29 Jan
2009
29 Jan
'09
11:08 a.m.
On Jan 29, 2009, at 10:54 AM, Tim McNerney wrote:
At 01:00 PM 1/28/2009, Tim McNerney wrote:
No matter how many times I hear stories like this, or lived
through stories like this, I still shake my head and can't
understand how the source code gets lost. All these
corporations, all the lawsuits, all the programmers, all the
marketing money, and so often no one, NO ONE,
preserves arguably the most important bits.
Oh, I completely understand how it happens (after having seen it
happen a few times). Typically all of the source for a project is
kept on a server(s) that run the SCM (source code management) system.
Yes, they are backed up but in many cases it's just a rotating backup
set (ie they get overwritten). When the project is EOL'd usually the
server(s) are either re-tasked (involving a complete re-install and
wipe) or scrapped (which involves wiping). The backups are usually
scrapped at that time as well. Almost no one does an archive of those
backups. If they do, then that's were to look because they are kept
in an off-site location for disaster recovery. In many cases the off-
site storage is mostly write-only (ie tapes/whatever go in but are
never asked for and not purged).
Most of this stuff gets "lost" because of apathy on the part of the
owners rather than malicious intent.
TTFN - Guy
I am
looking for volunteers to help reverse-engineer and
document Apple II VisiCalc.
I have three versions of the Apple II software. I
know one of
them still boots (1983?)
and have some confidence that the other two
versions (1979 and
1981) work too.
I have been in contact with both Bob Frankston and
Dan Bricklin.
Of course you ask, well then why do you need volunteers?
The answer is because no one can find the sources.
In the case of Personal Software, I believe Lotus bought them to
reduce competition,
not to preserve VisiCalc. At the time, people seldom think about
the historical significance
of source documents. The Intel 4004 schematics would have been lost
if it hadn't been for
some Intel engineer who realized they needed to be saved. True,
there are some engineers
who take stuff home and store it in their basement, but in the case
of the 4004, Federico
Faggin couldn't hold on to any schematics when he left Intel because
he was going off to
found Zilog and had to protect himself legally. The problem with
packrats is that, even
when you are unusually organized, like Bob Frankston, you still end
up with boxes that
are mis-labeled. Even Hewlett Packard has thrown out some important
engineering
documentation. When the calculator division moved from Santa Clara
to Corvalis,
the threw out dumpsters full of drawings on their very first RPN
calculator, the 9100.
--Tim