Signum!

To: dziewon@xs4all.nl
Cc: root@NZ5608.spb.edu
Message-Id: <AAbVnvuuH4@NZ5608.spb.edu>
Organization: 25
From: "Nikita A. Zinoviev" <root@NZ5608.spb.edu>
Date: Sun, 2 Apr 0 13:50:29 +0400
X-Mailer: BML [MS/DOS Beauty Mail v1.36H]
Return-Receipt-To: root@NZ5608.spb.edu
Subject: Atari + formats
Lines: 16
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-UIDL: 954671324.maildrop1.21226

Hello

I am an Atari 1040 STf owner since 1986 (it's still ticking) and now we encountered a problem of conversion old Signum2's documents to a PC readable form. You are the onliest person I found in Internet that says something about Signum's format. So if you know something about converters from Signum to PC (but I don't think they can handle russian) or if you can write about it's text and font formats (so I will be able to write a converter myself to suit our needs) PLEASE inform me. I am desperate and looking to disassembling Signum but the only emulator I have with such a possibility is Pacifist and mine is compiled as Non-debug and I can't get a Debug one (it's sites seem to be dead). So if you can helf with that please do it.

With best regards
Nikita Zinoviev

I recently managed to convert some old Signum!2 files for a Jugoslavian Atarian so I should be able to help you as well.

What I need to know is how your text-format is, what the alphabet looks like. Send me - as an attachment - an old Signum!2 file + a Signum!2 file that only contains the alphabet:

Russian char. = Latin description
a = a
b = b
c = c

= sjtsj
etc.

gtx, Rein

p.s.

do you use ISO-8859 standards or KOI-8??

To: dziewon@xs4all.nl
From: Nikita Zinoviev <root@NZ5608.spb.edu>
Subject: Atari + Signum2 + formats
Cc: root@NZ5608.spb.edu

Hello

I am sending you a zip file containing:

testm1.sdo // an old signum2 file
rusantic.sdo // a file with an alphabet (uses anticpyc.scr)
ruswin.txt // a txt file that contains the same alphabet
// in our windows encoding
FONTS\
mathem // mathematical font
griech // greek font
antickro // english font
// used in testm1.sdo
anticpyc // russian font used in all sdos
// I think you would like to see it

P.S. It seems that we are not using ISO-8859 because I can't read some symbols in your letter but may be I am wrong. I can read KOI-8

I am looking forward for your reply and please if it won't cause you much trouble send me a descriptions of Signum2 text and fonts formats.

Thanks a lot,
Nikita

thanks for the files you sent me. I had a look at your text with 'part 1' and 'diffusion', they look like Russian to me :)

It takes me some more time to figure out the conversion table but eventually it should be in Windows 1251 [that's how I read your ruswin.txt using Papyrus].

If everything works out alright I can sent you a Win1251 file tomorrow...

gtx, Rein

From: "Nikita Zinoviev" <root@NZ5608.spb.edu>
X-UUCPGate: SMTP->UUCP, MailServer, ver. 2.12
To: "R.C. Bakhuizen van den Brink [Rein]" <dziewon@xs4all.nl>
References: <200004031947.VAA14368@smtp1.xs4all.nl>
Subject: Re: Atari + Signum2 + formats
Date: Tue, 4 Apr 2000 17:36:37 +0400
MIME-Version: 1.0
charset="iso-8859-2"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2615.200
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Lines: 40
Content-Type: text/plain; charset=us-ascii
X-UIDL: 954855642.maildrop2.21460

Ok It looks quite fine except ? is replaced with . and ? with +

That is not a big problem. But you've converted ALL of it into Russian independing on fonts. I could do this too using Signum2's option "save as ASCII" and then running a simple table based converter on the output file. The question is to make a more accurate converter that doesn't need Signum2. So I'd like to get as much information from you as possible.

Thanks A Lot

Nikita.

{\rtf0\pc \paperh15840\paperw31680\paperwauto\signumpixels
\margl0\margr0\margt42\margb426\fsauto\sl-293\tx1152\tx1424\tx1936\tx2304
\tx3088\tx3456\tx4448\tx5472\tx5840\tx6608\tx7264\tx8688\tx9648\tx10368
{\fonttbl{\f0\fnil ANTICPYC;}
{\f1\fnil ANTIKRO;}
{\f2\fnil MATHEM;}
{\f3\fnil GRIECH;}
{\f4\fnil GROTMIKR;}

\par

lky gjnjrf ghjbpdjkmzj ds,hfzzjuj rjvgjztznf \f1 K\expnd6 \f0 (\expnd0 \f1 J ) \f0 bvttn cktle.obq dbl`\par \li0\fi0\ri21744\sl-26\ql \m6848 \f4 K\par \li0\fi0\ri21744\sl-134\ql \m3584 \f2 \fs16\fsh24 *\par \li3040\fi-96\ri21744\sl-133\ql \f1 \fs16\fsh24 \fs24 J\m144 \f0 =\m48 -\m-16 \f1 D\m128 \f2 n\m64 \f1 N\m112 \f2 +\m48 \f1 N\m112 V\m192 \f0 \'E1\m3072 (1'1) \ql \f4 K K\expnd-23 K K K F\expnd-4 K \sl-106\ql \par

\par

Dear Nikita,

you're quite right saying that you could have done the ASCII export yourself and have it followed by a conversion....

The big problem is indeed not the Russian text but the formulas and Signum!2 micropositioning, they get intertwined unless you know exactly how Signum!2 got the internal 'tags' organised. But I'm afraid that's still a ASH trade-secret as long as they're selling some version of Signum! You would have run into the same problem 'saving it as ASCII', I'm afraid ;)

I'll have some more thoughts about it since it's a rather intriguing problem! The Leiden University has similar problems with texts being a mixture of Dutch and Russian all in Signum!2 files....

gtx, Rein

P.S.

Not knowing the internal format of Signum!2 it is possible to read the SDO-file with Papyrus provided you have the right Signum!2 fonts [I didn't get the 'Grotmikr' font from you]. Papyrus [version 7.58b] has a converter written by Andreas Pirner that translates the SDO file into RTF [rich text format]. Inside the RTF-file you will find all the necessary font-tags [plus a whole lot of rubbish]. After having translated your Russian file into RTF I was able to re-read it in Papyrus and everything - including the formulas - looked more or less alright although the micropositioning seemed to me a bit 'foggy' Your keyboard-file looked like this:

{\rtf1\pc{\fonttbl{\f1\fnil ANTICPYC;}
{\f2\fnil GRIECH;}
}{\colortbl;}\paperw31680\paperh12960\margl0\margr0\margt1473\margb1473\pgnstart1
\ftntj\ftnstart1\makeback
\sectd\pgndec\headery1080\footery1080\cols1\colsx567\pgndec
\sl-326\tx16\tx1152\tx2304\tx3456\tx4608\tx5760\tx6912\tx8064\tx9216\tx10368\f2\fs24 \par
\f1 1\expnd3 234\expnd6 5\expnd3 678\expnd6 9\expnd3 0\expnd6 ? QWERTZUIOP\expnd3\'9a\expnd6 * ASDFGHJKL\'99\'8e \'c6\expnd3 Y\expnd6 XCVBNM;: qwert\expnd3 z\expnd6 u\expnd3 i\expnd6 op\expnd3\'81\expnd6 + asd\expnd3 fg\expnd6 hjk\expnd3 l\expnd6\'94\'84 @\expnd3 y\expnd6 xc\expnd3 v\expnd6 bn\expnd3 m,\expnd6 . }

I seems to me that when you get rid of the '\expend3 ' overhead [?] you may easily convert 'qwertz' into the appropiate Russian characters....

From: "Nikita A. Zinoviev" <root@NZ5608.spb.edu>
X-UUCPGate: SMTP->UUCP, MailServer, ver. 2.12
To: "Rein Bakhuizen van den Brink" <dziewon@xs4all.nl>
References: <aMail1.26b.20000407003240.2078917053@somehost>
Subject: Re: Signum!2
Date: Sun, 9 Apr 2000 00:19:42 +0400
MIME-Version: 1.0
charset="iso-8859-2"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2615.200
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Lines: 40
Content-Type: text/plain; charset=us-ascii
X-UIDL: 955225742.maildrop2.3421

Hello

THANKS! I've got your file RUSANTIC.ZIP but I didn't manage to find out what it is. I can't read converted testm1.rtf because I have no ANTICPYC font. Also what is rusantic.rtf (just converted rusantic.sdo as it seems to be or something else). But most mysterious is textconv.prg - I was to lazy to run it on ST (also now I have problems with PC floppy) it didn't run under PaCifiST but it runs perfectly under TosBox (rather amazing fact ...). And I can only guess what it is asking (an input .sdo filename , and output rtf and e24 fonts - but which of them??) . And really because I had an 9 pin Gemini printer I had only .scr and .p9 fonts no .e24 at least by hand. Also I thought converter would be a PC program or you are using Mac (if so - you're lucky) ? And what to do with font convertion?

Where did you get the textconv.prg from? I'd like sources if you have it - then I could compile it under PC.

Also: is the problem with Signum2 and it's format quite solved?

Now I've got all necessary software (IDApro that handles 68000) to disassemble it or better sprint.prg - it's much shorter - rather illegal move but if there is nothing else to do ?! But this would be rather hard work, is it really needed? (you mentioned Leiden University - may be they are interested in a sort of development ?) And what to do with fonts?

So I am waiting for your explanations and suggestions.

yours truly,

Nikita

Dear Nikita,

I can imagine that you don't quite know what it was. The 'testm1.rtf' is a conversion of your 'testm1.sdo' file made by testconv.prg. When it asks for 'e24' it suffices to have your 'scr' files renamed into 'e24'. Of course all necesssary Signum!2 fontfiles should be available.

In the below fragment of a rtf-file only the untagged characters should be converted according to the appropriate charset-table; as long as '\f0' is valid use the ANTICRYC to ISO-8859-5 table; as soon as '\f2' is the MATHEM to Mathematic font table, etc.

Change the Signum!2 fontnames in the definition lines {\f1\fnil ANTIKRO;} into your own font-names, and you should be able to import this file into MS Word or any other PC-programme that understands RTF.

I understood that you could do some programming yourself so it shouldn't be too difficult.

I don't think there is another way!! It seems to me quite awkward to have to skip the RTF-tags... Textconv.prg goes with Papyrus 7, but should be available separately [?].

Have a try at the RTF-files and see whether you can convert them as I wrote above.

gtx, Rein

Also I have problems with getting Papyrus (that's why I haven't replied et) - nobody knows what it is. I found it's homepage in Internet but there are only 7.0xx versions also they say that 8.0 is available only for Mac. I've looked at their format libraries but can't see any sdo. (may be it is not easy to find...) so that's it.

From: "Nikita A. Zinoviev" <root@NZ5608.spb.edu>
X-UUCPGate: SMTP->UUCP, MailServer, ver. 2.12
To: "R.C. Bakhuizen van den Brink [Rein]" <dziewon@xs4all.nl>
References: <200004082118.XAA25661@smtp8.xs4all.nl>
Subject: Re: Signum!2
Date: Sun, 9 Apr 2000 14:02:06 +0400
MIME-Version: 1.0
charset="iso-8859-2"
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2615.200
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200
Lines: 109
Content-Type: text/plain; charset=us-ascii
X-UIDL: 955274742.maildrop2.15735

Dear Rein,

Ok, I maneged to write a shifter and it works rather fine - converts only Russian fonts leaving English unchanged. It was nice to see that text attributes(italics, etc.) are preserved. I've surprised to see that some of Signum characters textconv converts to \'xx and this must be handled manually (now all is OK) . But all this additional converters wouldn't be neccessary if there is a way to "convert" Signum fonts to PC, what can you say about it? (as I remember your site was concerned with Signum fonts, not documents). Textconv ommits Signum's images, sometimes spacing is incorrect (if narrow fonts are used ("somethingl iket his" - rather curious effect or "likethis"). And so on. If I can help you with something or if you know other problems here that need solving, or you have additional info

please write me.

Thanks A Lot

Nikita

Dear Nikita,

I'm glad you have managed alright sofar. As said before the Russian text has never been the problem, the combination of two different fonts plus two completely different charsets [Russian, Mathematic] requires the use of tags [as Signum!2 does internally]. So a conversion into RTF is absolutely necessary.

Did you get the Mathematic symbols alright??

The Signum!2 fonts can be converted into GEM-fonts which is of no help to you on a PC unless you have the old GEM-desktop + Ventura DTP [GEM not DOS or Windows version]. You should take into consideration that the Signum!2 fonts are bitmap fonts and I know of no font-converter that converts into vector[ized] fonts. All converters go from vector to bitmap. All modern computerprograms use vector-fonts [TrueType, PostScript, Speedo]. What you could do is adapt a TrueType font that contains Cyrillic characters and change their position with Fontographer or Fontmonger according to the position inside your original Signum!2 fonts. But I think that's trying to get the world upsidedown. We should accept that fact that 15 years ago the Signum!2 concept was avantgarde. Even the RTF concept is getting outdated now as it needs a 8-bit character encoding [only 0-255 are accepted]. With multilingual texts a conversion into 16-bits character encoding [Unicode UCS-16, UTF-8] is the next step...

In what language have you written the 'shifter'?? If it's in a language that can be used on an Atari, I wouldn't mind having a look at the source code... Papyrus has a Windows version but I don't think it will handle Signum!2 fonts...

gtx, Rein

Dear Rein,

Really (to keep things easy) I thought that my program must be used several times one conversion at time , but I don't try to convert mathematical symbols because I don't know what to replace them with.

Yours, Nikita

what a pity!

Should you have access to MS Word 97 [or higher] I would suggest make use of their Unicode format. When 'shifting' you might also try to use the Unicode/HTML 4.0 entities:

like
&#8704 forall
&#8706 partial differential
&#8707 there exists
&#8709 empty set
&#8711 nabla
etc

The same goes for your Russian characters....

Do you need lists or pictures of those ???

gtx, Rein

P.S.

Papyrus can read above entities ;) in HTML-files

Dear Rein,

After a great deal of thought I've nearly got your idea. I am not familiar with Unicode format but I have access to MS Word . Sorry, but what to do with these img files? What are they and where are they from? I suppose, I need lists and pictures.

Thanks, yours,

Nikita

I hope you can display the IMG pictures on your Atari, If not I can send you GIF's instead....

The images are screendumps I made today to show you
-1 the effect of the HTML-entities in Papyrus: the actual math symbols became visible
-2 the range in Unicode for Cyrillic
-3 the range in Unicode for math symbols

Lucida Sans Unicode is the font that would cover most of your requirements but also Cyberbase should do:

Bitstream CyberBase - 1249 glyphs in version beta v1.0
Platforms: Windows
Unicode ranges: Basic Latin; Latin-1 Supplement; Latin Extended-A; Latin Extended-B; Spacing Modifier Letters; Combining Diacritical Marks; Greek; Cyrillic; Hebrew; Arabic; Thai; Latin Extended Additional; General Punctuation; Superscripts and Subscripts; Currency Symbols; Letterlike Symbols; Number Forms; Arrows; Mathematical Operators; Miscellaneous Technical; Box Drawing; Block Elements; Geometric Shapes; Private Use Area; Alphabetic Presentation Forms; Arabic Presentation Forms-A; Arabic Presentation Forms-B

Availability: Free download from ftp://ftp.netscape.com/pub/communicator/extras/fonts/windows/

gtx, Rein