Survival rates in computers with the eye on archiving.

[CheetahHugger]

Complicated one, let's say i have a need to keep certain programs/data
in running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.
I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i
think aging on stored components could be a problem as well sudden shoch
when becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones
from a recycler and strip them down, pack and vacuum seal the parts and
stock them?
I was thinking of aquiring a few older generation workstations as i
assume those ones were still built to proper industrial specs.
Any suggestions?

Cheetah


"have you counted your spots today?"

 

[[email protected]]

Complicated one, let's say i have a need to keep certain programs/data
in running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.
I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i
think aging on stored components could be a problem as well sudden shoch
when becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones
from a recycler and strip them down, pack and vacuum seal the parts and
stock them?
I was thinking of aquiring a few older generation workstations as i
assume those ones were still built to proper industrial specs.
Any suggestions?

Cheetah

"have you counted your spots today?"

PIII machines, heck, I keep checkin' on my ol' 333MHz Celeron and it
runs fine. I'm guessing a Workstation would be the way to go since
they are tested to last. But are you serious about this?

 

[James Sweet]

Complicated one, let's say i have a need to keep certain programs/data in
running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.
I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i think
aging on stored components could be a problem as well sudden shoch when
becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones from
a recycler and strip them down, pack and vacuum seal the parts and stock
them?
I was thinking of aquiring a few older generation workstations as i assume
those ones were still built to proper industrial specs.
Any suggestions?

Cheetah

For a number of years, bad capacitors were cropping up all over the place
and killing motherboards right and left. Stuff made before that and stuff
made after that should be fine, aside from faulty capacitors, fans, and
mechanical storage devices, I've never had a computer component fail, they
just become obsolete.

If you really need to keep something going forever, I'd advise you to just
go out and buy a pile of identical inexpensive modern parts, you should
average at *least* 5 years out of a motherboard, but 10+ would not surprise
me in the least.

 

[CheetahHugger]

But are you serious about this?


I'm afraid that the current situation may force me to do such thing, yes.
The monitors do worry me as well, although they could be more repairable.

Cheetah

 

[James Sweet]

I'm afraid that the current situation may force me to do such thing, yes.
The monitors do worry me as well, although they could be more repairable.

Cheetah

What would prevent you from replacing the monitors with something else?

 

[William R. Walsh]

Hi!

Complicated one, let's say i have a need to keep certain programs/data
in running order for the next 50 years.

Theoretically speaking, it's doable. What history has shown us so far is
that the reality may be a little different. Semiconductors ("chips") are
themselves very reliable and can last an extremely long time in a well
designed circuit running within specifications or close to. The problematic
areas include mechanical parts (disk storage), power supplies and some
supporting components like capacitors.

That said, I have a lot of 286 and 386 machines around that are still in
fairly regular use.

You also have to factor in exteral influences...lightning strikes, natural
disasters, etc. These things can wipe out otherwise healthy and properly
functioning equipment almost instantly. And don't forget the upkeep (human)
factor as well. If everything else goes well, the systems will still need
periodic restarting, cleaning and other work.

Any suggestions?

Truthfully, I think you'd do very well to consider building a system that
could be upgraded and moved to newer equipment as time goes on and things
wear out. Sure, you could stockpile many parts and systems in order to keep
the same old thing running along, but leaving yourself an option to upgrade
to more modern hardware over time is a much better plan. You can reasonably
assume that new computer hardware will be available. You can't assume that
any spare parts will be in serviceable condition by the time you need them.

William

 

[N Cook]

Complicated one, let's say i have a need to keep certain programs/data
in running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.
I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i
think aging on stored components could be a problem as well sudden shoch
when becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones
from a recycler and strip them down, pack and vacuum seal the parts and
stock them?
I was thinking of aquiring a few older generation workstations as i
assume those ones were still built to proper industrial specs.
Any suggestions?

Cheetah


"have you counted your spots today?"

No one has considered the software in this thread.
The ferrite material , whether on a hard disc platten, floppy or tape
streamer tape can debond from the backing.
With optical media the code carying metal surface can oxidise/ part company
from the plastic media,
Unless anyone knows of the ideal storage media that is incorruptible, is
this the best procedure ?
Store the original media plus copies on perhaps more futureproof media in an
evacuated container. Then another repeat of that collection at another site
, in case of fire or flood at one site.
Is long term optical storage more reliable as far as it cannot spontaneously
change magnetic state by coalescence of neighbouring magnetic poles or
whatever the corrupting process is.

 

[N Cook]

No one has considered the software in this thread.
The ferrite material , whether on a hard disc platten, floppy or tape
streamer tape can debond from the backing.
With optical media the code carying metal surface can oxidise/ part company
from the plastic media,
Unless anyone knows of the ideal storage media that is incorruptible, is
this the best procedure ?
Store the original media plus copies on perhaps more futureproof media in an
evacuated container. Then another repeat of that collection at another site
, in case of fire or flood at one site.
Is long term optical storage more reliable as far as it cannot spontaneously
change magnetic state by coalescence of neighbouring magnetic poles or
whatever the corrupting process is.

A recent newspaper article on this matter
http://www.guardian.co.uk/technology/2007/oct/11/guardianweeklytechnologysec
tion.news
part quote
"I always end up doing really interesting work with the data conversion," he
says. He spent almost five years working with treasure hunters on his own
salvage mission, trying to access photos of tens of thousands of Spanish
coins retrieved from a wreck off the coast of Florida.

"There were 65,000 of these doubloons that were pulled up, and the
archaeologists took digital photos of the front and back of each individual
coin because they were all unique," he says. "It was very early digital
technology back in the mid-80s, and they had all these tapes and discs that
they couldn't access. All this was before jpegs and gifs and today's
standard formats."

On another job Ismail had to go back in time to resurrect the dead from a
huge graveyard in California. The cemetery records stretched back to the
1950s and 60s on 50,000 punch cards, and staff no longer knew who was buried
where and what plots were already taken.

"None of the data was printed on the top of these cards, so you could not
have had a human doing the job," Ismail says. Instead, he hooked up his
technological time machine and came up with a system that could read up to
200 cards a minute.

Resurrecting classics

Ismail's talents stretch to his own computer workshop, where he has built
replicas of some of the earliest machines from scratch. "The one that's in
demand is Digital Equipment Corporation's PDP 1 from 1961," he says. "It
stands for program data processor and it cost $120,000 (£60,000) new, but it
wasn't called a computer. Budgets were scrutinised and if it had said
'computer' on the request it would've been turned down."

The PDP 1 featured the first ever video game, Spacewar!, before there was
such a thing as a gaming industry. Ismail has made a replica for a Japanese
display on the history of computers and gaming, and he is working on another
one for a touring exhibition of videogames.
end quote

Presumably Sellam Ismail
http://www.sellam.com/
would be worth contacting as he is going to know more than most , how to
archive/futureproof hardware and software

 

[Jumpster Jiver]

Complicated one, let's say i have a need to keep certain programs/data
in running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.
I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i
think aging on stored components could be a problem as well sudden shoch
when becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones
from a recycler and strip them down, pack and vacuum seal the parts and
stock them?
I was thinking of aquiring a few older generation workstations as i
assume those ones were still built to proper industrial specs.
Any suggestions?

Cheetah


"have you counted your spots today?"

Please don't think I am against modern technology in any way but I think
by trusting any modern form of storage alone, we're asking for trouble.

Magnetic media such as hard drives are fragile. Water, shock,
lightning, and other factors can easily damage a hard drive beyond
repair. Some or all of the data may be retrievable by an expert using
expensive methods and equipment, but it's a gamble.
Optical media is very fragile. Discs can easily be broken, and a single
deep-enough scratch on the reflective (top) side of a disc will
destroy some of the data.
Older methods that are proven to work:
Paper - writing on paper with permanent ink can survive hundreds of
years and can still be legible after being torn, floods, etc.
Parchment, papyrus, etc, may last thousands of years!
Stone carvings last even longer, but are a problem when shock is
involved such as earthquakes, landslides, etc.

I'm not suggesting that we go back to the last century and store
everything on paper or back to the stone age...
I'm suggesting that we find a way to store our data in a way that can
last hundreds to thousands of years, while making it repairable and
retrievable, and able to be understood by the equipment we will have a
century from now.
At this point I believe that everything that is stored in some elctronic
way should be constantly updated and translated into every possible file
format, especially the newest formats as they are created.

 

[msg]

Jumpster Jiver wrote:

I'm not suggesting that we go back to the last century and store
everything on paper or back to the stone age...
I'm suggesting that we find a way to store our data in a way that can
last hundreds to thousands of years, while making it repairable and
retrievable, and able to be understood by the equipment we will have a
century from now.

Perhaps take a lesson from Voyager; encode baseband data by engraving
spiraling grooves onto a noble-metal disc (phonograph record). The
recordings attached to the spacecraft are intended to endure the
journey between stars Closer to home and seriously, home recordings
made on disc machines from the 1930's are often as good now as when
originally made; consider this technology using existing equipment
and add robust error correcting coding. Also, some photographic films
have good longevity; consider microfiche printouts for data archiving.

In terms of supporting technology, if personnel are available throughout
the archiving period, a periodic refresh of data onto contemporary
media is in order using updated equipment at each cycle. If this is
not possible, one of the archiving methods described above would require
only basic technology to reconstruct the data by future societies.

Regards,

Michael

 

[[email protected]]

Complicated one, let's say i have a need to keep certain programs/ data
in running order for the next 50 years.
Don't discuss upgrading, converting ect, it's not applicable in this
situation as the actual structure hard and software have to match each
other and you cannot emulate.

I don't believe it. If it's that important, everything can be
emulated.

I see a lot of survivors among early PIII slot 1 machines, may sound
strange but i have seen quite a lot of those that are still in use and
work just fine.
So if i need a computer for the following 50 years based on either PII
PIII PIV or similar (AMD/CYRIX...) (isn't that called I865 architecture)
what machines/brands of components i would source that could last that
long, storing units for spares is possible until certain level but i
think aging on stored components could be a problem as well sudden shoch
when becoming operational again?
Or am i paranoia and should just get myself a batch of PIII machones
from a recycler and strip them down, pack and vacuum seal the parts and
stock them?

Yes do that

I was thinking of aquiring a few older generation workstations as i
assume those ones were still built to proper industrial specs.
Any suggestions?

Cheetah

"have you counted your spots today?"

Radium, is that you?

GG

 

[CheetahHugger]

I think you see my problem, reliabillity goes down and modernization
comes with new systems every few months, we sit with over 2TB on private
pics and videos and come to the conclusion my granny's super 8 films
last longer then any of today's media.
Blank dvd's corrupted in a month, harddisks last a year, a pc 2 years if
you're lucky.
How can anyone base a buisness on such low quality standards, failing
capacitors being one example.
I (and a few other people) refuse to upgrade anymore and rather stock up
parts to keep existing equipment going.
Due to high volumes and recycling problems, older machines can often be
had for practically nothing.
Conversions are good but reliabillity of current hardware is poor, so
poor that i don't even consider using cd's for music anymore, i run
everything off solid state, music on a fileserver accessable in the
entire house, same for video and pictures, no DRM and no milky way nonsense.
No future headaches on what version of xp will run what without any new
restrictions to make the life of users to a hell, linux on a p3 running
off solid state memory might be a way out, and all data on triple
mirrored harddisks.

Cheetah

 

[CheetahHugger]

Why you think there has been such a massive campaign to have people send
in equipment for recycling and at the same time promote digital as the
new future?
because they know it won't last, my father's records all play, my old
tapes from 1980 still play, my cd's with pictures burned in 2004 are
unreadable, i do have backups.
The cd's were bought from TDK and advertised as archive grade.
guess who didn't made the standard.

Cheetah

 

[CheetahHugger]

Radium, is that you?

I don't get that, i was asking this question with serious eye on archiving.
Thanks for the input everybody, i will draw conclusions.

Cheetah

 

[[email protected]]

Why you think there has been such a massive campaign to have people send
in equipment for recycling and at the same time promote digital as the
new future?
because they know it won't last, my father's records all play, my old
tapes from 1980 still play, my cd's with pictures burned in 2004 are
unreadable, i do have backups.
The cd's were bought from TDK and advertised as archive grade.
guess who didn't made the standard.

Cheetah

Well I have CDRs nearly 10 years old that are still OK. We bake tapes
at work (135 F for 12 hours) to get them to play on the 1978 Ampex 2"
quadruplex VTR. My turntable is OK but you can't get a replacement
stylus (Shure V15 type V) anymore. 50 years is a LONG time to expect
electronics to keep working. You might be dead by then, I certainly
will be and its pretty arrogant to assume whatever task those old
machines do couldn't be done by newer people and machinery.

GG

 

[N Cook]

Well I have CDRs nearly 10 years old that are still OK. We bake tapes
at work (135 F for 12 hours) to get them to play on the 1978 Ampex 2"
quadruplex VTR. My turntable is OK but you can't get a replacement
stylus (Shure V15 type V) anymore. 50 years is a LONG time to expect
electronics to keep working. You might be dead by then, I certainly
will be and its pretty arrogant to assume whatever task those old
machines do couldn't be done by newer people and machinery.

GG

If you stored items of hardware in evacuated containers are there components
that will fail if subjected to 0 psi ?
Will hermetically sealed crystal oscillators explode for example ?

 

[bz]

Do NOT vacuum seal them!

That will draw out any volatile or semi volatile materials in any
components.

You probably should seal them with 1 or 1.5 atmospheres of DRY nitrogen.

Even then, some plastics will out gas plasticizers over time and may
crumble.

No one has considered the software in this thread.
The ferrite material , whether on a hard disc platten, floppy or tape
streamer tape can debond from the backing.
With optical media the code carying metal surface can oxidise/ part
company from the plastic media,

Again slightly above atmospheric pressure, a dry, inert gas such as
nitrogen, will give the best protection.

Unless anyone knows of the ideal storage media that is incorruptible, is
this the best procedure ?
Store the original media plus copies on perhaps more futureproof media
in an evacuated container. Then another repeat of that collection at
another site , in case of fire or flood at one site.
Is long term optical storage more reliable as far as it cannot
spontaneously change magnetic state by coalescence of neighbouring
magnetic poles or whatever the corrupting process is.

--
bz 73 de N5BZ k

please pardon my infinite ignorance, the set-of-things-I-do-not-know is an
infinite set.

[email protected] remove ch100-5 to avoid spam trap

 

[CheetahHugger]

If you stored items of hardware in evacuated containers are there components
that will fail if subjected to 0 psi ?
Will hermetically sealed crystal oscillators explode for example ?

Good question, maybe just vacuum packing is better, and the reason for
archiving is a personal one, if i die the matter is no longer relevant.

Cheetah

 

[James Sweet]

If you stored items of hardware in evacuated containers are there
components
that will fail if subjected to 0 psi ?
Will hermetically sealed crystal oscillators explode for example ?

I doubt the oscillators will be a problem, but I would be concerned about
electrolytic capacitors. An atmosphere of dry nitrogen or other inert gas at
approximately atmospheric pressure or even a slight positive pressure would
be better. Some hermetically sealed military subcircuits are pressurized
with nitrogen.

 

[William R. Walsh]

Hi!

Optical media is very fragile. Discs can easily be broken

Fragile? Easily?

With all due respect, have you tried this?

I got a bunch of America Online CDs a long time ago. It's a long story, but
it involved a display being taken down. Despite the fact that the software
was unchanged (same version), the CDs were "useless" and would be thrown out
with the rest of the display. I asked the woman tearing down the stand if I
could have them...as I was thinking thoughts of "what wastefulness", "maybe
I could do something with these" or "anything would be better than throwing
most of them away".

Once I ran out of ideas (hanging them in trees to annoy birds, using them to
reinstall AOL software on AOL-users computers after they'd messed something
up--why they make it so hard to download a simple one-shot installer is
beyond me, using them in art projects, etc)...there was really only one
thing left to do, and that was to find out what kind of abusive handling it
would take to break them--both in and out of their paper sleeves.

I used a few implements of destruction--an office chair, a table, a PS/2
model 85 (!!!), one file cabinet drawer and a vehicle. Of all of these
things, only the file cabinet drawer could break a disc into pieces on the
first try, and that was only by "cheating" and holding the disc in the path
of the moving drawer. One *slam* and it was all over for that disc. It
didn't matter if the disc was in its sleeve or not. I could run over the
disc almost to my heart's content with the chair, slap it against the PS/2's
heavy steel case as hard as I could manage to do so or run it over
repeatedly with the vehicle. If the disc was in its sleeve...it would
survive. If the disc was out...it got scratched but was at least partially
readable.

William