User Plea Means EISA Support Not Removed From Linux

jones_supa writes A patch was proposed to the Linux Kernel Mailing List to drop support for the old EISA bus. However a user chimed in: "Well, I'd like to keep my x86 box up and alive, to support EISA FDDI equipment I maintain if nothing else — which in particular means the current head version of Linux, not some ancient branch." Linus Torvalds was friendly about the case: "So if we actually have a user, and it works, then no, we're not removing EISA support. It's not like it hurts us or is in some way fundamentally broken, like the old i386 code was (i386 kernel page fault semantics really were broken, and the lack of some instructions made it more painful to maintain than needed — not like EISA at all, which is just a pure add-on on the side)." In addition to Intel 80386, recent years have also seen MCA bus support being removed from the kernel. Linux generally strives to keep support even for crusty hardware if there provably is still user(s) of the particular gear.

Comments

Crusty Hardware

[Anonymous Coward]

I find it hard to believe that anyone is using EISA still. It got almost no traction in desktops and the only systems that ever had EISA slots were 386-486 era servers before the VL-BUS and PCI bus started to gain traction in late 486's.

If someone actually has a working EISA system, I'd like to see a photo. I had never managed to see more than one of these systems in my lifetime, and only saw one because it was being replaced in 1997 by a Pentium desktop.

I've actually seen more MCA systems than I've ever seen EISA.

Re:Crusty Hardware

[Penguinisto]

I'm surprised to see someone still using it myself (I've seen a few of them in the distant past, though...)

On the other hand, this is some hella stark contrast to certain other OS makers, who go out of their way to dump support for something as soon as they can in order to keep you on that upgrade treadmill - even if it means being forced to buy new hardware.

Re:Crusty Hardware

[__aaclcg7560]

My first tech job ten years ago was a Token Ring to Ethernet network conversion at a local branch office for a Wall Street firm. That was the first and last time I saw Token Ring in the wilds. Some Wall Street firms today are still using Token Ring.

Re:Crusty Hardware

[Carewolf]

I find it hard to believe that anyone is using EISA still. It got almost no traction in desktops and the only systems that ever had EISA slots were 386-486 era servers before the VL-BUS and PCI bus started to gain traction in late 486's.

If someone actually has a working EISA system, I'd like to see a photo. I had never managed to see more than one of these systems in my lifetime, and only saw one because it was being replaced in 1997 by a Pentium desktop.

I've actually seen more MCA systems than I've ever seen EISA.

EISA was parallel with VL-BUS for a long time, where consumer hardware used VL-BUS and enterprise or server hardware used EISA. PCI replaced both that wasn't until the mid 1990s. Still 20 years ago though.

Re:Crusty Hardware

[ArcadeMan]

On the other hand, this is some hella stark contrast to certain other OS makers, who go out of their way to dump support for something as soon as they can in order to keep you on that upgrade treadmill - even if it means being forced to buy new hardware.

You're talking about IBM's OS/2, right?

Sent from my PowerPC Mac mini which cannot run Yosemite.

Re:Crusty Hardware

[ArcadeMan]

I've seen a firm where they still use a Tolkien Ring. It's their most prized possession.

Re:Crusty Hardware

[Penguinisto]

I meant both Apple and Microsoft (albeit to a slightly lesser extent for Microsoft, though they seem to be embracing what Apple has going in that department.)

Re:Crusty Hardware

[Anonymous Coward]

Same here.

Re: Crusty Hardware

[Anonymous Coward]

If I remember correctly there were motherboards designed for industrial control systems that used EISA. They supported I think atleast pentium 3 CPUs. Though it's been a while so my memory could be wrong.

Re:Crusty Hardware

[MadKeithV]

I've seen a firm where they still use a Tolkien Ring. It's their most prized possession.

I'm sure you meant to say their most precious possession.

Re:Crusty Hardware

[cheesybagel]

Vesa local bus was kind of like AGP in that people mostly used it for graphics cards. PCs back then still had ISA slots for everything else.

Since these couldn't autoconfigure back then you had to mess with IRQs and shit. I hated that after having come from the Amiga. Thank God they fixed that mess.

Re:Crusty Hardware

[hughbar]

This way, sanity lies too, the resource cost of manufacturing new PC is enormous: http://en.wikipedia.org/wiki/G... so doing it just to put money in the pockets of Microsoft, Apple etc. is an aberration.

I recently changed my desktop but it was about 10 years old and I've given it to a recycling company. I've been able to do that because I'm a Linux user, it would have worked for BSD etc. too.

Re:Crusty Hardware

[HBI]

AC, the Compaq Deskpro line implemented EISA. While these systems rarely saw use in homes, banks were full of the things as were other firms. Deskpros were built like tanks up until about 1996. I installed lots of EISA NICs, in particular.

My point is that there were more of these systems out there than you think.

Re:Crusty Hardware

[satsuke]

Not entirely true ..

Back in those days, you didn't have serial/parallel, floppy, or IDE/disk controllers built into the motherboard,. At least not on non-Dell/Gateway/Micron/IBM computers.

All of those functions were built onto a VESA slot board alongside the video card.

You _could_ have that stuff on an ISA board, but you lost quite a bit of performance.

Re:Crusty Hardware

[cheesybagel]

People stopped using all that crap once the Pentium Triton chipset came out with PCI support. PCs back then still had some ISA slots but VLB and EISA basically vanished.

So does Linux have i486 support?

Re:Crusty Hardware

[Anonymous Coward]

and that is token ring with 1 ring per port on a switch?

Re:Crusty Hardware

[BitZtream]

So basically you want Apple and Microsoft to keep supporting your crusty old hardware with new features that it doesn't support anyway?

Both Apple and Microsoft provide 'support' longer than pretty much any other software maker on the planet, including any Linux vendor you can name.

Get a clue fanboy.

Re:Crusty Hardware

[HBI]

Bad form (replying to myself) but, I imagine many readers weren't even out of diapers when the last EISA system shipped. EISA slots supported ISA cards as well - the slots were dual use and had special keyed connectors that reached down lower in the slot when operating in 32-bit EISA mode. To an 8 or 16 bit ISA card, the slots were standard ISA slots for all intents and purposes, except that you had to run the EISA config utility (kind of like a BIOS config, but software rather than firmware) to get the system on board with what you were doing.

Re:Crusty Hardware

[morgauxo]

"I had never managed to see more than one of these systems in my lifetime"

When I was in college I worked for the campus (NPR, not student) radio station. I wasn't a DJ, did tech work there. Anyway.. that was around the same time you were replacing that EISA system.

We were on a tight budget (hey, it's public radio, what do you expect). During my time there we put the first computers in the studios as well as everyone's offices. Previously only a few managers had computers.

Anyway... to accomplish this we used to get old computers from other university departments. I'm talking real old, like original IBM ATs and knockoffs from the same era. We would hack the case with a nibbler and pop-rivet gun to fit a standard motherboard. We kept the power supply, floppy drives and sometimes they had usable hard drives. Yes.. we installed Windows 95 (and later 98) on MFM hard drives!

My boss, the engineer had a source for 486 motherboard, CPU + RAM combos. Depending on CPU speed and amount of RAM they were either $10 or $15! This all sounds horribly obsolete but at the time it was plenty good enough to run the then current web browsers, Office and email. Anything beyond that they probably weren't supposed to be doing anyway. In an era when new PCs were still $1500+ we were equiping people for under $50.

Well.. back to the point.. these motherboards had EISA slots and somehow the boss managed to get ahold of a handful of EISA video cards. Those computers that got the EISA cards did seem to run a lot faster!

So.. long story short.. I have seen more EISA cards than you. :-P

Re:Crusty Hardware

[Anonymous Coward]

The last time I saw an EISA machine was a Compaq Pro (before the HP-Compaq merger) back in the mid 1990s which sported both that and a VLB slot. This was a pretty expensive machine.

Other than that one line of machines, I've not seen EISA. Either there was a VLB slot, the box was all ISA... or eventually PCI came around and put the final nail into that coffin.

Of course, there was MCA, but by then, the cat was out of the bag and there were hundreds to thousands of white box makers [1], and since MCA was IBM only, the industry pretty much stayed with ISA with VLB as a workaround for faster video until PC makers agreed on PCI (and successors) as a bus standard.

[1]:

I do miss the days where you could buy a Computer Shopper mag, which was about the size of a newspaper and had 500-800 pages, with an insane number of white box PC companies. Pretty much most used the same computer case with the little square being different for each maker's logo. Even stick and brick PC shops were interesting, almost like mom and pop bike shops, with people stopping in for random conversations.

Re:Crusty Hardware

[tlhIngan]

Vesa local bus was kind of like AGP in that people mostly used it for graphics cards. PCs back then still had ISA slots for everything else.

Since these couldn't autoconfigure back then you had to mess with IRQs and shit. I hated that after having come from the Amiga. Thank God they fixed that mess.

VLB was an aberration. It basically was a direct link to the 486 processor - it was so heavily tied that the Pentium basically ended it (since it wasn't a 486 and used a new bus).

VLB was fast, but it was basically a stopgap - the 486 was so fast, ISA was so slow (but had a ton of peripherals) that there was no expansion bus technology available (I suppose they could use NuBus, which was TI (and used by Apple), but that was the wrong endian) as PCI was not quite done yet.

The Pentium basically introduced PCI.

AGP was an offshoot of PCI meant to offer higher bandwidth because PCI, despite having 66+MHz and 64 bit extensions, pretty much everything was tied to the 32-bit, 33MHz variant.

AGP appeared to the system as a regular PCI bus when enumerated (it basically WAS PCI, just signalled faster) and lasted until PCI Express came out. Interestingly, we've not exceeded the capabilities of PCIe yet as you can often use fewer lanes and not see a drop in graphics performance.

Re: Crusty Hardware

[Anonymous Coward]

So you could say it's precious to them?

Re:Crusty Hardware

[morgauxo]

I loved being able to set the IRQs and memory addresses. Ok, not really but the for the first decade or so I HATED plug and play, or plug and pray as we called it in my office. Half the time it didn't work! It would try to put things right on top of one another. When possible I would disable it and just use the jumpers. Once you got used to it it wasn't THAT hard and it was great knowing that your soundcard was on IRQ X and your video on IRQY and never the two would conflict. Early plug and play seemed to randomly decide to reshuffle things and the next time you boot it may not still work.

That was of course a long time ago and things work well on any reasonable hardware today. But.. I still cringe when people complain about setting IRQs. It's not hard to move a jumper and its not that complicated of a concept to know that to things probably shouldn't try to use the same resource. Not being able to set those things caused me far far more frustration than having to set them ever did! Even though things are great now I'm not entirely sure that decade of pain was worth it to get here.

Re:Crusty Hardware

[jeffmflanagan]

Token Ring was fun. You could convince noobs that they'd let the token out of an open connection, and ruined the network.

Re:Crusty Hardware

[Penguinisto]

So basically you want Apple and Microsoft to keep supporting your crusty old hardware with new features that it doesn't support anyway?

You're approaching it from the hobbyist/end-user viewpoint - turn in your geek card, please. The corporate/enterprise side of things will actually keep hardware around a whole hell of a lot longer, and industrial use cases keep old crap around the longest of all.

Example? No problem, I got a ton of those, including this little gem I dealt with a couple of years ago: Company spent millions on a certain specialized (solar cell) wafering machine whose computer still uses a parallel port (remember those?)/ It's a year or two out from ROI when it breaks down, but the manufacturer won't update or repair anything w/o the company spending millions on a new machine. Why? Because they stopped issuing patches/drivers for the machine long ago when Microsoft decided to drop their OS support, and the old stuff won't support USB enough to allow for a USB/Parport adapter.

This has fuck-all to do with fanboy ideology, and everything with having to keep systems up in situations where they need to.

Re:Crusty Hardware

[ArcadeMan]

I used a synonym on purpose so it wouldn't be too obvious.

Re:Crusty Hardware

[operagost]

I had an AMD 486DX-40 system with two VLB slots. I had both VLB graphics and multi-I/O. Before the Pentium era, a motherboard didn't usually have all the basic I/O on it, so you had to buy a card with serial, parallel, and usually an IDE controller on it. Anyway, the second most annoying thing with VLB is that it ran at your FSB speed, so if you had a 25 MHz FSB you could have three cards, 33 MHz you could have two. I had 40, so I was only supposed to have one. It was surprisingly stable, but when I had the occasional video glitch I understood why it happened and could deal. It got better when I dumped the Computer Shopper white-box special for a DX2/80 with a decent ASUS board.

The most annoying thing with VLB being, of course, that every card was almost full-length because the VLB part sat behind the ISA slot. They popped out all the time.

Re:Crusty Hardware

[Iniamyen]

Aren't they running Frodo?

Re:Crusty Hardware

[morgauxo]

What the fuck are you talking about?

I've watched my parents throw away perfectly good printer/scanner combos that were only a few years old because there were no drivers beyond XP.

I have dozens of network and video adapters on a shelf in my garage that work great in Linux but have no Windows drivers beyond XP.

Until recently even a 386 could run Linux!

Linux vendor? I wouldn't know. I've never used one. I can install my own software thank you!

Re:Crusty Hardware

[freeze128]

The EISA spec is a lot like the color implementation of NTSC video. It was something that was added on top of ISA so that it would work just fine with OLDER ISA-only cards. The EISA setup programs of the time were especially clunky. This was just before the widespread use of flashable BIOS ROMS on the motherboard, so there was a system partition on the hard drive that contained the program (think of a rudimentary EFI EPS partition) and the settings. One of the major features was that hardware settings for add-in cards could be configured via this setup partition rather than via jumpers or dipswitches.

When was the last time you set a jumper or dipswitch on your motherboard or add-in card?

Re:Crusty Hardware

[Anonymous Coward]

I owned an EISA system once: 486DX-50 (not DX/2) with an Adaptec 1742 EISA SCSI controller. At the time, that was a seriously fast system between having the fastest non-clock-multiplied 486 ever made and bus-mastering disk I/O on a full 32-bit bus. I ran both OS/2 and Linux (kernel 0.99pl12) on that box.

Yep, I'm getting old.

Re:Crusty Hardware

[Anonymous Coward]

It was really fun when you plugged the D Sub 9 Token Ring plug into a CGA video socket and brought down the entire network!

Re:Crusty Hardware

[__aaclcg7560]

Replacing the ESC button with a red PANIC button on the keyboard and convincing the noobs not to press that button under any circumstances was funnier.

Re:Crusty Hardware

[Anonymous Coward]

Arcnet was better - 1TB/s on line lengths up to 2000 miles, over barbed wire!

Re:Crusty Hardware

[FreonTrip]

Yes, for the time being i486 support's still maintained. Debian and Slackware both nominally support it, and I suppose a very patient person could also flog their 486 nearly to death with Gentoo. Though I'd think a sane person would cross-compile...

Re:Crusty Hardware

[FreonTrip]

You can still occasionally find a DIP switch on overclocker-friendly motherboards to ratchet up clock speeds and apply a corresponding voltage bump; the vagaries of that are handled by the BIOS/UEFI. But the only jumpers I ever see are for CMOS flashing, and maybe once in a blue moon to enable or disable an integrated component. It's definitely not 1995 any more.

Re:Crusty Hardware

[Anonymous Coward]

Plenty of 486 systems had EISA. VESA Local Bus was only useful for graphics adapters. PCI was way, way late in the 486 era -- by that time, people were using Pentiums.

Re:Crusty Hardware

[jandrese]

ISA Plug and Play was a horrendous hack. It was always a roll of the dice. It's really no surprise that people were so willing to jump on PCI as soon as it came out.

Re: Crusty Hardware

[Anonymous Coward]

I really loved Arcnet at the time. I did plenty of installations. It's certainly not missed though.

Re:Crusty Hardware

[FreonTrip]

That's CMOS resetting, my bad. *cough*

Re: Crusty Hardware

[Bigbuzzman]

Starting around VLB time I started to run into computers with the cards hot glued to the sockets.

Re: Crusty Hardware

[dunkelfalke]

I had a dual pentium motherboard with eisa slots once, made by Asus iirc. Used to sit in a server of a German content management software company.

Re:Crusty Hardware

[sconeu]

Back in the early to mid-90's, I admined a 486/33 (no bloody SX/DX suffix). It ran SCO ODT 2.0, and functioned as a departmental development server, supporting about 20 devs.

It had an Adaptec 1740 EISA SCSI-2 adapter, and an early Diamond S3(?) video card.

I *LOVED* that box, and wish I had tried to purchase it after they removed the classified media and decommissioned it. The thing was a frickin' tank, with a steel enclosure.

Re:Crusty Hardware

[Anonymous Coward]

> the manufacturer won't update or repair anything w/o the company spending millions on a new machine

Intellectually dishonest much? That's obviously a problem with the manufacturer, not the software. The same thing would've happened with an equally old RHEL or other Linux distro from that era, except it would've happened much, much sooner than with "certain other OS makers".

Re:Crusty Hardware

[armanox]

Apple and Microsoft are far behind in terms of support from what I see.

SGI IRIX 6.5: 1998-2014
Solaris 10: 2005-2021
Solaris 11: 2011-2024
If I want long term support, I know I'm not going for a Windows systems or a Linux box. Know where to go for it (I'm sure that IBM and HP have some long lifecycles too....but I'm less versed in AIX, i, HP-UX, and OpenVMS)

Re:Crusty Hardware

[ZorinLynx]

I remember VIP 486 boards. VESA-LB, ISA and PCI in one motherboard. Very awesome if you had a really expensive VLB video card that you didn't want to part with quite yet, but wanted to upgrade to a system with PCI. :)

Re: Crusty Hardware

[ZorinLynx]

The VLB card connector was really long and would sometimes make a poor connection or start to pop out. Hot glue was a common hack to make sure the card stayed in, and it was reversible (just heat the glue and pop it out).

It truly was a dark age. But we all loved it. :)

Re:Crusty Hardware

[ZorinLynx]

That was the reliable way to run the DX50, because VLB was unreliable at 50MHz. You were lucky to be able to afford something so kick-ass back then!

I bet you ran DOOM at 60FPS! ;)

Re:Crusty Hardware

[NJRoadfan]

I somehow managed to land up with 3 EISA/VLBus 486 motherboards. They appear to be workstation class boards as 2 of them have 16(!) 30pin SIMM slots and take 256MB of RAM. The hardware is out there, but its fairly uncommon.The latest boards I have seen with EISA were 440FX based Pentium Pro and Pentium II boards (almost always SMP server boards). Any support for it seems to have been dropped starting with the 440LX chipset, so PCI effectively killed it off. The latest motherboard made with EISA dates to around 1996-97, so either way that guy is running some old hardware. I wonder what is special about that FDDI card though?

Re:Crusty Hardware

[__aaclcg7560]

On the Token Ring to EtherNet job I did, the fresh out of high school techs plugged the Ethernet cable into the Token Ring card (which took both coaxial and twisted-pair cables) and didn't bother to run the network TV app to catch their error on 200 PCs. I made an extra four hours in OT pay because the project manager let them go without checking their work.

Re:Crusty Hardware

[MightyYar]

I've been able to do that because I'm a Linux user, it would have worked for BSD etc. too.

I have a 10-year-old Mac running an older version of MacOS. I'd still have an 8-year old XP machine if the hard drive hadn't crapped out. It's nice that Linux continues to be updated for older hardware, but you can certainly keep an old Mac/PC running for a long time.

Re:Crusty Hardware

[jcdr]

There is still embedded systems running Linux that use ISA bus. My company maintain one such system that run every day in a public transportation system. At the time (around 2005) the system was designed (by a now defunct company), it was a fast project that reused schematics from previous successful projects to lower the risk.

You can find a photo of the system CPU board in this manual: support.elmark.com.pl/advantech/pdf/SOM-4481man.pdf

Re:Crusty Hardware

[David_Hart]

What the fuck are you talking about?

I've watched my parents throw away perfectly good printer/scanner combos that were only a few years old because there were no drivers beyond XP.

I have dozens of network and video adapters on a shelf in my garage that work great in Linux but have no Windows drivers beyond XP.

Until recently even a 386 could run Linux!

Linux vendor? I wouldn't know. I've never used one. I can install my own software thank you!

What are you talking about? You do realize that Microsoft has no control over vendor hardware or their creation of drivers? It's not Microsoft's fault if the vendor dropped support.

I agree that it's annoying when vendors dropped the ball on developing drivers for the new driver model in Vista/Win 7. That being said, there are tricks that allow you to install the Windows NT drivers for older hardware on Windows 7, 8, 8.1. Most older hardware, that had XP drivers, also had NT drivers.

Re:Crusty Hardware

[hackertourist]

Meanwhile in that era, us Apple users plugged in & played with our NuBus cards without ever having to fiddle with IRQs. Doing things right the first time saved us a lot of aggravation.

(Get off my lawn, etc. )

Re:Crusty Hardware

[Anonymous Coward]

Same here, same system, late 80s or early 90s IIRC. I remember running Yggdrasil and Slackware on it among others -- and hacking the kernel to make my CD drive work correctly. Fond memory. And I am not feeling old at all.

Re:Crusty Hardware

[cyberspittle]

I have a NICE Super EISA with plenty of EISA cards in it. 2MB ATI Mach32 EISA. 3COM 10/100 EISA, Adaptec EISA, additional Mach64 on the 1 VLB slot. RAM maxed out to 256MBs (16 16MB SIMMS) with interleave enabled. It was meant to run Netware 3.12, but I used it for IBM OS/2 Warp (version 3), until I downloaded original Slackware floppy images via the ISA 33.6 MODEM (in EISA lost). Have a Crystal Semi sound card (ISA in EISA slot). CPU been upgraded to a Trinity Stacker 133Mhz from the original 486DX2/66, which was upgraded from a 486SX25. I keep this old beast in a full AT tower (has the large L-shaped power supply).

Re:Crusty Hardware

[Anonymous Coward]

You had hardware manufacturers manufacturing for a 1% marketshare back then?

That's probably why your PnP worked so well back then. You could only install like 1 other thing. LOL

Re:Crusty Hardware

[JackieBrown]

I think what he's saying is that linux manages to retain support for these while Windows 7,8,8.1 needs active vendor support to keep it working

Re:Crusty Hardware

[hughbar]

Yes, agree, but the problem also with XP is end-of-life. Shame because it was one of their best ones too!

Re:Crusty Hardware

[MightyYar]

EOL doesn't really have any practical effect on a 10-year-old system, though. I mean, unless you have it hooked directly to the internet or use IE. I don't think even a totally unpatched XP would be a problem if you stuck to an alternative browser and surfed behind a NAT. Pehaps in a few more years when even software stops supporting XP.... that's what made me leave 98SE, and then later 2000. I only dumped XP because once the hard drive went, I couldn't really justify putting money into such an old system when I could get all-new guts for under $200. And once you have new guts you really need to reinstall Windows. And once you are doing that, why not use 7? Now I'm good for the next 10 years :)

Re:Crusty Hardware

[sjames]

MS set up the deal between themselves and the hardware vendors. Clearly long term support wasn't a priority.

Linux has a different arrangement and makes it work.

Re:Crusty Hardware

[sjames]

The real frustration of setting jumpers was that too many cards and drivers only had a couple options. You knew very well they could support any arbitrary IRQ but simply didn't do it.

Plug'N'Pray was obnoxious since it rarely worked right (or at all) and often offered no manual fallback.

Re:Crusty Hardware

[lowen]

At least one HP Proliant dual Pentium (Pentium I) had EISA. Some DEC AlphaServers had EISA.

Re:Crusty Hardware

[morgauxo]

If you were on a budget and the upgrade to PCI had to come in stages that just meant you had a mixed system. Now you had plug and play PCI devices that couldn't understand why some resources were just not available since they were already taken by the old ISA hardware. That started to improve once BIOSs got smart enough to let you block off resources from the PnP pool.

Re:Crusty Hardware

[ArcadeMan]

Things like scanners and printers shouldn't require special drivers in 2015. When we plug a keyboard or mouse into our computers, it just works because they're standard devices with standard drivers.

Re:Crusty Hardware

[unixisc]

I had that problem once, and got solved by XP-mode in Windows 7

Re:Crusty Hardware

[unixisc]

Wasn't EISA a substitute for ISA on servers, where presumably 32-bit instead of 16-bit buses were needed?

Re:Crusty Hardware

[qubezz]

This is also completely Microsoft's fault. In Vista they decided to kiss the ass of big media companies in order to play Blu-Ray content, which required encrypted end-to-end data transport, mandating the rewriting of the driver stack for everything from video and sound cards to imaging devices and audio mixing. They should have just given them the finger.

What Microsoft didn't have to do was just completely discard gameport support. Microsoft blatantly removed the code to support 15 pin gameports from the OS. In Vista 32 bit, it could be partially put back by driver hacks of old dlls, but that hack was made impossible in win7. You could literally buy joysticks at the same CompUSA that would not work on the Vista shitboxes they were selling.

Re:Crusty Hardware

[allaunjsilverfox2]

Things like scanners and printers shouldn't require special drivers in 2015. When we plug a keyboard or mouse into our computers, it just works because they're standard devices with standard drivers.

I would argue a LOT of the Wal-Mart specials don't run without driver support. And before you say, "Then don't buy from Wal-Mart!", Your conjecture is all printers should run without intervention. This is simply not the case. Big box stores provide brands that prefer you have vendor lock in. They can't force you to view the, BUY X BRAND INK NOW! ads without the bundled driver.

Re:Crusty Hardware

[phantomfive]

Exactly. There is no reason to replace hardware that works when the newer product will cost money and perform no better (and could even perform worse).

Re:Crusty Hardware

[qubezz]

I specified and owned an EISA system, a rare 486-50 (not double-clocked DX2), with 16MB memory, $4000 or so spent.

EISA is a very odd beast, if you recall the original ISA bus that had jumpers you had to set on each card to non-conflicting IRQ, Address IO, and DMA values, then you will see the "brilliance" of EISA, which had a floppy disk config program for every card you bought to set the bus values. Seeing anyone that still has the matching and required EISA setup disks for their hardware is going to be the rare thing to find.

Re:Crusty Hardware

[__aaclcg7560]

I dumped the Computer Shopper white-box special for a DX2/80 with a decent ASUS board.

Most white-box specials had biege cases during the era. Non-biege cases came much later. My very first PC-AT came in an green industrial case that bolted to the floor and weighed 80 pounds. That was a conversation piece.

Re:Crusty Hardware

[arglebargle_xiv]

I've watched my parents throw away perfectly good printer/scanner combos that were only a few years old because there were no drivers beyond XP.

It's not so much printer/scanners as standalone scanners, it seems to be a requirement that a model of a scanner introduced during the heyday of a particular OS never gets updated drivers beyond that OS. Canon are particularly bad in this regard, if you look at your favourite online auction site you can find Canon scanners going by the trailerload because the only way to move to a new OS is to buy a new scanner (after carefully checking that its drivers work with the newer OS, which isn't guaranteed even with a brand-new scanner).The one notable exception to this is Epson, you can take a scanner you bought ten-plus years ago and download the latest Windows 8.1 drivers for it.

In general, printer/scanner combos tend to be better-supported, but for standalone scanners it seems to be a variation of the printer-cartridge planned-obsolescence scam.

Re:Crusty Hardware

[Anonymous Coward]

Printers have PCL and/or PostScript. Scanners have TWAIN. If your printer or scanner don't support those, then why did you buy it?

PCL and PS and TWAIN allow these devices to work in predictable and interchangeable ways. They work exactly in the manner you describe.

Re:Crusty Hardware

[maestroX]

parallel still available, http://www.newegg.com/Product/... or try ebay.

Re:Crusty Hardware

[maestroX]

I've watched my parents throw away perfectly good printer/scanner combos that were only a few years old because there were no drivers beyond XP.

In Windows 7, run driver setup with compatibility settings to WinXP and as Administrator.
This usually works.
Or get VueScan http://www.hamrick.com/.

Re:Crusty Hardware

[Grog6]

With a "good" (lol) mobo, I was able to run my VLB board at 50MHz, with certain combinations of cards and memory.

I ran a 486-DX4-100 for a couple of years with that mobo, a SCSI card, an Orchid Fahrenheit video card, with a 2x multiplier and a 50MHz fsb.

I still have this somewhere, lol.

It got replaced with a TX430 mobo, lol.

Re:Crusty Hardware

[Anonymous Coward]

Printers have PCL and/or PostScript. Scanners have TWAIN.

But... I thought... never the TWAIN shall meet?

Re:Crusty Hardware

[Grog6]

Yes, it was primarily for high bandwidth communications; SCSI, Fiber, etc.

DataAq was secondary, but the cards I have have a 50pS jitter spec, and it holds across four cards in one box; try buying that for less than $20k today.

The best cards I see these days won't hold that jitter spec; only systems.

Re:Crusty Hardware

[Bert64]

This is one of the inherent problems of being closed source, support for niche and older hardware will be lacking.
Microsoft have to break compatibility from time to time in order to progress, either due to hardware changes (64bit) or software changes (new video driver stack etc).

The problem is that with closed drivers, only the original authors of those drivers can change them and hardware manufacturers have little or no incentive to continue supporting old hardware as they want to sell you new kit. With open source drivers, all it takes is for one of the users to either update it themselves, or hire someone capable of doing so. In some cases updates aren't even necessary, eg a lot of linux drivers written for 32bit x86 compile just fine on 64bit or even other architectures like ARM.

The same is true of niche hardware, a lot of hardware was intended by the manufacturer to be connected to x86 systems but uses standard cross platform buses like pci or usb... While the number of people using linux on ppc, alpha, sparc or arm etc might be too low for the manufacturer to bother providing official support, the drivers will often just work. I used to run an alphastation on linux with all manner of pci and usb devices which were never intended to be used on alpha based hardware.

Re:Crusty Hardware

[Bert64]

Well you get what you pay for when buying such devices...
There are standards for printers, scanners and various other hardware. I wouldn't ever buy a printer which didn't support Postscript, and i never install the official drivers as they're often extremely bloated and probably full of ads. Sure printers which support postscript generally cost more, but they're usually higher quality, older ones are still available cheaply and the toner/ink is likely to remain available for far longer.

Re:Crusty Hardware

[Anonymous Coward]

Come wander round the cheap end of industry sometime. I just narrowly avoided buying a CMM which had a 486 era control pc still quietly running its bespoke software that relied on bespoke hardware cards on the bus (don't know details to know if theyre eisa or other nor what OS was underneath), and one of my other machines uses a PC98 japanse computer running some weird ported version of DOS 6.0 and a completely different bus standard (C-BUS). I would kill for that to take the relative sanity and easier to find EISA cards...
One of my other machines is running linuxcnc, and that had a deskpro with eisa in it before I revamped it this year with a shuttle media pc inside the cabinet.
You don't scrap a couple of ton of rigid machine tool that can still do a good job because its married to some obsolete control computer. There's plenty of reasons for them to still be around, in all sorts of odd places.

Re:Crusty Hardware

[Anonymous Coward]

Get out.

Re:Crusty Hardware

[drinkypoo]

In Windows 7, run driver setup with compatibility settings to WinXP and as Administrator.
This usually works.

This usually doesn't work

Or get VueScan http://www.hamrick.com/.

they want $90 for slide scanning, or I can scan under Linux for free

Re:Crusty Hardware

[drinkypoo]

Half the time it didn't work! It would try to put things right on top of one another.

Yeah, that was because ISA register sharing was supposed to be working by then, but it wasn't. PCI register sharing didn't even really work reliably until after PCI-E had already come out, it's pretty bulletproof now but it still caused awful performance degradation in the early days and having to force your SCSI card and your GPU onto different IRQs to get peak performance was definitely a thing.

Even though things are great now I'm not entirely sure that decade of pain was worth it to get here.

To be fair, the Macintosh has had working autoconfiguration for a very long time. The Amiga did, too.

Re: Crusty Hardware

[drinkypoo]

The VLB card connector was really long and would sometimes make a poor connection or start to pop out.

VLB controllers were often just garbage. There was really no point to having more than two VLB slots and little point to having more than one, as often you could get your storage controller or your video card working but not both.

Re:Crusty Hardware

[drinkypoo]

That's CMOS resetting, my bad. *cough*

Some boards don't even have that, you have to pull a battery. But some boards actually have a flash enable jumper still, which is a pretty cool feature and a good thing to wire to a keyswitch or turbo button in a vintage case install.

Re:Crusty Hardware

[P0ck3tR4wk3t]

PCL and PostScript are interpreted VERY differently by EVERY printer manufacturer on the planet. Every printer out there has their own additions to the language. All of their own hints and crap. People make LOTS of money selling custom page counting software that accounts for these differences. Their drivers are hand-rolled and do all sorts of crazy stuff to ensure that their driver is used with THEIR printer.

Re:Crusty Hardware

[Anonymous Coward]

You said Tolkien and ring, you muppet. It can't get any more obvious.

Re:Crusty Hardware

[Ben Hutchings]

Debian accidentally dropped support for 486 by turning on kernel stack smashing protection, which uses RDTSC. This was done before Debian 6.0 (squeeze) but we only recently received a report that it failed to boot on a 486.

Re:Crusty Hardware

[Anonymous Coward]

Parallel ports will continue to be available until at least PCIe EOLs. People sell PCIe parallel cards, which are usually PCI parallel chipsets hanging off PCIe-PCI bridges.

When PCIe dies, someboday will have a DOSBox or QEMU IO breakout box on the market.

There are companies which sell PDP11 emulators which are just Intel i3 machines with QBUS slots. It's unlikely these things will ever go away, at least until the fleet of Fission reactors using PDP11, VAX and who knows what else is retired. Simple machines like turret mills can have their PDPs and VAXen stripped out and replaced with Mach3 or whatever; replacing complex industrial plant controllers is much, much more expensive than buying a new emulator.

Re:Crusty Hardware

[Anonymous Coward]

You _could_ have that stuff on an ISA board, but you lost quite a bit of performance.

Yea, but on every non-PCI 486 I ever worked on, those things were attached to 16-bit ISA slots, damn the performance. The only VLB cards I've ever seen were video cards. I'm sure other cards existed, but they weren't common.

Re:Crusty Hardware

[meta-monkey]

My favorite back in 386 days was remapping the ANSI definitions when one of the other kids in my programming class wasn't looking so when he tried to press the backspace key it would print "DELETION DENIED!!!!!"

Re:Crusty Hardware

[david_thornley]

Several years back, I had a network printer that was getting flaky, and would no longer talk to a computer by anything but Appletalk. This was about the time Apple was dropping it. Fortunately, I had a Linux box that could still use it.

Re:Crusty Hardware

[allston]

That is not completely right, look up support pack either for the audigy or live sound cards. I have a SB Live 5.1 and a Microsoft Sidewinder Force FeedBack Pro joystick. They both work great in Windows 7.

Re:Crusty Hardware

[kjs3]

I've probably had ~1000 EISA machines pass new through my hands in the old days, but almost exclusively servers, and mostly Pentium. There were also a number of Alpha-based machines that used EISA, and the HP 9000 HP-PA-based machines used it. I *think* I've seen a MIPS-based EISA machine from back when we thought we'd be running WinNT on MIPS. Good times. For a relatively short period of time, it was extremely popular at the high end. It was never really direct competition for VL-Bus, which was more "we need more bandwidth for consumer level video and PCI isn't here yet" than "new bus arch for high performance computing". I don't have one now, but did within the last 5 years. FWIW...EISA is why PCI isn't as closed and license heavy as MCA.

Re:Crusty Hardware

[toddestan]

Well, at least you don't have to deal with licensing issues. Microsoft won't sell you a license for older versions of Windows (such as XP) so you're stuck dealing with old hardware so you can use the license that's attached to it.

Re:Crusty Hardware

[toddestan]

Actually, you would probably need SP2 and likely SP3. Support for pre-SP2 is pretty spotty nowadays, and a lot of stuff expects SP3. Also, I would stay well away from Flash or Java since that's the most likely way you'll get pwn3d even if you decided to use IE anyway.

Re:Crusty Hardware

[Anonymous Coward]

Dipshit

You can still run that old hardware with any modern distro.

The end of an era.

[Anonymous Coward]

"Linus Torvalds was friendly about the case"

They did it, they neutered him.

Re:The end of an era.

[Anonymous Coward]

"Linus Torvalds was friendly about the case"

They did it, they neutered him.

Torvalds is in general very reasonable. It's just when people push him with unreasonable requests that he bites back hard.
Like when that guy wanted the kernel to be ported to C++ for no other reason than object orientation being the fad of the year.
That was hilarious.

Re:The end of an era.

[Penguinisto]

They did it, they neutered him.

Bullshit. He's harsh on coders who fuck up (and rightfully so), but I have never seen him unleash the Kraken on any reasonable user request.

Re:The end of an era.

[ArcadeMan]

Godamn Kraken, all I want is to reach Selbina in peace.

And when it's not a Kraken it's godamn sea pirates with skeletons.

Re: The end of an era.

[Anonymous Coward]

Stay below deck

The kraken can solo on rdm45 or so

Re:The end of an era.

[Anonymous Coward]

I could use a shot of The Kraken.

Re: The end of an era.

[ArcadeMan]

I miss my cute little Mithra. :(

Re:The end of an era.

[Anonymous Coward]

who what?!

Re:The end of an era.

[ZorinLynx]

They're referring to the recent drama where Social Justice Warriors are pissed at Linus because he said he "didn't care about diversity."

What he meant is that he doesn't care who the code comes from, just that it's good code. Which is the right way to do things.

Re:The end of an era.

[Anonymous Coward]

the kernel to be ported to C++ for no other reason than object orientation being the fad of the year.

There is quite a bit of object oriented C code in the wild, that request just shows that the guy had no idea what he was talking about. While there are a lot of nice (and not so nice) things you can get from C++ if object orientation is all you want its overkill.

Re:The end of an era.

[gweihir]

Engineering is in general gender/race/whatever-neutral. For code, it is possible to judge the merit of it without even knowing who or what produced it. That gives any good engineer a high-level of BS resilience and Linus is just one that acts it out on those trying to feed him BS. Nothing wrong with that, his work has merit in spades, he is able to communicate well and he has no problems recognizing and appreciating good work by others. It is really not hard to see why his pet-project became huge.

Re:The end of an era.

[meta-monkey]

Also, he doesn't get salty until they've refused to fix or even acknowledge their error multiple times. It's mostly a third-strike thing.

Re:The end of an era.

[rdnetto]

There is quite a bit of object oriented C code in the wild, that request just shows that the guy had no idea what he was talking about. While there are a lot of nice (and not so nice) things you can get from C++ if object orientation is all you want its overkill.

Amusingly, the kernel is probably one of the better examples of how to do OOP in C - the VFS code is a good example of this. (The file_operations struct is basically a vtable.) So given that they already had a working solution, they wouldn't have gained anything from C++ except additional complexity.

Which is fair

[davydagger]

Also fair, is that if its an old piece of code that is causing reggressions elsewhere, the person who wants it, should be either coding it themselves, or arranging for someone else to maintain the code.

As for the 386, I can't fathom why a new kernel would really need support for the 386, the use cases for supporting it are very slim. Industrial machines that last decades most likely will use the same software, or patched versions of the same software written around the time the machines where put online, and most likely have a very specialized distribution built around the kernel. consumer 386 hardware is next to worthless on any OS for modern computing tasks. For legacy computing tasks, legacy software is better suited for the job.

There is really no reason you need kernel 3.x 1980s hardware x86 HW. None at all.

Re:Which is fair

[ArcadeMan]

By a strange coincidence “None at all” is exactly how much suspicion the ape-descendant Arthur Dent had that one of his closest friends was not descended from an ape, but was, in fact, from a small planet somewhere in the vicinity of Betelgeuse. Arthur Dent's failure to suspect this reflects the care with which his friend blended himself into human society - after a fairly shaky start. When he first arrived fifteen years ago, the minimal research he had done had suggested to him that the name ‘Ford Prefect' would be nicely inconspicuous. He will enter our story in thirty-five seconds and say “Hello, Arthur.” The ape-descendant will greet him in return, but in deference to a million years of evolution, he will not attempt to pick fleas off him; Earthmen are not proud of their ancestors and never invite them round to dinner.

Re:Which is fair

[tnk1]

If I've been using a piece of hardware from the 1980s and it still works perfectly for me, and it costs little or nothing at all to maintain it in the current kernel, I don't see why you wouldn't.

At some point, there will be a reckoning where they have to make a big change and the EISA stuff will not be reasonable to maintain any more. Until then, there's no reason to remove it just for the sake of "it's old". I presume once that point comes, Linus will ruthlessly shut down anyone who wants to keep it.

Re:Which is fair

[morgauxo]

I remember reading (back in Pentium I days) that NASA still used 386s for a lot of things that went into space. The larger transistors inside made the chip more radiation resistant.

I wonder if they ever found a way to make modern processors more radiation resistant or if they just added more shielding or maybe even still use the old stuff...

Re:Which is fair

[Anonymous Coward]

There is really no reason you need kernel 3.x 1980s hardware x86 HW. None at all.

Have you actually enumerated the individual computing needs of a hundred million people? It's a big world out there, and tunnel vision biases will cripple you.

Re:Which is fair

[drinkypoo]

There is really no reason you need kernel 3.x 1980s hardware x86 HW. None at all.

I found it sad to lose 386 support because there are a bunch of cheap embedded boards with 386s on them still floating around out there. But not sad enough to do anything about it... Regardless, if I can't have a modern kernel on something then I mostly won't mess with it. It's just easier to not have to think your way around multiple versions.

Re:Which is fair

[Anonymous Coward]

There is really no reason you need kernel 3.x 1980s hardware x86 HW. None at all.

Portability has more advantages than just running on "1980s hardware"

1) you can run the kernel on emulators if real hardware is too costly / hard to track down;

if you have an old app that needs "1980s hardware" a good migration step is to get it into a virtual machine;

a maintained, recent OS (preferably with source) is preferred to some ancient, closed source, vendor-long-gone OS,
virtual machine or no;

things like binary emulation (free/net/openBSD does this, Linux too I believe -- e.g. on Sparc NetBSD, you can run Solaris
binaries, just add the proper libraries...on Alpha, you can run things like the HP/compaq C compiler, which being vendor-supplied
is likely much better than gcc since it has been tuned for a single CPU family only...)

2) code portability saves trouble later.
in the long run, it is less work. if your code is 16-bit clean, then was made 32-bit clean, then moving to 64-bit
should likely not be a problem. if you drop 16-bit support in the meantime, then 32-bit-dependent things are likely
to creep in.

if you run on 64-bit alpha today, by the time x86_64 comes out, your code is ideally already 64-bit ready.

this is easier said than done, but C is historically not tied to bit widths for various types; why are you writing a kernel in C
if it is just going to be tied to specific CPUs with registers of specific sizes?

3) portable code finds bugs you may not find otherwise; you can weed out code that "seemed to work" but was actually subtlely
breaking e.g. a C standard, or invoking undefined behavior, but just happened to work on some particular hardware/compiler.

if you don't test on other hardware / other compilers, you may never know; this makes the code more robust, not less

4) arbitrarily dropping support for things is obnoxious; Xen/VMWare/etc. used to not require x86 virtualization featues in the CPU;
magically one day they dropped support if you did not have a new enough CPU; no change to anything, they could have supported
both and not required special HW (they were already doing this for a number of years); the code ends up more brittle, because
now you are tied to CPU-specific instructions/features subject to change

Now, in some cases (say, MMU support) you may say either "finding suitable hardware" or maintaining a separate code path is arguably lots of work...even then, flexibility is a good thing.

this is 2014, not 1965. we have objects and classes (or equiv. abstactions, even in C, hell even in assembly).

if your "kernel" is tied to specific hardware, some abstraction is fundamentally broken.

that is why there are drivers and even dynamically-loadable kernel modules; the kernel itself should be very generic and flexible
EVEN IF IT ONLY RUNS ON ONE MACHINE, EVER.

it is bad coding style to tie a generic kernel to specific hardware; exceptions might be real-time applications, something that is
generated from generic source e.g. for speed, etc.

it is a code smell to do otherwise; it is bad, lousy code if the kernel is tied to particular hardware.

even if you drop 386 support.....it should be clear and easy "this is how/where one would add it back" ... what is more strange
than anything is the idea "386 support" would break any other platform...this is ridiculous, signs of bad abstractions. droppi

Re:Which is fair

[Anonymous Coward]

this is 2014, not 1965

2015 even, for fuck's sake man!

Re: Which is fair

[Ford Prefect]

Hello, Arthur.

Re:Which is fair

[Anonymous Coward]

No, it's more complex than that.

When space agencies put things in orbit, they use radiation hardened versions of CPUs. It takes considerably more time to design a CPU so that it has redundant voting paths throughout such that there are no SEUs (single event latchups). This means that in a rad-hard CPU, any single particle collision cannot introduce an operational fault (though multiple events in a clock interval might).

It's not like they are sticking old commercial grade CPUs in space, they are instead taking commercial CPUs and redesigning them to spread the logic out across a larger area and introduce majority voting logic so the chip functions reliably even in the higher radiation flux found in orbit.

Recently NASA has been experimenting with "phone sats" which are commercial cell phone SoCs in space, and found that while they do mostly work, they also have bizarre failures as a result of SEUs. I suspect NASA and other agencies will continue to use radiation hardened electronics for important missions.

Re:Which is fair

[davydagger]

yes. At the same time, the real question is, why do you use a modern kernel? If you have hardware from the 1980s, why can't you also use software from the 1980s?

Re:Which is fair

[davydagger]

they still do actually(I think they are up to p1/p2 era for space). That said, a modern Linux kernel is not really the optimal choice, try something like an embedded OS.

What a cheap fucker

[ArchieBunker]

Just looked on Ebay and you can get PCI FDDI cards for less than $20. How much time is spent maintaining this 20 year old code? Total economics fail.

Re:What a cheap fucker

[Anonymous Coward]

Yeah and how reliable is that $20 PCI card compared to what they are currently running?

You generally do not replace a proven reliable system "just because".

Re:What a cheap fucker

[ArcadeMan]

How much time is spent maintaining this 20 year old code?

None at all.

"So if we actually have a user, and it works, then no, we're not removing EISA support. It's not like it hurts us or is in some way fundamentally broken." - Linus Torvalds

Re:What a cheap fucker

[mrchaotica]

You generally do not replace a proven reliable system "just because".

So why is it reasonable for this user insist on running a trunk kernel, instead of an old branch that only gets security updates?

Re:What a cheap fucker

[geert]

Because he's the maintainer.

http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/MAINTAINERS#n2998

DEFXX FDDI NETWORK DRIVER
M: "Maciej W. Rozycki"
S: Maintained
F: drivers/net/fddi/defxx.*

He's supposed to run the latest kernel, and keep this driver working...

Re: What a cheap fucker

[LeePriorCollier]

Maybe, but I would hazard a guess that where this old tech is being used at all is likely to be in a resource poor environment. The person supporting the network might not even be on the same continent. It's highly unlikely this is being used in a cheap-ass company who just don't want to upgrade a machine.

Re: What a cheap fucker

[Shimbo]

Maybe, but I would hazard a guess that where this old tech is being used at all is likely to be in a resource poor environment.

Alternatively, one with long term support requirements and onerous change control: e.g. military, nuclear, medical.

Re: What a cheap fucker

[Anonymous Coward]

Or the chain reaction of the upgrade will end up costing millions of dollars. /anon

Re:What a cheap fucker

[MightyYar]

Didn't your momma teach you not to pick scabs?

Re: What a cheap fucker

[Anonymous Coward]

If there is onerous change control, then why would they also insist on a modern kernel which would be subject to the same qualification?

Re:What a cheap fucker

[sjames]

Probably not much time at all, if any.

It's just along for the ride now. Total economics fail would be spending time ripping it out.

Note that the '386 code was demanding actual effort to maintain so it's gone.

Re: What a cheap fucker

[gweihir]

And maybe the users of these cards just have very high reliability needs and do not want to buy a new card that has a 10% chance to crap out in the first year. There may be a lot of other valid reasons. "New" does not imply "good". And if it works, do not break it.

FDDI?

[Virtucon]

1Gbps and 10Gbps SFP+ cards are cheap. Upgrade Now!

Re:FDDI?

[kjs3]

I've still got a little bit of FDDI, as it's the only 100Mb tech for things VAXen and older Sun. But I suppose that kinda proves the point of "leave old crusty tech behind".

Re:FDDI?

[Anonymous Coward]

FDDI was close to being obsolete as soon as 100Base-FX came out. I remember putting large MAN class FDDI rings and removing them about a year later with 100Base-FX. We didn't deploy FDDI NICs except in routers/switches and 100Base-FX was about 1/5th the cost initially.

Re:FDDI?

[gweihir]

Somehow you seem to have missed the part where that was not a sane option....

Re: FDDI?

[Anonymous Coward]

Don't support old crap. Doing so is insane

Re:FDDI?

[Anonymous Coward]

Well, I still know places using real hardware VAX in production. I've suggested they replace them with Charon or one of the hardware VAX emulators, but they seem pretty convinced that keeping machines from the 1980s running is the right way forward.

Re: FDDI?

[gweihir]

This is old, but not crap. Equating "old" and "crap" just shows that you have no clue about competent engineering.

Re:Overhead?

[CimmerianX]

Well that's the point. He's basically saying "It's no skin of our backs" to keep it supported, so he does.

I think that's kinda awesome where 1 guy says he still need it supported and it stays.

Damn it, this ruins my plans

[Anonymous Coward]

I was planning on using the lack of EISA support going forward to force an upgrade past the boss.

Now what excuse will I use with my wife? Somebody think of something or I'll have to deflate her!

Re:Damn it, this ruins my plans

[Anonymous Coward]

Now what excuse will I use with my wife? Somebody think of something or I'll have to deflate her!

Don't know why you posted as an AC.

We all know who you are, Bill Belichick.

Re:Damn it, this ruins my plans

[Anonymous Coward]

Windshield wipers on a billy goats ass.

When you looking for excuses..one is the same as the next.

Remember if it ain't broke, don't fix it.

Which is fair

[Anonymous Coward]

"Here are a couple of small but admittedly valid reasons for having x86 HW. ... There is really no reason you need it. None at all."

Old hardware / technology never dies,...

[rstanley]

It just gets renewed, reused, and recycled, not only in more wealthy Western countries, and Third World countries, but by poorer people all over the world. (Hopefully with Linux and other FLOSS software installed!) There was an article in Slashdot in 2013 about an IBM 402 with punch cards, still in use!!! (I wonder if it could run Linux?) ;^)

But seriously, even though most of us are using computer less than 5 years old, a lot of old computers are still in use. What about 16 bit embedded systems, many running Linux! I have to agree with Linus, if the old technology in the kernel, does not adversely affect newer technologies, and people are still using it, then there is a legitimate reason for leaving it in the kernel. I trust his opinion.

IMHO, I think the FLOSS community has an obligation to continue to support older hardware & technologies that certain other proprietary O/S manufactures have long ago abandoned. Isn't that one of the reasons the Free Software and Open Source Software communities, and software were created in the first place?

Re:Old hardware / technology never dies,...

[Anonymous Coward]

IMHO, I think the FLOSS community has an obligation to continue to support older hardware & technologies that certain other proprietary O/S manufactures have long ago abandoned. Isn't that one of the reasons the Free Software and Open Source Software communities, and software were created in the first place?

You are the FLOSS community. If doing that scratches your itch, go for it.

Re:Old hardware / technology never dies,...

[Anonymous Coward]

What about 16 bit embedded systems, many running Linux!

There are no 16 bit embedded systems running Linux. Linux requires a paged MMU, and the only 16-bit system I can think of that has a paged MMU is the PDP11-XA, which doesn't have a Linux port.

Also I'm almost certain Linux requires a 32-bit processor.

You might have been thinking of ucLinux. NuttX is a much superior option. Neither is Linux.

Re:Old hardware / technology never dies,...

[rstanley]

ELKS:

http://en.wikipedia.org/wiki/E...

http://elks.sourceforge.net/in...

Linux should be scalable to any processor.

Pretty good actually

[Anonymous Coward]

They are DEC branded cards so they should work well. Not some Realtek chipsets with the crab logo.

EISA?

[Anonymous Coward]

I never heard of EISA before. Thanks for sharing the link.

Re:EISA?

[ledow]

Get off our lawn.

In the real world, installed base 'just works".

[DutchUncle]

It is obvious that nobody would install a brand new system with such old technology. It should be equally obvious, though, that just as one expects old buildings to "just stay up" (with a little maintenance), there are plenty of old technology systems still up and running just fine for whatever they do. Lots of people in big cities have 75-year-old telephone wiring which works fine for what it always did (though it can't handle DSL), and the same thing will happen to the brandy-newest fiber optic cable when someone comes up with an LED laser frequency that needs a different glass with different chromatic aberration. There are lots of industrial and scientific devices out there that never got updated drivers past (whatever release of whatever system), and they cost a lot of money at the time, and they still work. (Though I admit that, while they may be worth maintaining, at some point one has to give up on trying to update them.)

What about ISA?

[morgauxo]

I was assuming that EISA was just a special case inside of the same code as ISA and that what was proposed was to remove all ISA support. Is that what was going to happen?

ISA is old but I am sure there is quite a bit more than just one person out there with some sort of legacy hardware using it. I have a little bit of ISA hardware myself that I would like to use but not quite enough to build up a legacy PC. Every now and then I search the internet for ISA to USB adapters. There actually IS one company selling such a beast but it is way to expensive to be worthwile for me. But.. if I had some expensive piece of lab equipment or something like that with a proprietary ISA adapter... it would make sense.

Re:What about ISA?

[drinkypoo]

I was assuming that EISA was just a special case inside of the same code as ISA and that what was proposed was to remove all ISA support. Is that what was going to happen?

I doubt it. ISA buses are still present even in some modern hardware, even when they're not exposed. It's less common these days than it used to be, but it's not over.

Re:What about ISA?

[Anonymous Coward]

VMs. The virtual hardware is usually reported as ISA with an overdriven bus.

Re:What about ISA?

[gweihir]

You can also get them easily on industrial PCs. One reason is that there are quite a few custom interface cards based on ISA around, as it is so easy to interface with it. The other thing is that there are cards around that have quite a bit of remaining lifetime and would be expensive and problematic on software-side to replace adequately.

For example, PC104 is ISA with a different connector and it is far from dead.

Re:What about ISA?

[Blaskowicz]

About every PC still emulates ISA and even if you don't use the legacy I/O such as PS/2 you need it at least to store BIOS setting (not sure what UEFI motherboards do), or even for that hated Trusted Platform Module!

http://en.wikipedia.org/wiki/L...

Re:What about ISA?

[petermgreen]

From a software perspective what sets memory/IO mapped busses apart from each other is how the OS reads and/or sets their configuration, not what they are like on an electrical level. ISA was either manually configured or (later) used a hacked on plug and play mechanism. EISA had it's own configuration system.

Re:What about ISA?

[Anonymous Coward]

One reason is that there are quite a few custom interface cards based on ISA around, as it is so easy to interface with it.

Maybe this isn't a surprise to you, but actually people are still designing new ISA cards. They are cheaper to manufacture than PCI cards, and much cheaper to design. When you already have a PC based control system, and you need to expand it with some new controller, and you need better latency than USB and better bandwidth than a serial or parallel port, Adding an ISA card to the system is much less expensive than designing a PCI or PCIe card and gets the job done just as well. Because the ISA bus runs at 8MHz, it's quite easy to implement bus transactions on a cheap microprocessor like an AVR or PIC32, or on a cheap Flash FPGA.

Also while a PCI domain supports only 4 slots (without expanders), you can get completely passive ISA backplanes with up to 18 slots. These are also coarse pitched, so if it gets exposed to corrosive materials, you can pull the chassis out and resolder it without having to disassemble the chassis, which saves a bunch of time in the field.

And ISA is much less expensive than VME, which is the other standard that just won't go away. Of course VME has *much* higher performance.

Re:What about ISA?

[gweihir]

Indeed. "Simple and works" beats "shiny and new" every time in competently done engineering.

I think part of this "throw it away because it is old" mindset is from people that do not understand that many things are a mature and fine solution for a specific problem-set. This seems to be a mixture of incompetence and a misguided belief in perpetual advancement in all things electronic.

Re:What about ISA?Hip and Trending

[Anonymous Coward]

Hipsters have taken over linux. This shit about removing suppot is total crap. Fuck linux if this is what they're about

We need more women coders too, and girls not brides. SystemD yeah!

Learn, Mozilla!

[fraxinus-tree]

In unrelated news, Mozilla cuts off digital signatures. Even if they have millions of users.

NI A-2000's...

[Grog6]

I have a bunch of these, and some top of th line IBM servers (c1991) to run them.

They are better than everything newer except the Aquaris... :)

MCA removal

[jones_supa]

recent years have also seen MCA bus support being removed from the kernel

Just for reference, here's also the original discussion on MCA support removal from 2012 in LKML.

Re:MCA removal

[myrrdyn]

Just for reference, Linus' reply to that:
"Maybe we could some day remove EISA support too.."
https://lkml.org/lkml/2012/5/1...

The thing I remember about EISA?

[m.dillon]

I remember that every time I changed a card out the machine took 30 minutes to reconfigure itself, because some doochebag of a programmer wrote the #$%#$% configurator that all the vendors used. An operation that could have been done in 5 seconds if written properly. That was the first ... and last EISA machine I ever bought.

-Matt

Re:The thing I remember about EISA?

[Anonymous Coward]

I had the same experience. It seemed odd it was such a hassle.

Working with state agencies in the '90s

[aussersterne]

I saw a lot of EISA systems. It was a reasonable performer and physically robust (not as sensitive as PCI cards to positioning in slots, etc.). I'd say that EISA hardware was generally of very good quality, but high-end enough that most consumers wouldn't run into it despite being a commodity standard, sort of like PCI-X.

The systems I had experience with were running Linux, even then. :-)

recommend moving EISA suport to systemd

[Anonymous Coward]

perfect for a small, optional, add-on to systemd that you dont have to use if you dont want

DVB cards

[phorm]

I've got a bunch of PCI DVB/capture cards that are in the same boat. They *could* be useful if I had drivers for them, but alas, they do not.

My personal disfavorite is software that depends on dongles which have OS-specific drivers. The software *WOULD* work on the newer OS if the dongle had a driver that allowed it to authenticate (of course, dongles suck in general).

Re:DVB cards

[Bert64]

What is likely to work however, is a cracked version of that software.

Thank you, Linus!

[MagickalMyst]

It's nice to see that Linux is still being developed by the people, for the people.

OFFTOPIC: Re:Learn, Mozilla!

[higuita]

You really don't want CA that aren't trusted in your browser... even if it breaks some sites... a broken CA can break ALL your https sites

We should have killed SSL3 years ago, but to avoid breaking some sites, we found later that SSL3 bug could break all sites, even if they had already TLS1.2
When something should die, let it die or you will pay for it later

Re:OFFTOPIC: Re:Learn, Mozilla!

[fraxinus-tree]

Well, I meant crypto.signText.

Actually there are two users

[Anonymous Coward]

I have it also.

Spelling

[Anonymous Coward]

This article provably should have been spell checked.

Also, this is not exactly riveting news the geek world needs to know.

I think I'll go back to reading TheRegister.co.uk, so I can see what will be on Slashdot next week.

Banyan vines

[Anonymous Coward]

I still have a HP Net Server LH II with 6 ~10Gb HDD's.
Running Banyan Vines... which i have on floppy.

I'm not going to replace it until I'm no longer able to find drives for it.
It's Very Loud and it eats Electricity likes its going out of style but has always been faithful to me since ~1995-1996.
I have never had to replace either one of the power supply's yet.

It looks like this.
https://www.oneeyedman.net/school-archive/images/hp-netserver.jpg

Everything still works, I access it via the client software running in DOSBOX from my current Phenom II Ubuntu Box.

I also still have complete sets of floppy's for win 2.0/2.1/3.0/3.1/3.11 and every every other game/application you could think of from Office to Oregon trail...

And they all still work had to break them out a couple years ago to reinstall 2.1 on a machine for a client of mine who owns a grocery store that's still running its inventory on a original Pentium PC that he says he had somebody upgrade him from a even older machine!!! some time in the 90's.

EISA in NT RISC workstations

[unixisc]

EISA was also used in the earliest RISC workstations that supported Windows NT 3.1. There was the MIPS Magnum R4000, and the DECpc AXP 150, both of which had EISA buses for NT, and equivalent Turbochannel based workstations for UNIX (Irix/RISC-OS for the Magnum R3000 and Ultrix for the DEC 3000 AXP). MIPS abandoned the model after its foray into the NT market ended quickly, while DEC replaced both the above models w/ the PCI based AlphaStations and AlphaServers.

parallel still available

[Blaskowicz]

You have many motherboards options with latest gen hardware and a parallel port, still.
That may work if your software is so backwards it needs to think the parallel port is attached to the ISA bus.

http://www.asrock.com/mb/Intel...
http://www.asrock.com/mb/AMD/A...
http://www.asrock.com/mb/Intel...

Linux for the .... middle?

[taikedz]

So two thoughts come to me after reading the summary and the comments

1/ This clearly demonstrates that Linux is the right technology to use when supporting business syetms with legacy hardware (I'll second the opinions that enterprise keeps hardware around the longest - I supported users on obsolete AS/400s which, whilst not as old as what some people here talk about, still mean we're frequently learning old technology in Support); and the point that the leadership (Linus right now, hopefully the same with whomever comes to replace him eventually!) can be more ameanable to keeping up support for old hardware is great. I dream of desktop uptake, but enterprise and research are where it's at.

2/ However I also wonder - isn't this an offshoot problem of the fact that Linux is a monolothic kernel? Can this kind of interface-specific support not be modularized? Say, an API/ABI (in-kernel)standard that allows the kernel to plough on with currently evolving requirements, whilst maintaining a stable interface for previously integrated kernel features that have been split off into modules...?

(and no, I'm not at all familiar with the ins and outs of kernel development and architecture - I just read newsposts and Wikipedia ...)

PC104 ?

[dargaud]

Isn't EISA (or ISA) still used on CURRENT industrial embedded systems like PC104 ? I'll soon know, I've just been asked to work on some... In that case it's very good if a modern kernel can support it.

Good to see

[chris_clay]

It's good to know that the developers are listening to the community. Not sure how this particular user knew of EISA support being dropped but good that we can make sure the Linux kernel and GNU/Linux operating system survives as community-driven software which will always outdo proprietary software where it is driven by a single company's wishes. GNU/Linux is excellent at supporting new and old hardware and retaining backwards compatibility. Thanks again to all of the developers and for making GNU/Linux such a huge success.