[z-machine] [Inform mntce] The Z-machine and Informs 6 and 7

[Johnny Billquist]

On Sun, 23 Oct 2005, fredrik at ramsberg.net wrote:

> Quoting Johnny Billquist <bqt at Update.UU.SE>:
>
> > On Thu, 20 Oct 2005, Rikard Peterson wrote:
> >
> >> fredrik at ramsberg.net skrev:
> >>
> >> >Unfortunately, there isn't much interest in supporting accented
> >> >characters in terps, especially in user input. I haven't had the
> >> >chance to examine any of the Palm terps with regard to this, but
> >> >I would be surprised if any of them had acceptable support for
> >> >accented characters. And I'm just talking about Latin-1
> >> >characters here. I expect the situation is a lot worse for
> >> >Cyrillic, Greek etc.
> >> >
> >> >
> >> I haven't yet played any Swedish IF on it, but I have Frobnitz installed
> >> on my Palm (a Treo 600) and I loaded up "Den Mörka Manteln" to it to
> >> check how it handled it. I'm pleased to report that Frobnitz handles
> >> both input and output without problems. It only messes up in the game
> >> listing where the game is shown as "Den M~rka Manteln".
> >
> > Well, I can report that my Z-machine implementation for the PDP-11 handles
> > this game just fine as well... :-)
>
> That's really good news! I also want to thank Rikard for his
> investigation.
>
> If you, or anyone else, would care to spend ten minutes making a more
> complete test, please download this file:
>
> http://www.microheaven.com/swedish/Swedish_testfiles.zip
>
> The zip archive includes a small test game in z5 and ulx format
> and a transcript. The game first writes out three words, with
> the first character capitalised. Then the game starts and you
> need to type in (or cut-and-paste) five commands. If the output
> is the same as in the transcript, the terp has passed the test.
> If it passes the test, the terp can...
>
> * print accented characters contained in strings.
>
> * print accented characters using "print (char)".
>
> * handle a completely new Zscii character table, defined
> with Zcharacter (Z-machine only).
>
> * accept accented characters in user input and store
> them using the correct charcodes in the buffer.
>
> * properly match words containing accented characters in
> player input against dictionary words.
>
> * downcase accented characters in user input (Z-machine only).
>
> I'd be very happy to get reports about how this test went,
> with a download link to the specific version of the terp used,
> and will publish the results.

Ha! Thanks. That progam uncovered two bugs in my code, that I now fixed.
One was that I wasn't doing any case conversion of input, and the other
was that I was doing ZSCII -> Latin1 conversions when printing to stream
3.

All works fine now. The terp is ZEMU and the new version is V1.9. I'll
place it on my site tomorrow or so. It can be found at
ftp://ftp.update.uu.se/pub/pdp11/zemu.tar

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol