4 ms·

Realistically how long would it take to modernize the software while keeping the old hardware?

Dalewyn 3y ago
Far more than anyone is willing to pay.
rightbyte 3y ago
Why would they? The TUI seems to be fine. An modernization would probably be way worse given today's "UX" trends. Maybe they should replace the computer and cassettes before they run of out them.
Adamantcheese 3y ago
Assuming the computer talks over a standard protocol like serial or parallel, I don't see how it wouldn't be fairly straightforward to move it to a newer machine. Might even be able to run it on a virtual machine for better compatibility.
tivert 3y ago
> Realistically how long would it take to modernize the software while keeping the old hardware? What value would that provide (than make-work for a couple of software engineers)? Probably the most bang-for-the-buck effort would be get their existing software running on an emulator, and get that integrated with the hole-punch machine. An even then, I think the main value there would be to have a backup in hand if they have trouble maintaining that old computer hardware.
- Freedom2 3y ago
  This sounds like indoctrination to me. I generally have a different worldview that it would provide a lot of value.
  - rightbyte 3y ago
    Maybe you would. But nowadays, it would probably be an agile disaster with two or three layers of managers and sales people, maybe even on different firms, if done by best practice, between you and the customer. Forget talking to the operators. KISS is a beauty of the past. There is no way the simple and straightforward text based solution would be made. It is more likely it would be a bloated mess with clouds or what not.
  - tivert 3y ago
    > This sounds like indoctrination to me. I generally have a different worldview that it would provide a lot of value. That's a weird way of putting it, but I honestly think the "modernize" all the things crowd are probably the ones that "sound indoctrinated." So, if your "different worldview" says reflexively redeveloping the software would provide "a lot of value," then name the value it would provide. They've got a process that's worked for them for 40+ years, what exactly is the value in re-doing software that's only a minor part of that process? Does that "value" justify the cost? Honestly, what they've got can probably be understood more as an appliance than as an IT system, and anything developed today would almost certainly take the form of an IT system, with all the recurring cost and complexity that entails.
- WJW 3y ago
  I think characterizing it just as "make-work for software engineers" is too harsh. Just from a business continuity perspective there is a real benefit to being on a well-supported platform for which it is cheap and easy to source new parts. I don't know offhand how many punch card reader manufacturers still exist, but the competition is going to be a lot less than it is in (say) USB stick manufacturing. It's a difficult thing to quantify, but access to stable supply lines is surely worth more than zero.
  - Symbiote 3y ago
    The punch cards will be directly controlling the loom, it looks like it is read electro-mechanically. Maintaining the card reader is presumably a minor part of maintenance for the whole loom. One person probably has the skills to maintain the weaving part and the reading part. If you replaced the input with USB or similar, you now also need someone with IT skills if that part goes wrong.
- zokier 3y ago
  > What value would that provide (than make-work for a couple of software engineers)? In the lingo, business continuity and risk management. When, not if but when, the current system fails, there is high chance of it being unrecoverable. Even if it were technically recoverable, can they find someone to do that work? How long will the factory stand still while the system is being recovered or replaced? In general it is more cost-effective to replace things before they fail rather than after the fact. Doubly so in these sort of cases where using the old system (in working state!) as a reference would most likely be beneficial.
amiga-workbench 3y ago
Can you think of any modern platform that would have the longevity, stability and lack of maintenance burden of that Sharp computer over another 40 years? The only thing I can think of that would make their current setup more robust would be having the program burned into ROM and not on tape.
- jrockway 3y ago
  > Can you think of any modern platform that would have the longevity, stability and lack of maintenance burden of that Sharp computer over another 40 years? I mean, probably any UNIX program written in C? Bill Joy vi can run on modern computers. What makes this software substantially different than a text editor?
  - hulitu 3y ago
    Just try to compile a vi from 20 years ago on a modern computer.
    - jrockway 3y ago
      According to Wikipedia that was a problem ... 20 years ago. That version compiles today. It's called "Traditional vi" and the Sourceforge site where it's hosted doesn't have a valid certificate so I didn't visit it.
    - zokier 3y ago
      20 years ago was 2003, four years after C99. First x86-64 CPUs were released 20 years ago. Compiling 20 year old C codebase should not be an issue at all if it was half-decent to begin with. Compiling pre-ANSI-C and pre-POSIX codebases might have challenges, but standardization has really gotten us lot of stability here. I'd expect conservatively written C99/POSIX code to be usable certainly for another 20 years, there is just so much inertia there https://en.wikipedia.org/wiki/Lindy_effect https://en.wikipedia.org/wiki/Lindy_effect
- poplet 3y ago
  I wonder how old these tapes are
ohuf 3y ago
...that would void the warranty! ;)
hulitu 3y ago
An eternity. They wouldn't have finished the first version when they start working on the second.
lost_tourist 3y ago
You could probably find an old timer who could bridge the gap fairly easily and lead a team of a couple of engineers to do it in a reasonable amount of time. I'm guessing that program is written in basic or assembly which wasn't too hard to learn on those old machines.
nuc1e0n 3y ago
The z88dk toolchain supports the MZ-80K as a target. So you could write new software in C or any other language that can output C code.

Far more than anyone is willing to pay.

Why would they? The TUI seems to be fine. An modernization would probably be way worse given today's "UX" trends. Maybe they should replace the computer and cassettes before they run of out them.

Adamantcheese 3y ago

Assuming the computer talks over a standard protocol like serial or parallel, I don't see how it wouldn't be fairly straightforward to move it to a newer machine. Might even be able to run it on a virtual machine for better compatibility.

tivert 3y ago

> Realistically how long would it take to modernize the software while keeping the old hardware? What value would that provide (than make-work for a couple of software engineers)? Probably the most bang-for-the-buck effort would be get their existing software running on an emulator, and get that integrated with the hole-punch machine. An even then, I think the main value there would be to have a backup in hand if they have trouble maintaining that old computer hardware.

Freedom2 3y ago
This sounds like indoctrination to me. I generally have a different worldview that it would provide a lot of value.
- rightbyte 3y ago
  Maybe you would. But nowadays, it would probably be an agile disaster with two or three layers of managers and sales people, maybe even on different firms, if done by best practice, between you and the customer. Forget talking to the operators. KISS is a beauty of the past. There is no way the simple and straightforward text based solution would be made. It is more likely it would be a bloated mess with clouds or what not.
- tivert 3y ago
  > This sounds like indoctrination to me. I generally have a different worldview that it would provide a lot of value. That's a weird way of putting it, but I honestly think the "modernize" all the things crowd are probably the ones that "sound indoctrinated." So, if your "different worldview" says reflexively redeveloping the software would provide "a lot of value," then name the value it would provide. They've got a process that's worked for them for 40+ years, what exactly is the value in re-doing software that's only a minor part of that process? Does that "value" justify the cost? Honestly, what they've got can probably be understood more as an appliance than as an IT system, and anything developed today would almost certainly take the form of an IT system, with all the recurring cost and complexity that entails.
WJW 3y ago
I think characterizing it just as "make-work for software engineers" is too harsh. Just from a business continuity perspective there is a real benefit to being on a well-supported platform for which it is cheap and easy to source new parts. I don't know offhand how many punch card reader manufacturers still exist, but the competition is going to be a lot less than it is in (say) USB stick manufacturing. It's a difficult thing to quantify, but access to stable supply lines is surely worth more than zero.
- Symbiote 3y ago
  The punch cards will be directly controlling the loom, it looks like it is read electro-mechanically. Maintaining the card reader is presumably a minor part of maintenance for the whole loom. One person probably has the skills to maintain the weaving part and the reading part. If you replaced the input with USB or similar, you now also need someone with IT skills if that part goes wrong.
zokier 3y ago
> What value would that provide (than make-work for a couple of software engineers)? In the lingo, business continuity and risk management. When, not if but when, the current system fails, there is high chance of it being unrecoverable. Even if it were technically recoverable, can they find someone to do that work? How long will the factory stand still while the system is being recovered or replaced? In general it is more cost-effective to replace things before they fail rather than after the fact. Doubly so in these sort of cases where using the old system (in working state!) as a reference would most likely be beneficial.

amiga-workbench 3y ago

Can you think of any modern platform that would have the longevity, stability and lack of maintenance burden of that Sharp computer over another 40 years? The only thing I can think of that would make their current setup more robust would be having the program burned into ROM and not on tape.

jrockway 3y ago
> Can you think of any modern platform that would have the longevity, stability and lack of maintenance burden of that Sharp computer over another 40 years? I mean, probably any UNIX program written in C? Bill Joy vi can run on modern computers. What makes this software substantially different than a text editor?
- hulitu 3y ago
  Just try to compile a vi from 20 years ago on a modern computer.
  - jrockway 3y ago
    According to Wikipedia that was a problem ... 20 years ago. That version compiles today. It's called "Traditional vi" and the Sourceforge site where it's hosted doesn't have a valid certificate so I didn't visit it.
  - zokier 3y ago
    20 years ago was 2003, four years after C99. First x86-64 CPUs were released 20 years ago. Compiling 20 year old C codebase should not be an issue at all if it was half-decent to begin with. Compiling pre-ANSI-C and pre-POSIX codebases might have challenges, but standardization has really gotten us lot of stability here. I'd expect conservatively written C99/POSIX code to be usable certainly for another 20 years, there is just so much inertia there https://en.wikipedia.org/wiki/Lindy_effect https://en.wikipedia.org/wiki/Lindy_effect
poplet 3y ago
I wonder how old these tapes are

ohuf 3y ago

...that would void the warranty! ;)

hulitu 3y ago

An eternity. They wouldn't have finished the first version when they start working on the second.

lost_tourist 3y ago

You could probably find an old timer who could bridge the gap fairly easily and lead a team of a couple of engineers to do it in a reasonable amount of time. I'm guessing that program is written in basic or assembly which wasn't too hard to learn on those old machines.

nuc1e0n 3y ago

The z88dk toolchain supports the MZ-80K as a target. So you could write new software in C or any other language that can output C code.