NETWORK@FRSAC11.BITNET.UUCP (02/12/87)
NETWORK at FRSAC11
To: INFO-ATA at SCORE.ST
After spending some time on the retrieval of software on the net, I must
say that:
STOP! using brain damaged obsolete methods on the ATARI ST: The ST is a
8 bit international character set system (keyboard, screen, printer, and
software) so STOP doing ... & 0x7F ... on input... this makes totally
useless programs like LESS and MicroGnuEmacs (version MG1)... If you
live in a country where US-ASCII is NOT the national thing...
I have even seen an emacs that used Scan codes for the keyboard, and was
therefore totally unable to be used on most of the European STs.
I had to spend many evenings and nights debugging MicroEmacs v30 (The father
of MG) to remove all restrictions, and make it a usable editor (Took
more time to do so than to implement the most useful UN*X V8 regular
expression search & replace, I cant live without them; too bad they
are not implemented in MG, as the use of function keys on the ST.)
So as soon as I have the sources of MicroGnu I will do the same.
(I have the complete Mod.sources of MG, but it is for the Amiga or UN*X or
VMS, unlike the original that included the ST. But I am lacking the ST
specific parts, fellow Europeans will wait...)
As I had mail about it: For CScheme on the ST: The microcode is ported
using MWC (The best overall C compiler as far as I am concerned), It is
a 170K binary, but in order to load the runtime (band) you need a lot more:
the MIT keep on telling me that under 2Meg it is quite impossible, one
would have to strip down the thing, and I am not good enough to begin
an important runtime Scheme job. So wait until I and you have 2 or 4 megs
(under $1000 the upgrade ?). By that time the new CScheme will be out
with a 68000 compiler that should put to shame a lot of language...
Last for this time: I have tested/used a lot of uudecode that came on the net,
Not a single one is able to decode all that is posted without having to
edit the file beforehand. So, tired and bored, i rewrote a uudecode that
worked on all files posted, with ability to deal automatically with:
Files with a table header (That was an excellent idea, to put a table)
Files with include commands (the same source as above...)
Files with missing trailing blanks (I am on BITNET, and this is obvious)
Files with tilde instead of circonflex (not common but there is some)
Files with grave accent (backquote) instead of blanks (more common)
Files that have any reasonable combination of the above...
Files that have the usenet header and trailing signatures... It is in the
specifications published this/last week.
It is writen in MWC, quite fast (50 seconds,everything included to decode
UNITERM 1.6g, on a SH204), will give some diagnostic if your file is too weird.
Currently I just double click on the .uue files freshly kermited on the ST,
and there it is, no fuss. (Thanks to the guys that gave information on the
desktop.inf, now I have a line ... C:\XBIN\UUDECODE.TTP *.UUE@ ...)
I Will send it for posting this/next week, unless you tell me not too.
(14K binary)
Yes you guessed it... I am in France.
+--------------------------------------------------+
| Jean-Pierre H. Dumas |
| |
| network@frsac11 (bitnet) |
| network%frsac11.bitnet@wiscvm.wisc.edu (arpanet) |
| ..!ihnp4!frsac11.bitnet!network (usenet ?) |
| dumas@sumex-aim.stanford.edu (arpanet) |
+--------------------------------------------------+