How long can a computer last?

[audion_1908]

Originally Posted by stevehertz

Surely with computers the issue is not just hardware, but their inability to keep up with up to date apps and stuff? These all require a never ending increase in computing power that old computers just don't have. They become 'outdated', slow and unusable.

Some banks retained the use of ancient computers well beyond their commercial lifespan, and used more modern computers as the front end, now they offen use software emulation to run the original software, sometimes with two or more layers of emulation! Stacked emulators scare me!
I think many Banks only got rid of their heirloom computers due to y2k! Didn't the internet use pdp-11 for DNS until late 1999?

[SiriusHardware]

Originally Posted by audion_1908

How do they hide the code? And how many times can an eprom or eeprom be rewritten?

If we only propose to rewrite the contents every ten years, the number of possible rewrites is to all intents and purposes infinite.

With Microcontrollers the default behaviour is that you can read the code out of them as easily as you can program it in, but devices which feature 'code protection' have a 'flag' which can be set at programming time to stop the 'read out' feature from working. In this case the programmer programs the code, reads it back (verifies it) to confirm that it has programmed correctly and then sets the 'protect' flag so that the code can no longer be read out by a device reader/programmer. This feature is intended to stop the code, which may have taken many man-hours to write and debug, from being stolen by a competitor. They may be able to clone your hardware but without the code to programme into the microcontroller, the copy will be useless.

Usually the only way to clear the protection flag is to erase the device, which allows you to read it once more. Of course if you do that, there is no longer any code to read out of the device.

[G6Tanuki]

I know that there were some early "Fusible Link" type PROMs that had an issue with a sort of re-growth effect whereby a 'blown' location would sometimes become un-blown after a few years. We had some RACAL-Milgo leased-line modems built in the 70s which suffered badly with this.

[audion_1908]

Originally Posted by TonyDuell

Yes, just like a TV22 will not receive digital television broadcasts, a DAC90A will not receive VHF FM broadcasts, a Decola will not play CDs, etc.

With an adapter the tv22 can display digital TV just in black and white and mono sound and a screen size that will likely puzzle the new generation! Lol but with a direct RF feed probably a very nice picture! And there where oil-filled plastic magnifiers, I have one in my wardrobe!

I had to Google the Decola, it would probably make an excellent amp hooked to a CD player, it has so meny small speakers! and must weigh a ton!
The Decola hooked to a good bluetooth adapter could be controlled by your phone and have a nearly infinite playlist!

Some valve radios have a pa input that can be used with a simple adapter made with a resistor and 2 y class caps, or a plug in Bluetooth receiver, thow the DAC90A dose not, but a Pico am transmitter can be used to connect it to anything!

Old teck can be extended

[audion_1908]

Originally Posted by G6Tanuki

I know that there were some early "Fusible Link" type PROMs that had an issue with a sort of re-growth effect whereby a 'blown' location would sometimes become un-blown after a few years. We had some RACAL-Milgo leased-line modems built in the 70s which suffered badly with this.

Is that an issue with the the burning voltage being too low? I needed a replacement eprom a few years ago, but found it way cheaper buying an obsolete eprom burner that ran off an ISA cards, I think it could burn up to 21 or 24v volts, and 8 eproms at a go!

But I can set the voltage, I was surprised to be able to find the original dos software!

I think you can just re blow the Fusible Link" type PROMs but best at a higher voltage and current, as long as the codes the same.

How did you fix it?

[audion_1908]

Originally Posted by SiriusHardware

If we only propose to rewrite the contents every ten years, the number of possible rewrites is to all intents and purposes infinite.

With Microcontrollers the default behaviour is that you can read the code out of them as easily as you can program it in, but devices which feature 'code protection' have a 'flag' which can be set at programming time to stop the 'read out' feature from working. In this case the programmer programs the code, reads it back (verifies it) to confirm that it has programmed correctly and then sets the 'protect' flag so that the code can no longer be read out by a device reader/programmer. This feature is intended to stop the code, which may have taken many man-hours to write and debug, from being stolen by a competitor. They may be able to clone your hardware but without the code to programme into the microcontroller, the copy will be useless.

Usually the only way to clear the protection flag is to erase the device, which allows you to read it once more. Of course if you do that, there is no longer any code to read out of the device.

As the company that made it is likely long gone, could the code be probed out or analysed what it's doing and write an equivalent?

How common is it to add copy protection?

[Scimitar]

Originally Posted by audion_1908

CID? Why did the PSU die?

Caller ID. The server has two 'phone interfaces connected and when a call comes in, it logs the call and sends that and any associated pre-logged information over the network to the other computers running Virtual-RPC.

Can't say why the PSU died, but my guess is electrolytics failure being a SMPSU.
I'm dreading hard drive failure because I know life is going to become interesting then and there is no equivalent PC application that I have found.

[audion_1908]

Can custom firmware be used to convert multi bit per cell SSD to 1 bit per cell for much greater data retention reliability, but also a large drop in capacity?

[SiriusHardware]

I know that there were some early "Fusible Link" type PROMs that had an issue with a sort of re-growth effect whereby a 'blown' location would sometimes become un-blown after a few years.

Yes, Bipolar fusible-link PROMs. I have a retro computer board, the OS of which is held in that type of PROM. I was just congratulating myself that at least the OS would never die from EPROM - style data fade, and then I heard about this habit they have of occasionally re-growing the blown links. I was not happy.

I know this does happen, I have helped several people 'rescue' old computer systems by programming replacement bipolar PROMs for them after the original ones had failed.

Which brings us to another problem.

You have a replacement blank device <check>
You have the code which need to go into it <check>

Now, how do you get the code into the device? Taking the example of the Fusible-Link PROMs, these are now truly historic and almost no modern, reasonably priced programmer can program them. (There are about three or four eye-wateringly expensive programmers in current production which still support them).

If you have the need to program such devices every now and again, you may find yourself having to acquire and maintain the entire equipment infrastructure necessary to do the job - an ancient device programmer and an ancient PC to run its support software. Both of these items (the ancient programmer and the ancient PC) may themselves contain antique programmed devices which may fail at any moment... and so on and so forth.

How common is it to add copy protection?

In commercial (as opposed to hobby) applications, very. Not protecting the code would usually be considered careless.

[G6Tanuki]

Originally Posted by audion_1908

How did you fix it?

After complaining to Racal-Milgo, who reminded us that the offending modems were now several years out of support, we 'fixed it' by replacing all the modems with some current-production ones (which took up less rack-space and also had vastly better automatic line-equalisation).

Problem solved!

[audion_1908]

Originally Posted by Scimitar

Caller ID. The server has two 'phone interfaces connected and when a call comes in, it logs the call and sends that and any associated pre-logged information over the network to the other computers running Virtual-RPC.

Can't say why the PSU died, but my guess is electrolytics failure being a SMPSU.
I'm dreading hard drive failure because I know life is going to become interesting then and there is no equivalent PC application that I have found.

What type of hard drive dose it take? If ide you can use a cf flash card and cheap passive adaptor, or an ide to sata adapter and a standard sata SSD, I think there are also SCSI to sata adaptors

[audion_1908]

Originally Posted by SiriusHardware

Yes, Bipolar fusible-link PROMs. I have a retro computer board, the OS of which is held in that type of PROM. I was just congratulating myself that at least the OS would never die from EPROM - style data fade, and then I heard about this habit they have of occasionally re-growing the blown links. I was not happy.

I know this does happen, I have helped several people 'rescue' old computer systems by programming replacement bipolar PROMs for them after the original ones had failed.

Which brings us to another problem.

You have a replacement blank device <check>
You have the code which need to go into it <check>

Now, how do you get the code into the device? Taking the example of the Fusible-Link PROMs, these are now truly historic and almost no modern, reasonably priced programmer can program them. (There are about three or four eye-wateringly expensive programmers in current production which still support them).

If you have the need to program such devices every now and again, you may find yourself having to acquire and maintain the entire equipment infrastructure necessary to do the job - an ancient device programmer and an ancient PC to run its support software. Both of these items (the ancient programmer and the ancient PC) may themselves contain antique programmed devices which may fail at any moment... and so on and so forth.

In commercial (as opposed to hobby) applications, very. Not protecting the code would usually be considered careless.

How long dose a company keep the original source code?

[audion_1908]

Originally Posted by G6Tanuki

After complaining to Racal-Milgo, who reminded us that the offending modems were now several years out of support, we 'fixed it' by replacing all the modems with some current-production ones (which took up less rack-space and also had vastly better automatic line-equalisation).

Problem solved!

But sadly the modems dumped.
That's how a lot is fixed now, dump and replace, as it's cheaper but more wasteful, and not very green

[SiriusHardware]

Originally Posted by audion_1908

As the company that made it is likely long gone, could the code be probed out or analysed what it's doing and write an equivalent?

The short answer is No. The long answer is Maybe.

As mentioned before, the code protect feature on microcontrollers has not always worked perfectly, ways have sometimes been found to make the microcontroller cough up its internal code despite it being 'protected'. I would imagine most modern microcontrollers will not respond to that sort of attack.

As for reverse engineering software, there are some tremendously clever people around who, starting with the circuit diagram, the datasheets of all the ICs and full knowledge of what the product is supposed to do, can create replacement firmware with equivalent, sometimes even improved functionality from scratch.

For this to happen, there either has to be a lot of demand, or there has to be some kind of reward / bounty involved for the programmers concerned, or the programmers have to have some kind of obsessive love for the equipment and a desire to get it working again no matter what. Amazingly, the latter scenario is much more common than you might imagine.

[SiriusHardware]

Originally Posted by audion_1908

How long does a company keep the original source code?

The company I work for still has archived copies of every version of every bit of firmware they have ever produced. They have been going since the mid 70s, I think.

As a repair technician I still draw on those archives from time to time, if we get an old piece of equipment back for repair with a failed EPROM, Microcontroller or PLD.

[duncanlowe]

Regarding getting code out of protected controllers, there have been extreme ways, including machining the tops off the IC's and using weird voltages on programming pins. I guess that these days ways of bypassing the protection are significantly harder as protecting the intellectual property is very important. I did once come across an application, PIC based, where the originator hadn't set any code protection. This though was a non commercial product. It was useful that they hadn't because they dissapeared completely, but it was possible to make a few more by simply reproducing the circuit and cloning the firmware. The information was freely available to create the device from scratch, but having the code left freely available saved a lot of time and testing!

[audion_1908]

Originally Posted by SiriusHardware

Yes, Bipolar fusible-link PROMs. I have a retro computer board, the OS of which is held in that type of PROM. I was just congratulating myself that at least the OS would never die from EPROM - style data fade, and then I heard about this habit they have of occasionally re-growing the blown links. I was not happy.

I know this does happen, I have helped several people 'rescue' old computer systems by programming replacement bipolar PROMs for them after the original ones had failed.

Which brings us to another problem.

You have a replacement blank device <check>
You have the code which need to go into it <check>

Now, how do you get the code into the device? Taking the example of the Fusible-Link PROMs, these are now truly historic and almost no modern, reasonably priced programmer can program them. (There are about three or four eye-wateringly expensive programmers in current production which still support them).

If you have the need to program such devices every now and again, you may find yourself having to acquire and maintain the entire equipment infrastructure necessary to do the job - an ancient device programmer and an ancient PC to run its support software. Both of these items (the ancient programmer and the ancient PC) may themselves contain antique programmed devices which may fail at any moment... and so on and so forth.

In commercial (as opposed to hobby) applications, very. Not protecting the code would usually be considered careless.

Lol a big can of worms! But I think my 80 or 90s EPROM burner can do a range of proms, the 8 bit ISA card is in a Pentium 1, I think 100mhz, it uses a full length PCB packed full of TTL chips, and cost me very little and was surprisingly easy to set up, I was worrying as I think the software was written for the IBM XT! and lots of early software wouldn't run on pentiums but worked flawlessly, I suppose my first computers all used dos helped, in the bios I had the set the ISA bus to slow lol, I got the computer out of a skip years ago and that was it's only use! it may still be lurking under the table. A Pentium 100 is nothing now but way over spek for the task, I may have dos Tetris? If not it should!

I think the eproms I needed for my old commodore need 12 volts and the newer the eprom burner the lower the voltage it could output, but why are modern EPROM programmers that support the old proms so expensive?

[audion_1908]

Originally Posted by SiriusHardware

The company I work for still has archived copies of every version of every bit of firmware they have ever produced. They have been going since the mid 70s, I think.

As a repair technician I still draw on those archives from time to time, if we get an old piece of equipment back for repair with a failed EPROM, Microcontroller or PLD.

They should like a good company to do business with, my phone I am typing this on is 4 years old so now unsupported and obsolete!

[audion_1908]

Originally Posted by duncanlowe

Regarding getting code out of protected controllers, there have been extreme ways, including machining the tops off the IC's and using weird voltages on programming pins. I guess that these days ways of bypassing the protection are significantly harder as protecting the intellectual property is very important. I did once come across an application, PIC based, where the originator hadn't set any code protection. This though was a non commercial product. It was useful that they hadn't because they dissapeared completely, but it was possible to make a few more by simply reproducing the circuit and cloning the firmware. The information was freely available to create the device from scratch, but having the code left freely available saved a lot of time and testing!

That was lucky!

[SiriusHardware]

why are modern EPROM programmers that support the old proms so expensive?

I think because only a tiny minority of people now need to program things like Bipolar PROMs, so to get that functionality you have to buy the top of the range programmer which can programme absolutely everything else as well. Creating the programming profiles for each and every single device that the programmer can handle takes time, and the man-hours involved are reflected in the price.

My high-end device programmer also dates from the mid-nineties and I bought that one specifically because it programmed Bipolar PROMs, which, even then, only the high end models still did. The support software is DOS, so I keep not just one, but two DOS computers working to ensure that I will be able to keep using it for the forseeable future.

I've noticed that some of the newest, cheap 'universal' EPROM programmers can not generate a VPP (programming voltage) of 25V. The designers are assuming that virtually no-one wants to be able to programme first and second generation EPROMs any more.