At 05:05 PM 2/18/2015, Ethan Dicks wrote:
>I snark about Pascal all the time. I encountered it in a professional
>capacity in 1987. [...]
>Utterly trivial in languages that trust the programmer to handle
>unformatted I/O.
All true, but doesn't it make you wonder why Turbo Pascal
was such a popular development environment for the PC for so long?
Was it the sheer will power and marketing of Borland, or was
it the volume of developers who didn't need intensive low-level I/O?
By '85, there were several C compilers for the PC and I even
remember using Gimpel C-terp, a C interpreter that made development
and debugging easier.
- John
They seem to have broken it sufficiently now that nothing is returned after the end of October.
Is there anyone indexing Usenet that has a clue? It seems like all that is left is for-pay
services for searching alt.binaries.
I'm seeking input from the various people that have in recent times
made PCBs to plug into old microcomputers and minicomputers using PCB
edge card connectors. There are quite a few companies (and coops)
that offer very inexpensive PCB fab service, even in low volumes,
particular for double-sided PCBs (vs. multilayer). However, I haven't
found a PCB fab house that offers any reasonable deal on low volume
runs of PCBs with plating of hard-gold over nickel edge connectors.
Has anyone found a vendor that does?
I was somewhat surprised at the difficulty in tracking down good
information on best practices for PCB edge fingers. Several of the PCB
companies suggested just using an ENIG (electroless nickel immersion
gold) on the whole PCB, including the edge card connector, but that
isn't really suitable. According to an excerpt of a draft of the
IPC-4522 standard, ENIG gold is typically only 0.025 to 0.05 microns,
and is not suitable for edge finger use for more than five
insert/remove cycles. (It would be perfectly acceptable to have ENIG,
HASL, or tin finish on the rest of the board.)
A document from AMP recommends 0.4, 0.8, or 1.3 microns of hard gold
over nickel, to withstand 200, 1000, and 2000 cycles before failure,
respectively.
The IPC-2221 standard suggests three categories:
Class 1: 0.8 micron gold over 2.0 micron nickel, for general
electronic products
Class 2: 0.8 micron gold over 2.5 micron nickel, for dedicated
service electronic products
Class 3: 1.3 micron gold over 2.5 micron nickel, for high-reliability
electronic products.
The PCIe specs require 0.7 micron gold over 1.2 micron nickel.
I'm perfectly willing to build prototypes for my own personal use
using ENIG or even HASL with nothing special for the edge fingers, and
be resigned to them not withstanding many cycles. However, if I were
to offer a product for sale to other hobbyists, I don't think that
would be acceptable.
I suspect that the original PCI specs had a similar requirement to
PCIe. I estimate that a 32-bit PCI edge card requires about 1300 mm^2
of gold over nickel (counting both sides of the PCB), which works out
to a little less than 1 mm^3 overall, which at recent gold pricing
would be about USD $0.80 in gold. An Apple II card would probably have
a roughly similar amount of gold, while S-100, Qbus, Omnibus,
Multibus, etc. cards would require several times that. I'd be
delighted if in small volume I could pay less than an additional $25
each to get the hard gold over nickel edge connector, but so far I
haven't found a PCB fab comes even close to that.
Eric
In case anyone is wondering, I still have a P112 kit for sale. $197
shipped in the US.
--
David Griffith
dave at 661.org
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
I love this loop construct. I was just talking to a friend the other day
about how nice the GO language is, but I lamented that no one has ever
copied the elegance of the single loop verb in COBOL.
On 2/18/2015 8:50 AM, Chuck Guzis wrote:
> Is there any other language (save PL/I for obvious reasons) that has
> an equivalent to the COBOL PERFORM...THRU statement?
>
> That one always struck me as being a bit offbeat. Example: Consider a
> section of a program with 4 consecutive paragraphs, named PARA-1
> through PARA-4. One can say in the same program:
>
> PERFORM PARA-1 23 TIMES.
> PERFORM PARA-1 THRU PARA-4.
> PERFORM PARA-1 THRU PARA-2.
> PERFORM PARA-2 THRU PARA-3 WITH TEST AFTER VARYING INDEX-1 FROM 10 BY
> 3 UNTIL FLAG-1 > 0.
>
> It's that variable scope of a PERFORM that I don't recall seeing in
> any other language.
>
> --CHuck
>
>
>
>
so, what is easier to read and write?
IF STRING1 = STRING2
or
if !(strncmp(string1,string2,10) {
I've rewritten cobol into C a few times when I was a big enthusiast of C
and the C code was longer and harder to maintain.
On 2/17/2015 1:29 PM, John Foust wrote:
> At 01:02 PM 2/17/2015, Ethan Dicks wrote:
>
>> What I read horrified me. How could anything be that mind-crushingly
>> verbose?
>>
> As compared to purposefully terse C one-liners? :-)
>
> - John
>
>
>
>
I reccived this message from Jose - they want to rent 1980s computers for a movie.?They are located near Montreal, I think.
Please respond below:
----------------------------------------------------------------------------------
I'm working on a feature film for Fox Entertainment, in Canada and we are looking for computers dating back to 1983.
Ii was wondering if renting some of them would interest you.
Thank you.
Best regards.
--
Jose Varela
jfvarela at me.com
514 756 5699
Fox Quebec Productions
1777, Carrie Derick, 3e etage, suite 311
Montreal, Quebec H3C 6G2
?514 613 5852?
I set up a poll to select a moderator to replace the inactive moderation
on the Catweasel group. I know the owner / creator is probably on here,
and I am doing this per Yahoo instructions to get rid of the spam.
I'm posting here briefly as I'm sure a lot have filtered out the group
due to the spam.
Here is the poll link:
https://groups.yahoo.com/neo/groups/catweasel/polls/poll/3206349?pollType=E…
The information about this is in a couple of messages on the Yahoo
groups site. This is partly an effort to salvage the Catweasel group
there, and partly an effort to see if Yahoo customer service follows
thru. So far they are communicating with me. I'll let you all know how
it goes.
I'll end the poll for a new moderator around the end of February, per
their suggestion.
Thanks
Jim