Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.os.linux.misc > #90011 > unrolled thread
| Started by | c186282 <c186282@nnada.net> |
|---|---|
| First post | 2026-08-17 22:25 -0400 |
| Last post | 2026-08-19 11:48 -0400 |
| Articles | 20 on this page of 83 — 13 participants |
Back to article view | Back to comp.os.linux.misc
This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by
below is the oldest one visible, not the original post.
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-17 22:25 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-18 03:05 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-18 00:58 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-18 10:21 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Rich <rich@example.invalid> - 2026-08-18 13:55 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Robert Riches <spamtrap42@jacob21819.net> - 2026-08-18 21:24 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-19 14:41 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-19 19:00 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Rich <rich@example.invalid> - 2026-08-18 13:53 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2026-08-18 16:30 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-08-18 23:02 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-18 23:02 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Rich <rich@example.invalid> - 2026-08-19 12:53 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-19 19:25 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-19 21:37 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-20 03:24 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-20 12:32 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-20 20:53 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-20 22:14 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-21 04:29 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-21 01:17 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Rich <rich@example.invalid> - 2026-08-20 13:55 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-20 13:04 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2026-08-20 18:04 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-20 21:02 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-20 22:30 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-21 04:24 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-21 01:12 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! TJ <TJ@noneofyour.business> - 2026-08-21 08:22 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Marc Haber <mh+usenetspam2616@zugschl.us> - 2026-08-21 16:01 +0200
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-21 15:22 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-21 23:55 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-21 18:18 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-08-22 01:56 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Richard Kettlewell <invalid@invalid.invalid> - 2026-08-22 10:17 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-21 23:31 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! TJ <TJ@noneofyour.business> - 2026-08-22 08:36 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-22 13:44 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! TJ <TJ@noneofyour.business> - 2026-08-22 17:06 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-08-22 22:35 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-23 02:57 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! "Carlos E.R." <robin_listas@es.invalid> - 2026-08-23 13:32 +0200
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-23 13:10 +0100
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! "Carlos E.R." <robin_listas@es.invalid> - 2026-08-23 14:47 +0200
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-23 14:26 +0100
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-23 18:34 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! "Carlos E.R." <robin_listas@es.invalid> - 2026-08-23 21:40 +0200
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-24 01:24 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-24 02:21 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-24 21:05 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-24 00:54 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! "Carlos E.R." <robin_listas@es.invalid> - 2026-08-24 09:18 +0200
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-24 21:13 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-25 00:01 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 11:38 +0100
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-25 22:55 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! "Carlos E.R." <robin_listas@es.invalid> - 2026-08-25 14:08 +0200
Legacy computer in embedded industial systems (Re: Oh Gods ...) Lars Poulsen <lars@beagle-ears.com> - 2026-08-23 06:20 -0700
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) The Natural Philosopher <tnp@invalid.invalid> - 2026-08-23 14:31 +0100
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) Lars Poulsen <lars@beagle-ears.com> - 2026-08-23 11:28 -0700
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) The Natural Philosopher <tnp@invalid.invalid> - 2026-08-23 19:58 +0100
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-24 01:43 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) "Carlos E.R." <robin_listas@es.invalid> - 2026-08-24 09:21 +0200
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-25 00:33 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) "Carlos E.R." <robin_listas@es.invalid> - 2026-08-25 14:11 +0200
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-26 03:02 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) rbowman <bowman@montana.com> - 2026-08-26 17:53 +0000
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-27 02:57 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) rbowman <bowman@montana.com> - 2026-08-27 15:08 +0000
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-24 01:06 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) Anssi Saari <anssi.saari@usenet.mail.kapsi.fi> - 2026-08-26 13:59 +0300
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) The Natural Philosopher <tnp@invalid.invalid> - 2026-08-26 13:53 +0100
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-27 01:28 -0400
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) "Carlos E.R." <robin_listas@es.invalid> - 2026-08-27 08:32 +0200
Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) c186282 <c186282@nnada.net> - 2026-08-24 01:01 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-23 23:19 -0400
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-23 12:46 +0100
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! rbowman <bowman@montana.com> - 2026-08-23 18:41 +0000
Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-24 00:19 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! c186282 <c186282@nnada.net> - 2026-08-21 21:54 -0400
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! The Natural Philosopher <tnp@invalid.invalid> - 2026-08-21 13:27 +0100
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! Rich <rich@example.invalid> - 2026-08-22 04:15 +0000
Re: Oh Gawd - Now MX Linux "Security Verification" - NO !!! TJ <TJ@noneofyour.business> - 2026-08-19 11:48 -0400
Page 4 of 5 — ← Prev page 1 2 3 [4] 5 Next page →
| From | The Natural Philosopher <tnp@invalid.invalid> |
|---|---|
| Date | 2026-08-23 19:58 +0100 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <116fft9$25kdt$3@dont-email.me> |
| In reply to | #90403 |
On 23/08/2026 19:28, Lars Poulsen wrote: > On 2026-08-23 06:31, The Natural Philosopher wrote: >> On 23/08/2026 14:20, Lars Poulsen wrote: >>> On 2026-08-23 04:32, Carlos E.R. wrote: >>>> On 2026-08-23 08:57, c186282 wrote: >>>>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>>>> >>>>>>> It goes against the grain, for me at least, to send perfectly good >>>>>>> running hardware off to the recycler, just because it is >>>>>>> "experienced." >>>>>> >>>>>> CPU generations (and entire architectures) get dropped from the Linux >>>>>> kernel because they can’t find anybody to step forward and maintain >>>>>> the support for them. >>>>>> >>>>>> That is how they tell that there are so few instances of those >>>>>> particular kinds of hardware around, that essentially nobody cares >>>>>> about them any more. >>>>> >>>>> Sad, but quite correct. >>>>> >>>>> Really not much call to run Linux-ish stuff >>>>> on a 386/486 these days. How many people >>>>> still USE those old boards for ANYTHING ? >>>>> >>>>> MAYBE a few "industrial" monitor/control >>>>> systems ... but REALLY they oughtta upgrade. >>>> > > However SOME won't because it might cost a >>>>> few dollars. They'll run the old hardware >>>>> until it literally bursts into flames - and >>>>> then be SCREWED for days/weeks/months. But, >>>>> the bean-counters will TELL them they're >>>>> saving SO much money, now ....... >>>> >>>> Industrial computers may have boards, like data acquisition boards, >>>> to interface with motors and sensors (or cameras). If you upgrade >>>> the computer, it comes with a new operating system, so their >>>> expensive software does not run on it. The cards probably can not be >>>> plugged in, so they also need replacing; the new boards will >>>> probably not be compatible with the expensive software. So finally >>>> they need to recreate the software (hopefully the company that >>>> created it still exists and has the source files). And maybe there >>>> are a bunch of computers to replace. >>>> >>>> The price is no longer minor. >>> >>> What you have to do is make sure you have one or two backup hardware >>> packages, because when the computer breaks, you won't be able to buy >>> one that works. (Just try to find a motherboard with ISA slots today!) >> >> https://www.directindustry.com/industrial-manufacturer/isa- >> motherboard-255106.html >> >>> And with custom plugin boards, it is kinda too late to have them >>> fabbed when the system is already down. And fire it up once a year to >>> make sure it still runs. >>> >>> That may not guarantee that you can recover, but it vastly increases >>> your chances. >>> >>> And when you put the reserve system into service, it is time to >>> expedite developing a replacement! >>> >>> For PC systems, the big hurdle is that old 16-bit systems allowed >>> programs to do I/O to ISA bus peripherals without kernel programming, >>> device drivers and OS hooks. When you "modernize" that ability is >>> lost, and you have to rebuild with much more complicated programming. >>> >> That is a function of the operating system, not the chipset > > Good luck finding a motherboard with ISA bus slots. I already showed you a link to where a dozen or so are available For a very long > time, motherboards had ISA bus parallel ports and serial ports, where > all you had to do to access them directly was to populate the footprint > for the connector. I think those went away around 2018. > No, they became niche rather than mainstream, thats all -- I would rather have questions that cannot be answered... ...than to have answers that cannot be questioned Richard Feynman
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-24 01:43 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <l9icnfrKBY3CRxb3nZ2dnZfqnPGdnZ2d@giganews.com> |
| In reply to | #90403 |
On 8/23/26 14:28, Lars Poulsen wrote: > On 2026-08-23 06:31, The Natural Philosopher wrote: >> On 23/08/2026 14:20, Lars Poulsen wrote: >>> On 2026-08-23 04:32, Carlos E.R. wrote: >>>> On 2026-08-23 08:57, c186282 wrote: >>>>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>>>> >>>>>>> It goes against the grain, for me at least, to send perfectly good >>>>>>> running hardware off to the recycler, just because it is >>>>>>> "experienced." >>>>>> >>>>>> CPU generations (and entire architectures) get dropped from the Linux >>>>>> kernel because they can’t find anybody to step forward and maintain >>>>>> the support for them. >>>>>> >>>>>> That is how they tell that there are so few instances of those >>>>>> particular kinds of hardware around, that essentially nobody cares >>>>>> about them any more. >>>>> >>>>> Sad, but quite correct. >>>>> >>>>> Really not much call to run Linux-ish stuff >>>>> on a 386/486 these days. How many people >>>>> still USE those old boards for ANYTHING ? >>>>> >>>>> MAYBE a few "industrial" monitor/control >>>>> systems ... but REALLY they oughtta upgrade. >>>> > > However SOME won't because it might cost a >>>>> few dollars. They'll run the old hardware >>>>> until it literally bursts into flames - and >>>>> then be SCREWED for days/weeks/months. But, >>>>> the bean-counters will TELL them they're >>>>> saving SO much money, now ....... >>>> >>>> Industrial computers may have boards, like data acquisition boards, >>>> to interface with motors and sensors (or cameras). If you upgrade >>>> the computer, it comes with a new operating system, so their >>>> expensive software does not run on it. The cards probably can not be >>>> plugged in, so they also need replacing; the new boards will >>>> probably not be compatible with the expensive software. So finally >>>> they need to recreate the software (hopefully the company that >>>> created it still exists and has the source files). And maybe there >>>> are a bunch of computers to replace. >>>> >>>> The price is no longer minor. >>> >>> What you have to do is make sure you have one or two backup hardware >>> packages, because when the computer breaks, you won't be able to buy >>> one that works. (Just try to find a motherboard with ISA slots today!) >> >> https://www.directindustry.com/industrial-manufacturer/isa- >> motherboard-255106.html >> >>> And with custom plugin boards, it is kinda too late to have them >>> fabbed when the system is already down. And fire it up once a year to >>> make sure it still runs. >>> >>> That may not guarantee that you can recover, but it vastly increases >>> your chances. >>> >>> And when you put the reserve system into service, it is time to >>> expedite developing a replacement! >>> >>> For PC systems, the big hurdle is that old 16-bit systems allowed >>> programs to do I/O to ISA bus peripherals without kernel programming, >>> device drivers and OS hooks. When you "modernize" that ability is >>> lost, and you have to rebuild with much more complicated programming. >>> >> That is a function of the operating system, not the chipset > > Good luck finding a motherboard with ISA bus slots. > Yea ... threw out my last ISA-bus boards a long time ago. Everyone did. Maybe a bit short-sighted ........ > For a very long > time, motherboards had ISA bus parallel ports and serial ports, where > all you had to do to access them directly was to populate the footprint > for the connector. I think those went away around 2018. > > I got a handful of machines with footprint for a legacy parallel port as > refurbs after corporate desktop replacement rounds, and brought them to > my PCB rework guy. I was very disappointed to learn that the PCB > manufacturer had reduced the footprint from 2.5 mm pin spacing to 2 mm > pin spacing, making the ribbon cables used for this application not fit. You MIGHT be able to find add-on, maybe USB, adapters and drivers that WILL properly emulate parallel/serial ports from the old days. There ARE cheap dongles to fake 9-pin serial ports, but haven't seen any of late that'll do the older 25-pin ones. If you're next level you could MAKE them with microcontrollers and hand-made drivers (TOO old now to get that deep in again). The huge old "wide"-Centronics printer interface, good fuckin' luck. Too bad, had a great old Epson wide pin-printer with that interface - slow but insanely reliable, went through endless giant boxes of the green- stripe wide paper without a glitch. As for pin spacing .... built a bunch of devices on the old Rabbit-II boards. Rabbit-III boards went from 3mm spacing to 2mm spacing - too tiny for mere humans to work with conveniently with ordinary dremel-presses and soldering irons. Sad, the Rabbits were GOOD boards (STILL sold BTW) but had to drop 'em.
[toc] | [prev] | [next] | [standalone]
| From | "Carlos E.R." <robin_listas@es.invalid> |
|---|---|
| Date | 2026-08-24 09:21 +0200 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <r08tlmxeb9.ln2@Telcontar.valinor> |
| In reply to | #90431 |
On 2026-08-24 07:43, c186282 wrote: > On 8/23/26 14:28, Lars Poulsen wrote: >> On 2026-08-23 06:31, The Natural Philosopher wrote: >>> On 23/08/2026 14:20, Lars Poulsen wrote: >>>> On 2026-08-23 04:32, Carlos E.R. wrote: >>>>> On 2026-08-23 08:57, c186282 wrote: >>>>>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>>>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>>>>> >>>>>>>> It goes against the grain, for me at least, to send perfectly good >>>>>>>> running hardware off to the recycler, just because it is >>>>>>>> "experienced." >>>>>>> >>>>>>> CPU generations (and entire architectures) get dropped from the >>>>>>> Linux >>>>>>> kernel because they can’t find anybody to step forward and maintain >>>>>>> the support for them. >>>>>>> >>>>>>> That is how they tell that there are so few instances of those >>>>>>> particular kinds of hardware around, that essentially nobody cares >>>>>>> about them any more. >>>>>> >>>>>> Sad, but quite correct. >>>>>> >>>>>> Really not much call to run Linux-ish stuff >>>>>> on a 386/486 these days. How many people >>>>>> still USE those old boards for ANYTHING ? >>>>>> >>>>>> MAYBE a few "industrial" monitor/control >>>>>> systems ... but REALLY they oughtta upgrade. >>>>> > > However SOME won't because it might cost a >>>>>> few dollars. They'll run the old hardware >>>>>> until it literally bursts into flames - and >>>>>> then be SCREWED for days/weeks/months. But, >>>>>> the bean-counters will TELL them they're >>>>>> saving SO much money, now ....... >>>>> >>>>> Industrial computers may have boards, like data acquisition boards, >>>>> to interface with motors and sensors (or cameras). If you upgrade >>>>> the computer, it comes with a new operating system, so their >>>>> expensive software does not run on it. The cards probably can not >>>>> be plugged in, so they also need replacing; the new boards will >>>>> probably not be compatible with the expensive software. So finally >>>>> they need to recreate the software (hopefully the company that >>>>> created it still exists and has the source files). And maybe there >>>>> are a bunch of computers to replace. >>>>> >>>>> The price is no longer minor. >>>> >>>> What you have to do is make sure you have one or two backup hardware >>>> packages, because when the computer breaks, you won't be able to buy >>>> one that works. (Just try to find a motherboard with ISA slots today!) >>> >>> https://www.directindustry.com/industrial-manufacturer/isa- >>> motherboard-255106.html >>> >>>> And with custom plugin boards, it is kinda too late to have them >>>> fabbed when the system is already down. And fire it up once a year >>>> to make sure it still runs. >>>> >>>> That may not guarantee that you can recover, but it vastly increases >>>> your chances. >>>> >>>> And when you put the reserve system into service, it is time to >>>> expedite developing a replacement! >>>> >>>> For PC systems, the big hurdle is that old 16-bit systems allowed >>>> programs to do I/O to ISA bus peripherals without kernel >>>> programming, device drivers and OS hooks. When you "modernize" that >>>> ability is lost, and you have to rebuild with much more complicated >>>> programming. >>>> >>> That is a function of the operating system, not the chipset >> >> Good luck finding a motherboard with ISA bus slots. > > > > Yea ... threw out my last ISA-bus boards a > long time ago. Everyone did. Maybe a bit > short-sighted ........ > >> For a very long time, motherboards had ISA bus parallel ports and >> serial ports, where all you had to do to access them directly was to >> populate the footprint for the connector. I think those went away >> around 2018. >> >> I got a handful of machines with footprint for a legacy parallel port >> as refurbs after corporate desktop replacement rounds, and brought >> them to my PCB rework guy. I was very disappointed to learn that the >> PCB manufacturer had reduced the footprint from 2.5 mm pin spacing to >> 2 mm pin spacing, making the ribbon cables used for this application >> not fit. > > You MIGHT be able to find add-on, maybe USB, adapters > and drivers that WILL properly emulate parallel/serial > ports from the old days. There ARE cheap dongles to > fake 9-pin serial ports, but haven't seen any of late > that'll do the older 25-pin ones. If you're next level > you could MAKE them with microcontrollers and hand-made > drivers (TOO old now to get that deep in again). > > The huge old "wide"-Centronics printer interface, good > fuckin' luck. Too bad, had a great old Epson wide > pin-printer with that interface - slow but insanely > reliable, went through endless giant boxes of the green- > stripe wide paper without a glitch. > > As for pin spacing .... built a bunch of devices on > the old Rabbit-II boards. Rabbit-III boards went from > 3mm spacing to 2mm spacing - too tiny for mere humans > to work with conveniently with ordinary dremel-presses > and soldering irons. > > Sad, the Rabbits were GOOD boards (STILL sold BTW) > but had to drop 'em. But those usb adapters can't do something the real rs232 or parallel port can do: directly interrupt and get processed within a known time, so much so as to sync clock to a GPS receiver. I wonder if you can connect two machines via parallel port cable and emulated port. -- Cheers, Carlos. ES🇪🇸, EU🇪🇺;
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-25 00:33 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <4R-dndxP7cUahhD3nZ2dnZfqn_adnZ2d@giganews.com> |
| In reply to | #90439 |
On 8/24/26 03:21, Carlos E.R. wrote: > On 2026-08-24 07:43, c186282 wrote: >> On 8/23/26 14:28, Lars Poulsen wrote: >>> On 2026-08-23 06:31, The Natural Philosopher wrote: >>>> On 23/08/2026 14:20, Lars Poulsen wrote: >>>>> On 2026-08-23 04:32, Carlos E.R. wrote: >>>>>> On 2026-08-23 08:57, c186282 wrote: >>>>>>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>>>>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>>>>>> >>>>>>>>> It goes against the grain, for me at least, to send perfectly good >>>>>>>>> running hardware off to the recycler, just because it is >>>>>>>>> "experienced." >>>>>>>> >>>>>>>> CPU generations (and entire architectures) get dropped from the >>>>>>>> Linux >>>>>>>> kernel because they can’t find anybody to step forward and maintain >>>>>>>> the support for them. >>>>>>>> >>>>>>>> That is how they tell that there are so few instances of those >>>>>>>> particular kinds of hardware around, that essentially nobody cares >>>>>>>> about them any more. >>>>>>> >>>>>>> Sad, but quite correct. >>>>>>> >>>>>>> Really not much call to run Linux-ish stuff >>>>>>> on a 386/486 these days. How many people >>>>>>> still USE those old boards for ANYTHING ? >>>>>>> >>>>>>> MAYBE a few "industrial" monitor/control >>>>>>> systems ... but REALLY they oughtta upgrade. >>>>>> > > However SOME won't because it might cost a >>>>>>> few dollars. They'll run the old hardware >>>>>>> until it literally bursts into flames - and >>>>>>> then be SCREWED for days/weeks/months. But, >>>>>>> the bean-counters will TELL them they're >>>>>>> saving SO much money, now ....... >>>>>> >>>>>> Industrial computers may have boards, like data acquisition >>>>>> boards, to interface with motors and sensors (or cameras). If you >>>>>> upgrade the computer, it comes with a new operating system, so >>>>>> their expensive software does not run on it. The cards probably >>>>>> can not be plugged in, so they also need replacing; the new boards >>>>>> will probably not be compatible with the expensive software. So >>>>>> finally they need to recreate the software (hopefully the company >>>>>> that created it still exists and has the source files). And maybe >>>>>> there are a bunch of computers to replace. >>>>>> >>>>>> The price is no longer minor. >>>>> >>>>> What you have to do is make sure you have one or two backup >>>>> hardware packages, because when the computer breaks, you won't be >>>>> able to buy one that works. (Just try to find a motherboard with >>>>> ISA slots today!) >>>> >>>> https://www.directindustry.com/industrial-manufacturer/isa- >>>> motherboard-255106.html >>>> >>>>> And with custom plugin boards, it is kinda too late to have them >>>>> fabbed when the system is already down. And fire it up once a year >>>>> to make sure it still runs. >>>>> >>>>> That may not guarantee that you can recover, but it vastly >>>>> increases your chances. >>>>> >>>>> And when you put the reserve system into service, it is time to >>>>> expedite developing a replacement! >>>>> >>>>> For PC systems, the big hurdle is that old 16-bit systems allowed >>>>> programs to do I/O to ISA bus peripherals without kernel >>>>> programming, device drivers and OS hooks. When you "modernize" that >>>>> ability is lost, and you have to rebuild with much more complicated >>>>> programming. >>>>> >>>> That is a function of the operating system, not the chipset >>> >>> Good luck finding a motherboard with ISA bus slots. >> > >> >> Yea ... threw out my last ISA-bus boards a >> long time ago. Everyone did. Maybe a bit >> short-sighted ........ >> >>> For a very long time, motherboards had ISA bus parallel ports and >>> serial ports, where all you had to do to access them directly was to >>> populate the footprint for the connector. I think those went away >>> around 2018. >>> >>> I got a handful of machines with footprint for a legacy parallel port >>> as refurbs after corporate desktop replacement rounds, and brought >>> them to my PCB rework guy. I was very disappointed to learn that the >>> PCB manufacturer had reduced the footprint from 2.5 mm pin spacing to >>> 2 mm pin spacing, making the ribbon cables used for this application >>> not fit. >> >> You MIGHT be able to find add-on, maybe USB, adapters >> and drivers that WILL properly emulate parallel/serial >> ports from the old days. There ARE cheap dongles to >> fake 9-pin serial ports, but haven't seen any of late >> that'll do the older 25-pin ones. If you're next level >> you could MAKE them with microcontrollers and hand-made >> drivers (TOO old now to get that deep in again). >> >> The huge old "wide"-Centronics printer interface, good >> fuckin' luck. Too bad, had a great old Epson wide >> pin-printer with that interface - slow but insanely >> reliable, went through endless giant boxes of the green- >> stripe wide paper without a glitch. >> >> As for pin spacing .... built a bunch of devices on >> the old Rabbit-II boards. Rabbit-III boards went from >> 3mm spacing to 2mm spacing - too tiny for mere humans >> to work with conveniently with ordinary dremel-presses >> and soldering irons. >> >> Sad, the Rabbits were GOOD boards (STILL sold BTW) >> but had to drop 'em. > > But those usb adapters can't do something the real rs232 or parallel > port can do: directly interrupt and get processed within a known time, > so much so as to sync clock to a GPS receiver. Quite true ... you're a slave to the USB interface. Often that doesn't matter much - but SOMETIMES ... > I wonder if you can connect two machines via parallel port cable and > emulated port. You could connect them, within reason, using the older parallel ports - basically invent your own 'network'. Done it. Had to xfer lots of stuff from some old DECs to the newfangled PC. Xmodem sucked. LOW speed though, but EVERYTHING was slow back in the day. Serial ports worked. Parallel ports worked. Lots of old detailed mag articles - search. The best timeframe is a few years before/after the original IBM-PCs. Now with USB emulators ??? Would not work as well. Sometimes no substitute for those old direct interrupt pins and UARTS. Hmmm ... come across any serial terminals ? I see a few on E-Bay, but don't trust 'em to work. Linux/BSD does have everything to support multiple terminals though. Might have to fab something from PI-3s - but the old all-in-one terminals with built-in keyboards .... HeathKit used to sell some nice ones. Used one, with an old phone modem, to connect to Compuserve Forums back in the day. No PC required. "ATTD ..."
[toc] | [prev] | [next] | [standalone]
| From | "Carlos E.R." <robin_listas@es.invalid> |
|---|---|
| Date | 2026-08-25 14:11 +0200 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <qcd0mmxb3s.ln2@Telcontar.valinor> |
| In reply to | #90482 |
On 2026-08-25 06:33, c186282 wrote: > On 8/24/26 03:21, Carlos E.R. wrote: >> On 2026-08-24 07:43, c186282 wrote: >>> On 8/23/26 14:28, Lars Poulsen wrote: >>>> On 2026-08-23 06:31, The Natural Philosopher wrote: >>>>> On 23/08/2026 14:20, Lars Poulsen wrote: ... >> But those usb adapters can't do something the real rs232 or parallel >> port can do: directly interrupt and get processed within a known time, >> so much so as to sync clock to a GPS receiver. > > Quite true ... you're a slave to the USB interface. > Often that doesn't matter much - but SOMETIMES ... > >> I wonder if you can connect two machines via parallel port cable and >> emulated port. > > You could connect them, within reason, using the > older parallel ports - basically invent your own > 'network'. Done it. Had to xfer lots of stuff from > some old DECs to the newfangled PC. Xmodem sucked. > LOW speed though, but EVERYTHING was slow back in > the day. Serial ports worked. Parallel ports worked. > Lots of old detailed mag articles - search. The best > timeframe is a few years before/after the original > IBM-PCs. > > Now with USB emulators ??? Would not work as well. > Sometimes no substitute for those old direct interrupt > pins and UARTS. > > Hmmm ... come across any serial terminals ? I see a few > on E-Bay, but don't trust 'em to work. Linux/BSD does > have everything to support multiple terminals though. > Might have to fab something from PI-3s - but the old > all-in-one terminals with built-in keyboards .... > > HeathKit used to sell some nice ones. Used one, with > an old phone modem, to connect to Compuserve Forums > back in the day. No PC required. "ATTD ..." > Some boards do have a serial port interface, ie, the pins. -- Cheers, Carlos. ES🇪🇸, EU🇪🇺;
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-26 03:02 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <ZfqcnXJt_dRBEhP3nZ2dnZfqnPWdnZ2d@giganews.com> |
| In reply to | #90533 |
On 8/25/26 08:11, Carlos E.R. wrote: > On 2026-08-25 06:33, c186282 wrote: >> On 8/24/26 03:21, Carlos E.R. wrote: >>> On 2026-08-24 07:43, c186282 wrote: >>>> On 8/23/26 14:28, Lars Poulsen wrote: >>>>> On 2026-08-23 06:31, The Natural Philosopher wrote: >>>>>> On 23/08/2026 14:20, Lars Poulsen wrote: > > ... > >>> But those usb adapters can't do something the real rs232 or parallel >>> port can do: directly interrupt and get processed within a known >>> time, so much so as to sync clock to a GPS receiver. >> >> Quite true ... you're a slave to the USB interface. >> Often that doesn't matter much - but SOMETIMES ... >> >>> I wonder if you can connect two machines via parallel port cable and >>> emulated port. >> >> You could connect them, within reason, using the >> older parallel ports - basically invent your own >> 'network'. Done it. Had to xfer lots of stuff from >> some old DECs to the newfangled PC. Xmodem sucked. >> LOW speed though, but EVERYTHING was slow back in >> the day. Serial ports worked. Parallel ports worked. >> Lots of old detailed mag articles - search. The best >> timeframe is a few years before/after the original >> IBM-PCs. >> >> Now with USB emulators ??? Would not work as well. >> Sometimes no substitute for those old direct interrupt >> pins and UARTS. >> >> Hmmm ... come across any serial terminals ? I see a few >> on E-Bay, but don't trust 'em to work. Linux/BSD does >> have everything to support multiple terminals though. >> Might have to fab something from PI-3s - but the old >> all-in-one terminals with built-in keyboards .... >> >> HeathKit used to sell some nice ones. Used one, with >> an old phone modem, to connect to Compuserve Forums >> back in the day. No PC required. "ATTD ..." >> > > Some boards do have a serial port interface, ie, the pins. I think even PIs have a dedicated UART in there somewhere. However, with better CPU speed, 'soft' serial is more commonly used. But it's not QUITE the same as with real UARTs and their hardware interrupts.
[toc] | [prev] | [next] | [standalone]
| From | rbowman <bowman@montana.com> |
|---|---|
| Date | 2026-08-26 17:53 +0000 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <nf8nhnFprcuU5@mid.individual.net> |
| In reply to | #90600 |
On Wed, 26 Aug 2026 03:02:41 -0400, c186282 wrote: > I think even PIs have a dedicated UART in there somewhere. GPIO 14/15. The Picos have two. https://docs.micropython.org/en/latest/rp2/quickref.html#uart-serial-bus You can enable it in raspi-config.
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-27 02:57 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <G6-dnQJsRICUfRL3nZ2dnZfqnPadnZ2d@giganews.com> |
| In reply to | #90638 |
On 8/26/26 13:53, rbowman wrote: > On Wed, 26 Aug 2026 03:02:41 -0400, c186282 wrote: > >> I think even PIs have a dedicated UART in there somewhere. > > GPIO 14/15. The Picos have two. > > https://docs.micropython.org/en/latest/rp2/quickref.html#uart-serial-bus > > You can enable it in raspi-config. Checked. Not sure HOW WELL they act like the olde-tyme UARTS however. But, for MOST things, seems "well enough". Old UARTS ... direct dedicated interrupt lines and such. Lots of neat-o stuff was thus possible. The more 'evolution' of serial comms though, the fewer those neat-o options seem to become. Real UARTS become "software-emulated" UARTS. Gee ... try to find something to do the old 25-pin serial. There were lots of other functions you DON'T see for the new 9-pin emulations. Old serial comms methods ... they are becoming rather obsolete almost everywhere, except in certain 'industrial' applications. RS-485/RS-422 are STILL relevant. NOW it's mostly all TCP or at least CAN-bus. Fatter and far more complex. You don't ALWAYS need super-speed ... "slow and SURE" can be JUST as important.
[toc] | [prev] | [next] | [standalone]
| From | rbowman <bowman@montana.com> |
|---|---|
| Date | 2026-08-27 15:08 +0000 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <nfb283FprcuU21@mid.individual.net> |
| In reply to | #90671 |
On Thu, 27 Aug 2026 02:57:57 -0400, c186282 wrote: > Gee ... try to find something to do the old 25-pin serial. There were > lots of other functions you DON'T see for the new 9-pin emulations. RS-232 specified the DB-25 always was overkill for most serial connections. The printer port made better use since you didn't need all 36 pins on the Centronics connector end. https://www.analog.com/en/resources/technical-articles/fundamentals-of- rs232-serial-communications.html That page gives the full pin assignment. For a computer to computer cable TD, RD, and the ground works and that's reflected in the Pi and Pico's use of two GPIO lines. I supposed if you really wanted to you could implement the rest and write the software. https://microcontrollerslab.com/neo-6m-gps-module-raspberry-pi-pico- micropython/ I assume if you went digging into 'UART' you would find some of that PIO magic. The GPS does work as advertised.
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-24 01:06 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <ViCdnRdVJp6uTBb3nZ2dnZfqnPudnZ2d@giganews.com> |
| In reply to | #90396 |
On 8/23/26 09:31, The Natural Philosopher wrote: > On 23/08/2026 14:20, Lars Poulsen wrote: >> On 2026-08-23 04:32, Carlos E.R. wrote: >>> On 2026-08-23 08:57, c186282 wrote: >>>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>>> >>>>>> It goes against the grain, for me at least, to send perfectly good >>>>>> running hardware off to the recycler, just because it is >>>>>> "experienced." >>>>> >>>>> CPU generations (and entire architectures) get dropped from the Linux >>>>> kernel because they can’t find anybody to step forward and maintain >>>>> the support for them. >>>>> >>>>> That is how they tell that there are so few instances of those >>>>> particular kinds of hardware around, that essentially nobody cares >>>>> about them any more. >>>> >>>> Sad, but quite correct. >>>> >>>> Really not much call to run Linux-ish stuff >>>> on a 386/486 these days. How many people >>>> still USE those old boards for ANYTHING ? >>>> >>>> MAYBE a few "industrial" monitor/control >>>> systems ... but REALLY they oughtta upgrade. >>> > > However SOME won't because it might cost a >>>> few dollars. They'll run the old hardware >>>> until it literally bursts into flames - and >>>> then be SCREWED for days/weeks/months. But, >>>> the bean-counters will TELL them they're >>>> saving SO much money, now ....... >>> >>> Industrial computers may have boards, like data acquisition boards, >>> to interface with motors and sensors (or cameras). If you upgrade the >>> computer, it comes with a new operating system, so their expensive >>> software does not run on it. The cards probably can not be plugged >>> in, so they also need replacing; the new boards will probably not be >>> compatible with the expensive software. So finally they need to >>> recreate the software (hopefully the company that created it still >>> exists and has the source files). And maybe there are a bunch of >>> computers to replace. >>> >>> The price is no longer minor. >> >> What you have to do is make sure you have one or two backup hardware >> packages, because when the computer breaks, you won't be able to buy >> one that works. (Just try to find a motherboard with ISA slots today!) > > https://www.directindustry.com/industrial-manufacturer/isa- > motherboard-255106.html > >> And with custom plugin boards, it is kinda too late to have them >> fabbed when the system is already down. And fire it up once a year to >> make sure it still runs. >> >> That may not guarantee that you can recover, but it vastly increases >> your chances. >> >> And when you put the reserve system into service, it is time to >> expedite developing a replacement! >> >> For PC systems, the big hurdle is that old 16-bit systems allowed >> programs to do I/O to ISA bus peripherals without kernel programming, >> device drivers and OS hooks. When you "modernize" that ability is >> lost, and you have to rebuild with much more complicated programming. >> > That is a function of the operating system, not the chipset Correct. You don't HAVE to use Linux for such purposes. For really old stuff, a board that will run DOS may be quite adequate. There are also systems intentionally designed for the described purposes. OS-9 is still sold, for example, and is kinda-sorta "UNIX-ish".
[toc] | [prev] | [next] | [standalone]
| From | Anssi Saari <anssi.saari@usenet.mail.kapsi.fi> |
|---|---|
| Date | 2026-08-26 13:59 +0300 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <sm01pblkv09.fsf@lakka.kapsi.fi> |
| In reply to | #90396 |
The Natural Philosopher <tnp@invalid.invalid> writes: > https://www.directindustry.com/industrial-manufacturer/isa-motherboard-255106.html Wow. I had no idea, last time I saw a modern board with an ISA slot was around a couple of decades ago, when Pentium 4 was a thing. I mean sure, Intel renamed ISA to LPC and I think some common motherboard sensor chips use that even today but still. I guess what this endurance really shows is that the ISA bus was easy to interface to, so loads of proprietary interface/acquisition/control/whatever cards were done and those old designs live on. Probably dirt cheap to manufacture now too.
[toc] | [prev] | [next] | [standalone]
| From | The Natural Philosopher <tnp@invalid.invalid> |
|---|---|
| Date | 2026-08-26 13:53 +0100 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <116mnka$kn29$1@dont-email.me> |
| In reply to | #90628 |
On 26/08/2026 11:59, Anssi Saari wrote: > The Natural Philosopher <tnp@invalid.invalid> writes: > >> https://www.directindustry.com/industrial-manufacturer/isa-motherboard-255106.html > > Wow. I had no idea, last time I saw a modern board with an ISA slot was > around a couple of decades ago, when Pentium 4 was a thing. > > I mean sure, Intel renamed ISA to LPC and I think some common > motherboard sensor chips use that even today but still. I guess what > this endurance really shows is that the ISA bus was easy to interface > to, so loads of proprietary interface/acquisition/control/whatever cards > were done and those old designs live on. Probably dirt cheap to > manufacture now too. ISA was easy to talk to. But TBH i2C and USB are just as well supported -- "If you don’t read the news paper, you are un-informed. If you read the news paper, you are mis-informed." Mark Twain
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-27 01:28 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <G6-dnQFsRICeVhL3nZ2dnZfqnPadnZ2d@giganews.com> |
| In reply to | #90628 |
On 8/26/26 06:59, Anssi Saari wrote: > The Natural Philosopher <tnp@invalid.invalid> writes: > >> https://www.directindustry.com/industrial-manufacturer/isa-motherboard-255106.html > > Wow. I had no idea, last time I saw a modern board with an ISA slot was > around a couple of decades ago, when Pentium 4 was a thing. > > I mean sure, Intel renamed ISA to LPC and I think some common > motherboard sensor chips use that even today but still. I guess what > this endurance really shows is that the ISA bus was easy to interface > to, so loads of proprietary interface/acquisition/control/whatever cards > were done and those old designs live on. Probably dirt cheap to > manufacture now too. But nobody DOES make 'em. ISA was perfectly adequate for a LOT of plug-in boards, at least In The Day. However it was a BIG plug, and didn't have everything higher-speed boards might need - so it vanished. Many things people used ISA boards for, indeed many PCI(e) boards, are now on built-in chips. There's some little port on the side of yer lap or PI for that. No, NOT as good/versatile often, but for Joe Consumer "good enough" and the price is right. Ease and economics OFTEN beats "functional" and "versatile" alas. Hmm ... bought a couple SuperMicro mini-board computers - the use was to run a software router/firewall. Nice little units ! However on checking they had a LOT of really weird "industrial"-type connectors - some of which even I'd never heard of. Indeed 'super' boards for a vast variety of purposes. NOT sure you can get those anymore. DID look lately, but they seem to be DEEPLY buried, can't directly search for them. DID seem to be built/documented with Linux in mind however. The best one was the mini-board IN a little metal, dual fan cooled, box. JUST JUST enough room to add in a laptop HDD, now would be a SDD in the same profile. Used those. Very successful. Ran my whole office's net for years. Even wrote a little Python graphics pgm on an attached PI to monitor I/O and errors and such and show 'em on a little touch screen. You could know what/who was at fault at a single glance and, on touch, get more detailed info. Good Shit ! Made the world go 'round. The PI monitored both OUR side of the firewall AND the external ISPs 'modem' side ... figuring our WHICH was at fault quickly had long been a problem.
[toc] | [prev] | [next] | [standalone]
| From | "Carlos E.R." <robin_listas@es.invalid> |
|---|---|
| Date | 2026-08-27 08:32 +0200 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <m825mmx6b8.ln2@Telcontar.valinor> |
| In reply to | #90659 |
On 2026-08-27 07:28, c186282 wrote: > On 8/26/26 06:59, Anssi Saari wrote: >> The Natural Philosopher <tnp@invalid.invalid> writes: >> >>> https://www.directindustry.com/industrial-manufacturer/isa- >>> motherboard-255106.html >> >> Wow. I had no idea, last time I saw a modern board with an ISA slot was >> around a couple of decades ago, when Pentium 4 was a thing. >> >> I mean sure, Intel renamed ISA to LPC and I think some common >> motherboard sensor chips use that even today but still. I guess what >> this endurance really shows is that the ISA bus was easy to interface >> to, so loads of proprietary interface/acquisition/control/whatever cards >> were done and those old designs live on. Probably dirt cheap to >> manufacture now too. > > But nobody DOES make 'em. Did you see the link above? They are making today PC boards with ISA sockets. If that exists, it is probably that others are making ISA boards, so I googled, and they do. <https://www.impulse-embedded.co.uk/computing/boards/isa-sbc> <https://icop-shop.com/product-category/embeddedboard/halfsize/> <https://www.tindie.com/products/easypcb/isa-8-bit-prototype-board-pc-xt-8088-pcb/> <https://www.wdlsystems.com/embedded/sbc/custitem_ff_form_factor/ISA?page=1> <https://www.fieldworks.nl/industrial-automation/pc-based-i-o-board/isa-bus-i-o-boards.html> -- Cheers, Carlos. ES🇪🇸, EU🇪🇺;
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-24 01:01 -0400 |
| Subject | Re: Legacy computer in embedded industial systems (Re: Oh Gods ...) |
| Message-ID | <ViCdnRRVJp6-TRb3nZ2dnZfqnPudnZ2d@giganews.com> |
| In reply to | #90394 |
On 8/23/26 09:20, Lars Poulsen wrote: > On 2026-08-23 04:32, Carlos E.R. wrote: >> On 2026-08-23 08:57, c186282 wrote: >>> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>>> >>>>> It goes against the grain, for me at least, to send perfectly good >>>>> running hardware off to the recycler, just because it is >>>>> "experienced." >>>> >>>> CPU generations (and entire architectures) get dropped from the Linux >>>> kernel because they can’t find anybody to step forward and maintain >>>> the support for them. >>>> >>>> That is how they tell that there are so few instances of those >>>> particular kinds of hardware around, that essentially nobody cares >>>> about them any more. >>> >>> Sad, but quite correct. >>> >>> Really not much call to run Linux-ish stuff >>> on a 386/486 these days. How many people >>> still USE those old boards for ANYTHING ? >>> >>> MAYBE a few "industrial" monitor/control >>> systems ... but REALLY they oughtta upgrade. >> > > However SOME won't because it might cost a >>> few dollars. They'll run the old hardware >>> until it literally bursts into flames - and >>> then be SCREWED for days/weeks/months. But, >>> the bean-counters will TELL them they're >>> saving SO much money, now ....... >> >> Industrial computers may have boards, like data acquisition boards, to >> interface with motors and sensors (or cameras). If you upgrade the >> computer, it comes with a new operating system, so their expensive >> software does not run on it. The cards probably can not be plugged in, >> so they also need replacing; the new boards will probably not be >> compatible with the expensive software. So finally they need to >> recreate the software (hopefully the company that created it still >> exists and has the source files). And maybe there are a bunch of >> computers to replace. >> >> The price is no longer minor. > > What you have to do is make sure you have one or two backup hardware > packages, because when the computer breaks, you won't be able to buy one > that works. (Just try to find a motherboard with ISA slots today!) > And with custom plugin boards, it is kinda too late to have them fabbed > when the system is already down. And fire it up once a year to make sure > it still runs. > > That may not guarantee that you can recover, but it vastly increases > your chances. Well, your supply of back-ups won't last forever ... and a lot of control-oriented boards are only made for a year or two. Did used to use PC-104 industrial boards to run machines. However the maker would either drop the series or go out of business in a year or two. > And when you put the reserve system into service, it is time to expedite > developing a replacement! > > For PC systems, the big hurdle is that old 16-bit systems allowed > programs to do I/O to ISA bus peripherals without kernel programming, > device drivers and OS hooks. When you "modernize" that ability is lost, > and you have to rebuild with much more complicated programming. Again, there may be good use for "AI" in translating old software to modern langs/systems. Humans don't want to bother, but the AIs don't care.
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-23 23:19 -0400 |
| Subject | Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! |
| Message-ID | <ViCdnRtVJp66JRb3nZ2dnZfqnPudnZ2d@giganews.com> |
| In reply to | #90383 |
On 8/23/26 07:32, Carlos E.R. wrote: > On 2026-08-23 08:57, c186282 wrote: >> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>> >>>> It goes against the grain, for me at least, to send perfectly good >>>> running hardware off to the recycler, just because it is >>>> "experienced." >>> >>> CPU generations (and entire architectures) get dropped from the Linux >>> kernel because they can’t find anybody to step forward and maintain >>> the support for them. >>> >>> That is how they tell that there are so few instances of those >>> particular kinds of hardware around, that essentially nobody cares >>> about them any more. >> >> Sad, but quite correct. >> >> Really not much call to run Linux-ish stuff >> on a 386/486 these days. How many people >> still USE those old boards for ANYTHING ? >> >> MAYBE a few "industrial" monitor/control >> systems ... but REALLY they oughtta upgrade. > > > However SOME won't because it might cost a >> few dollars. They'll run the old hardware >> until it literally bursts into flames - and >> then be SCREWED for days/weeks/months. But, >> the bean-counters will TELL them they're >> saving SO much money, now ....... > > Industrial computers may have boards, like data acquisition boards, to > interface with motors and sensors (or cameras). If you upgrade the > computer, it comes with a new operating system, so their expensive > software does not run on it. The cards probably can not be plugged in, > so they also need replacing; the new boards will probably not be > compatible with the expensive software. So finally they need to recreate > the software (hopefully the company that created it still exists and has > the source files). And maybe there are a bunch of computers to replace. > > The price is no longer minor. Well, it's not the HARDWARE anymore. It's the SOFTWARE - and perhaps all the LEGAL protections/shields/contracts involved in modern biz ventures. As I suggested somewhere, porting old useful apps to more modern hardware/systems may be a good job for "AI". Humans don't care to do it, but AI just doesn't give shit. Used to have a PC-104 board, i386. It ran DOS. There was a toolkit. Used it to run/monitor a big machine. Nothing TOO complicated, some switches, a few sensors and timers, safety and hazard interlock stuff. Never really taxed the thing, just seemed a cool way to run the machine and it was a 'standard' industrial bus back then, so I did it. But what if 'the machine' was STILL being used for important stuff 25+ years later ? Just wait for the little board to self-destruct for a bunch of possible reasons, maybe be DEPRIVED of that highly profitable machine for a long time ? Even something cheap like an Ard2560 could probably do everything that old board did. A Pi or BBB with a cheap hat or two, fer sure. They are CHEAP hardware. But who will translate the SOFTWARE to them ? There is where the $$$ comes in. You'd think only one in a billion could translate some old BASIC-ish lang into Python or 'C' or some PCL - and you WILL get charged $$$$$$+ for that assumption. Heh heh ... long back one of our guys was a good STATS guy. He had a collection of old FORTRAN routines but no good/affordable way to port them to the original IBM-PC. A couple of us, mostly me, had to re-do them in IBM/M$ BASIC - and hand code, in DATA statements, the stuff to make the math processor do its thing. (the 8087 was like a CPU designed by autistic aliens) What a PAIN IN THE ASS ! Yea, a FORTRAN compiler DID come out pretty soon, but not quite THEN - and the company didn't wanna buy one anyhow. I'd still have been stuck porting HIS version of FORTRAN to the M$ compiler - always a lot of nuances. NOW I would not wanna do it ... easier/faster to write a whole NEW app instead. But those were the Do It Yourself days. So, an AI agent may be the better approach, leave the rip-off/bored humans out of the equation (I can say that since I'm retired now :-) Alt - find a way to hire/pay over-caffeinated 12 year olds to do it.
[toc] | [prev] | [next] | [standalone]
| From | The Natural Philosopher <tnp@invalid.invalid> |
|---|---|
| Date | 2026-08-23 12:46 +0100 |
| Subject | Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! |
| Message-ID | <116emic$1sako$3@dont-email.me> |
| In reply to | #90373 |
On 23/08/2026 07:57, c186282 wrote: > On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >> >>> It goes against the grain, for me at least, to send perfectly good >>> running hardware off to the recycler, just because it is >>> "experienced." >> >> CPU generations (and entire architectures) get dropped from the Linux >> kernel because they can’t find anybody to step forward and maintain >> the support for them. >> >> That is how they tell that there are so few instances of those >> particular kinds of hardware around, that essentially nobody cares >> about them any more. > > Sad, but quite correct. > > Really not much call to run Linux-ish stuff > on a 386/486 these days. How many people > still USE those old boards for ANYTHING ? > Not many. Some retro computer people keep them alive for interest, running the original OSs Some are in embedded applications still running DIS with turbopascal apps in them and no networking at all > MAYBE a few "industrial" monitor/control > systems ... but REALLY they oughtta upgrade. > If running well leave well alone. > However SOME won't because it might cost a > few dollars. They'll run the old hardware > until it literally bursts into flames - and > then be SCREWED for days/weeks/months. But, > the bean-counters will TELL them they're > saving SO much money, now ....... > Any sane business is run on cost benefit lines. As kit gets older the cost of maintaining it starts to rise, but the cost of replacing it also starts to be considered. At a given point you start allocating budget for inevitable replacement, get someone in to replicate the old code, run the systems in parallel and then once everything is hunky dory (that's English for copacetic) you switch the old system off. You might then store it a couple of years in case the new one has issues. > There was a case a few years back, Paris airport. > Suddenly the entire ground-control system went > down hard. Soon found, it had been running on > a Win-98 box for MANY years. Since it worked > nobody bothered at looking to modernize. Not > spending $10,000 probably cost them $10,000,000 :-) > > Now here is a case where a properly tuned "AI" > just MIGHT be of good use. Calculate and do what > the humans don't want to do anymore. What IS > involved in making an old app work on a 20 year > newer board ? Humans CAN do it - but don't WANT > to do it. The "AI"s don't care about any of that > and WILL offer/script solutions. > Big companies stumble on from crisis to crisis. Crises that would break a smaller one. Because big companies get infested with gold diggers, status seekers, arse coverers and dead wood. The system carries them, until it doesn't -- It is the folly of too many to mistake the echo of a London coffee-house for the voice of the kingdom. Jonathan Swift
[toc] | [prev] | [next] | [standalone]
| From | rbowman <bowman@montana.com> |
|---|---|
| Date | 2026-08-23 18:41 +0000 |
| Subject | Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! |
| Message-ID | <nf0t76F5rktU22@mid.individual.net> |
| In reply to | #90384 |
On Sun, 23 Aug 2026 12:46:20 +0100, The Natural Philosopher wrote: > Any sane business is run on cost benefit lines. As kit gets older the > cost of maintaining it starts to rise, but the cost of replacing it also > starts to be considered. At least in the area I'm familiar with Y2K was a watershed. IBM patched their current OS but it wouldn't run on the previous RS6000 systems. The sites looked at the cost of replacing the IBM gear versus PCs and rolled the dice. 911 (999) call centers aren't the best funded operations.
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-24 00:19 -0400 |
| Subject | Re: Oh Gods - Now MX Linux "Security Verification" - NO !!! |
| Message-ID | <ViCdnRpVJp6CWxb3nZ2dnZfqnPsAAAAA@giganews.com> |
| In reply to | #90384 |
On 8/23/26 07:46, The Natural Philosopher wrote: > On 23/08/2026 07:57, c186282 wrote: >> On 8/22/26 18:35, Lawrence D’Oliveiro wrote: >>> On Sat, 22 Aug 2026 08:36:42 -0400, TJ wrote: >>> >>>> It goes against the grain, for me at least, to send perfectly good >>>> running hardware off to the recycler, just because it is >>>> "experienced." >>> >>> CPU generations (and entire architectures) get dropped from the Linux >>> kernel because they can’t find anybody to step forward and maintain >>> the support for them. >>> >>> That is how they tell that there are so few instances of those >>> particular kinds of hardware around, that essentially nobody cares >>> about them any more. >> >> Sad, but quite correct. >> >> Really not much call to run Linux-ish stuff >> on a 386/486 these days. How many people >> still USE those old boards for ANYTHING ? >> > Not many. Some retro computer people keep them alive for interest, > running the original OSs > Some are in embedded applications still running DIS with turbopascal > apps in them and no networking at all I made an old 386 PC-104 board run a semi-complex machine. In my case the machine became obsolete before the board died - but that won't always be true for everybody. Huge $$$ might be lost. Yep, the software COULD pretty easily be moved to some new/modern board/language. Python on a PI or BBB would have served. But who DOES that ? A 'perception of rarity' has been cultivated, like only on in a billion can translate the sense of an old process-control system to anything newer. Thus the PRICE of doing so becomes prohibitive. >> MAYBE a few "industrial" monitor/control >> systems ... but REALLY they oughtta upgrade. >> > If running well leave well alone. BUT ... if yer highly-profitable machine running on the old board STOPS WORKING because the ancient board has died then you LOSE $$$$$$ ! As such, there may indeed be a motive to upgrade periodically. Even an Ard2560 or BBB or PI, with a hat or two, could do everything that PC-104 was doing. The hardware is CHEAP. It's the effort to move the practical sensibility of the old software over to a new system/lang. You'll get RIPPED OFF for that now (unless you have an in-house geek). >> However SOME won't because it might cost a >> few dollars. They'll run the old hardware >> until it literally bursts into flames - and >> then be SCREWED for days/weeks/months. But, >> the bean-counters will TELL them they're >> saving SO much money, now ....... >> > Any sane business is run on cost benefit lines. As kit gets older the > cost of maintaining it starts to rise, but the cost of replacing it also > starts to be considered. Sounds so reasonable ... but, depending, such delays might cost you yer entire biz. If YOU suddenly can't make Polyvinyl Acetate then everybody will shift over to some OTHER maker - and probably stick with them. Bean-counters should NEVER run a business. > At a given point you start allocating budget for inevitable replacement, > get someone in to replicate the old code, run the systems in parallel > and then once everything is hunky dory (that's English for copacetic) > you switch the old system off. > > You might then store it a couple of years in case the new one has issues. > >> There was a case a few years back, Paris airport. >> Suddenly the entire ground-control system went >> down hard. Soon found, it had been running on >> a Win-98 box for MANY years. Since it worked >> nobody bothered at looking to modernize. Not >> spending $10,000 probably cost them $10,000,000 :-) >> >> Now here is a case where a properly tuned "AI" >> just MIGHT be of good use. Calculate and do what >> the humans don't want to do anymore. What IS >> involved in making an old app work on a 20 year >> newer board ? Humans CAN do it - but don't WANT >> to do it. The "AI"s don't care about any of that >> and WILL offer/script solutions. >> > Big companies stumble on from crisis to crisis. Crises that would break > a smaller one. Because big companies get infested with gold diggers, > status seekers, arse coverers and dead wood. > > The system carries them, until it doesn't Yep ... Until It Doesn't. Oh well, the high-ups get their golden parachutes and WHO CARES about anyone/anything else eh ? :-) According to Machiavelli, the 'golden parachute' idea was born in old Rome. By insulating the high-up generals and such from the consequences of failure such individuals could afford to be BOLD. Apparently this tended to pay off on average. But SHOULD it be applied to mere "business" ?
[toc] | [prev] | [next] | [standalone]
| From | c186282 <c186282@nnada.net> |
|---|---|
| Date | 2026-08-21 21:54 -0400 |
| Message-ID | <Nd-dnZAOl66mnBT3nZ2dnZfqnPudnZ2d@giganews.com> |
| In reply to | #90271 |
On 8/21/26 08:22, TJ wrote: > On 2026-08-20 22:30, c186282 wrote: >> >> 32-bit versions of anything are RARE now alas ... > > Mageia 10, recently released, has an i686 version. It doesn't have > everything the 64-bit version has, because a lot of upstream support for > 32-bit has disappeared, but it IS supported. ANTIX has a 32-bit, but also warns that software that'll run on 32 is becoming more and more rare. I encounter 32-bit distros here and there, but it's gonna be the same problem. Still pissed Intel chips won't run by old 8/16 stuff anymore. Last one that'd do it was the Core2-Quad (have one :-)
[toc] | [prev] | [next] | [standalone]
Page 4 of 5 — ← Prev page 1 2 3 [4] 5 Next page →
Back to top | Article view | comp.os.linux.misc
csiph-web