated on the code for a while. Looks all perfectly good.

Finally, I tested in good old #CMD. And it worked perfectly well.

So, #Powershell is messing with the encoding of my stdout stream on redirects. Really? Really??? WTF, #Microsoft? How should that ever do any good? 🤯

https://github.com/Zirias/dos2ansi/releases/tag/v0.6.1

Meanwhile released v0.6.2 with a few more (less important) bugfixes ...

... and working on a new feature, parse #SAUCE metadata.

#lldb (which comes with #FreeBSD base) is quite nice for some first testing 😎

SAUCE ref: https://github.com/radman1/sauce

Yep, progress ... unfortunately I don't own any "sauced" ANSI files making use of all possibilities of sauce, so it will be impossible to verify more things I'll add ....

If you have #MSDOS #ANSI files containing #SAUCE using the ansiflags or the fontname properties and/or containing additional comments, please let me know 😎 (maybe someone following #retrocomputing?)

There will probably be a "preview release" soon containing what I have so far ...

Did a "normal" release: dos2ansi v0.7

#SAUCE support isn't complete yet and parts of it are unfortunately untested (any help from #retrocomputing fans? 🙏), but it's still a step forward, so, why not just release it 🙈

https://github.com/Zirias/dos2ansi/releases/tag/v0.7

And dos2ansi v0.8 is here!

Not only because #SAUCE indeed had a few bugs in 0.7 .... there are also quite a few improvements.

For SAUCE, I now added #comment support ... and as I don't own any authentic #MSDOS file with that, I added some fake comment according to the spec to test it ... which is ... not so great ... please, if you have files using some more SAUCE features, contact me 🙏

Download source and #win32 binary here:
https://github.com/Zirias/dos2ansi/releases/tag/v0.8

Found a pretty nice font to go with #dos2ansi: #BlockZone.

It seems to contain a pretty complete set of glyphs (arabic and cyrillic work fine), and approximates the original #VGA looks very well! 👍

(Screenshots from #konsole on #FreeBSD again)

https://github.com/ansilove/BlockZone

#retrocomputing #MSDOS

Few features, but a lot of refactoring and internal improvements ... and #Windows users now get a program using "native" APIs ...:

here's #dos2ansi v0.9! 😎
https://github.com/Zirias/dos2ansi/releases/tag/v0.9

Screenshot shows displaying some random #SAUCE metadata, and the new (finally, lol) flag to display version info (also note the license matching my favorite OS, #FreeBSD)

#retrocomputing #MSDOS #ansiart

This will probably be the last release of #dos2ansi for quite a while, v1.1:
https://github.com/Zirias/dos2ansi/releases/tag/v1.1

Reason: It now has all the features I could think of initially ... and some nasty #win32-api-braindamage-related bug (only affecting #Windows of course) plus a few minor things I missed in v1.0 were fixed/improved. 🙈

Sure, if someone has a great idea what should be added, I'm open to suggestions 😎

#MSDOS #VGA #ansiart #retrocomputing

Never say something is finished 🙈

Browsing https://16colo.rs/ I found some #ansiart file that immediately had my attention because it crashed my #terminfo writer. Turns out attempting fallback to #termcap names makes no sense with terminfo ... although it shouldn't hurt either, but it did with "xterm-256color" on #FreeBSD for the "blink"/"mb" capability, "mb" returned something, but something "else" 🙄 ... so I fixed that.

This file was strange enough to use it for more testing, it had "blinking" characters (according to #SAUCE), although ignoring that and using bright colors instead looked much more sane, see two screenshots for comparison (oh wow, #xterm can actually blink? although not like #VGA would...)

Anyways, I found a lot more things to optimize and fix in both my "plain ANSI" and "terminfo" color writers!

So, there will be #dos2ansi v1.2 soon. Will probably add some small feature first, just "because" 😂

And here's the upcoming "feature" for #dos2ansi v1.2: Attempt to format #SAUCE comments nicely. 🙈

It seems to be common practice to wrap comment lines in the middle of a word, so here's an attempt to be "smart" with this, adding an (optional) formatting heuristic:

* Empty lines are alway printed as is
* Same for lines containing "drawing" characters (anything >= 0xb0 in #cp437)
* Otherwise, if the last character of a line *and* the first of the next line are non-space, it's assumed to be one word
* Formatted lines are filled with as many complete words as possible, preserving spaces except at the end of the line
* More than two spaces at the end of a (raw) line are interpreted as a line break

Here are two real examples of the result, while the second already shows some classic shit-in-shit-out case (to fix this, I'd probably need a language model 😂)

And here it is, #dos2ansi v1.2:
https://github.com/Zirias/dos2ansi/releases/tag/v1.2

It's mostly optimizing the output and some subtle little bugfixes ... plus this feature to re-format #SAUCE comments 😎

So now I enjoy watching my #xterm blink on #FreeBSD 😂

#MSDOS #ansiart #VGA #retrocomputing

And #pixelperfect (minus aspect ratio correction) #VGA display is possible in #xterm using bitmap(!) fonts I found here:
https://github.com/farsil/ibmfonts

As always, it wasn't *that* simple ... xterm expects a complete set of #VT100 drawing characters to be available, otherwise it will render its own which leads to funny color bugs at least with the "medium shade" used a lot in #ansiart ...

... so, let's patch the missing ones in (using dummy empty glyphs), which I did here in a #FreeBSD port for these fonts:
https://github.com/Zirias/zfbsd-ports/tree/local/x11-fonts/ibmfonts

Screenshots show #dos2ansi render some ansiart and its #SAUCE info, once in VGA-9x16 and for comparison in VGA-8x16 😎

One more: #dos2ansi rendering some #ansiart in #xterm on #FreeBSD using bitmap IBM fonts, here VGA-8x16.

#retrocomputing

I'm about to prepare yet another release, #dos2ansi v1.3.

Finding these bitmap fonts triggered two things:

1. I realized dos2ansi needs a mode to NOT attempt to mimic #VGA looks by chosing more similarly looking characters: A font recreating the orignal IBM fonts probably won't have glyphs for these other characters (and the original ones will look, well, original 😜 ...)

2. It gave me the idea that I now had everything in place for some sort of "ANSI viewer", just needs a script to tie together dos2ansi, #xterm, #less and the #ibmfonts. So, here it is: showansi. It should work with other X terminals, pagers, fonts, for configuration see the updated README:
https://github.com/Zirias/dos2ansi

The script first calls dos2ansi to query some #SAUCE metadata it then uses to pick the most appropriate font, window size and title for xterm.

The screenshot uses completely unconfigured Xorg and xterm, to show the script sets everything as needed.

#ansiart #retrocomputing

Released #dos2ansi v1.4:
https://github.com/Zirias/dos2ansi/releases/tag/v1.4

Once again I have the feeling it's finally "feature complete" ... I hope it is this time 🙈

This release mainly improves the "showansi" script. it doesn't use environment variables any more (for security reasons, running `eval` on stuff partially obtained from the environment is a bit risky indeed) but uses configuration files instead. It also avoids creating a temporary FIFO for reading #SAUCE metadata now.

Screenshot: showansi automatically picking the VGA-9x16 font as indicated in SAUCE.

#retrocomputing #MSDOS #ANSIart #VGA

Oh freakin hell, of course this happens directly *after* release: I found some #ansiart file that really uses every obscure display-related feature of ANSI.SYS, and as a result, #dos2ansi v1.4 just renders garbage 😠

It now works with the latest commit, I had to implement:

- absolute cursor positioning
- cursor position saving and restoring
- all the erasing (line/screen) commands
- and additionally at least ignoring sequences to set the screen mode 🤯

Why is anyone doing that? First I thought save bandwidth for BBS, but this file has 7620 bytes, and processed with -AId flags (so only sequences used also available in ANSI.SYS), it shrinks down to 5906 bytes even using #UTF8... I really don't understand that 🙈

Here's this file: https://16colo.rs/pack/arc-17/M0D-RMS1.ASC

#dos2ansi v1.5 will be there soon I guess...

It's done, #dos2ansi v1.5 is here:
https://github.com/Zirias/dos2ansi/releases/tag/v1.5

Apart from supporting all the cursor movement and screen erasing sequences that were needed for that silly file I found, there's now also support for the screen mode setting from ANSI.SYS (with non-text modes silently ignored).

There are also tons of little improvements to the "showansi" script, which should now work very well as a generic DOS ANSIart viewer picking reasonable defaults (and making it easy to override them if really necessary).

Example screenshot includes rendering this silly file again 😂

Have fun, and if you need some files to display, have a look at https://16colo.rs/

#MSDOS #ANSIart #retrocomputing

#dos2ansi v1.6:
https://github.com/Zirias/dos2ansi/releases/tag/v1.6

* Fix cursor movement again, it never wraps or scrolls.
* Many smaller improvements to the #showansi script
* Add an example configuration using scalable fonts from https://int10h.org/oldschool-pc-fonts/ instead of the default bitmap fonts

Scalable fonts are a tradeoff here. You get freely configurable sizes and correction of aspect ratio. You lose pixel-perfect (perfectly sharp) display. You can't have both, unfortunately.

Screenshot shows scalable fonts 😉

#MSDOS #ANSIart #retrocomputing

#dos2ansi v1.6.1 (a pure bugfix/service release):
https://github.com/Zirias/dos2ansi/releases/tag/v1.6.1

* Fix possible crash (🙈) when querying for font or codepage on a file with a non-text #SAUCE type.
* Try to parse something meaningful even out of partially broken SAUCE data

Yep, the possible crash was the reason to create a release even for a pretty "small" bugfix. 🙄 Chances are nobody would have noticed though, the SAUCE *display* was not affected, only the machine-readable query (-q).

Screenshot: The "offending" file. It even had literal ",1" written in the month field of the SAUCE date, but obviously meaning January here 😂

#MSDOS #ANSIart #retrocomputing

And now there's even #dos2ansi v1.6.2: https://github.com/Zirias/dos2ansi/releases/tag/v1.6.2

The good thing is: The issues I'm fixing here are getting more and more obscure 😂

First I noticed #xterm doesn't accept all #utf8 for its '-title' option, no matter what. But it *does* support the escape sequence to set a title, so, why not use this instead.

Having added support for that (and also optionally for setting the window size using an escape sequence), I had a look at quoting again. Indeed, #showansi failed for e.g. file names containing #shell special characters.

Therefore added a *safe* quote function. The downside is, it requires #sed now. Well ....

Screenshot shows both improvements, I picked a "maximum weird" file name and used a file that has an 'ö' in its title 🥳 -- in the showansi debugging output, you can see the quoting in action 🤯

#MSDOS #ANSIart #retrocomupting

Yet another revision, #dos2ansi v1.6.3
https://github.com/Zirias/dos2ansi/releases/tag/v1.6.3

--- brings yet another subtle bugfix in #showansi: To set the title with an escape sequence, it uses "printf %b" with its \-notation for control characters, so literal backslashes in the title must be escaped as well 🙈

Apart from that, added nice "portability bikeshed":

- Removed the dependency on 'sed' again, quoting/escaping strings for 'eval' *can* be done in pure #POSIX #shell script.

- Added proper detection of the POSIX shell to my #zimk build system, along with variables to override the choice. Used that to generate the shebang for #dos2ansi. Really an awesome way to write /bin/sh without writing /bin/sh 😂 (rumors say there *are* POSIX systems with the compliant shell in a different path ...)

Screenshot: Yet another file with "weird" title from #SAUCE

#ANSIart #MSDOS #retrocomupting

I just released #dos2ansi v1.7, which includes one final bugfix: Quoted output of #SAUCE metadata should use proper "shell-style" quoting, so it can be directly consumed by #showansi, which it does now. Plus #showansi got a (tiny) bit of optimization.

https://github.com/Zirias/dos2ansi/releases/tag/v1.7

Well, just after releasing, I found more stuff to tackle. No, not new bugs at least.

But more weirdly broken ANSIart files. Example:
https://16colo.rs/pack/root0297/FU-B%26B.ANS

This one *has* SAUCE metadata, but it's not properly "hidden" behind a DOS EOF character (0x1a), which means even 'type' on a real DOS machine will output the SAUCE gibberish below the pic 🙈

So, there *might* be a v1.8 adding more heuristics to detect and parse more broken SAUCE ... 😂

#MSDOS #ANSIart #retrocomputing

#dos2ansi v1.8 released!
https://github.com/Zirias/dos2ansi/releases/tag/v1.8

There's some major rework inside adding support for seekable streams, which was necessary to implement the logic detecting #SAUCE even when the preceding DOS EOF character is missing.

Quite some improvements to the #showansi script as well, now comes with "fontsets" (explained in the comments in these and the config file) 😎

Plus added a manpage. Wanted a single source of truth (shared with the help output) and didn't find a good tool for that, so I quickly came up with my own:
https://github.com/Zirias/mkclidoc
It's bundled with dos2ansi now. Can produce man in both classic troff and BSD mandoc. The default choice is based on `uname` for now, tested on #FreeBSD and #Debian.

#ANSIart #MSDOS #retrocomputing

Released: #dos2ansi v2.0
https://github.com/Zirias/dos2ansi/releases/tag/v2.0

The real "visible change" is documentation. #showansi now got a manpage as well, and the one for #dos2ansi improved a lot. Also better build instructions and some updates/corrections in the README. With these docs, you can hopefully make it do exactly what you want 😉

Also, the build system (my own homebrewn #GNU #make framework) got lots of improvements and fixes.
https://github.com/Zirias/zimk/

#ANSIart #MSDOS #retrocomputing