FidoNews · Vol 2, No 7 · 18 Mar 85
fidonews -- 18 Mar 85 18:52:30 Page 1
Volume 2, Number 6 15 March 85
+----------------------------------------------------------+
| _ |
| / \ |
| - FidoNews - /|oo \ |
| (_| /_) |
| Fido and FidoNet _`@/_ \ _ |
| Users Group | | \ \\ |
| Newsletter | (*) | \ )) |
| ______ |__U__| / \// |
| / FIDO \ _//|| _\ / |
| (________) (_/(_|(____/ |
| (jm) |
+----------------------------------------------------------+
Publisher: Fido #1
Chief Procrastinator: Tom Jennings
Disclaimer or dont-blame-me:
The contents of the articles contained here are not
my responsibility, nor do I necessarily agree with them;
everything here is subject to debate. I publish EVERYTHING
received.
You can take this to mean anything you want, but
hopefully as an invitation to comment, make suggestions, or
write articles of your own.
.........................................
HOT NEWS
This is my last editorial, yeah! Thom Henderson at
FidoNode #375 is taking over as Head Procrastinator. Thom
has an article here.
Please note that this means you submit articles to
him, not me. You can FidoNet them, or upload them manually.
The latter will probably be a lot easier than spending hours
and hours trying to autodial into Fido #1.
Good luck to Thom, and thanks. Thanks also to the
others who volunteered for Procrastinator duty, it is
appreciated greatly.
No editorializing here today; I wrote a whole
article instead. There are a lot of changes coming soon.
This also means that I'll be able to WRITE ARTICLES
instead of having to MUNGE ARTICLES. Currently, I cringe
when I think about the newsletter; usually it's 10 PM on
Monday night when I think about it, and have to work on it
then. Ecch! I'll probably write more now that I don't have
to. (?)
fidonews -- 18 Mar 85 18:52:37 Page 2
NEWS
ARTICLE SUBMISSION
Note that articles go to Fido #375, not #1. Please
make a note of it, and tell others.
All articles you see in this issue are written by
users and sysops, and have one way or another managed to
consume disk space on Fido #375. In order to get rid of
them, and free up my precious disk space, I include them
here, then quickly delete them. Then they are YOUR problem.
EDITORIAL CONTENT:
Totally up to you; I publish anything at all.
Article are generally Fido or BBS related; this is by no
means a decision on my part, nor a requirement.
FOR SALE, WANTED, NOTICES:
Pretty much self explanatory. Commercial ads are
welcomed, if of reasonable length. These will be run each
issue, or until I lose them. (Dont laugh, Ive already lost a
few.)
SUBMITTING AN ARTICLE:
Manage to get a copy of your article to Fido #375,
preferably by Fidonet mail, or by uploading. If manual
upload, please put it in the MAILFILE area, so that I can
find them all at once.
ARTICLE FORMAT: VERY IMPORTANT! PLEASE READ!
The requirements are a little tighter in this
department, due to purely practical constraints. I cannot
devote hours to converting every text format in the world to
the one I use.
1. NO LEFT MARGINS! Flush left please. These are
extremely difficult to remove.
2. RIGHT MARGIN AT COLUMN 60 OR LESS! Less is OK,
more is definitely not. This includes fancy
boxes, dotted lines, etc.
3. KEEP FUNNY CONTROL CHARACTERS TO A MINIMUM!
This includes formfeeds and other oddities.
These areent as dificult to handle, but on long
files still takes a while.
4. WORDSTAR IS FINE IF THERE ARE NO LEFT MARGINS!
5. PLAIN ASCII (EDLIN, ED, etc) IS FINE IF LINE LENGTH
IS UNDER 60 COLUMNS!
fidonews -- 18 Mar 85 18:52:42 Page 3
6. NO GRAPHIC CHARACTERS! Believe it or not, not every
one in the world has an IBM PC. My computer
understands printable characters from 20 hex to
7e hex. (Space to tilde) This is ASCII; "American
Standard Code for Information Interchange". We are
"Interchanging Information". Everything else is
GARBAGE. ASCII is universal; Graphics are not.
7. TOTAL ARTICLE LENGTH: Up to you; note, however, that
I will probably avoid publishing dictionaries,
bibles translated into NAPLPS, and ASCII encoded
LANDSAT pictures of Russian wheat farms.
8. WHERE ON EARTH IS THIS ARTICLE FROM? Well ... good
question! A good idea to identify yourself
somewhere, unless you wish to remain anonymous.
Thats OK too, but I may balk at publishing rude
or otherwise racy submissions.
9. AN EXAMPLE:
Article from Tom Jennings
Fido #375
Here is what an article could look like for
submission. Very simple: plain old dumb ASCII, WordStar,
PMATE, Vedit, or whatever. WordStar Document mode is fine.
Note no left column; that is done when FidoNews is
published. Note the total width is only 60 columns; it looks
narrow on your screen, but fine when formatted later.
END irrational tirade on text files ... on to bigger
and better things!
--------------------------------
Thom Henderson
Fido node 375
System Enhancement Associates
Change of Command
Back when I was in the military, they made a big deal out of
"Change of Command". That is, when new leaders took over
from the old. Well, new leaders are taking over from the
old here at Fidonews, because we at node 375 are going to be
publishing Fidonews in the future, but as far as I am
concerned, there is no big deal. We plan on following Tom
Jenning's policies, and the only difference you need worry
about is to send articles to node 375 instead of node 1.
Needless to say, I do hope that you will continue to send
articles to Fidonews, since the strength of this publication
is its user support. This is a newsletter by, of, and for
the users. This is YOUR newsletter. It is no more and no
less than what YOU make it.
fidonews -- 18 Mar 85 18:52:48 Page 4
The primary policy I intend to pursue is to make my
editorials short and to the point, and this one is no
exception.
------------------------
As you all probably know by now, Fidonet is growing
faster than any one of us ever imagined. What was once a
simple group of people having fun has become a very large
and complex group with very large and complex needs.
The file FIDOHIST.TXT available on most boards
covers all the details; if you're not familiar with it
please read it. I'll assume you have read it for the rest of
this article.
There are now 200 nodes in FidoNet. FidoNet requires
much more background work than is apparent; it's one of
those cases that you don't see it until it collapses. The
software requires continuous support, and if Ken Kaplan et
al quit doing the node list, I guarentee FidoNet would
collapse into total disarray within 6 weeks. I cannot
continue to be the only source of FidoNet software, and Ken
Kaplan cannot be the only person managing the network.
This article isn't to make anyone feel guilty, no!
Fidonet is supposed to be fun, and useful second. If it
wasn't for those 200 sysops, well, what would the point be?
This article is about growing pains.
This may come as bad news, but Fidonet is now large
enough that it needs some sort of formal type organization.
This is not a joke, it's not something for the future, it is
needed NOW. Something on the order of the ARRL (American
Radio Relay League, ie. ameteur radio) I would guess sounds
good. Just barely formal enough to keep everyone in touch,
but not stiflingly so. Probably many sysops will drop out
instantly; it is probably unavoidable. However, if we do
not, everyone will be forced out, since there will no longer
be a FidoNet.
The advantages will be the much smoother operation,
and much better information available to new sysops
especially. The newsletter is a start (see the other article
somewhere about newsletter changes) but not enough. It's
readership isn't wide enough.
FidoNet expertise is now being scattered across the
net, as opposed to being centered in one spot, previously
me. Ken Kaplan (St. Louis) on node list organizational
areas, John Warren (Riverside CA), Kurt Reisler (VA), John
Anderson (OH), and many many more for utilities, Allen
Miller (OH) for Multilink/multiline Fido, Ben Baker (St.
Louis) on network routing, are just a small sampling of
people with Fidonet expertise, that I can recall off the tip
fidonews -- 18 Mar 85 18:52:56 Page 5
of my tongue. There is no way that anyone could know of this
expertise floating around, except by accident.
Friday 15 Mar, on Fido #1, I answered 400 messages;
150 Fidonet, 150 Comments-to-the-sysop, and the rest
sprinkled around other areas. This was ten days worth of
messages, and is a typical load. Almost all of these fall
into a few categories:
1. Questions on Fido operation
2. Bug complaints, ie. outright failures
3. Annoyances, (checking for mail, etc)
4. Suggestions for changes
5. Misc. comments, questions, etc
Most, but certainly not all, questions are from new
sysops. The same questions are repeated over and over, as is
to be expected. The Fido manuals stink. Also, many things
just don't make sense, and are totally arbitrary.
Bug complaints are almost without exception all real
and serious enough to report. Annoyances are usually "it
works fine, but this one thing drives me crazy". I repair
bugs as I can, and fix annoyances where possible.
Suggestions for changes, well, how I handle these
has changed. I used to respond to all of them, then some of
them, now almost none of them. I just dont have time. If
there were more developers, this would be a whole different
story.
Before I continue, let me state the reasons that the
Fido program source is not public: Fido was written using
many proprieatry sources and libraries which I routinely use
in my work; not all are my property, and even those that are
are used in many commercial programs, Ptel included. I
cannot under any circumstances give out Fido sources.
Back to the software: this issue depends on whether
you see Fido/FidoNet as a program, or as a network. I see it
as a network first, the Fido software second. (Many may not
agree ...)
To get down to it, we need a second FidoNet
compatible program soon. If for no other reason than "What
if TJ drops dead?" To repeat, I cannot give out source, do
not ask. Fido now consumes 20 hrs a week, now being dropped
to under 10 by the expedient method of not answering all my
mail. (I hear grumbling ...)
If there were source available for a FidoNet
ocmpatible program, no matter how limited at first, it would
(1) take pressure of me (selfish ...) (2) provide a start
for people who need customized systems and (3) allow hackers
to hack, from whence all good things happen.
Fido is an evolutionary dead end; it works of
fidonews -- 18 Mar 85 18:53:03 Page 6
course, but has many serious drawbacks. It is huge,
cumbersome, and consumes disk space like a congressional
subcommitee consumes money. It is like that ancient reptile
Apteryx, it flies, but only with great effort. Time for
someone to create an eagle with what we've learned.
WHAT THIS ORGANIZATION NEEDS TO DO:
Basically, nothing that's not being done now. These
are pretty straightforward tasks.
DISEMINATE SOFTWARE:
A place where latest versions are kept, and possibly
an alternate method of distribution. (Diskettes, etc).
Probably a system almost devoted to this.
HELPING NEW SYSOPS:
Providing them with the information needed to start
a system. People who can answer messages about problems,
questions, etc. Maybe someone monitoring for the 100 most
frequently asked questions.
SOFTWARE MAINTENAINCE:
Maybe the same node that passes out software, a
group of people who repair the software, implement changes,
listen to ideas and generally keep it working.
NETWORK TOOLS:
Things like NODELIST, ROUTEGEN and CONVERT save
hours or days of work. Many people have never heard of them.
NETWORK MAINTENAINCE:
This covers adding new nodes to the node list,
checking the phone numbers for accuracy, verifying that
nodes really exist, and referring new sysops to a useful
place.
REGIONAL NETWORKS:
FidoNet is splitting naturally into local nets, as
in Boston, LA, St. Louis, etc. This is good, and should be
encouraged further, even to the point of almost totally
independent groups to help new sysops, etc. Basically, it
works this way now informally, so it would not be much of a
change.
RESEARCH:
There are a lot of things out there that just arent
addressed, such as interfacing to the AMSAT *amateur*,
*privately owned and operated* packet radio satellite, new
network architectures, low cost nodes, high volume nodes,
fidonews -- 18 Mar 85 18:53:08 Page 7
third world information transfer, you name it, there is a
need for it.
I repeat, this is not a harebrained fantasy, nor a
wish list. It must become reality, and quickly. I am not
going to set up this organization; I'm too busy fighting
fires in the short term. Ken Kaplan is too busy as well.
This is a grassroots, low cost, hobbiest network,
the first and only one of it's kind. It is truly the
forefront of technology; you can ride along for free and see
it collapse by Fall, or you can start things going, even
better than it is.
---------------------------------------------------------
From: Jon Tara on FIDO #92 Subject: Usenet, uucp, etc.
In reply to Tom Kenny's comments in the last
newsletter:
I am still planning a FIDOnet-Usenet gateway. I am
still waiting for the release of a commercial uucp package
for MSDOS. (for info, send uucp mail to ihnp4!vortex!lauren
or on ARPA to vortex!lauren@RAND-UNIX.ARPA) The package
will probably sell for about $350. This is the same package
that will be (aparently) supplied with Coherent and possibly
other UNIX-clones in the future. It's the only commercial
non-ATT uucp implementation I know of.
First, some basic information on uucp and Usenet:
1) uucp generally runs on minicomputers that run UNIX.
I assume that's what was meant by Tom's "mini" comment -
not that it is less than "full scale".
2) usenet does indeed have routing. The difference is
that the user has to explicitly specify the route to be
used, in most cases. (Some sites do indeed have "smart
routers.") FIDO does some things that uucp does not, as well
as vice- versa. I wouldn't say that either one is "better"
than the other.
3) uucp is a proprietary program written by ATT as part
of the UNIX system. It is NOT public domain, and indeed ATT
has been VERY touchy about it. Anyone selling a commercial
version of uucp had better be able to prove that they didn't
use uucp source code, or face prosecution. The program and
protocol(s) are sketchily documented in a few Bell Labs
memos and journal articles. There is apparently no source
of complete documentation. I've heard that the source code
itself is impossible to follow.
4) uucp is a PROGRAM. It stands for unix-to-unix copy.
Think of it as the Xmodem of the Unix world, with some
additional capabilities. The original intent of uucp was to
fidonews -- 18 Mar 85 18:53:15 Page 8
allow ATT to easily update UNIX software at their remote
sites.
5) Usenet is a NETWORK of sites that use the uucp
program. It is of an ad-hoc organization. There are a few
"backbone" sites - mainly within ATT and at large
universities. The largest site is ihnp4 at ATT in Indian
Hill, Illinois. ihnp4 talks directly to at least 200 sites.
How much traffic each node is willing to pass on, the nature
of the traffic, etc. is determined by whoever owns that
system. Many of the big sites have dedicated lines to other
sites, so it costs them nothing to pass on messages - it
only becomes a problem if their capacity is exceeded.
Others have to make phone calls, which could range anywhere
to internal corporate lines (again, "no cost") to normal ATT
dial-up.
6) Superimposed on all this is Usenet news. The
terminology is still unclear to me, but you often see
references to "the uucp net" and the "usenet net". The uucp
net is simply the collection of sites using uucp which are
able to pass mail. The Usenet net is a subset of these that
"get the news." I think. Like I said, I'm unclear.
7) "The news" consists of something like 100 seperate
"news groups". Most of the sites receiving the news receive
ALL of it. (Necessary to be able to pass it on.) The news
amounts to something like 1MB A DAY. Much of it is quite
useful - for example there is a newsgroup for distribution
of public-domain source code UNIX utilities. A lot of it is
garbage consisting of random flames and endless requests for
the same thing. ("Does anyone know about a PLUGH for the
XYZZY machines?)
While I've been waiting for uucp, and because not too
many FIDO sysops are likely to want to shell out $350 for
uucp, I've been thinking about other ways of doing it. I'm
now leaning toward writing a program to be called xxcp.
xxcp would be put in the public domain, and would be
installed on both your FIDO system and a "friendly" usenet
site. xxcp would send mail back and forth using the Xmodem
protocol, instead of the proprietary (and, for all practical
purposes, undocumented) protocol used by uucp. You would
need a "very" friendly usenet site, since this would require
them to install the xxcp program, make entries in their cron
table, etc. On the other hand, you can offer THEM the
ability to communicate with any MSDOS machines they may
have. There is some precedent for this: there are sites on
the net that are using Kermit for a link protocol, instead
of uucp. I choose Xmodem instead, since it's more popular
with micro users, and I have source for a pair of programs
called SB and RB for UNIX which implement the Modem7 "send
batch" and "receive batch" commands. It looks easier to work
with than to try to cut only the necessary portions out of
the Kermit source.
There may be some political problems associated with
fidonews -- 18 Mar 85 18:53:24 Page 9
FIDOs going onto the net. There has been some resistance to
"hobbyists" participating in the net. People are worried
about floods of junk mail and obscene, etc. messages.
Personally, I think that BBS users are more likely to be
well-behaved than the shipping clerk who has access to a
UNIX system and just learned that he can send mail to "the
rest of the world"... The MSDOS users of today are
tomorrow's UNIX users. The net is going to have to face
explosive growth eventually - better it should learn to cope
with a few BBSs now, than to suddenly face thousands who
just discovered the "uucp" command in their shiny new Xenix
package later.
---------------------------------------------------------
From: Jon Tara on FIDO #92 Subject: oops...
An article that was published in FIDOnet newsletter
#203 on making the Anchor Signalman MKXII work with FIDO was
incorrect. I forgot to observe Heath's law:
"Any kit should be tested by a complete idiot before
printing the manual. Even so, there will still be at least
one mistake that slips through. Furthermore, at least one
idiot will get it right with the WRONG instructions."
(Must explain the messages of thanks I got...)
Anyway, thanks to one person who proved ME an idiot for
not trying to build this from my own instructions.
The INCORRECT step follows:
8) Solder the MIDDLE wire of the transistor to the
cut trace, on the other side of the cut.
This should be corrected to read:
8) Solder the LEFT-hand wire of the transistor to
the cut trace, on the other side of the cut.
(in step 9, a resistor gets soldered to "the remaining
wire" of the transistor. This is the MIDDLE wire)
Sorry for any inconvenience and/or hair-pulling this
has caused. For anyone who missed the original article, and
wants to use an Anchor with FIDO, I'll be glad to FIDOmail
copies of the correct procedure.
----------------------------------------------------------------
Recently, here in Cherry Hill, NJ, a BBS system by
the name "Hackers Heaven" was closed down by police. The
sysop(s) and users of this system were trading pirated
programs, unauthorized MCI access codes, passwords for
several computers in the area, etc. Unlike what happended
on the west coast where a system was used to post a single
PACIFIC BELL access code without the sysop's knowledge, this
fidonews -- 18 Mar 85 18:53:30 Page 10
sysop was well aware of the situation, and as far as I know,
even supported it. (I assume this by rumor, and by what the
name suggests). My viewpoint differs from the situation out
on the west coast; in this case, it appears to be the sysops
own damned fault! How can someone be as nieve as that! I'm
glad the sysop out west was released without charges, but
these kids should go to jail! It's shit like this that gives
us sysops a bad name! The system was being run on an Apple
IIe and with only a simple questionaire, users were granted
access to use. Rumor also says that the FBI was also
involved with some fake numbers of some sort. Charges seem
to be based however on theft of access codes from Dial
America; a local long-distance company.
-Brian Sietz- Sysop FIDO#82
The following is a reprint from an article by David
Lee Preston appearing in the March 12 issue of the
Philadelphia Inquirer. Editorial notes [] by myself appear
within.
About $35,000 worth of computer equipment has been
confiscated from two homes in Cherry Hill and one in
Voorhees Township as part of an investigation into an
alleged ring of youthful computer hackers, Camden County
Prosecutor Samuel Asbell said yesterday. Investigators with
search warrants seized computers, modems, printers, disk
drives, software and other equipment, which four residents
of the houses allegedly had used to gain access to Dial
America, a two-year old long distance telephone service
based in Camden, Asbell said. Although no arrests were made
during the Saturday sweep, Asbell said his office planned to
charge about 20 South Jersey youths with using personal
computers to make long-distance calls through Dial America.
Most of the members of the alleged ring live in Camden
County, all are males, and the oldest is 20, he said.
Hackers are personal-computer enthusiasts, often in their
teens, who use their technical skills to gain illegal entry
into private or corporate computer systems.
[Ed: Bullshit! That is only the derrogitory term for
"Hackers". The real meaning is more subtle and not so
clearly defined. Hackers are generally computer
enthusiasts, most often not associated with anything
harmful. Using this definition, I myself am a "hacker"]
Gade Kreckel, president of Dial America, said Dial
America began an investigation in November after a customer
reported that there were several calls on his bill he had
not made. The firm, in the Wilson Building on Broadway, used
its computers to identify calls being made with stolen
account numbers and transferred those calls into fake
accounts, he said. "Before the calls ever got onto our
customers' bills, we transferred them into a fraud
account," he said. "In other words, customers get a clean
bill." The firm turned the case over the prosecutor's office
about three weeks ago, Kreckel said. He said the fraudulent
fidonews -- 18 Mar 85 18:53:39 Page 11
calls made during the investigation cost his firm as much as
$6,000. Asbell said the alleged hackers would telephone
other hackers on computers around the country and trade
information such as computer access codes. "We have taped
conversations between the computer hackers," Asbell said.
"We have calls throughout probably most of the United
States... Once you have the access code to Dial America, you
can go trade it to somebody in the state of Oregon for the
access code to VISA or Mastercard." Dial America says it has
4,000 customers in Burlington, Gloucester and Camden
Counties. "Unless you are from South Jersey, you couldn't
use our service," Kreckel said. "They would get on there and
make long-distance calls. For example, one says, 'Do you
have Pac-Man? OK, transmit it to me.' And in return, he
might give him Donkey Kong, then pirate video games and send
them back and forth." The alleged hackers also were trading
AT&T credit-card numbers, Kreckel said. The first seizure
took place about 7:30AM at the home of a Cherry Hill
juvenile, Asbell said. Later that morning, investigators
seized more equipment from a juvenile in Voorhees, he said.
About 11PM, they executed a third search warrant at the
Cherry Hill home of two brothers, ages 18 and 20, he said.
At each house, Asbell said, the parents reacted with "actual
shock and dismay as to what was taking place." He said the
parents were "very cooperative in the first two instances"
but that the Cherry HIll brothers' father initially
resisted the investigators' efforts. "They don't realize
that stealing from a Dial America, or a Sprint or MCI, is no
different from walking into a 7-Eleven and stealing five or
six thousand dollars," Kreckel said. The ability of long-
distance companies to trace fraudulent usage has improved
measurably in the last year, Kreckel said. "When the long
distance-industry was in a different stage a year ago,
computer hackers found it pretty easy to get into systems
without being detected," he said. "But nowadays, it's such
a stupid crime because it's so easy to catch them." "The sad
part about it is it's almost 95 percent minors that are
involved in this thing. And that's the case with almost all
the computer hackers."
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
!! S P E C I A L E D I T I O N !!
or
The Complete FIDONET Nodebook
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
--------------------------------
|Submitted by: |
|Christopher von Schilling |
|Sysop FIDO NODE 388 |
|Keene, NH - 300/1200 - 24 hrs. |
|(603) 357-2090 |
--------------------------------
fidonews -- 18 Mar 85 18:53:47 Page 12
Here I was, "mailing" another cry for help to T.J.
somewhere in California (where it was warm and not
snowing). I had fallen into the habit of mailing
"urgent help messages" to Tom a few weeks back. Anyone
ever use a NOVATION SMART CAT Plus modem? Well,
neither has he. I had accessed his system frequently
in the past and browsed through it. After all, how
expensive could it really be? I found out. I also
found out that everyone with problems ranging from "I'd
like to use the scheduler to ring my alarm clock at
8:00 AM. Any suggestions?" to "What's an Orphan
message?" would leave these kind of questions on FIDO
1. So, I fell into the habit of bouncing my questions
off FIDO 1 too.
All went well until, one morning, I found a reply on my
board to one of my "urgent" problems from Fido 1, which
in effect told me "I don't know". HA! He must be
kidding, right? This is a joke?! If he doesn't know,
who does? And that's it! An incredible thought
occured to me at that moment. Never mind that I hadn't
really been awake for more than an hour yet, I
immediately hit the keyboard. My message was the
longest I've ever "mailed". Did you know that a
message form can only hold 46 lines? Trivia. But I
squeezed it all in there. Every last bit. I listed
the message one more time, to make sure there weren`t
any mistakes in it, and re-read my work of ingenuity.
I was asking Tom Jennings what he thought of my idea of
a SPECIAL EDITION NEWSLETTER which would appear
whenever we had enough material to publish an issue.
This newsletter would contain detailed listings of all
the FIDO NODES currently on-line. What type of
hardware they were using, what their sysops were most
interested and knowledgable in and what user base their
FIDO targeted. Also, telephone numbers, hours of
operation, baud rate and locations. Like the nodelist
but completely different.
With that kind of detailed information at your
fingertips, you could try to have your technical
questions answerd by another NODE nearest you which
uses the same hardware and whose sysop would be willing
to help you. This would stimulate communications
between sysops and users who have access to the special
edition(s). There'd be a new edition every time
there's enough information to compile for a new
publication. Sorted by Hardware? Who knows. This is
what I suggested. Ho Ho Ho. I could use that kind of
information. I'm sure many other sysops would like it
too. Would he find someone to do it?
Exactly one week later I received my reply from Tom
Jennings himself: "That's a very good idea. I'm too
busy but many people have requested just that kind of
information. We need this type of listing. Go ahead
and compile it!". I went out and had lunch. Something
fidonews -- 18 Mar 85 18:53:55 Page 13
terrible had happend. My message was too long and got
garbled in transit. Me? Compile THAT? How many NODES
times how many lines of information? That would equal
a lot of work. Wouldn't it?
I hereby volunteer for just this job. (It looks rather
good to read this on WordStar. Will it make FidoNews?
We'll see). If all NODES cooperate and send me the
information below in not TOO detailed form to FIDO 388
(see top of article) I think I can pull it off. (What
am I talking about?)
If you would like to participate and be listed in this
kind of format, please send this information. You must
be willing to answer other sysops' questions regarding
any problems. If you are not interested in
corresponding with other sysops or users, please don't
participate.
a) THE NAME OF YOUR BOARD
b) TYPE OF COMPUTER YOU USE
c) TYPE OF MODEM YOU USE
d) WHICH VERSION OF FIDO
e) YOUR (SYSOPS') NAME
f) YOUR SPECIAL INTERESTS AND KNOWLEDGE
g) IS THERE A SPECIAL INTEREST GROUP YOUR BOARD SERVES?
h) WHERE CAN YOU (SYSOP) OFFER HELP
i) TELEPHONE NUMBER
j) PARAMETERS (BAUD RATE ETC.)
k) LOCATION
l) TIME OF OPERATION
m) MISCELLANEOUS THINGS OF INTEREST TO OTHERS
Please keep the information you'll send limited to this
general form unless you have something of importance to
add. I hope to distribute the compiled list in the
same fashion as FIDONEWS.
I wonder if this is something of interest and I hope
many NODES will participate. I believe that this type
of publication will tie all of us closer together and
de-centralize the exchange of information by allowing
us to communicate directly without bogging down FIDO 1.
Finally, if you have any suggestions or problems
regarding the above, please mail your correspondence to
FIDO 388 as listed at the beginning of this article.
I'm looking forward to receiving lots of listings from
all sysops out there!
Thank you very much,
--Christopher von Schilling
Sysop Fido Node 388
*=*=*=*=*=*==*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
fidonews -- 18 Mar 85 18:54:01 Page 14
NOTICES
FOR SALE
-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-
M E M O R Y C H I P S A L E !
-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-
I have the following Dynamic RAM available at the lowest
price.
256k 120ns FUJITSU 256Kx1.........@ $25.00/chip
256k 150ns NEC 256Kx1.........@ $20.00/ "
256k 150ns HITACHI/TOSHIBA 256Kx1.........@ $20.00/ "
128k 150ns OKI 128Kx1.........@ $25.00/ "
64k 120ns NEC 64Kx1.........@ $ 8.50/ "
64k 150ns NEC 64Kx1.........@ $ 8.00/ "
64k 150ns FUJITSU 64Kx1.........@ $ 5.00/ "
64k 200ns NEC 64Kx1.........@ $ 6.00/ "
64k 200ns MATSUSHITA 64Kx1.........@ $ 5.50/ "
Large Scale Integration devices:
8087-3 6MHZ INTEL Math Co-processor for IBM PC $250.00
8087-2 8MHZ INTEL Math Co-processor for IBM PC $475.00
These co-processors are very difficult to obtain. Which
means delivery time is INDEFINITE. A $100 non-refundable
deposit required to order co-processors.
Add $3.00 for UPS Blue Label Service
Add $6.00 for Federal Express
Digital Equipment Corportation sells the 256k upgrade
option (PC1XX-AZ) for $695. You can get the SAME set
here for $180.
Send orders (Only money orders accepted, checks will
be returned.) to:
Advanced Software Applications
5258 Vickie Drive
San Diego, California 92109
Please indicate on your order: BRAND, SIZE, QUANTITY &
SHIPPING PREFERRED and of course your name and address.
Further inquiries may be made via Fidonet to: Node #350.
Addressed to: ASA (PRIVATE)
Now a freebie! A dissertation on the differences between
150ns and 200ns Dynamic RAM.
fidonews -- 18 Mar 85 18:54:06 Page 15
You might be asking yourself if its ok to put 150ns 64k
or 256k Dynamic RAMs into an IBM PC. Its not only OK,
its preferred!
Also, there is NO problem with mixing 150ns chips with
200ns chips.
ICs are made from extremely pure silicon, first refined in
the form of a large ingot, usually shaped like a large 4"
diameter sausage. A thin slice of ultra-pure silicon wafer
is first coated with a photographically sensitive coating.
The photosensitised wafer is exposed to
light through a photo-reduced negative, which created the
pattern of one layer of many transistors, with line widths
typically of 2 to 3 microns. A micron is 1/1,000,000 of
an inch. Next the photographic coating is developed like
a negative. Where the light strikes or doesn't strike, the
coating is removed. In effect, little openings are made in
the coating, exposing the top surface of the ultra-pure
silicon wafer in only a few, small locations. The wafer is
then placed into a diffusion furnace where 'impurity' atoms
of Phosphorous or Boron are permitted to diffuse into the
surface of the silicon to a precisely calculated depth with
calibrated oven temperature and time controls. The process
of applying photographic sensitive coatings, exposure to
light, and exposure to 'impurity' diffusion gases is
repeated six to a dozen times. The final step will result in
the application of minature deposits of aluminum on the top
of the wafer, connecting as many as thousands of
transistors into an Integrated Circuit (IC).
At this point the wafer is cut into individual IC chips,
called dies. On a 4" wafer you can expect around 144 chips
of Dynamic RAMs.
Now the dies are tested to see which ones turn out to be
150ns, 200ns, 250ns etc. and which don't work.
A reason for variations of speed across a wafer is the ex-
treme difficulty of optical alignment with the different
photographic exposures required. Remember, line width tol-
erances are only 2 to 3 microns. Also, quality varies be-
cause of the random nature of the impurity doping
diffusions, and the random nature of other impurities.
Even contaminants within the refined silicon will cause a
variation in the quality.
Now you can see that all Dynamic RAMs of the same brand are
'identical twins', designed, fabricated, and processed the
same way.
Access time is only one parameter measured. Although it is
a measurement of the overall quality of the chip. All the
ICs are meant to test out to 150ns or better. So, when you
buy higher speed memory ICs, you are nearly always buying
higher quality devices.
fidonews -- 18 Mar 85 18:54:14 Page 16
An IBM PC memory board can tolerate Dynamic RAMs as slow as
250ns, however, when you use chips slower than 200ns, you
are running with borderline defective IC's. Most likely
problems will insue.
I strongly recommend you use 150ns chips. They are better
quality! Your computer will not run any faster if its not
designed to operate on 150ns parts, but, you will most
likely have fewer problems overall. Your computer will not
run faster with faster memory because the processor
dictates the execution speed. If your processor is designed
to wait at least 200ns for something, faster memory won't
affect the wait time.
Dynamic RAMs are NMOS devices slow down as they get hot.
The IBM PC has a very bad air flow design, indicated
by the extremely warm temperatures found inside the IBM PC.
Therefore with 150ns devices, you have more room for timing
syncronization when the PC gets warm.
Have no fear of mixing 150ns chips with 200ns chips as
long as your computer circuit was designed to handle slower
devices.
Comments on brands of ICs:
The ranking of 64k Dynamic RAM by quaility in Japan is now:
1. NEC
2. FUJITSU or HITACHI
3. MITSUBISHI
4. TOSHIBA
5. MATSUSHITA or OKI
Originally it was thought that HITACHI was the number one
quality product because it was the brand IBM initially
purchased and installed into their PCs. Now it is the
author's understanding that IBM has contracted to buy the
256K RAMs from both NEC and HITACHI in such large
quantities that NEC & HITACHI are modifying their prod-
uction lines to go from 64k to 256k to satisfy this new
requirement.
If you believe there are inaccuracies in this report, feel
free to direct them to Fido #350, user: ASA. There is no
intent to mislead anyone.
Author: Rick Eliopoulos, Sysop #350
-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-
Download original FidoNews · Volume 2 (1985) · ← Previous · Next →