Re: RIP Jerry Pournelle - Firsts

[Lars Brinkhoff via cctalk]

Charles Dickman via cctalk  writes:
>> He seems to have been the first to mention ARPANET in a popular
>> hobbyist-type context like BYTE. (Leading him to get kicked off
>> ARPANET!)
> Yes I remember reading something like that too. I would like to know
> the story of that.

You should probably ask Leigh Klotz.

http://klotz.me/

Re: Convex C220 lives

[Charles Anthony via cctalk]

On Mon, Sep 11, 2017 at 5:04 PM, Laurens Vets via cctalk <
cctalk@classiccmp.org> wrote:

>
>>
> There's something seriously wrong with your site
> http://www.vaxbarn.com/index.php/other-bits/603-convex-c220 unless the
> title of your work is indeed "Feel Like Having Sex Tonight | Best Legal
> Viagra Uk" :)
>
>
The URL works for me.

$ host www.vaxbarn.com
www.vaxbarn.com has address 46.30.215.19

-- Charles

Re: markup vs. word processing [ was RE: RIP Jerry Pournelle, the first author to write a novel on a computer Earl...]

[Ethan Dicks via cctalk]

On Mon, Sep 11, 2017 at 3:05 PM, Rich Alderson via cctalk
 wrote:
> I didn't think of it as word processing, but I did something similar at Ohio
> State in 1974...

Hi, Rich,

Did we ever discuss when you were at OSU?  I know I've had
conversations with Herb Johnson about when he worked at OSU.  I ask
(in case we have not) because I lived on 12th Ave from 1970 through
1978, originally because my dad was a photographer for the Vet School,
then we just stayed when he got a job as a police photographer (taking
pictures of football riots 2 blocks from our house) until the Mormon
Church bought several houses in a row, including ours (we were the
last to sell), and built a small college on our block.

I attended OSU from 1984-1989, but it was practically our "park" at
the end of the block, so I wandered the halls long before then, hung
out in the Union, saw Star Wars at the theater next to the McDonalds,
etc., etc.,

-ethan

markup vs. word processing [ was RE: RIP Jerry Pournelle, the first author to write a novel on a computer Earl...]

[Rich Alderson via cctalk]

I met Jerry once, at a BYTE-sponsored conference held at McCormick Place
in Chicago.  Colorful.  I already knew his writing, of course.  He will be
missed.

Now:

From: Guy Sotomayor Jr
Sent: Sunday, September 10, 2017 4:59 PM

> I think it all depends upon how you define “word processing”.  For me I
> absolutely detest things like MS Word.  Probably because I started with
> markup languages.

> The first one was one that I wrote for the IBM 1130 so I could do a high
> school research paper (1974).  It was written in Fortran (sorry long gone)
> and the “paper” was all on punch cards and printed on a 1403 printer.  I did
> it mainly because it was a pain to keep track of how to format for footnotes
> and attributions.

I didn't think of it as word processing, but I did something similar at Ohio
State in 1974.  I worked on the consulting team (= "help desk") for the College
of Administrative Science, hired because I (a) knew COBOL from my high school
adventures and (b) knew Coursewriter III from my time at the University of
Texas Computer-Assisted Instruction laboratory.  I was supposed to write an
on-line Quick (Introduction to) COBOL.

At UT, I had seen text input for the Coursewriter II system (IBM 1500) done on
punched cards, and probably for the 360/50 running Coursewriter III but I
didn't have access to that computer room.  I thought that that was a really
good way to do it, and convinced my boss to let me create a program for the
same purpose using PL/I (my favorite language at the time).  A grad student
wrote the COBOL course, and 2 other undergraduates put his text onto coding
sheets in Coursewriter III markup for our 2 professional keypunch operators to
enter.

At first I had the ladies (and they were) entering multipunch TAB and similar
characters, but then I got an idea from the single Hazeltine 1500 in our
terminal room:  A CONTROL key!  I used 2 shift characters, a cent sign for
CONTROL and a dollar sign for SHIFT, with the pair in opposite orders for
characters themselves in course text.

Overall, it saved about 6 months over the previous regime of having
undergraduates typing course material in on 2741 terminals.  I was rewarded
with a big raise, from $3.35/hr to $3.50/hr.  Not bad for a college kid in 1974.

As it happens, I had tried using Coursewriter III for word processing (though I
did not know that's what I was doing) in the fall of 1973, entering the text of
"The Goblin Market" from the pages of (IIRC) the September 1973 issue of
_Playboy_.  I was not satisfied with the result because the fixed spacing on
the 2741 did not work well to mimic the fancy printing job in the magazine.

> At CMU I used Scribe that output to the XGP (Xerox Graphics Printer driven by
> a PDP-11/45).  This was the first time I used something where there were
> selectable fonts (1976).  At IBM *everything* was done with various versions
> of SCRIPT.  At this point I can’t recall but I believe a number of the IBM
> manuals were all done in SCRIPT.

I didn't encounter SCRIPT until I was in graduate school at UChicago, working
at the Computation Center.  My first programming job there (after I was
promoted from the help desk) was to update the program from using Wylbur-format
text internally to using SuperWYLBUR when the Comp Center went to the more
capable commercial product.  I used SCRIPT to create my Advanced Dungeons &
Dragons materials for a campaign I was running.

I believe that all the IBM manuals which use the 1403 TN train were done using
SCRIPT to produce camera-ready copy, but that's from hearsay.

> I then used Interleaf (a *high* end document publishing/management system)
> and then FrameMaker (before Adobe completely screwed it up and finally killed
> it).

Used FrameMaker at Cisco.  Loved the output.  Hated the proprietary file format,
which meant that nothing I did could be reused on another computer.

> I currently use LaTex for producing anything more complicated than an email.

Although I encountered TeX at UChicago (one of my office mates came from
Stanford and knew DEK), I didn't learn and use it until I got to XKL, where the
Toad-1 manual was done using LaTeX with local mods (thus TeX was necessary).  I
did the work on updating the Tops-20 JSYS manuals there, although they were
never released while I was working for Len.

(Oh.  I just remembered.  I *did* used infotex once at Stanford, for some
 manual I wanted in hardcopy, but I can't remember what for.  Different from
 LaTex just to be different, I think.)

I still prefer markup languages to word processing systems, but I'm kind of
stuck with Word at work, for obvious reasons.  While I was still a contractor,
back in 2003, I wrote all my reports in LaTeX and created documents with
pdflatex, which I consider a gift from the gods.

Rich

Re: Convex C220 lives

[Laurens Vets via cctalk]

On 2017-09-11 08:47, Camiel Vanderhoeven via cctech wrote:

For a change, rather than a request for help, here¹s a success story: I
managed to bring a Convex C220 (dual vector CPU mini supercomputer from
1988) back to life. Both CPUs are working, but I¹m running with a 
single

CPU because of the power it draws with two CPUs. Next challenges: the
Convex C1, and quad vector processor C240 (not before I¹ve upgraded the
power feed).

Running ConvexOS 11.5.1, it has FORTRAN 7.0.1 installed; I ran a little
benchmark, and with a single CPU the system clocks in at 49.1 MFLOPS on 
a
big multiply-add loop (advertised peak performance was 50 MFLOPS per 
CPU).


Getting the system to the state where it is now was quite a journey
(though nowhere near as bad as it might have been). If you¹re 
interested
in the details, I have a (somewhat long) report of my work on my 
website;
if you go to 
http://www.vaxbarn.com/index.php/other-bits/603-convex-c220,
there are some links at the bottom that have much more details, as well 
as

photos of the system and the boards.

Now I¹m looking for some FORTRAN code that would typically have run on
this kind of computer so I can show people what this kind of system was
used for.


There's something seriously wrong with your site 
http://www.vaxbarn.com/index.php/other-bits/603-convex-c220 unless the 
title of your work is indeed "Feel Like Having Sex Tonight | Best Legal 
Viagra Uk" :)

Re: Convex C220 lives

[Holm Tiffe via cctalk]

Camiel Vanderhoeven via cctech wrote:

> For a change, rather than a request for help, here¹s a success story: I
> managed to bring a Convex C220 (dual vector CPU mini supercomputer from
> 1988) back to life. Both CPUs are working, but I¹m running with a single
> CPU because of the power it draws with two CPUs. Next challenges: the
> Convex C1, and quad vector processor C240 (not before I¹ve upgraded the
> power feed).
> 
> Running ConvexOS 11.5.1, it has FORTRAN 7.0.1 installed; I ran a little
> benchmark, and with a single CPU the system clocks in at 49.1 MFLOPS on a
> big multiply-add loop (advertised peak performance was 50 MFLOPS per CPU).
> 
> Getting the system to the state where it is now was quite a journey
> (though nowhere near as bad as it might have been). If you¹re interested
> in the details, I have a (somewhat long) report of my work on my website;
> if you go to http://www.vaxbarn.com/index.php/other-bits/603-convex-c220,
> there are some links at the bottom that have much more details, as well as
> photos of the system and the boards.
> 
> Now I¹m looking for some FORTRAN code that would typically have run on
> this kind of computer so I can show people what this kind of system was
> used for.
> 
> Camiel
> 

Congrats Camiel!

We had an C34 and an C220 in the Computing Centre of out university here
in Freiberg (Saxony). I don't had much todo with it, but I've used The Metrum 
Robot with two RSP2150 later for Backups at a small ISP.

I have still some quarter Inch Tapes labeled with golden Convex Labes
here laying around..but the contents are'nt original anymore.

Regards,

Holm

-- 
  Technik Service u. Handel Tiffe, www.tsht.de, Holm Tiffe, 
 Freiberger Straße 42, 09600 Oberschöna, USt-Id: DE253710583
i...@tsht.de Fax +49 3731 74200 Tel +49 3731 74222 Mobil: 0172 8790 741

Re: Convex C220 lives

[Ethan via cctalk]

managed to bring a Convex C220 (dual vector CPU mini supercomputer from
1988) back to life. Both CPUs are working, but I¹m running with a single
CPU because of the power it draws with two CPUs. Next challenges: the
Convex C1, and quad vector processor C240 (not before I¹ve upgraded the
power feed).


This is amazing! Kudos and congrats, those systems are so rare. I've never 
heard anyone mention in the hobbyist community ever. Never thought I would 
hear of one running!



--
: Ethan O'Toole

Re: Subject: Re: RIP Jerry Pournelle, the first author to write a novel on a microcomputer

[Chuck Guzis via cctalk]

On 09/11/2017 07:28 PM, Murray McCullough via cctalk wrote:
> Was Jerry Pournelle the FIRST to write a PUBLISHED NOVEL on a
> MICROCOMPUTER? Yikes! Talk about SHOUTING. As a historian is it worth
> the aggravation to please everyone? Does revisionism take away the
> honour(Cdn. Here!) Jerry Pournelle rightly deserves for pioneering
> work in our beloved computing genre? Are the passing of our early
> microcomputing pioneers
> leaving a hole in our history?

Murray, I thought I'd mentioned that the first published novel known to
have been prepared on a word processor (an IBM MT/ST) was  Len
Deighton's Bomber, published in 1970.

So, likely the prep was 1969.  So claims Wikipedia.

--Chuck

Re: ETAOIN SHRDLU (Was: RIP Jerry Pournelle, the first author to write a

[Chuck Guzis via cctalk]

On 09/11/2017 08:35 PM, Fred Cisin via cctalk wrote:
> Etaoin Shrdlu was most importantly one of Walt Kelly's characters in Pogo.
> 
> ETAOIN SHRDLU   ETA name
> On Mon, 11 Sep 2017, Chris Elmquist via cctalk wrote:
>> I was told that the name came from this string which could be found in
>> printed works and that people had always seen it but just read past it
>> because it didn't fit or make sense.
> 
> That is canonically "loren ipsum", the filler "greeking" text used to
> set up layout.
> Although admittedly ETAOIN SHRDLU was also sometimes used for greeking.

It also inadvertently crept into printed copy.  If the Linotype operator
made an error in a line, it was common to just fill the line out with
ETAOIN SHRDLU so that the copy editor would know to discard the line.

--Chuck

Re: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Chuck Guzis via cctalk]

On 09/11/2017 07:22 PM, Chris Elmquist wrote:
> Lincoln had ETAOIN on his personalized MN license plate (on the very well 
> worn Ford full size van he drove) and another guy had SHRDLU on his plates.
> 
> I was told that the name came from this string which could be found in 
> printed works and that people had always seen it but just read past it 
> because it didn't fit or make sense.  I believe it was Neil's youngest son 
> Brian who offered up this name but I'm not remembering why anymore.
> 
> There was also a Lincoln story that ETA was like The Eta, a Spanish terrorist 
> group and all of us at ETA were going to terrorize the supercomputer industry.

I'd say that the second was apocryphal.

It was around 1983 when Neil dropped by to talk to us that he told the
ETAOIN SHRDLU source of the ETA name and his son's part in the naming.

Hey, that was almost 35 years ago.  I have trouble remembering where I
put my glasses (sometimes they're on my head...)

--Chuck

Re: ETAOIN SHRDLU (Was: RIP Jerry Pournelle, the first author to write a

[Ethan Dicks via cctalk]

On Mon, Sep 11, 2017 at 11:35 PM, Fred Cisin via cctalk
 wrote:
> (It is sometimes stated that Sholes deliberately chose QWERTY to slow down
> typists to the speeds that the early typewriters could handle; that story
> has been challenged, but a replacement story of how that layout was chosen
> has not been advanced)

Two things I have heard that contributed to the QWERTY layout was
_not_ that it "slowed down typists", but that with common english
words, it produced a cadence that alternated left and right hands to
minimize jamming of the hammers which led to _faster_ typing than
earlier arrangements - the typists formerly had to slow down either to
prevent hammers from entangling or to take time to unjam them and get
back into the flow.

The other facet of the arrangement was a sales gimmick to be able to
type TYPEWRITER on the top row in customer demonstrations.

-ethan

ETAOIN SHRDLU (Was: RIP Jerry Pournelle, the first author to write a

[Fred Cisin via cctalk]

Etaoin Shrdlu was most importantly one of Walt Kelly's characters in 
Pogo.


ETAOIN SHRDLU   ETA name
On Mon, 11 Sep 2017, Chris Elmquist via cctalk wrote:
I was told that the name came from this string which could be found in 
printed works and that people had always seen it but just read past it 
because it didn't fit or make sense.


That is canonically "loren ipsum", the filler "greeking" text used to set 
up layout.

Although admittedly ETAOIN SHRDLU was also sometimes used for greeking.


ETAOIN SHRDLU
is one of the common versions of the list of the relative frequency of 
letter in English language text.

'e' is the most common letter
't' is the next most common letter
'a' is the third most common letter, etc.
The first dozen letters are generally agreed to have pretty much that 
relative frequency, with some disagreements, largely  based on the samples 
used for the statistical data.
The further down the list that you go, the more likely that you will find 
different sequences, particularly in different genres.
If you count letters in Shakespeare, Gideon bible, New York Times, etc., 
by the time that you get to the end of those dozen letters, you may start 
to see some discrepancies.


In very low level cryptography, any sufficiently long message coded with a 
"Caesar cypher" can be easily broken just by assigning 'e' to the most 
common letter, 't' to the next most common, and paying attention to the 
limited number of one and two letter words.
There are a few other fine points, such as a different sequence for what 
are the most common letters to be the first letter of a word (look at the 
thickness of different letters in the dictionary, or to be the last letter 
of a word.   (Compaq and news of the last few decades have dramatically 
increased the incidence of 'Q' as last letter)


Notice also, that in Morse code and other variable length encodings, the 
most common letters are generally assigned the shortest sequence of 
symbols.



"Etaoin shrdlu definition: the letters produced by running the finger down 
the first two vertical rows of keys at the left of the keyboard of a 
Linotype machine: used . . . "
Do you suppose that the designers of the machine knew about or cared 
about the frequency of use of letterss?
(It is sometimes stated that Sholes deliberately chose QWERTY to slow down 
typists to the speeds that the early typewriters could handle; that story 
has been challenged, but a replacement story of how that layout was chosen 
has not been advanced)


The Linotype layout may be why that was sometimes more convenient to input 
in a hurry than "Ipso Lorem"



"Correlation V causation":
Have those letters BECOME the most commonly used in English due to the 
influence of typesetting? :-)

Re: RIP Jerry Pournelle - Firsts

[Charles Dickman via cctalk]

On Mon, Sep 11, 2017 at 1:15 PM, Shoppa, Tim via cctalk
 wrote:
> He seems to have been the first to mention ARPANET in a popular hobbyist-type 
> context like BYTE. (Leading him to get kicked off ARPANET!)

Yes I remember reading something like that too. I would like to know
the story of that.

Much more interesting to read about than that nobody was every really
first at anything etc, etc, etc.

-chuck

Re: Subject: Re: RIP Jerry Pournelle, the first author to write a novel on a microcomputer

[Lyndon Nerenberg via cctalk]

> On Sep 11, 2017, at 7:38 PM, Fred Cisin via cctalk  
> wrote:
> 
>> Was Jerry Pournelle the FIRST to write a PUBLISHED NOVEL on a
>> MICROCOMPUTER? Yikes! Talk about SHOUTING.
> 
> It is EMPHASIS of individual words and phrases.
> THIS IS SHOUTING!  FIGURE OUT THE DIFFERENCE!

Zippy the Pinhead HAS ESCAPED! I'm wearing a ZOOT SUIT with a REET PLEAT!!!

--emacs

Re: Subject: Re: RIP Jerry Pournelle, the first author to write a novel on a microcomputer

[Fred Cisin via cctalk]

On Mon, 11 Sep 2017, Murray McCullough via cctalk wrote:

Was Jerry Pournelle the FIRST to write a PUBLISHED NOVEL on a
MICROCOMPUTER? Yikes! Talk about SHOUTING.


It is EMPHASIS of individual words and phrases.
THIS IS SHOUTING!  FIGURE OUT THE DIFFERENCE!

As a historian is it worth the aggravation to please everyone? Does 
revisionism take away the honour(Cdn. Here!) Jerry Pournelle rightly 
deserves for pioneering work in our beloved computing genre? Are the 
passing of our early microcomputing pioneers leaving a hole in our 
history?


Do you consider it to be "revisionism" to get the historical record 
correct, rather than continue popular myths?


As a historian, do you think that it is correct behavior to continue to 
declare Columbus to be the "first" "discoverer" of America?
As a historian, do you think that it is correct behavior to let it 
slide when somebody declares that Steve Jobs invented computers?
As a historian, do you think that it is correct behavior to let it 
slide when somebody declares that Bill Gates invented software?


Jerry did good work.  Let's not credit him with the wrong things.


--
Grumpy Ol' Fred ci...@xenosoft.com

Subject: Re: RIP Jerry Pournelle, the first author to write a novel on a microcomputer

[Murray McCullough via cctalk]

Was Jerry Pournelle the FIRST to write a PUBLISHED NOVEL on a
MICROCOMPUTER? Yikes! Talk about SHOUTING. As a historian is it worth
the aggravation to please everyone? Does revisionism take away the
honour(Cdn. Here!) Jerry Pournelle rightly deserves for pioneering
work in our beloved computing genre? Are the passing of our early
microcomputing pioneers
leaving a hole in our history?

Happy computing all!

Murray  :)

Re: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Chris Elmquist via cctalk]

Lincoln had ETAOIN on his personalized MN license plate (on the very well worn 
Ford full size van he drove) and another guy had SHRDLU on his plates.

I was told that the name came from this string which could be found in printed 
works and that people had always seen it but just read past it because it 
didn't fit or make sense.  I believe it was Neil's youngest son Brian who 
offered up this name but I'm not remembering why anymore.

There was also a Lincoln story that ETA was like The Eta, a Spanish terrorist 
group and all of us at ETA were going to terrorize the supercomputer industry.

Chris

On September 10, 2017 9:55:02 PM CDT, Chuck Guzis via cctalk 
 wrote:
>On 09/10/2017 06:25 PM, Tapley, Mark via cctalk wrote:
>
>> There was one of those machines in my Junior High School shop
>> classroom. I saw it run once (not well enough to successfully set a
>> line of type, but nearly).
>> 
>> I endorse Mark’s assessment of its safety characteristics...
>
>I knew a fellow who had one of them in his barn--and he set the local
>freebie weekly newspaper with it.   Open gas flame, hot type metal
>that's mostly lead, lots of open whirling parts--what's not to like?
>
>Running one is definitely a real skill.  ETAOIN SHRDLU CMFWYP...
>
>Neil Lincoln once told me that the name of ETA Systems back in the 80s
>was suggested by his son.  Neil knew about the Linotype order, but it
>was unclear to me if his son got it from a literary work (there were
>several) or from the actual machine.  Chris Elmquist might know.
>
>FWIW, the "assembler" in a Linotype machine is where the type matrices
>drop down in a row, ready for "kerning".  Another non-computer use of
>the word.
>
>--Chuck

-- 
Chris Elmquist

Re: Tips For Soldering a Surface Mount PLCC Socket

[Jon Elson via cctalk]

On 09/11/2017 01:52 PM, shad via cctalk wrote:

leaded solder = AgPb alloy

Uh, no, that would be Silver-Lead.  Regular old solder was 
SnPb, that is Tin-Lead.


Jon

Re: Tips For Soldering a Surface Mount PLCC Socket

[Jim Brain via cctalk]

On 9/11/2017 1:38 PM, Geoffrey Oltmans via cctalk wrote:

The way a lady at my work did them was to crack out the bottom piece of the
socket to give better access to the pins. She didn't typically put the
bottom piece back in since we used the PLCC socket to plug into an emulator
pod so there was no need for it (shoulder around the POD connector wouldn't
allow it to fit too deeply into the socket. I suppose if you just dropped
the bottom of the socket back in place before plugging in the chip that
would probably be sufficient.
I just did a plcc68 smt socket in the same way, and it worked fine. I 
put the bottom piece back in, and it looks fine.

Re: Tips For Soldering a Surface Mount PLCC Socket

[shadoooo via cctalk]

leaded solder = AgPb alloy

Re: Tips For Soldering a Surface Mount PLCC Socket

[Paul Koning via cctalk]

> On Sep 11, 2017, at 2:23 PM, shad via cctalk  
> wrote:
> 
> Hello,
> in my experience SMD components doesn't need a presoldering operation at
> all, because pins are already stained or golded, thus very solderable
> 
> Note: use lower temperatures (<380C, less is better), good quality flux and
> AgPb are required for comfortable operation with low risk of damage.

I would suggest one change: avoid lead-free solder, use real solder (63/37 
solder is best).  It works just fine with modern politically correct parts, and 
it's much easier to use.  I followed this advice from a professional PCB tech a 
couple of years ago when working on a fine pitch home project -- she was 
absolutely right.

paul

Re: Tips For Soldering a Surface Mount PLCC Socket

[Geoffrey Oltmans via cctalk]

The way a lady at my work did them was to crack out the bottom piece of the
socket to give better access to the pins. She didn't typically put the
bottom piece back in since we used the PLCC socket to plug into an emulator
pod so there was no need for it (shoulder around the POD connector wouldn't
allow it to fit too deeply into the socket. I suppose if you just dropped
the bottom of the socket back in place before plugging in the chip that
would probably be sufficient.



On Mon, Sep 11, 2017 at 1:23 PM, shad via cctalk 
wrote:

> Hello,
> in my experience SMD components doesn't need a presoldering operation at
> all, because pins are already stained or golded, thus very solderable.
> Plus the extra heat could let the pin move inside the plastic, or add
> somehow excessive thickness when positioned on the PCB.
> In my experience the best option is to clean well the PCB with a
> desoldering wick, then add some flux and touch the pads with a thin
> soldering iron and some AgPb wire, to dilute the no-Pb alloy on the pads,
> which is harder to melt than Pb alloy.
> Thanks to flux, surface tension will be low and Agpb alloy will form a thin
> layer with no oxide.
> Now position the new component, it should seat well in position, without
> excessive height over the PCB.
> Now with thin iron, heat two opposite pins letting the alloy to melt, and
> the component will be held in place.
> Now proceed in order with all pins.
>
> Note: use lower temperatures (<380C, less is better), good quality flux and
> AgPb are required for comfortable operation with low risk of damage.
>
> Andrea
>

Re: Tips For Soldering a Surface Mount PLCC Socket

[shadoooo via cctalk]

Hello,
in my experience SMD components doesn't need a presoldering operation at
all, because pins are already stained or golded, thus very solderable.
Plus the extra heat could let the pin move inside the plastic, or add
somehow excessive thickness when positioned on the PCB.
In my experience the best option is to clean well the PCB with a
desoldering wick, then add some flux and touch the pads with a thin
soldering iron and some AgPb wire, to dilute the no-Pb alloy on the pads,
which is harder to melt than Pb alloy.
Thanks to flux, surface tension will be low and Agpb alloy will form a thin
layer with no oxide.
Now position the new component, it should seat well in position, without
excessive height over the PCB.
Now with thin iron, heat two opposite pins letting the alloy to melt, and
the component will be held in place.
Now proceed in order with all pins.

Note: use lower temperatures (<380C, less is better), good quality flux and
AgPb are required for comfortable operation with low risk of damage.

Andrea

Re: Convex C220 lives

[Paul Koning via cctalk]

> On Sep 11, 2017, at 2:15 PM, Chuck Guzis via cctalk  
> wrote:
> 
> On 09/11/2017 08:47 AM, Camiel Vanderhoeven via cctalk wrote:
> 
>> Now I¹m looking for some FORTRAN code that would typically have run
>> on this kind of computer so I can show people what this kind of
>> system was used for.
> 
> 
> Congratulations!
> 
> I'd recommend starting with LINPACK:
> 
> http://www.netlib.org/linpack/
> 
> Just to get an idea of the performance of the thing.

My favorite substantial FORTRAN program is NEC2 -- an antenna modeling program. 
 Public domain, I believe (since it's a US Government work).

paul

Re: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Chuck Guzis via cctalk]

On 09/11/2017 11:09 AM, Paul Koning wrote:

> No, not kerning.  Justifying.  Linotype machines can't kern; when that's done 
> in metal typesetting, which is rare, it involves cutting bevels onto the 
> sides of the type blocks to allow them to partly overlap.

You're correct; I misstated the process.

--Chuck

Re: Convex C220 lives

[Chuck Guzis via cctalk]

On 09/11/2017 08:47 AM, Camiel Vanderhoeven via cctalk wrote:

> Now I¹m looking for some FORTRAN code that would typically have run
> on this kind of computer so I can show people what this kind of
> system was used for.


Congratulations!

I'd recommend starting with LINPACK:

http://www.netlib.org/linpack/

Just to get an idea of the performance of the thing.

--Chuck

Re: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Paul Koning via cctalk]

> On Sep 10, 2017, at 10:55 PM, Chuck Guzis via cctalk  
> wrote:
> 
> On 09/10/2017 06:25 PM, Tapley, Mark via cctalk wrote:
> 
>> There was one of those machines in my Junior High School shop
>> classroom. I saw it run once (not well enough to successfully set a
>> line of type, but nearly).
>> 
>> I endorse Mark’s assessment of its safety characteristics...
> 
> I knew a fellow who had one of them in his barn--and he set the local
> freebie weekly newspaper with it.   Open gas flame, hot type metal
> that's mostly lead, lots of open whirling parts--what's not to like?

They did offer a safer heater (electric) as an alternative.

> Running one is definitely a real skill.  ETAOIN SHRDLU CMFWYP...

There are some nice training videos (Italian, mostly showing a Linotype clone 
(Intertype?)) on the web.

> Neil Lincoln once told me that the name of ETA Systems back in the 80s
> was suggested by his son.  Neil knew about the Linotype order, but it
> was unclear to me if his son got it from a literary work (there were
> several) or from the actual machine.  Chris Elmquist might know.
> 
> FWIW, the "assembler" in a Linotype machine is where the type matrices
> drop down in a row, ready for "kerning".  Another non-computer use of
> the word.

No, not kerning.  Justifying.  Linotype machines can't kern; when that's done 
in metal typesetting, which is rare, it involves cutting bevels onto the sides 
of the type blocks to allow them to partly overlap.

paul

Re: Convex C220 lives

[Ian S. King via cctalk]

On Mon, Sep 11, 2017 at 8:47 AM, Camiel Vanderhoeven via cctech <
cct...@classiccmp.org> wrote:

> For a change, rather than a request for help, here¹s a success story: I
> managed to bring a Convex C220 (dual vector CPU mini supercomputer from
> 1988) back to life. Both CPUs are working, but I¹m running with a single
> CPU because of the power it draws with two CPUs. Next challenges: the
> Convex C1, and quad vector processor C240 (not before I¹ve upgraded the
> power feed).
>
> Running ConvexOS 11.5.1, it has FORTRAN 7.0.1 installed; I ran a little
> benchmark, and with a single CPU the system clocks in at 49.1 MFLOPS on a
> big multiply-add loop (advertised peak performance was 50 MFLOPS per CPU).
>
> Getting the system to the state where it is now was quite a journey
> (though nowhere near as bad as it might have been). If you¹re interested
> in the details, I have a (somewhat long) report of my work on my website;
> if you go to http://www.vaxbarn.com/index.php/other-bits/603-convex-c220,
> there are some links at the bottom that have much more details, as well as
> photos of the system and the boards.
>
> Now I¹m looking for some FORTRAN code that would typically have run on
> this kind of computer so I can show people what this kind of system was
> used for.
>
>
Congratulations!  I've been too busy to do any restorations, so for now I
live vicariously through people like you.  :-)   -- Ian

-- 
Ian S. King, MSIS, MSCS, Ph.D. Candidate
The Information School 
Dissertation: "Why the Conversation Mattered: Constructing a Sociotechnical
Narrative Through a Design Lens

Principal Investigator, "Reflections on Early Computing and Social Change",
UW IRB #42619

Archivist, Voices From the Rwanda Tribunal 
Value Sensitive Design Research Lab 

University of Washington

There is an old Vulcan saying: "Only Nixon could go to China."

The Airport VIC

[Jim Brain via cctalk]

https://www.computerworld.com/article/3223131/it-industry/nice-work-if-you-can-get-it.html


--
Jim Brain
br...@jbrain.com
www.jbrain.com

RIP Jerry Pournelle - Firsts

[Shoppa, Tim via cctalk]

He seems to have been the first to mention ARPANET in a popular hobbyist-type 
context like BYTE. (Leading him to get kicked off ARPANET!)

Tim

RE: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Rob Jarratt via cctalk]

> -Original Message-
> From: cctalk [mailto:cctalk-boun...@classiccmp.org] On Behalf Of Dave Wade
> via cctalk
> Sent: 11 September 2017 10:28
> To: 'Chuck Guzis' ; 'General Discussion: On-Topic and
Off-
> Topic Posts' 
> Subject: RE: RIP Jerry Pournelle, the first author to write a novel on a
computer
> Early adopt
> 
> Alan Turing used the Teletypes on the Manchester MK1 to write letters. I
doubt
> he had any software, just typed, perhaps using paper tape to edit.
> 


Talking of Alan Turing, seen these?

http://www.bbc.co.uk/news/technology-41082391
http://www.manchester.ac.uk/discover/news/lost-turing-letters-give-unique-in
sight-into-his-academic-life-prior-to-death

I think they found this when looking for stuff for me for MU5. I got to see
them still in the filing cabinet a few months ago when I went there
researching MU5.

Regards

Rob

Convex C220 lives

[Camiel Vanderhoeven via cctalk]

For a change, rather than a request for help, here¹s a success story: I
managed to bring a Convex C220 (dual vector CPU mini supercomputer from
1988) back to life. Both CPUs are working, but I¹m running with a single
CPU because of the power it draws with two CPUs. Next challenges: the
Convex C1, and quad vector processor C240 (not before I¹ve upgraded the
power feed).

Running ConvexOS 11.5.1, it has FORTRAN 7.0.1 installed; I ran a little
benchmark, and with a single CPU the system clocks in at 49.1 MFLOPS on a
big multiply-add loop (advertised peak performance was 50 MFLOPS per CPU).

Getting the system to the state where it is now was quite a journey
(though nowhere near as bad as it might have been). If you¹re interested
in the details, I have a (somewhat long) report of my work on my website;
if you go to http://www.vaxbarn.com/index.php/other-bits/603-convex-c220,
there are some links at the bottom that have much more details, as well as
photos of the system and the boards.

Now I¹m looking for some FORTRAN code that would typically have run on
this kind of computer so I can show people what this kind of system was
used for.

Camiel

RE: RIP Jerry Pournelle, the first author to write a novel on a computer Early adopt

[Dave Wade via cctalk]

Alan Turing used the Teletypes on the Manchester MK1 to write letters. I
doubt he had any software, just typed, perhaps using paper tape to edit. 

Dave

> -Original Message-
> From: cctalk [mailto:cctalk-boun...@classiccmp.org] On Behalf Of Chuck
> Guzis via cctalk
> Sent: 11 September 2017 03:55
> To: Tapley, Mark via cctalk 
> Subject: Re: RIP Jerry Pournelle, the first author to write a novel on a
> computer Early adopt
> 
> On 09/10/2017 06:25 PM, Tapley, Mark via cctalk wrote:
> 
> > There was one of those machines in my Junior High School shop
> > classroom. I saw it run once (not well enough to successfully set a
> > line of type, but nearly).
> >
> > I endorse Mark's assessment of its safety characteristics...
> 
> I knew a fellow who had one of them in his barn--and he set the local
> freebie weekly newspaper with it.   Open gas flame, hot type metal
> that's mostly lead, lots of open whirling parts--what's not to like?
> 
> Running one is definitely a real skill.  ETAOIN SHRDLU CMFWYP...
> 
> Neil Lincoln once told me that the name of ETA Systems back in the 80s was
> suggested by his son.  Neil knew about the Linotype order, but it was
unclear
> to me if his son got it from a literary work (there were
> several) or from the actual machine.  Chris Elmquist might know.
> 
> FWIW, the "assembler" in a Linotype machine is where the type matrices
> drop down in a row, ready for "kerning".  Another non-computer use of the
> word.
> 
> --Chuck
> 
>