Info-IBMPC@USC-ISIB.ARPA (Info-IBMPC Digest) (11/08/85)
Info-IBMPC Digest Friday, 8 November 1985 Volume 4 : Issue 129
This Week's Editor: Richard Nelson
Today's Topics:
STRIP15.ASM Added to INFO-IBMPC Library
SIMTEL-20 Public Domain Repositories Moved
IBM-PC to Symbolics Ethernet Connection
FORTH
Epsilon
Express Systems Tape Backup
FastBack
Disk Optimizer
Man-Machine Interfaces for the Blind
Whetstone on Cray = 1/200 PC-AT
Hidden Files
Beta Test Sites for TransNet
----------------------------------------------------------------------
Date: 2 Nov 1985 17:01:43 PST
Subject: STRIP15.ASM Added to INFO-IBMPC Library
From: Koji Okazaki <swg.KOJI@USC-ISIB.ARPA>
To: Info-IBMPC@USC-ISIB.ARPA, David_and_Toad <ABN.ISCAMS@USC-ISID.ARPA>
cc: Koji@USC-ISIB.ARPA
The following files have been added to our public domain software
library. Arpanet hamsters can FTP them from host USC-ISIB, directory
<INFO-IBMPC>. For our latest program listing, FTP PROGRAM-LIBRARY.LIST.
If you're new at this sort of thing and you need more *G R A P H I C*
help, send your messages directly to me (Koji@USC-ISIB.ARPA)...
STRIP15.ASM This useful hack strips Wordstar-unique characters
strip15.txt and replaces four or more consecutive spaces with
strip15.doc tabs (something that some other tab programs didn't
do correctly). That is, it converts WS document
files to regular ASCII format. David used the guts
of TABS (another public domain file alignment
utility) whilst writing STRIP15. STRIP15.DOC is
in David's words, "a typical horrible 8-bit WS
document file." STRIP15.TXT is the same file --
this program's documentation -- after STRIP15
worked on it. If you know that you're not going
to be particularly thrilled by this demonstration,
and you just want the program and its docs, then
just FTP the .ASM and .TXT files.
<TABS author: Vernon Buerg>
<David Kirschbaum, ABN.ISCAMS@USC-ISID> 11/2/85
------------------------------
Date: Sun, 3 Nov 1985 15:38 MST
From: "Frank J. Wancho" <WANCHO@SIMTEL20.ARPA>
To: INFO-CPM@AMSAA, INFO-MICRO@BRL, ADA-SW@SIMTEL20.ARPA,
UNIX-SW@SIMTEL20.ARPA
Cc: INFO-IBMPC@USC-ISIB, INFO-HZ100@RADC-TOPS20
Subject: SIMTEL-20 Public Domain Repositories Moved
We now have our new RP07 disk drive installed as PD: and all the
public domain repositories have been moved from their former homes on
MICRO: and PS: to PD:. PD: is now 43% full, and I suspect will start
growing again. The new releases from SIG/M and PC/BLUE will be
uploaded shortly after we receive them. Look for announcements of
availability.
Former Home New Home
----------- --------
PS:<ADA*> PD:<ADA*>
MICRO:<CPM.*> PD:<CPM.*>
MICRO:<CPMUG*> PD:<CPMUG*>
MICRO:<SIGM*> PD:<SIGM*>
MICRO:<PC-BLUE*> PD:<PC-BLUE*>
PS:<UNIX*> PD:<UNIX*>
The CRC Lists will be updated to show PD: instead of MICRO: as we get
to them.
--Frank
------------------------------
Date: 4 Nov 1985 14:13:43 PST
Subject: IBM-PC to Symbolics Ethernet Connection
From: Billy <BRACKENRIDGE@USC-ISIB.ARPA>
Last week's PC week had an announcement from the people who make Golden
Common Lisp for the IBM-PC that they have an ethernet based link to
Symbolics machines that allows a PC to remotely login. I don't have any
details other than it requires the 3Comm Ethernet cards.
I would assume one could make remote function calls from Common Lisp in
the PC to a Symbolics Machine. They also announced their large memory
model Lisp for the AT.
Has anyone tried this code?
------------------------------
Date: Mon, 4 Nov 85 19:28 EST
From: SECRIST%OAK.SAINET.MFENET@LLL-MFE.ARPA
Subject: FORTH
TELL THE NET ABOUT YOUR FREEBIE FORTHS !
I would like to compile a list of FORTH implementations that are available
on a not-for-profit basis on net or publicly accessible BBS systems for
several net lists (INFO-CPM, INFO-APPLE, UMFORTH [aka FIGIL, bitnet],
INFO-IBMPC and any others that are appropriate or suggested.)
If you have downloaded FORTH from ANYPLACE and found it to be usable and
intact, please send me details on the version of FORTH, what machine it is
for, which standard it follows, THE SYSTEM YOU ACQUIRED IT FROM, the date
you snarfed it or saw it there last, and anything else you care to say
about it. (Sources could include SIMTEL20, CompuServe, BBSes, etc.. The
list will include multiple sources since people do not all have access to
the same places.)
I will post the results in a couple of weeks. Please send info to:
SECRIST%OAK.SAInet.MFEnet@LLL-MFE.Arpa, with thanks.
-- Richard Secrist, President
East Tennessee FORTH Interest Group
------------------------------
Date: Wednesday, 6 November 1985 16:16:34 EST
From: Joe.Newcomer@a.sei.cmu.edu
To: info-ibmpc@usc-isib.arpa
Subject: Epsilon
Allow me to join the growing voices of Epsilon 3.0 beta-sites.
I've got some good news and some bad news.
The good news is that Epsilon 3.0 is the absolutely spiffiest EMACS-class
editor I have ever used, period, including TOPS-20 EMACS, various
Gosling EMACS clones, and GNU. It is easily programmed (I'm not quite
at the level of throwing up at C code, currently having written about a
megabyte of it in the last year), 3.0 seems to have almost all the primitives
I want, and it is ***FAST*** even for some slow operations.
The bad news is it is addicting. I spent the first week I had the beta test
reprogramming my entire universe. Some of this was some changes (mostly
matters of personal taste) to the standard commands (e.g., C-U-C-V scrolls up
4 lines, not 4 screens...a subtle and pervasive misinterpretation of the EMACS
specification); a lot were extensions to the standard commands; and most were
entirely new pieces of functionality. An automatic change-log maintainer, a
semi-fancy C-mode (I haven't finished it yet), compare-windows (which they
rewrote in the 3.0 release to be an order of magnitude faster than my
version), buffer menus, etc., etc., etc. I got nothing else done that week,
but I got a really super editor as a result.
It is nice to know that if I don't like how something works, I can go in and
change it. No more grovelling to the TECO wizards; no more attempting to
decrypt completely undocumented variables in MockLisp code; no more waiting
65 seconds while EMACS loads on my PC, no more running out of memory (love
that automatic swapping!), ...
Anyway, as a consequence of having Epsilon, I ordered an IBM PC for my office
to replace a MicroVax, because the PC environment is now a super editing
environment, and since I do a lot more text now than I'd like, I want the
best editor that I can get.
In addition, my AT has a 1MB extended memory, so I assign the Epsilon swap
device to E:, a 1MB VDISK in extended memory, and it really screams along
even when I've got most of my source library in buffers.
The process window feature is so mind-bogglingly wonderful that I hardly
ever use straight DOS...just for screen-oriented programs (which I try to
run as a subprocess). Everything else is in a process window. Being
able to scroll back, or be reading one part of the DOS output while entering
new commands, is super.
And process windows work RIGHT!!!! Unlike the bogus process windows I had
to contend with on the Vax, where typeahead is lost or confused, Epsilon
has the notion of the input cursor and it is *RIGHT*.
Needless to say, I have no connection whatsoever with these people except as
an exceptionally pleased beta test site. But, when somebody does the job
right for a change, they deserved to be pointed out to all and sundry.
While you may argue about the virtues of EMACS-style editors vs. mouse-based
what-you-see-is-what-you-get style editors, I've never found a mouse-based
editor yet that was worth the powder to blow away its distribution disk.
Most of them don't work very well for constructing C code, and none I've seen
give me features like auto-indent of C code and alignment of matching {}s,
done precisely the way *I* want it done.
If you like EMACS-class editors, rush out and get this one...it is the very
best of its kind.
joe
------------------------------
Date: Monday, 4 November 1985 12:16:45 EST
From: Joe.Newcomer@a.sei.cmu.edu
To: info-ibmpc@usc-isib.arpa
Subject: Express Systems Tape Backup
I've been evaluating tape backup systems recently. I had a chance to
see the Express Systems tape backup software in operation last week.
These guys have got it right. No pre-formatting of the tape, multiple
save-sets, incremental backups, full file-by-file backup, selective restore,
directory printouts, file searches, etc., etc. I used it and asked a lot
of questions and every question was answered not with "sure we can do that"
but with a demonstration of the solution.
I will probably buy this unit, because it seems to do everything I need. I
would urge anyone seriously considering a tape backup to look at this one. It
comes as both a built-in half-height unit and as an external unit; I will get
the external unit and they even will sell me two more controller cards for our
other machines ($400 each) so we only need to move the tape drive around once
a month for the full monthly backups.
See also my review of FastBack.
------------------------------
Date: Monday, 4 November 1985 12:23:49 EST
From: Joe.Newcomer@a.sei.cmu.edu
To: info-ibmpc@usc-isib.arpa
Subject: FastBack
I recently purchased 5th Generation System's "FastBack" file backup
utility. I was faced with the prospect of having to do a full backup
of my 17Mb of data, which was depressing. So I had 17 incremental
1.2Mb disks...
I got it, I love it. It did the backup in 25 minutes; no formatting of
the diskettes is required. It formats them along with writing the data.
It took 25 minutes because I forgot to update the "buffers=" declaration
in my config.sys file, and that omission really showed up (they told me
to do it, but I forgot). It should have taken more like 16 minutes.
It was advertised in the special IBMPC issue of Byte, at a special
introductory price of $149. It is copy-protected and requires putting the
master disk in the drive to activate it, but for $25 they will sell a
non-protected license upgrade, which I will probably get.
BACKUP/RESTORE, goodbye.
------------------------------
Date: Monday, 4 November 1985 12:35:08 EST
From: Joe.Newcomer@a.sei.cmu.edu
To: info-ibmpc@usc-isib.arpa
Subject: Disk Optimizer
I also purchased the Disk Optimizer program from SoftLogic (the makers of
DoubleDos). It has several utilities, including a file encryptor. One of the
utilities, "analyze", looks at disk fragmentation and gives you a percentage
value which is a quality measure of how contiguous files are. Depressing;
Lattice C, the Lattice libraries, dBASE III, and other things I use a lot were
*** 0% *** meaning they were scattered to hellangone over the disk. I ran the
optimizer after doing a full backup (see FastBack review). The purpose of the
optimizer is to rearrange the files so that they are all contiguous. It also
sorts the directories so that subdirectories appear at the front, and,
apparently (from looking at a disk dump) the "deleted file" entries are moved
to the back). It said "will run perhaps 30 minutes the first time it is run".
Well, after three hours I gave up and went to bed; at 9am the next morning (8
hours later) it was still "optimizing" the disk (the head was freebing every
few seconds). I have no idea what it was doing; I stopped it. I then ran
"analyze" again, and every file was listed as 100% contiguous and the overall
disk optimization was given as 99%. No, I don't know why.
I then rebuilt my C-based major system and it took about 5 minutes less.
(37:35 vs 42:10) so it really did have an effect. It is worth noting that the
C sources were on the D: drive and not optimized, so all this saving was in
the compiler load, linking, etc. over the 85 modules comprising the system (I
use E: VDISK for the intermediate files so no disk activity occurs there at
all).
Copy-protected, but installable so the floppy is not required. $49 and
a bargain at the price.
------------------------------
To: Info-Kermit@cu20b.ARPA
Cc: Info-IBMPC@usc-isib.ARPA, Info-Micro@brl-vgr.ARPA
Subject: Man-Machine Interfaces for the Blind
Date: 02 Nov 85 19:52:51 EST (Sat)
From: dave@mimsy.umd.edu
the following companies have experts in understanding the specific
problems of developing and using man-machine interfaces for the blind.
Automated Functions, Inc. Washington, D.C.
Maryland Computer Services Gambrills, Maryland
Talking Computers Arlington, Virginia
Telesensory Systems Inc. Palo Alto, California
I own a Macintosh, but find it very difficult to use myself. I own
it because, nevertheless, it is the best tool for the application for which
I got it. I have heard of a blind Mac user who reads the screen with an
OPTACON.
Dave Stoffel
Amber Research Group, Inc.
------------------------------
Date: Sun, 3 Nov 85 21:09:10 mst
From: dwf%f@LANL.ARPA (Dave Forslund)
Subject: Whetstone on Cray = 1/200 PC-AT
To: info-iBMPC@USC-ISIB.ARPA
I ran the Whetstone benchmark from <info-ibmpc> on a Cray-1S here at
Los Alamos and got a time of .05 seconds. Increasing I to 100 gave .5
seconds so the overhead was not to bad. That figures at 20 million
whetstones/second and there was no vectorization by the compiler; all
the code was scalar
David Forslund (dwf@lanl)
[This compares with a 15MHz AT time of 10.0 seconds, so the AT works
out to be 1/200 of a Cray.]
------------------------------
Date: 5-Nov-85 5:49PM-PST (Tue)
From: Mark Zbikowski <Markz@Gonzo>
Subject: Hidden Files
To: Uw-Beaver!Usc-Isib.Arpa!Info-Ibmpc@Xenix
A comment on a recent INFO-PC article:
...
Hidden Files
...
Date: 1 Nov 1985 14:25:01 PST
Subject: Hidden Files
From: Dwight Baker <SWG.DBAKER@USC-ISIB.ARPA>
To: info-ibmpc@USC-ISIB.ARPA
Someone recently requested help with hidden files. Here are some
instructions that might help.
...
If you don't have a disk utility tool like Norton's you can use Debug
to change the status. Before you work on a live disk try formatting
a new disk with the /s option and then examine the directory entries
of the DOS files. The following procedure will show you how to this
can be done using debug.
...
The method you espouse is not only incomplete (no instructions on how to do
this in subdirectories) but it is DANGEROUS. It only takes a little slip-up
to trash your root directory, or worse, your FAT.
You don't see people advocating directory listing via reading /dev/rrp07.
There is an operating system feature that is easy to use to get the same
effect.
System call 43h will allow you to change attributes (except VID and DIR) to
your heart's content:
To set file attributes to Hidden and System:
mov ax,4301h
mov cx,0004h + 0002h
mov dx,offset filename
int 21h
jc SetDidntWork
To get a file's current attributes:
mov ax,4300h
mov dx,offset filename
int 21h
jc GetDidntWork
This is MUCH easier than twiddling with raw sector I/O.
------------------------------
From: Ross Greenberg <greenber@phri.uucp>
Subject: Beta Test Sites for TransNet
Date: 31 Oct 85 13:12:27 GMT
To: info-ibmpc@usc-isib.ARPA
After a long development period, TransNet is finally just about done.
What is it?
Good question! TransNet allows you to do XMODEM transfers of programs
as a background process while you're busy in the foreground with
Lotus, or any other program.
If the machine that you call is running TransNet, it will appear as
a rudimentary upload/download BBS. If, in addition, your machine is
running TransNet, then the two machines will go about communicating
without you having to notice.
It presently runs on the IBM-PC and *true* clones.
Enhancements underway include mail capability (*not* Unix Mail as of
now, I'm sorry to say!), and remote interactive character-by-character
communications.
Anyway, I'm looking for some people to give it a shot, and to get back
to me with suggestions regarding enhancements as well as that one ( :-) )
bug that I couldn't find.
If you're interested, please respond to me at the below address,
and make sure to include your Snail address and telco number.
You can try it out for uploads and downloads on my development machine,
with login username "demo" and password "demo". I'll be using the
machine for development during normal working hours (8:00am - 2:00am EST),
but *that doesn't matter* as TransNet will be running in the background!
Thanks.
Ross M. Greenberg
ihnp4!allegra!phri!sysdes!greenber
[The phone number of the above mentioned board: (212)-889-6438 .]
(Hit a few returns to establish connection....only at 1200 for now)
------------------------------
End of Info-IBMPC Digest
************************
-------