Don't Nurse Old Hardware - Emulate It

gManZboy writes "Bob Supnik, former team lead for DEC's VAX microprossesor, has an article up on Queue about his Computer History Simulation Project and how emulating old servers may be a better way to keep them running that servicing the physical machines. So how many PDP-11's can you run on a Pentium 4 anyhow?"

Comments

Not a bad idea

[mirko]

I was even considering emulating existing hardware on beowulf clusters, I know it sounds like a troll or deja-vu joke but I mean it : if I have 1000 machines emulated on a beowulf of 1000 machines, then it'll be harder to get downtime if one machine physically crash.

Re:Not a bad idea

[rf0]

Or for testing you can run a Cluster on a single CPU...
http://openssi.org/ssiuml-howto

Rus

Re:Not a bad idea

[awx]

You can emulate a VAXcluster with multiple instances of SimH on a single-cpu x86 box with no problems.

Re:Not a bad idea

[photon317]

Yeah they have that already for x86 on x86, VMWare's high-end enterprise product allows you to run VMWare on several machines and transparently move system images between physical hosts without taking any (perceivable) downtime. So hardware maintenance can be done without interrupting your "servers". Of course if the hardware crashes, any system images running on that particular hardware go down hard, but they can fail over to another peice of hardware and come up and fsck (or the equivalent in your OS of choice) themselves and go online immediately on one of the remaining hardwares.

Re:Not a bad idea

[zeugma-amp]

allows you to run VMWare on several machines and transparently move system images between physical hosts without taking any (perceivable) downtime.

I've seen this in action, and it is very impressive. Imagine, if you will a MS-SQL (ugh) database running in a VMWare session. Let's say you need to perform some hardware maintenance on the system it is running one. Using their control console, you can "migrate" the entire emulated session while it is still taking transactions to another system with a barely perceptable pause (a second or so) between when one server stops executing and the next server starts.

Disclaimer: I don't have anything to do with VMWare other than the fact that I use it.

Re:Not a bad idea

[TheLink]

Does that require shared disk?

And what if you have a VM with 2GB of ram - would that take 2GB/network bandwidth seconds to migrate to another machine?

Re:Not a bad idea

[Anonymous Coward]

We need to emulate the old green color scheme for this section so my eyes don't crash.

Re:Not a bad idea

[Elwood P Dowd]

What they're describing would only make sense if VMWare could start transfering the machine's state while it was still running. So lower bandwidth would make the whole process take longer, but would not increase the (one or two second) amount of downtime. No, it shouldn't take 2GB/second bandwidth. But I don't know what I'm talking about.

Re:Not a bad idea

[danheskett]

The way it basically works is that process is started, and an approximation of entire image is copied over to the other machine. After a minute or so on some images, more on machines with many resources, the machines are basically close in contents. When the delta between the machines is fairly small, execution on the first is stopped, and all changes comitted to the second. Commands in the pipeline are moved to the second machine and the machine unpaused.

On virtual machines with big resources it does take a solid minute or more to get things to that point, but the result is flawless everytime. Having a shared fast disk array really helps, FYI.

Re:Not a bad idea

[Wow8agger]

Yep, you're talking about VMWare ESX and their VMotion product.
It is helluva cool - requires shared attached storage to work, but you can do a lot of neat things with it.

More information is available here:
http://www.vmware.com/products/vmanage/vc_features .html#vmotion

Disclaimer, I work for a company that sells (amongst other products) VMWare, and I do VMWare installs/consolidations.

Re:Not a bad idea

[photon317]

Correct. Once you initiate migration, it basically takes a snapshot of the state of the virtual system (memory copy, basically) and pushes it to the new host. While copying they keep a bitmap on the old host of what's been changed while they were busy copying (much like snapshotting a filesystem). When the copy is complete, they go back and copy over again just the stuff that changed while they were copying the first time. When that's done, repeat, repeat, until the changeset gets so small that it repeats the cycle in some small amount of time under a second. At that point it freezes the image on the old host, transfers the final small set of state changes, and fires up on the new host.

And yes, shared disk of some sort is a requirement too.

none.

[Anonymous Coward]

pdp-11 is much heavier than you think. even one would crach a pentium

Re:none.

[TWX]

"pdp-11 is much heavier than you think."

Not to anyone who's ever tried to pick one up!

Re:none.

[Anne Thwacks]

pdp-11 is much heavier than you think. Sometimes you need a heavyweight solution!

And for the rest of the time, there is always a PDP8!

Can't believe no one's thought of this

[Jonny_eh]

Everyone has been so busy emulating the GBA and Xbox that no one has thought about emulating these old servers?

Re:Can't believe no one's thought of this

[PhilHibbs]

It's not like there is some pool of emulator writers, constantly considering what the world needs in terms of emulators. The people that write GBA emulators are people that want a GBA emulator. Asking them nicely to write a Pr1me emulator is likely to get you nowhere. You need to talk to the people that want a Pr1me emulator.

Re:Can't believe no one's thought of this

[Saeed al-Sahaf]

Christ. The truth is, Slashdot is basically a Gamer site. It's quite abnormal to see a real computer story here.

Re:Can't believe no one's thought of this

[pommiekiwifruit]

You need to talk to the people that want a Pr1me emulator.

Like a C standard fanatic... "the standard does not specify all bits zero for a null pointer - you should not use memset on structs that include pointers" :-)

I hear that if code runs on a pr1me someone has actually read the standard/documentation instead of assuming various silly things...

Re:Can't believe no one's thought of this

[MoneyT]

Actualy people have thought about it. The problem is, emulators are all well and good, until you actualy need to access a periphrial.

Re:Can't believe no one's thought of this

[Anonymous Coward]

Not only have people thought of this, but they have been doing it for a long time. Search trailing-edge.org, the comp.os.vms newsgroup, and for CHARON-VAX. You may be surprised to learn that just because /. has posted an article on a subject doesn't mean it's not old news.

-Tom O'Toole

Re:Can't believe no one's thought of this

[FlutterVertigo(gmail]

agreed. I was ~ sophomore in college using a pdp 11/70 when we installed a vax[1,2]. i remember when there were rumors there would soon be a "vax on a chip" and we found that to be incredible. iirc, when ada's ("the green team" for the trivia buffs) first complete specs were made public and compilers were readied, the vax was the smallest machine capable of running/compiling ada - i think there was also a "janus ada" running on some apple hardware, but that was a very wispy version of ada. when it appeared the dod was pushing ada to the point of pushing cobol out the door, everyone felt the way to make a basket of dough was basically a cobol->ada translator or compiler.


[1] everyone keeps hinting at a full teco editor for windows, but is there one? data general dup'd it and called it "speed".
[2] has basic-plus-2 been ported to anything other than the dec world?

Re:Can't believe no one's thought of this

[Anonymous Coward]

Check out dbit.com.

Re:Can't believe no one's thought of this

[jallison]

everyone keeps hinting at a full teco editor for windows, but is there one?

Not sure if this is "full teco", but it claims to be teco: TECO for DOS.

Re:Can't believe no one's thought of this

[Eunuchswear]

Don't believe it.

Lots of people have thought of this.

See this guy for example.

Re:Can't believe no one's thought of this

[confused one]

There have been PDP and VAX emulation boards available for some time. I believe some of the other big iron is also emulated already. This isn't something new, just new to /.

Re:Can't believe no one's thought of this

[Riff6809]

I'm sure its being done as we speak. In 1998, I helped implement a Honeywell 316 emulator to replace a real H316 computer in two F15/F16 diagnostics test stations. The test executive and user programs were all written in H316 assembly language. Seeing that the H316 is no longer in production, the only choice was to emulate the machine given that porting the existing code would have taken an inordinately long time and porting code we don't know about is obviously impossible. The test stations themselves contain new hardware from the late 1990s and the emulator contains device handlers to communicate to them via GPIB and other busses. The test executive uses standard H316 IN and OUT instructions to access the hardware. I also implemented a TEK400 display emulator and interfaced with a VB app that replaced a big custom console table with dedicated operator buttons that would be used during the tests. The customer must have liked it because they followed up on the contract with a request for more. In that sense, I'd say emulation is definitely a reasonable option. Its not always the right option. The downside is that there is still have old code to maintain. The system is a little more understood now, but you still have old archaic code. The upside is that writing an emulator often takes significantly less time than the alternative of rewriting.

Re:Can't believe no one's thought of this

[pla]

The problem is, emulators are all well and good, until you actualy need to access a periphrial.

So you emulate those, as well... And usually, they take a lot less to emulate than the core system.

As a trivial example, consider the peripherals available even on previous-gen consoles... You have 3rd party joysticks, mice, keyboards, cameras, tape drives, printers(?), etc. All those eventually end up emulated, if enough people needed them.

The same goes for something like a PDP-11 or VAX 11/750 or the like. You have some odd storage devices (that store a tiny fraction of modern HDDs, thus you can emulate them with an image file). You have printers (emulatable with... printers!). Perhaps really ancient input devices such as a cardreader (scanner -> conversion tool -> file). No doubt other exotic peripherals exist, but you can somehow emulate them all.

The conceptual problem with dealing with peripherals I think lies in just how much we've advanced since the days of Big Iron... Even for emulating CPU-cycle-critical hardware interactions, you can deal with it in emulation, by pure brute-force. Consider the 11/780, which ran at a whopping 6MHz. On a modern P4, that means you have over 500 CPU cycles per emulated cycle (and while the P4 can push through more than two ops per clock, the VAX only managed one instruction per 6(?) clocks, meaning you realistically have over six thousand real instructions per emulated one, on average). With six thousand instructions to burn, you could emulate your VAX while still getting a good framerate playing Super Metroid on an emulated SuperNes in the foreground.

That I know of, only the humble old laserdisc has thus far resisted attempts at perfect emulation, due to using an analog encoding scheme (rather than storing bits, it actually encoded the raw NTSC or PAL signal handled by the TV. And depending on what sort of access to them you need, even that problem has a way around it, via an MPEG rip and a frame file (ala Daphne).

Re:Can't believe no one's thought of this

[Kedder]

Everyone has been so busy emulating the GBA and Xbox that no one has thought about emulating these old servers?

I think you're wrong:

$ apt-cache search pdp 11
simh - Emulators for 32 different computers
ts10 - Emulators for various old computers
pdp11-unix-v5 - Caldera UNIX V5 images for a PDP-11 emulator
pdp11-unix-v6 - Caldera UNIX V6 images for a PDP-11 emulator
pdp11-unix-v7 - Caldera UNIX V7 images for a PDP-11 emulator

Re:Can't believe no one's thought of this

[Carnildo]

The problem is, emulators are all well and good, until you actualy need to access a periphrial.

So you emulate those, as well... And usually, they take a lot less to emulate than the core system.

What about real peripherals? Ones you can't emulate?

I once worked at a mechanical testing lab. The "peripherals" were fatigue testing machines that connected using a 30-pin DIN connector. Some of the pins were digital input, some were analog input, and some were analog output. Further, some of the internal circuitry on the controller board that would need to be emulated was analog, while the overall controller logic was digital.

Have fun!

Re:Can't believe no one's thought of this

[pla]

What about real peripherals? Ones you can't emulate?

Fair enough point - But in that case, I would have to consider the computer secondary to the machine itself - I don't know your specific situation, but the computer most likely did data collection and analysis. The controller-proper I would consider part of the "peripheral", and search for the easiest-to-tap connection as the point to break away from reality and into emulation.

I'll admit, I hadn't considered that point. You can't emulate a fatigue testing machine. You can't emulate a GC/MS. You can't emulate a CNC. You can't emulate an MRI. But I see that as similar to saying you can't emulate a car... You could fake-out a car's computer, but it doesn't do much good without car itself. Apples and Oranges.

So no, emulation won't solve all obsolescence problems... But for those situations where you wouldn't consider the computer as secondary to the peripheral...


However - as I touched on above, you can almost certainly still replace the computer portion of the system. At some point, the computer just takes a digital input, and at that point, you make the split from reality. My car doesn't care if its own computer runs it, or a PC sitting in the passenger seat. That might not do much to reduce the maintenence costs of the primary machine involved, but it does mean you can have your data on something other than an ancient reel-to-reel... ;-)

Re:Can't believe no one's thought of this

[Robert The Coward]

You would have two choices. Depending on what you needed. If you needed to run testing using that equipment you would have to replace the device and drivers. I don't know how that custom hardware worked but if you knew how it works then it can be replaced. If it was simple 10 Digital inputs, 10 Analog inputs 2 Digital outputs, 5 Analog outputs, and 3 grounds. That you could get 1 Digital 16 channel IO board and 1 16 Channel Analog card and write drivers to interface those lines up in modem OS to emulated OS. Then create a cable to match those 30 Pin connectors. It can be done. Now if you had logic controls that said if this is High the switch this low in the hardware itself then that logic would have to be done in the driver itself.

What you are describing is very spefic hardware. How many units were made and how many companys would have used it. In the end it comes down to how much you know about the interface and how bad you need to replace it. Depending on how well documented the interface was you could hire someone to write that driver. This from someone who works with equipment that talks to phone systems.

Re:Can't believe no one's thought of this

[devilspgd]

Hell, just because the article was just posted to /. today doesn't mean it's new to /. either.

Re:Can't believe no one's thought of this

[jeremyp]

Your approach is far too naive. I can see how you could emulate a card reader, but what about a paper tape reader? You can't easily emulate that with a scanner. Or if my client has a jukebox full of 5.2Gb optical discs, how do I emulate that? Can't be done without real hardware. Or a half inch reel tape drive?

It's fine emulating those if the data is not on the associated media (but then why bother?) On the other hand if you have 200 quarter inch tapes, no qmount of emulation is going to help.

BTW big iron is not dead, it's bigger, and a VAX 11/780 was never counted as big iron. Even when new it was only a mini-computer.

Re:Can't believe no one's thought of this

[tricorn]

Absolutely. About the time the DEC Alpha came out, we decided it was the perfect platform to emulate a Cyber to run our custom-built system that had been in development for over 20 years at the time. PLATO is written in assembly language, which then implements a timesharing system running a custom-built programming language that the rest of the system was written in. The Alpha was ideal because the Cyber architecture is 60 bits. Handling the one's complement arithmetic was easy, the floating point was fussy but doable, and the result was a machine that ran 5 times faster than the original hardware. That's on a 300MHz Alpha. Today's are much faster. Fall on the floor with your jaw hanging down faster. Too bad that HPaq is dropping it in favor of Itanium.

For the emulator, we took the easy way out, only emulating the CPU instructions. Operating system calls were written in C, no hardware emulation or PPU emulation (PPUs were auxiliary processors that handled all the I/O; large amounts of the operating system were implemented in them, and in actuality they controlled the CPU side, which was effectively a very fast co-processor to the PPUs). To speed things up even more, I wrote a static translator to push out Alpha assembly code for much of the emulated code; that code was linked in as a dynamic library. A bit tricky, as Cyber code is inherently self-modifying, and standard programming practice was to use self-modifying code a fair bit. Using the "native translation" sped up the emulation by up to a factor of 10 times more. It's amazing that you can run around 5,000 simultaneous on-line interactive users on an emulated timesharing system running a pcode-style emulated language on a machine about as powerful as a Mac G5 (dual 1.8GHz, 1GB RAM).

Recently another group of people got together to work on their various Cyber emulators, with one of them (DtCyber) now being fairly mature. It takes a different approach to our "practical" approach, emulating PPUs (including some tricky timing issues), hardware, and all. It can bootstrap various versions of NOS and earlier Cyber operating systems. It isn't nearly as fast as the translated CPU-only emulation, but it is way cooler.

There's a small group of us bringing up a PLATO system on DtCyber for public use. For more information, see my journal entry.

Re:Can't believe no one's thought of this

[Anubis350]

this is only true if you read on the gaming section only.... besides you forgot linux, most linux stories get more posts than gaming ones as far as I can see... anyways, your comment was basically a troll anyway, and hear I am feeding it. ah well, such is life.

A guess

[NETHED]

I will have to guess, 7. Just because its a great number.

But seriously. Emulators are the way to go. I know that w/ Microsofts new initiative to let linux apps run in Windows, emulators are going mainstream. Yes, for all you anti-MS people, Windows is mainstream.

I was looking for the article about Windows Linux thing, but I cannot seem to find it at the moment.

Re:A guess

[doublem]

It's called Wintux

Re:A guess

[goldspider]

Correct me if I'm wrong here, but this doesn't sound real efficient. Here's how I'm seeing the system load being broken down (real simple breakdown):

1. system resources required to run modern OS and background processes
2. system resources required to run the emulator
3. system resources to run processes via the emulator.
So basically the emulator route would use roughly 2-3x the system resources compared to just running the original hardware/software. Would this really be cheaper/more efficient than just keeping the old stuff until it is replaced with a modern system?

Re:A guess

[Anonymous Coward]

Fool, Cygwin has been around for a long time. Apparantly some other yesman at MS said /new initiative/ and so now I suppose it's stable enough to take credit for some halfassed reimplementation?

Well, if you say emulators are the way to go, you must have it on good authority.

Re:A guess

[shrykk]

Yes.

Every year you can get better hardware for the same price. It could be far cheaper just to buy new hardware than to repair or replace parts for obsolete machines.

Re:A guess

[pilgrim23]

As I read this I am installing a dial-in Citrix Metaframe session on my iBook. But...Mac ICA Clients do not SUPPORT actual non-TCP modem connection so...I installed VPC, then installed Win2K Server (I had a spare unused license), then installed the client under Win2k all on a eternal firewire drive so it uses less on box space. Emulation is indeed the answer in those cases where the computer you can kick is not the one you need.

Re:A guess

[ChairmanMeow]

The first official release of WinTux will be three months after the introduction of Windows Whistler. Beta testers may apply for a special, low-cost, evaluation version at www.microsoft.com

Don't they mean Longhorn?

Re:A guess

[TedCheshireAcad]

iirc whistler was the code name for xp?

Re:A guess

[ChairmanMeow]

Yes, it was. The link doublem posted is a joke that needs to be updated, since XP/Whistler has been out for a while.

Re:A guess

[networkBoy]

Especially a machine as obsolete as a PDP8
-nB

SimH

[Anonymous Coward]

Bob's emulation software SimH is a *fantastic* bit of kit. Runs vanilla OpenVMS without modification - VMS doesn't even know it's in a sim until you tell it so when you licence it.

Re:SimH

[Mark Round]

If anyone is interested in running OpenVMS on it, take a look at http://www.wherry.com/gadgets/retrocomputing/vax-s imh.html.

It covers setting up the Linux side of things, adding network support and then installing OpenVMS. It's a great (geeky) exercise and provides a good way to learn an alternative to the Unix way of doing things.

While I've been using Solaris and other Unixes for too long to seriously contemplate "re-tooling", I can appreciate the elegance and unique approach OpenVMS brings to many tasks.

-Mark

Re:SimH

[ch-chuck]

simh really rocks - I can run a PDP10/TOPS10 system on my 233Mhz notebook - you can telnet into it over the network. I've also run Altair 8800, cpm2.0 and have worked up many vintage software projects including the Original Adventure game in FORTRAN (on the pdp10/tops10), Star Trek in MSBasic on an Altair, and the 8008 SCELBI BASIC interpreter (running on an 8080, but it was written for an Intel 8008!). Kind of amazing to see an 8008 BASIC crank thru the floating point calculations for 'Lunar Lander' ;)

Re:SimH

[awx]

Aye, i've emulated an entire VAXcluster on a single-cpu x86 box with no problems. Bob deserves beer, and lots of it :)

or how many

[Anonymous Coward]

pentium IV's can you emulate on a PowerPC?

Re:or how many

[CarrionBird]

Ehh about .75

Re:or how many

[Anonymous Coward]

Really, I get slightly more.

Re:or how many

[cheezfreek]

47. That's right, 47 very slow Pentium 4s.

Re:or how many

[bhtooefr]

Quite a lot, and viably too. However, the REAL question is: how many Pentium 4's does it take to emulate a PowerPC at a usable speed, with and without Altivec (G4 and G3 emulation)? PearPC appears to get about a 75MHz G3 equivalent speed on an A64 3200+. NOT good at all.

Re:or how many

[GoRK]

PearPC does not yet virtualize the PowerPC MMU, thus all memory read and write routines are still software based. The next big speed boost in PearPC will come from this.

The big advantage that the PowerPC CPU has in emulating the X86 (or any other CPU for that matter) is that it has far more registers and it can change endianness. I don't think anyone's disputing that PowerPC can emulate X86 better than X86 can emulate PowerPC :) The PowerPC is a really good processor for emulation in general.

Re:or how many

[gl4ss]

ok.. so you're saying that one powerpc cpu can easily run the worth of 'quite a lot' which must mean quite many more than the 'under 1' which is the reality.

Re:or how many

[caspper69]

Actually, with the G5 (PPC970) chip, IBM did away with the endian swapping feature. This is what has kept VirtualPC from running on the G5s. Wonder why they got rid of the feature? My guess is that die size limitations simply would not allow the required support logic anymore, especially when multiple 970 cores get lumped together on a single chip in the POWER family.

Re:or how many

[GoRK]

Damn! I didn't know the G5 couldn't change endianness. That's a pretty severe blow to the emulators. Do the Power5's?

Oh well.. have to wait for VPC 7 I guess to see how severely that product was crippled - or how much of a wiz microsoft is at writing emulation - the X86 emu in the version of windows for itanium doesn't really bode well for VPC!

Re:or how many

[Anonymous Coward]

You've obviously no clue. Go do your homework. A PPC can emulate a Pentium of roughly the same clock speed without any trouble via VPC right now, whereas a Pentium can emulate a PPC at about a tench of its own speed, in theory, and more like 1/100th of the clockspeed using the emulators that currently exist.

Re:or how many

[Christopher Thomas]

You've obviously no clue. Go do your homework. A PPC can emulate a Pentium of roughly the same clock speed without any trouble via VPC right now, whereas a Pentium can emulate a PPC at about a tench of its own speed, in theory, and more like 1/100th of the clockspeed using the emulators that currently exist.

I realize that asking for actual evidence from either side here is futile, but:

- Download SPEC benchmarks.
- Make them go on PPC machine natively with default and tweaked compiler options.
- Make them go on P4 natively with default and tweaked compiler options. Use MSVC/Windows and gcc/*nix, for a good spread - gcc's optimization was Not Good last time I checked. MSVC's was.
- Make the same P4 binaries go under VPC.

The results will probably be enlightening for both of you. I am deeply skeptical of *any* emuation being as fast as the real thing for code that's already optimized (i.e. won't benefit from emulator re-optimization).

You'll also be memory-bound on both machines for several types of task (same speed on both systems), clock-speed bound for several other types of task (favours the P4), register-bound on others (favours the PPC), and cache-bound on yet more (varies by machine configuration).

Apple's been reluctant to post SPEC numbers for a while now, so feel free to settle the issue with real benchmarks instead of a photoshop or emulator DSW.

Re:or how many

[Anonymous Coward]

It seems like Mr. Jobs got mod points today.

Re:or how many

[caspper69]

I personally think this has been the big delay in getting VPC7 to market. The performance must be abysmal. While the PPC970 can still perform endian swapping calculations quickly (as can the x86), the inability to mark specified regions of memory (code, data, stack) as big or little endian was of great importance. Now, every time an instruction is executed, the calculations must be performed. My guess is that MS is going to be very careful about performance on this issue. Everyone already has their eye out for their sinister intentions after having purchased Connectix, and imagine the ammo it'll give their detractors if VPC7 runs slower on G5s than it did on previous generation Apple hardware. I personally think the real reason for MS's purchase of Connectix was for their server virtualization technology which they needed in order to keep pace with VMWare's GSX Server and IBM's work on running multiple Linux kernels on big iron. But, I guess everywhere there's smoke, there's fire...

Re:or how many

[gl4ss]

*You've obviously no clue. Go do your homework. A PPC can emulate a Pentium of roughly the same clock speed without any trouble via VPC right now, whereas a Pentium can emulate a PPC at about a tench of its own speed, in theory, and more like 1/100th of the clockspeed using the emulators that currently exist.
*

uh? how about you do your homework, it's not quite that rosy, if it were it would have quite a bit of applications just for running pc programs(like games).

I wasn't talking just about clockspeed either, more like about the whole package(average powerpc cpu vs. average p4).

powerpc isn't some miracle chip some people make it out to be(in terms of performance per $)..

Re:or how many

[andreyw]

Architecutre-wise, its still easier to write an emulator for the PPC than for the IA-32. The IA-32 seriously has issues with the amounts of registers available - hence AMD increasing that amount in x86-64.

What do you think is easier to emulate? A low-register-count CPU on a high-register-count CPU, or vice versa? Similarly, its way easier to emulate a low-endian CPU on a low-endian CPU than on a big-endian. Its really pretty sad that the PPC-64 won't have the endian-switching instruction... but oh well.

Re:or how many

[Christopher Thomas]

Architecutre-wise, its still easier to write an emulator for [on] the PPC than for [on] the IA-32.

Part of the point of my previous post is that this fact a) means quite little when comparing the relative capabilities of the two processors, and b) is being used to make some rather silly claims in this thread. Both are good processors, and both have enthusiasts with more zeal than sense ready to bash the competition.

What do you think is easier to emulate? A low-register-count CPU on a high-register-count CPU, or vice versa?

Emulating a machine with many registers on a machine with few registers will cause spills to memory, which indeed slows things down, but there are other aspects of architecture which will slow things down far more. The x86 architecture is a crawling horror, full of operations with complex side effects and wierd idiosyncracies that are holdovers from bygone eras. I'm told that the PPC line has its own demons to worry about in this regard, before any of the Mac crowd start pointing fingers. Trying to emulate a crawling horror of an architecture requires a large amount of special-casing and time-consuming state tweaking after most operations, or else requires assuming that certain side effects won't be used most of the time and producing an emulator that only *mostly* supports the relevant code, or else requires a cross-compiler with the near-omniscience required to determine for all cases where features are needed, and produce code that does exactly what's needed and nothing more. My understanding is that it's the last approach that's used by modern emulators (with optimization thrown in for giggles), but it's far from perfect, and there is still substantial overhead both in execution time and program memory footprint (which causes extra I-cache stress, slowing things down even more).

In summary, the PPC will still have a hard time emulating an x86, and register count on the x86 is the least of its problems when trying to emulate a PPC.

Endianness is its own can of worms. You can relatively easily tweak endianness of code and immediate values during cross-compilation, but any memory buffers have to be munged at *run* time (because you don't know if the program treats them as being composed of 8-, 16-, 32-, or 64-bit elements, or all of the above). If the latest PPC chip doesn't support this in hardware, emulation becomes a tar-pit.

Anyways, this is why I'm cynically amused by all of these "My 3mul4tor can k1ck y0ur 3mul4tor's 4ss!" posts. And yes, IAACompEng.

Re:or how many

[andreyw]

Valid points, but don't insinuate on me being some sort of Mac zealot...

My first, only and likely last Mac ($$$ :-() is a dual-604e clone I saved from getting LARTed by some luser. Frankly, since its now a Yet-Another-Debian-Box, I almost forgot it was PPC.

I disagree

[Rosco P. Coltrane]

emulating old servers may be a better way to keep them running that servicing the physical machines.

I disagree. It's not the same thing.

-- Signed: your friendly PDP-11 system operator downstairs, 3 years from retirement.

Re:I disagree

[CommanderData]

You might want to retire early. Just take up smoking or drinking to reduce your lifespan by those 3 years if you don't have enough saved up yet.

Re:I disagree

[acebone]

Go for cheap brands, that'll reduce your lifespan more while hitting your cash reserve less, that's what I do !

Re:I disagree

[Jeremiah Cornelius]

Emulate the "wurlitzer" keys, to toggle a bootstrap into the registers!

I want to feel them [click]...

Re:I disagree

[Anonymous Coward]

Totally agreed --

How can you emulate the experience of getting a maintenance notice
in the mail from DEC that included a software patch on DECtape and
explicit instructions on how to patch the hardware via wire-wrap?

Or getting out the oscilloscope to set the baudrate on your PDP-11/05? And then
booting said 11/05 by
1) entering a program in octal via the front panel that is just
good enough to read a bootstrapper from paper tape,
2) jumping to the boostrapper from the front panel thus
3) reading a second boostrapper from paper tape
which in turn has a boostrapper to read from disk,
4) which in turn finally gets around to reading the bootblock
5) which might actually know something about booting RTS from that RK05 or RL10 or what-have-you.

Re:I disagree

[Rick.C]

Signed: your friendly PDP-11 system operator downstairs, 3 years from retirement

Being asked to train your outsource replacement is bad enough. Being asked to write your replacement is even worse.

But since you're the only one who really knows how it works, would you mind terribly?

Please?

Re:I disagree

[biobogonics]

And then booting said 11/05

Booting a PDP-10 was much more fun.

Re:I disagree

[SimonInOz]

Boy, that takes me back a bit!

All Well and good.

[jellomizer]

This is all well and good assuming there is an emulator for your mainframe. But what about lesser known mainframes such as the Prime. Where at their hay-day there were one of the big players but then they quickly moved to obscurity. There are a lot of other mainframes other then VAX and IBMs suff. And a lot of them are slightly obscure and near impossible to find emulators for.

Re:All Well and good.

[awx]

DEC never made mainframes - there was one "Mainframe-class" VAX, the 10000, but DEC always conceded that the VAXen were minicomputers.

Don't get me wrong - i'm not beating on them. I have three of my own...

Re:All Well and good.

[PhilHibbs]

You mean Pr1me?

Re:All Well and good.

[SQL Error]

The PDP-10, which preceded the VAX (and had a 36-bit word size) is considered by many to have been a mainframe.

Re:All Well and good.

[jellomizer]

Well kinda the company had 2 names Pr1me in like 1983 then later it became Prime. I am currently sitting behind about a hundred books that says "Prime" on it. I probably have enough resources available to make my own emulator if Only I had the time.

Re:All Well and good.

[Anonymous Coward]

s/quickly moved/crashed and burned/

And rightly so, I might add.....

Re:All Well and good.

[sphealey]

DEC never made mainframes

The DEC 36-bit systems were generally considered to be in competition with IBM's mainframes. DEC management did try to keep their perceived capaability just below the mainframe range to avoid provoking too strong a counter-attack from IBM - a strategy which many think was the beginning of the end for DEC.

sPh

Re:All Well and good.

[IPFreely]

Yeah, I used Primes for many years until they died. It would be a lot of fun to be able to bring up an old PrimOS prompt again.

I don't recall them ever publishing enough on their architecture to be able to reproduce it. There was an architectural guide, but it was mostly overview and behavior, and not so much instruction set detail. The ISA had so many really complex behaviors, like the stack frame and fault frames, not to mention four different instruction sets. It would be really hard to reverse engineer all that without some inside help.

I guess the question would be, what software of value is still needed that was only available on Primes?

Re:All Well and good.

[awx]

DEC also refused to call any VAX a Mainframe apart from the 10000, which it deemed "Mainframe Class".

Re:All Well and good.

[cmiller173]

Which is more likely, finding or creating and emulator vs. finding replacement parts for the old beast. And if you can find parts, at what expense?

Re:All Well and good.

[waynelorentz]

But what about lesser known mainframes such as the Prime.

How sad for me to find out that not only am I obsolete, I'm also obscure. I remember PrimeOS fondly if for no other reason than it was where I cut my teeth as a young hacker.

Would have been about 1984, dialing in to a univeristy's Prime in another state. Me on my Commodore 64 with my Microbits 300 baud modem (I was the fastest kid on the block. Everyone else had 110 or baudots). It was like NetHack, but in real life. Learned to get in. Learned to navigate. Learned about these great things called directories (unavailable on my Commodore 1541 disks). When I couldn't affort to buy new floppy disks, I started storing all of my programs and files on their mainframe. They seemed to have plenty of space. I was in there for more than a year before I got caught. Ah... those were the days.

Thank for you for the trip down memory lane.

Re:All Well and good.

[jellomizer]

Ohh we have parts! for the Prime. We are swimming in parts. And we don't see running out any time soon. We still have customers using the Prime and we still maintain them. But we would love to get our hands on some Prime emulators to help with the migration to a new platforms as well.

Re:All Well and good.

[jellomizer]

Well Prime is obscure compared to things like the VAX and other mainframes. It was widely used and still used by a lot of people oddly enough. But it is not like a the VAX or an RS6000 in term of popularity.

Re:All Well and good.

[DLWormwood]

Me on my Commodore 64 with my Microbits 300 baud modem (I was the fastest kid on the block. Everyone else had 110 or baudots). It was like NetHack, but in real life. Learned to get in. Learned to navigate. Learned about these great things called directories...

I do not think that word means what you think it means...

Re:All Well and good.

[doctorfaustus]

Yeah, I used Primes for many years until they died. It would be a lot of fun to be able to bring up an old PrimOS prompt again.

You think 10 years, 20 years from now we'll be saying this kind of thing about Windows?

Heh! I doubt it!

It depends...

[gschwim]

...on how many bugs your emulating under your Windows 2004 Emulator Edition.

Emulation is great ..

[z0ink]

.. but you forget the reason people dont upgrade is that it costs money to do so.

Re:Emulation is great ..

[Halo1]

The large cost of upgrading is not the hardware cost, it's the migration cost. If your new hardware can emulate exactly the same environment where you came from, this cost can be reduced immensely.

Re:Emulation is great ..

[Anonymous Coward]

But a lot of the cost involved is porting over old code from an obsolete system. Hardware is cheep. Software is cheep. Paying qualified people to do work - that's expensive.

Re:Emulation is great ..

[photon317]

Some of the large old systems cost huge $$$ just in run and maintain costs on a monthly basis though, which can provide sufficient impetus for finding a better solution when you show management how much it's costing you per year to stay on that hardware versus cheaper run and maintain costs elsewhere.

Re:Emulation is great ..

[Anonymous Coward]

people don't upgrade because it costs money
in general, businesses don't upgrade because they don't want shit to break in the process and fuck up a perfectly good system.

Good idea... but...

[AdeBaumann]

Before I even R'd TFA, I thought about one big problem:

How are you going to emulate a 5.25 inch drive to read old disks?

Re:Good idea... but...

[Anonymous Coward]

Either convert them yourself, or pay someone to do it.. The point is the efficiency of emulation rather than hardware compatability..

Re:Good idea... but...

[Anonymous Coward]

to add to the parent, companies like Intermedia will convert/image your disks for you.

Re:Good idea... but...

[antispam_ben]

How are you going to emulate a 5.25 inch drive to read old disks?

You can fold over an old 8-inch floppy disk and shove it into a modern 5 1/4" drive, but it still won't read it.

Re:Good idea... but...

[MajorDick]

5.25 ????
Damm are you young, when someone says 5.25 disks I dont even think too old.
Bring on the 8 Inchers, now those were the days they also fly great, late Friday afternoon we would all let loose by flying them around like frisbees, all fun and games till someone get a corner in the eye.

Re:Good idea... but...

[phoenix-gb]

Emulating a 5.25" isn't actually 100% necessary. The FDC that controls a 3.5" floppy is quite capable of controlling a 5.25" floppy in all three modes (DSDD, SSDD and SSSD) for reading any 'PC compatible' formatted disk. There are also a number of hardware options, such as the CatWeasel, that cen be used to drive a standard 5.25" drive to read non-PC Compatible disks.

Admittedly, you're still probably better off just using such a drive to create images of real world disks. Emulating drives with images is something that has been the core of almost all emulations of disk-enabled devices

The big problem, as far as I can see it, is reading 8" discs (as these do not, I believe, use the same controller as the others), or proprietry discs.

Re:Good idea... but...

[savagedome]

Bring on the 8 Inchers
late Friday afternoon we would all let loose by flying them around like frisbees

And then I saw your user id...

Re:Good idea... but...

[Jahf]

Honestly, is it such a big deal these days to dump your media onto a network store and -then- switch over to the emulated system?

Re:Good idea... but...

[aardwolf204]

With a 5.25 inch drive. They fit in 5.25 inch bays, right below your CD-ROM. Also, we've got External USB 3.5" drives, I see no reason why the same technology wouldnt work for 5.25" drives.

Re:Good idea... but...

[Anonymous Coward]

How are you going to emulate a 5.25 inch drive to read old disks?

Easy, it's an emulator. It will just multiply your 3.5" drive by 1.5.

Re:Good idea... but...

[LWATCDR]

A lot of the decs used 5 1/4" Hard sectored floppies. Where you would get a drive to read those or even blank disks these days?

Re:Good idea... but...

[awx]

A PC 5.25" drive won't read hard-sectored RX50 (DEC VAX) floppies.

You use your old VAX to make disk images for you before you give it to a collector.

Re:Good idea... but...

[ikegami]

Some programs create virtual CD-ROM drives when you give them an (.iso) image of a CD-ROM. If the computer you are emulating had a some kind of old disk drive, the drive itself would be emulated. The data would be read from a file on the emulator's host's hard drive. Or maybe it could even read 3.5" disks that were formatted to hold no more data than the old disks. Either way, you woulnd't actually use one of the old disks.

Re:Good idea... but...

[awx]

A PC 5.25" drive won't read hard-sectored RX50 (DEC VAX) floppies. The world is not a PC.

You use your old VAX to make disk images for you before you give it to a collector.

Re:Good idea... but...

[JohnnyGTO]

Try emulating a paper tape reader....

Re:Good idea... but...

[ikegami]

Some programs create virtual CD-ROM drives when you give them an (.iso) image of a CD-ROM. You could say that the said program is emulating a CD-ROM drive.

Similarly, if you were emulating an old computer which has some kind of old disk drive, the drive itself would be emulated along with the rest of the computer. The data would be read from a file on the emulator's host's hard drive. Or maybe it could even read 3.5" disks that were formatted to hold no more data than the old disks. Either way, you woulnd't actually use one of the old disks.

Re:Good idea... but...

[drinkypoo]

For most of those it should be pretty easy to build some kind of interface. Your average shop isn't going to do this but it doesn't make it impossible. There are interfaces for the PC to make it speak to mac floppy drives, for example, but mac floppy drives are relatively fast and complex compared to these old eight inchers and such. I should think that you could speak to it with a fairly simple microcontroller these days.

Re:Good idea... but...

[dmitriy]

> Try emulating a paper tape reader....

The article (which I actually read on paper) talks about taking the tape, scanning it in segments, splicing, and using custom image recoginition to convert little circles into sequences of bits and bytes.

8-inch floppy disks is a diferent story.

Re:Good idea... but...

[SillyNickName4me]

> Honestly, is it such a big deal these days to dump your media onto a network store and -then- switch over to the emulated system?

Lets say you have some ancient system with 8" floppies and no networking whatsoever. How are you going to dump those media on your network then?

(and no, not a troll, there are plenty old systems out there that did not network well or not at all)

Media conversion is the one biggest problem of emulation unless you can still access the original hardware and somehow connect it to a modern machine (as is done for exampel with Commodore's 1541 floppy drive) or alternatively, you get some modern reader to read ancient media.

Re:Good idea... but...

[SillyNickName4me]

Uh, there is a pdp-8 emulator aroudn that does this quite well.. the problem is more like.. where the hell do I insert that paper tape? :)

Re:Good idea... but...

[SillyNickName4me]

> Some programs create virtual CD-ROM drives when you give them an (.iso) image of a CD-ROM. You could say that the said program is emulating a CD-ROM drive.

Uh no, that is not the same at all.

An ISO image contains a filesystem that can be mounted using a loopback device or similar.
It in no way whatsoever tries to emulate a drive.

> Similarly, if you were emulating an old computer which has some kind of old disk drive, the drive itself would be emulated along with the rest of the computer. The data would be read from a file on the emulator's host's hard drive. Or maybe it could even read 3.5" disks that were formatted to hold no more data than the old disks. Either way, you woulnd't actually use one of the old disks.

Yes that works, but the problem is.. how do you get you old data in those files on your modern disk?

There is an aditional problem when you want to emulate devices like Commodore's 1541 and sfd 1001 drives, those are not like todays dumb floppy drives at all, they contain 1 or 2 (in case of the sfd 1001) cpus and you can run software on them. This feature was used quite extensively and you have to emulate that (which is done quite well by emulators like Vice for example)

Re:Good idea... but...

[jabberw0k]

how about a Serial RS-232 link... Kermit, that hoary standard, still is one of the best cross-platform transfer methods for machines of pre-Ethernet vintage.

At 9600bps you can transfer even the highest density 8" floppy in under half an hour!

(Ah, shades of my college job where our 'network' was oddball generic-MSDOS machines all with serial links to the VAX 11/750 in the back room with *Three* 30MB 14" Winchester drives, almost 100MB online, rah! Don't try and power all three up at once, though, or you'd blow the 100Amp breaker.)

Re:Good idea... but...

[biobogonics]

How are you going to emulate a 5.25 inch drive to read old disks?

Write really good software! Even the Apple II Disk II drives, which depended heavily on the CPU have been emulated at the hardware level. Rather than reading sectors from a file, some image formats contain the stream of bytes read by the drive hardware. [Data encodings, bytes per sector and address and data marks were defined in *software* rather than by a disk controller chip. There were a large number of screwball formats, particularly in copy protected game software.]

Re:Good idea... but...

[Hatta]

late Friday afternoon we would all let loose by flying them around like frisbees, all fun and games till someone get a corner in the eye. ...and then it's hilarious.

Re:Good idea... but...

[SillyNickName4me]

Uh yes, like I said, one of the 2 alternatives you have is usign the old hardware to read the original media (and then you transfer it with whatever communications link you can manage). Point is, in order to do this, you need the old hardware, so this is no option whatsoever for replacing a machine that has given up, it is only an option when you do it beforehand.

Re:Good idea... but...

[ikegami]

Yes that works, but the problem is.. how do you get you old data in those files on your modern disk?

Aye, that's very tricky. If the drive isn't compatible with modern hardware, even geting one's hands on the a drive (by buying, renting borrowing or using someone else's) won't be sufficient. You'd have to get your hands on the entire computer, write a program to read the data from the disk and send it to a PC, which is tricky in of itself. Does the old computer have a modem? a serial port? a parallel port? The local-ish nuclear power station has a PC emualuting a papertape machine, enabling communication between the PC and the old '72 computer that controls the reactor.

Re:Good idea... but...

[Jozer99]

Intense concentration, or just push the disk really hard into a slot loading CD-ROM drive.

Re:Good idea... but...

[SillyNickName4me]

Indeed. While this is usually not much of a problem when emulating a game system, it is a rather big problem when emulating more 'serious' systems where user data must be transfered to the emulator.

Re:Good idea... but...

[Jeremiah Cornelius]

O.K.

You get crucial "VAX" points. I missed the VMS generation from DEC, which seems to have been real cultish!

No - the point of this post is to give you KUDOS for your WEBLOG! The title alone, is worth the price of entry...

Re:Good idea... but...

[Anonymous Coward]

Have you been living undera rock. 5.25 inch is ancient, obsolete.

Re:Good idea... but...

[bhtooefr]

(Blatant plug for Apple II Oasis, IMO the best A2 emu except for two fatal flaws - doesn't run in Wine, and it's shareware)

A2O also does this, except it connects to a real Apple II, not a Disk II, and it transfers disk images via serial. It's actually sold as an emulation environment to migrate from Apple IIs to PCs.

Re:Good idea... but...

[SomeGuyFromCA]

> all fun and games till someone get a corner in the eye.

It's all fun and games till someone loses an eye.

Then it's a new game called find the eye.

Re:Good idea... but...

[FLEB]

I doubt it would be too hard to rig up a better reader than that... some motors (even a crank might work), some photocells, a few electronics, then rig it all up to a serial interface.

I'm not saying I personally have the skills, but I can't see it being much harder than things like homebrew barcode readers I've read about.

Re:Good idea... but...

[FortranDragon]

A PC 5.25" drive won't read hard-sectored RX50 (DEC VAX) floppies.

Have you tried the solutions mentioned on this page http://home.iae.nl/users/pb0aia/vax/rx50.html ? I remember using something similar for my DEC Rainbow 100.

Re:Good idea... but...

[Anonymous Coward]

> Some programs create virtual CD-ROM drives when you give them an (.iso) image of a CD-ROM. You could say that the said program is emulating a CD-ROM drive.

Uh no, that is not the same at all.

An ISO image contains a filesystem that can be mounted using a loopback device or similar.
It in no way whatsoever tries to emulate a drive.


Uh yes, the original poster is correct. There are *emulator* programs that use an ISO image file as the backend store to a complete CD-ROM interface emulation. This same thing is done with FAT filesystem images to emulate hard drives (freedos, Any of the MAME games with hard disk images). MPEG-2 video streams are often used to emulate laserdiscs (no, not a mistake). Laserdiscs are basically raw digitized NTSC video so there is obviously a conversion process required before hand, it is not exactly like using an ISO to emulate a CD/CD-ROM drive.

Loopback mounts are not the ONLY thing you can do with disk images.

Dickwads like yourself should take the time to consider that other posters might actually know something you don't.

Re:Good idea... but...

[Rick.C]

The big problem, as far as I can see it, is reading 8" discs (as these do not, I believe, use the same controller as the others), or proprietry discs.

An 8" controller chip will also handle 5.25" drives, but not always the other way 'round.

Most proprietary disks are only odd in the software/formatting area, although hard-sectored disks require special circuitry to handle the extra INDEX pulses.

The original Apple format was a totally different beast, hardware-wise. That's why an S100 disk controller card cost $350 back in the day and the Apple disk controller card cost only $80.

Re:Good idea... but...

[Anonymous Coward]

Aye, that's very tricky. If the drive isn't compatible with modern hardware, even geting one's hands on the a drive (by buying, renting borrowing or using someone else's) won't be sufficient.


No, then you haven't actually tried it. If I can obtain a working mechanical and electrical apparatus then it is usually no problem whatsoever to get the thing to work with modern hardware.


You'd have to get your hands on the entire computer


No, this is not necessary unless the hardware is very complex and documentation is unavailable. Since most of this older equipment uses discrete logic components it is generally no problem to hack up a interface to the appropriate port on your PC. Of course, it helps having one of these PCB milling devices.

Re:Good idea... but...

[Anonymous Coward]

Scan it with a scanner, and do something like OCR.

Re:Good idea... but...

[zeath]

Have you been living undera rock. 5.25 inch is ancient, obsolete.

That's the entire point. The PDP-11s and the other hardware they want to emulate are also ancient. And some of that hardware uses 5.25" or 8" disks as part of its daily operation. When emulating a processor they also need a way to emulate the hardware peripherals, including storage.

Re:Good idea... but...

[oO Peeping Tom Oo]

Easy. Turn all the disks into images.

Re:Good idea... but...

[SnarfQuest]

A PC 5.25" drive won't read hard-sectored RX50 (DEC VAX) floppies.

RX-50 floppies are not hard sectored, and there are several programs that allow reading them on PC's.

Re:Good idea... but...

[mauddib~]

I've heard of people making copies of them by placing them under a xerox. Maybe it works with a scanner too!

Re:Good idea... but...

[SillyNickName4me]

> Uh yes, the original poster is correct. There are *emulator* programs that use an ISO image file as the backend store to a complete CD-ROM interface emulation.

Rereading the original post I see this was indeed what the poster was refering to.

> Loopback mounts are not the ONLY thing you can do with disk images.

I never said so.

> Dickwads like yourself should take the time to consider that other posters might actually know something you don't.

Any reason you think you have to go calling names? (all the more funny since you prefer to stay anonymous). If you had taken a minute to read some of my other posts in this thread you'd see that I'm rather well aware of the concept of using a file together with an emulator to emulate both the media and the hardware, so typical case of pot and kettle I'd say.

Re:Good idea... but...

[WNight]

If the old system has any type of output at all you've got a few options. There's a system designed by Xerox that uses a system of alternating / and \ marks to store data. It scales well, you can either print it out with a standard font (doable by even a daisy-wheel of ball printer) or so small that it appears to be a uniform gray shading. The data is encoded such that you can lose any given percentage of it (you choose the redundancy) and still read the data. The software would need to be custom written for the specific obsolete machine, but it's pretty easy to do. Then you OCR it (the alternating slashes are trivially easy) and read it in on a PC.

If the machine or the storage device is dead, rent time on someone else's machine and do the same thing as above.

Sure, data is the important thing, but it's not impossible. I saw an apple 2 program that sent data to a PC's serial port by toggling the joystick port (one bit per port, four ports, I think) in a specific pattern over a custom cable that they provided a diagram of. You could always do a hex dump of the data and take a picture of each screen (automate it) and OCR it in...

Re:Good idea... but...

[anynameleft]

"How are you going to emulate a 5.25 inch drive to read old disks?"

Intuitively, I'd say the following, rather easy, procedure could be used:

Get a 386

Put rawread on a 3.5" diskette

Create a disk image of that 5.25" floppy

Transfer it to your new PC using the 3.5" drive

Therefore, I think other formats are much more challenging. For example:

How about the 800kib 3.5" floppies that the early Macintosh computers used? (has indeed been answered, there is PC hardware for them)

How about the 8" floppies?

And what to do with those 3" floppies, if you don't have a CPC ready?

Re:Good idea... but...

[SillyNickName4me]

> Sure, data is the important thing, but it's not impossible. I saw an apple 2 program that sent data to a PC's serial port by toggling the joystick port (one bit per port, four ports, I think) in a specific pattern over a custom cable that they provided a diagram of. You could always do a hex dump of the data and take a picture of each screen (automate it) and OCR it in...

Yes, it is quite possible for as long as you have access to the old hardware. As you point out correctly, it may require custom solutions, and what I was trying to point out is that it won't be a matter of 'just' dumping the data in a file in quite a few cases, because of what you say and because such hardware might not be that trivial to emulate correctly when it contains its own 'intelligence'.

I use a parallel port and a custom cable to connect an old Commodore 1541 to my PC, and an ieee-488 i/o card to connect a cbm sfd-1001 floppy drive, both for converting old media, but as always, it requires access to the old hardware ('intelligent floppy drives' in this case).

If access to the old hardware is relatively easy still then it saves a lot of hastle to just use the old hardware to run the software. Transfer solutions are still a rather good idea tho just in case you all of a sudden can no longer get the 'real' stuff and for backup purposes.

Re:Good idea... but...

[jeremyp]

OK....

What if all my data is *on* one of those old discs? How do I get it off if the hardware's packed it in?

Re:Good idea... but...

[jeremyp]

My PC doesn't have a 3.5inch drive.

I'd like to say that is going to be an issue, but I have had a laptop without a floppy drive for four years and never missed it.

Re:Good idea... but...

[SillyNickName4me]

Oh, and on another note..

> Laserdiscs are basically raw digitized NTSC video so there is obviously a conversion process required before hand.

Laserdisk video is analog, they may (in the last incarnatiomn of the format) contain digital audio. Also, the format was pioneered by Phillips and Magnavox, and while the format has never been very popular in Europe, seeing how Phillips is a Dutch company, I am pretty sure that the first disks around were actually pal disks, at any rate, pal/ntsc/secam are all possible.

It is of course no problem to 'emulate' a movie Laserdisk and player by means of an mpeg-2 file and player (or any other format you desire) if what you mean is providing a functional equivalent based on the same content. The pre-processing required includes a frame by frame capture of the analog video and then compressing it in your format of choise. From there the step to emulating the functionality of an interactive player + disk is not that big, tho I never saw it in practise. Happen to have a link regarding emulating a pc connected interactive player + disk by any chance?

With regards to the Laserdisk format, take a peek here: History of the Laserdisk format

The Stability of New Products vs Old

[Anonymous Coward]

Although this is a very good idea I question the stability of a new emulator vs an old proven system.

By using the original the kinks have already been worked out, quirks are known and understood, and everything just works.

By creating an emulator you have bugs to smash, that's just the way software is. Also keep in mind this seems to apply to big businesses (financial, medical) and large organizations (NASA) with legacy hardware. Since the stability of these systems is absolutely crucial why would they want to switch to a new, unproven, buggy system that stick with the old?

Re:The Stability of New Products vs Old

[Otter]

By using the original the kinks have already been worked out, quirks are known and understood, and everything just works.

And since that's precisely the reason why you're keeping that dinosaur on life support instead of replacing it with a Sun, it's a particularly important concern when you're considering replacing it with emulation on a Sun.

Re:The Stability of New Products vs Old

[GeoGreg]

Well, for one thing, the old, (formerly) stable hardware may be failing. It might be easier to get hold of a PDP/11 emulator being used (and, hopefully, improved) by multiple organizations than to attempt to translate the in-house PDP/11 assembly code into something that will run on a PIV Linux box. Especially if the people who wrote the legacy app are retired/laid off/dead.

Re:The Stability of New Products vs Old

[quarkscat]

There is no equivalence. For instance, DEC VMS
was used as the design "core" for Microsoft's
Windows NT. I have known of DEC VAX hardware
that ran continuously for 5 years without a
warm reboot, let alone a system shutdown. The
Microsoft OS often needed to be rebooted daily.

The hardware that Microsoft runs on is not as
reliable as the old DEC VAXs, as a rule. The
short term emulation of a legacy system is not
the same as replacing it. For exammple, an IBM
z/390 running MVS might be able to run 1000
linux servers, but in terms of reliability
(the proverbial 5 Nines), that z/390 could not
be replaced with 1000 linux boxes, or even 2000.

The old adage "They just don't make things the
way they used to." applies here. New hardware
costs are way down, as are HW/SW maintenence
costs, but the reliability of the new gear is
underwhelming.

Re:The Stability of New Products vs Old

[Anonymous Coward]

Well, yes you are right.

But emulation is at a very low level, and old computers were really not very complex.

If you have good documentation, it's not very hard to write an emulator, You'll only have problems with programs using undocumented features. And serious stuff tend not to do that.

Also, you can actually test the emulator while using the old computer. Often, there are tools written for the old computer for testing different things.

(I've done it before.)

Re:The Stability of New Products vs Old

[Karnak23]

I agree in principle, switching from old, reliable systems to new unproven systems is risky at best. The FAA however has done this for decades.

I recall an NPR article from the earyly 90's regarding FAA upgrades the the nations air traffic control systems. The article covered the fact that, the FAA had been using emulators on top of emulators for years because the hardware was not only kaput but, so was the company that made it and the plans/schematics were no where to be found.

Nice thought to keep in mind the next time you fly.

Re:The Stability of New Products vs Old

[gosand]

Although this is a very good idea I question the stability of a new emulator vs an old proven system.

Well, why not get the thing emulated now, so you can work the kinks out of the emulation while you still have a working system to compare it to? Before you know it, anyone who knows anything about these systems is going to be dead. (sorry, it's true) Would you want to hire anyone to work on an emulator for a system that was in use years before they were born? Or would you rather have people working on it who have used the systems?

It may seem kind of pointless to emulate these things now, but if you don't start on it now, when will you?

Re:The Stability of New Products vs Old

[Jeff DeMaagd]

I thought that Alpha sysems running OpenVMS can emulate the old VAX systems. Not sure if PDP is VAX though. DEC had been doing emulation on their new hardware to provide a smooth upgrade path to Alpha. They were pretty good at it too.

I would also think that rather than emulating old hardware to run on new hardware, it would be better to rewrite the software every couple of decades.

Re:The Stability of New Products vs Old

[mustangsal66]

...and large organizations (NASA) with legacy hardware

Like the Space Shuttle

Re:The Stability of New Products vs Old

[Anonymous Coward]

5 nines means 5 minutes of down time per year. Over 1,000 servers, however, this is 5 minutes per server which is more than enough time to get a new one online, and adds up to over 3 days per year for actual maintenance rather than just failover.

1,000 servers is FAR easier to get 5 nines on than any single server. You never heard of "single point of failure", bub?

Re:The Stability of New Products vs Old

[gardyloo]

"... Especially if the people who wrote the legacy app are retired/laid off/ dead ."

Ooh. That's going to be some downtime there, bub. Maybe we can emulate them.

Re:The Stability of New Products vs Old

[Anonymous Coward]

For instance, DEC VMS
was used as the design "core" for Microsoft's
Windows NT

I think it was actually Mica, Dave Cutler's next-gen VMS (or VMS+, if you will). Microsoft actually settled with Digital regarding the code they were using after he moved.

Re:The Stability of New Products vs Old

[Anonymous Coward]

small correction - a zSeries IBM mainframe can run 1000 or more Linux servers only if it is running the z/VM operating system, not z/OS (aka MVS) as the parent post stated.

legal issues?

[Turn-X Alphonse]

What about legal issues here, Nintendo own the rights to emulation if I recall correctly. Couldn't they pull the plug on this?

Emulating a server may also cross a few patents Microsoft own..

Re:legal issues?

[Anonymous Coward]

You mean emulating a Nintendo or emulation in general? The rights for emulation in general, if its possible to claim intelectual property on something that abstract, should be held by Mr. Turing, since he proved the emulation theorem and demonstrated how to do it, before there even were any computers.

Re:legal issues?

[Anonymous Coward]

Nintendo owns the rights to "emulation"? Emulating anything?

Re:legal issues?

[meringuoid]

Nintendo owns the rights to "emulation"? Emulating anything?

I think they patented some aspects of emulation, mainly to shut down people selling GBA emus for palmtops.

Given the current state of patent law, chances are that any universal Turing machine now owes Nintendo royalties.

Re:legal issues?

[Student_Tech]

I belive the Nintendo thing was emulating a handheld on another handheld GBA and optimizing it or something like that. /. article: Nintendo Patents Handheld Emulation, Cracks Down

Re:legal issues?

[Matheus Villela]

Emulation comes out with the first upgrade of a CPU.

If they have total rights of emulation(this isn't true, already mentioned in other replys) then they could stop the hardware industry.

An emulator is something wich does the some output or imput of the emulated hardware, all these videogames emulators are only "true emulators" if they output the same as the real hardware and receive the imput from acessories.

A more detailed view:
You will have an Super Nes emulator for the point of view of your TV if you output the grafics in your TV using the TV-OUT from your GPU.
You will have an Super Nes emulator from the perpective of the joystick if you adapt your joystick to use in the computer, then the joystick can send the same data to your computer that it sends to the real console.

Re:legal issues?

[SillyNickName4me]

> Given the current state of patent law, chances are that any universal Turing machine now owes Nintendo royalties.

I'd say.. prior art is undisputable in this case, so that would stand no chance whatsoever in court.

Re:legal issues?

[_KiTA_]

Well, that, and Nintendo's latest cash cow is emulated ports of older games. For example. NES games on GBA, NES roms in a EMU in Animal Crossing, etc. Of course they're gonna put up legal loopholes so Sony et all can't use their vast library of games from the 1980s to... er, wait, nevermind.

Re:legal issues?

[gosand]

Given the current state of patent law, chances are that any universal Turing machine now owes Nintendo royalties.

That would be one way to get a Turing machine to admit it is a machine - sue it.

Re:legal issues?

[Anonymous Coward]

Way to understand the joke, asshole

Re:legal issues?

[SillyNickName4me]

If only the joke didn't seem to be reflected by reality. On another note, stick your head in that what you think you have to call me.

DEC VAX -- why not port?!

[bunburyist]

If the DEC VAX systems are so antiquated, wouldn't it make more sense for businesses using these systems to simply upgrade to newer/better technology? I mean even if they have incredibly stability and "wow" factor, wouldn't it be easier than both solutions just to upgrade the software to newer systems. Uptime on many linux/unix systems are just fine for most usage, and QNX has their real time fail proof operating system, but I doubt people using VAX would even need something that powerful.

Re:DEC VAX -- why not port?!

[Jeff Mahoney]

If it works, don't fix it.

Re:DEC VAX -- why not port?!

[CarrionBird]

just upgrade the software?

Methinks you underestimate how badly software projects of that sort often go.

Re:DEC VAX -- why not port?!

[Mordaximus]

but I doubt people using VAX would even need something that powerful.

VAX users aren't typical, they're likely hospitals, banks or gonvernment institiutions that rely on the incredible stability. Besides, VMS is alive and kicking, why redo software more than necessary? Besides, Darl hasn't sued VAX owners yet!

Re:DEC VAX -- why not port?!

[bloggins02]

If I have a 1955 Chevy that still gets me to work everyday without breaking down on the side of the road, what point would there be in buying a new car?

Re:DEC VAX -- why not port?!

[Captain Rotundo]

bad example. With a newer car you would gain safety and economy, and reduce polution.

With a newer computer *if* the old one does what you need you get virtually nothing.

Re:DEC VAX -- why not port?!

[LWATCDR]

Well the only way to just "upgrade" the software is to run it under an emulator.

It may be possible to just "recompile" for VMS on new hardware but you have to have the source for that. You may run into some real issues during that recompile due to different versions of the OS.

Re:DEC VAX -- why not port?!

[PingPongBoy]

If something breaks or if the capacity is reached, what then? Are parts available? Can an upgrade be done economically?

An upgrade path could be way overdue

Re:DEC VAX -- why not port?!

[flabbergast]

There are a couple reasons, most mentioned previously.

1. Uptime on linux/unix is not good enough. Linux/Unix versus Microsoft, sure its got a better uptime, but against a mainframe that hasn't been turned off in 20-30 years? We're not talking commodity hardware here that'll break in a few years, or need constant replacing. Banks, hospitals, etc can't afford any down time.

2. The code is probably long gone, and the software was probably custom. So, how exactly are you going to "port" this software? Oh, I know, tell all those banks to use gnucash.

3. If it ain't broke, don't fix it. All their employees know how it works. All their sys admins know the quirks. If you move to a new system, you've got thousands of employees who have to learn the new system from sys admins who don't know the software.

When it comes to upgrading a mainframe, its not like upgrading your home computer or home file server. You don't simply copy your ~/ over to your new computer and start running firefox.

Re:DEC VAX -- why not port?!

[fishbowl]

"If I have a 1955 Chevy that still gets me to work everyday without breaking down on the side of the road, what point would there be in buying a new car?"

$3.00/gal gas in a car that gets 6 MPG.

I felt strongly enough about my '59 Impala that I set the costs aside for a few years, though.

Re:DEC VAX -- why not port?!

[Anonymous Coward]

If something breaks or if the capacity is reached, what then? Are parts available? Can an upgrade be done economically?

An upgrade path could be way overdue

Thus, the emulator was proposed.

Re:DEC VAX -- why not port?!

[mark_lybarger]

you're saying that newer computers aren't "safer" from attacks or whatnot, more economical to use, and don't reduce polution .vs. their 20 year old counter part?

interesting.

Re:DEC VAX -- why not port?!

[TedCheshireAcad]

But if it doesn't integrate with Office and Exchange Server, embrace and extend.

Re:DEC VAX -- why not port?!

[paganizer]

Safety for others, not for you. the chevy will not even be slowed much by impacts with most modern vehicles; you only have to worry about other vehicles actually made of metal and stationary objects.
as to economy, a well tuned big block with tons of performance will get 24-29mpg and low pollution if you keep your foot out of it, or more.

Re:DEC VAX -- why not port?!

[Anonymous Coward]

We replaced a bunch of 15-year-old (at the time) computers that controlled some complex industrial equipment. We upgraded the I/O hardware at the same time. The equipment itself was pretty much unchanged.

We went from a computer the size of a large file cabinet to a computer the size of our phone book (3/4" thick) that fit inside the I/O cabinet. The new I/O hardware takes up about half the space as the old hardware.

That's just the space savings. We figure we have 20 times the processing power than we did before, and this is buying low-end hardware.

Granted, the new system is not binary-compatible, but it came with a compiler that let us use the old code virtually unchanged. The hard work was replacing the green screen dumb terminals with a GUI interface.

Everyone's happy. We no longer need to find exotic spare parts or rebuild parts we just can't get any more. They don't break nearly as often. The realtime code is no longer maxing out its time slices. The developers can do updates over ethernet rather than hauling disk packs down the hall (walk the length of a football field, power up computer, spin up disk pack, boot, start program, test, exit program, spin down disk pack, power down computer, walk back down hall, make 1-line change to code, rebuild takes 45 minutes, spin up disk pack, copy changes, spin down disk pack, and walk back down the hall. Repeat as required. Got lots of exercise). The users have a pretty and easier to use interface. Backups onto DAT instead of 9-track reels.

Re:DEC VAX -- why not port?!

[Anonymous Coward]

VAX VMS -- the new UCSD P-System ^U er, java!

I wonder why...

[5m477m4n]

by Bob Supnik, Sun Microsystems

Gee, I wonder why he would be recommending buying new servers?

Re:I wonder why...

[Anonymous Coward]

Had this been written by someone from IBM, HP or Dell would you have posted the same comment?

Re:I wonder why...

[5m477m4n]

Actually yes. My point being, it's not suprising that a hardware manufacturer wants us to throw away older, most perfectly usable, equipment in favor of buying new stuff from them. It would be the same as Microsoft recommending that people upgrade their OS every 3 years... oh wait, they already do.

Re:I wonder why...

[kpansky]

"Perfectly usable"... VAX... VMS... hahaha hhahaaahahhahaha. Good. Got that out of my system.


HA HAHAHAHAH!

Re:I wonder why...

[SillyNickName4me]

> "Perfectly usable"... VAX... VMS... hahaha hhahaaahahhahaha. Good. Got that out of my system.

And while I hope it never happens to you, if you happen to get into hospital, there is quite some chance that your information will be registered on a.... VAX. that is right, an old, according to you obviously unusable VAX.

Next time you transfer some money, chances are quite good that your order will be processed by... again one of those unusable vaxen...

I suggest you delve a bit into the matter before laughing your ass off because this makes you look like someone who knows very little of what is being used.

Re:I wonder why...

[kpansky]

I am well aware that VAXen are still in use. I have personal experience with them. That is precisely where I got my opinion from. A large number of people on my program had to work with them and deal with them. No one had a good word to say about it. While it may "work" and be a reliable back end system, "perfectly usable" seems to imply human interaction; and in this regard, it is horrible sub par.

Re:I wonder why...

[SillyNickName4me]

Well, I'd not encourage people to 'use' them for interactive use, yet I do run my mailserver on a VAX (running NetBSD, not OpenVMS btw). It is not fast, but it does not slow down easily either.

At any rate, usable != interactive use.

Re:I wonder why...

[Anonymous Coward]

I've been working with Bob for a few years now. He's been into this *long* before we joined Sun Microsystems as part of the Nauticus Networks acquisition in February.
He's just a old school geek. He'll talk about this sh*t for days.

Re:I wonder why...

[Anonymous Coward]

He's not in that division of Sun. He came in as part of the Nauticus acquisition.

TX-1

[Euphonious Coward]

Is there a simulator for the TX-1? People talk about how great Sutherland's Sketchpad software was, and it ran on the TX-1. Is there even enough public information to simulate it, and does the Sketchpad code still exist?

More important question:

[Enahs]

How do you emulate an 8 inch drive?

Re:More important question:

[proj_2501]

fold the disk a couple times. duh!

Re:More important question:

[garcia]

Make a disk image of the original and read from that? Or like the C64 emulators (single example) do when they read from C64 floppies under emulation and hardware hacks?

Re:More important question:

[Jahf]

I don't have this problem ;)

Re:More important question:

[Anonymous Coward]

But how can you read the original in the first place is the gist of the question. You still need at least one 1541 drive somewhere in the chain.

Re:More important question:

[mopslik]

How do you emulate an 8 inch drive?

Simple, you just reply to some of those emails you ge-- oh, wait. I see now you said "disk" size.

Re:More important question:

[MindStalker]

I think their big thing is supporting legacy programs. Say your company spent millions of dollars back in the late 80's for some big iron, and some awesome server database stuff for that. Lets say these programs still meet your companies needs. But the hfardware itself is ailing, but you still need the software to run. Buy new machines to emulate the old machines. The actual data transfer can be done in many ways other than 1541 drives and I'm sure you can emulate the need to input data in old ways.

Re:More important question:

[Deagol]

Dude, he said emulate, not mutilate.

Re:More important question:

[rgmoore]

How about bend, fold, spindle, or staple?

Re:More important question:

[Anonymous Coward]

"But the hfardware itself is ailing"

I'm not trying to be an ass, your post is fine... but how did that f move over eighteen places? That's just freaky.

Re:More important question:

[MindStalker]

I'm a bit obsessive compulsive, I tend to highlight things with my mouse a lot, and sometimes drag letter or whole paragraphs around accidently, without noticing. Its really annoying.

Re:More important question:

[mindriot]

Make a disk image from the original

...using your wonderful 8-inch floppy drive emulation software.

SCNR...

Re:More important question:

[Crackez]

Dig out a TRS-80 Model II and use the serial port?

it had a Zilog z80, still a common cpu. I am certain that someone could write a small program to dump the disk to a serial port.

I used to have one of those when I was oh, about 10 or 11... I kept the disks, 3 of them. they're at my parents house... Damn 8" floppies were fast compared to that tape deck I had for my TI-99/4A...

I also had a ton of 5.25" floppies from my PC AT... Ah, my 286... I wish i still had that. 20MB hdd, 512K of RAM, 6Mhz CPU.... I used to write games on that thing in qbasic. you know, dumb stuff, like 2d shooters in 16 color EGA...

That machine was bad ass...

eventually i got a 486, but that thing doesn't evoke near the nostalgia of the TRS-80... Pong was way faster in Basic-80 on there then it was in TI-BASIC on the 99/4A...

oh i miss those machines... I used to code for days on those things, till my mom would yell at me to get some sleep... I remember when i first got the idea of how to do non blocking IO in basic... Oh the memories...

damn, basic sucked... Too bad I didn't know asm then...

Didn't work for Intel

[Anonymous Coward]

This didn't seem to work for Intel with the Itanium. On the other hand, nursing old hardware did pay off for Intel with the x86, but EPIC stood dead in the water because of lack of (good) x86 support and lack of native software.

Emulate death!

[BobLenon]

Awesome, imagine how many PDP-11 one can emulate to operate a therac-25. How many people one could kill!

amazing ;)

Re:Emulate death!

[malfunction54]

therac-25. sweeeet.

Re:Eww...

[Anonymous Coward]

No shit. This, the 503 errors and Slashdot trying to re-invent itself as an "IT" site. What a joke.

Win

[jmke]

if everything can be mapped correctly sure why not, it will do no harm to have a backup system in case the real hardware crashes, hell I have a Windows 2003 virtual server running on a Workstation for testing server applications :)

PDP-11s *still* in use!

[maynard]

I know of a PDP-11/73 which to this day is still cutting sheet metal for a duct factory. The damn thing just won't die. And they're not likely to emulate since the I/O board interface between the computer and their machine tools would be more expensive to implement on a custom PCI card, along with emulation drivers, than simply buying excess used PDP-11 parts. Someday they'll have to face the music and actually buy a commercial solution, but for the moment they continue maximizing their return on investment for a computer system originally purchased well over thirty years ago. And why the hell not? --M

Re:PDP-11s *still* in use!

[jmke]

a very valid point, those custom I/O boards are a pickle, but sure there must be a way to address those through the host system.

Re:PDP-11s *still* in use!

[Anonymous Coward]

/.

Nearly all modern thermonuclear delivery systems are tested at some point on DIDACS. The system is based on a PDP-11/34, except for the TVC controller which is a PDP-11/24.

Been running for thousands of tests without a hitch. I know, because I wrote the code about 15 years ago and periodically I still meet the techs who are still running it.

I sincerely doubt you could emulate a unibus machine running RSX or TSX (we're talking REALTIME operating systems here boyz) on any PC-type architecture. We had trouble emulating it on a VMS VAX with a massbus!

--Charlie

Re:PDP-11s *still* in use!

[maynard]

Yeah, but if you're going to build an I/O board and write custom emulation drivers, what's the cost compared to buying a new off-the-shelf system and tooling for the shop? Imagine the cost to hire contractors to tease apart the I/O specs on that old stuff just so they could then build a custom board and drivers to emulate the old software on a P4 or something. Compare that to reinvesting in a new system with tools. It's the new tooling which really costs. They've got an old system which works right now and they don't see the point in dumping a couple hundred grand or more to buy new stuff which does exactly what their old stuff still does perfectly well. Can't blame 'em. --M

Re:PDP-11s *still* in use!

[maynard]

I seem to remember they said they originally started out with an 11/44 in the early to mid seventies, and then upgraded along the way. That would mean a fairly large transition from a hex to quad backplane system (I think sometime in the late '80s), but the software should have continued running just fine. The system boots RT-11, and which then starts their control application. It was originally purchased as part of a package from a tools manufacturing company, which is either long out of business or at least not supporting that old stuff any longer. The owner is a friend and I occassionally go in to take a look at the box, then we toss some beers. There's rarely anything to do, the damn thing is a tank. It wouldn't surprise me one bit to see the government still relying on old PDP-11s and VAX-11s for critical real-time applications. These systems just don't die. Why risk change when something important still works? --M

Re:PDP-11s *still* in use!

[Anonymous Coward]

Actually you can cheat, there are PCI-to-Q and PCI-to-Unibus adapter cards available, so you can keep a PDP-11 card cage with one or two custom cards in it and still replace the other 90% of the system. Works great (but of course I would think that, I'm in the PDP-11 emulation business).

Re:PDP-11s *still* in use!

[Anonymous Coward]

They haven't upgraded a critical piece of equipment in 30 years? No wonder US manufacters are losing out in the global marketplace.

Re:PDP-11s *still* in use!

[acceleriter]

thermonuclear delivery systems

Best. Euphemism. Ever. :)

Re:PDP-11s *still* in use!

[geomon]

If foreign manufacturers made their products to last 30+ years they would be out of business just like Digital.

The cheap-ass stuff made by today's computer manufacturers means my Bic lighter will last longer than your Dell 3.6GHz desktop.

Re:PDP-11s *still* in use!

[Anonymous Coward]

/.

I loved the processor light on the old PDPs. You can actually tell how fast the box is "thinking" from across the room!

I hope your friend's machine still has that little red light...

--Charlie

Re:PDP-11s *still* in use!

[bigbigbison]

At my father's shop they finally had to replace one of their old machines. It was a huge pain for them because they had to rewrite all the tool paths for the machine since the old programs couldn't be converted to the new program's format.

Re:PDP-11s *still* in use!

[SillyNickName4me]

> Yeah, but if you're going to build an I/O board and write custom emulation drivers, what's the cost compared to buying a new off-the-shelf system and tooling for the shop? Imagine the cost to hire contractors to tease apart the I/O specs on that old stuff just so they could then build a custom board and drivers to emulate the old software on a P4 or something. Compare that to reinvesting in a new system with tools. It's the new tooling which really costs. They've got an old system which works right now and they don't see the point in dumping a couple hundred grand or more to buy new stuff which does exactly what their old stuff still does perfectly well. Can't blame 'em. --M

Add to that that while a commercial solution might in itself notbe that extremely expensive, the changes to processes, and need for retraining, the resulting initial mistakes that will occur and so on cause quite a lot of extra costs.

Re:PDP-11s *still* in use!

[Anonymous Coward]

thermonuclear delivery systems

Best. Euphemism. Ever. :)

The scary part is, we actually used it. That euphemism conveniently lumps together gas generators used by launch mechanisms (incidentally, watching a 50 foot steel pipe go from horizontal to vertical in less than half a second is pretty, um, eye-boggling) and for TVC (thrust vector control) with actual propulsion motors and ACMs (attitude control motors, used for steering at hypersonic speeds) into one three-word jargonism.

But yes, it means "Weapons of Mass Destruction" which must be the 2nd Best. Euphemism. Ever. ;)

--Charlie

Re:PDP-11s *still* in use!

[swb]

they continue maximizing their return on investment for a computer system originally purchased well over thirty years ago

No, they maximized the return on their investment, probably years ago. Maximizing implies getting unfulfilled value out of their system. Unless their PDP is doing more today than it was yesterday, it's probably already achieved its maximal value.

A newer system could probably do more than their current system does now, the question is whether they have the business skills to expand the business to get maximum value out of their new system.

Re:PDP-11s *still* in use!

[Anonymous Coward]

They haven't upgraded a critical piece of equipment in 30 years? No wonder US manufacters are losing out in the global marketplace.

That's right. Technological advancements in cutting sheet metal for ducts has left this business out in the cold.

Guess what? The cutting-edge technology for this sort of thing is to have a computer do it. It sounds like this company has been on the cutting-edge for more than thirty years.

Re:PDP-11s *still* in use!

[foidulus]

Just shows you how little you know about manufacturing, new does not necessarily imply quality just like old doesn't mean it's of poor quality.
At the mill I worked at our oldest major piece of equipment is about 80 or 90 years old. No real reason to upgrade it, it does it's job. The area of the mill that I worked at, the machines were about 40 years old, controlled by a 10 year old VAX who communicated with even older PLCs to do the real-time work, and yet the steel we produce is among the highest quality in the world...

Re:PDP-11s *still* in use!

[Duhavid]

What more could a newer system do?

The only advantage of something new would be that it might require less electricity and floor space.

They are getting more out of the system without doing more than provide it with electricity and maintainance, if that. How is that not maximizing their return?

Re:PDP-11s *still* in use!

[Anonymous Coward]

Nearly all modern thermonuclear delivery systems are tested at some point on DIDACS. Been running for thousands of tests without a hitch. I know, because I wrote the code about 15 years ago

When David Lightman visited your office, which one of the two guys were you?

Re:PDP-11s *still* in use!

[swb]

It's somewhat semantic, but maximizing your use of some technology means that the technology is doing all possible jobs for which it is capable.

Using it longer doesn't really count in that regard.

As far as what a newer system could do -- perhaps more elaborate sheet metal folds? Bigger shapes? More complex shapes?

Re:PDP-11s *still* in use!

[Anonymous Coward]

OTOH, if a pipe breaks and floods the PDP, they are fucked relative to competitors who can just order a new PC.

Re:PDP-11s *still* in use!

[Anonymous Coward]

PDP and VAX parts are still widely available on the used market. Though one day that will certainly change...

In answer to

[tekiegreg]

So how many PDP-11's can you run on a Pentium 4 anyhow?"

in finding This article and this article which go in to good detail about PDP-11 specs, I can't figure out how to translate them into Pentium 4 equivalent speed, anyone care to help me? Would be interesting to find out anyways.

Re:In answer to

[-brazil-]

Such a "translation" would be beyond meaningless. The architecture is fundamentally different, and most of the P4's cycles would have to be spent on emulating/accounting for quirky little hardware details or features. That's why it takes a 100 MHz Pentium (if not more) to properly emulate a 1 MHZ C64

SimH

[awx]

Bob's emulation software SimH is a *fantastic* bit of kit. Runs vanilla OpenVMS without modification - VMS doesn't even know it's in a sim until you tell it so when you licence it.

(re-posting my anon comment)

geek envy

[Anonymous Coward]

Seriously, how am I supposed to impress my geeky friends without my badass 25-year old servers? It's similar to how having the physical CD or record is preferable to merely having a digital copy. How am I supposed to show off?

Software License barriers

[ErroneousBee]

There are emulators for old IBM mainframes (S360 S370). Hercules is one of these.

Unfortunatly, the massive cost of liscencing the MVS (OS/390, zOS) operating systems means there is no way that a normal user can run a PC based mainframe. IBM employees can do it, of course.

I guess thats also true for the PDP-11 and many old Vaxen, its just cheaper to migrate to new hardware/OS.

Re:Software License barriers

[SEE]

Well, sure. If you want to run old IBM mainframe apps on new hardware, IBM wants you to buy a current eServer zSeries mainframe. Maintaining 40-year backward compatibility is a buisness strategy, after all.

Re:Software License barriers

[terminal.dk]

If you have a license for the OS, just run it in the emulator.

BTW: At works we use a mainframe z/OS machine with a capacity of 1000-1500 MIPS or more. The emulator gives you about 1 MIPS per GHz.

Re:Software License barriers

[pfaut]

If you go to the SimH page, there are pointers to downloadable software that can be used with various simulators. Among these you'll find RT-11 and RSTS/e kits (alas, someone else bought the rights to RSX-11M and it isn't available). Also, there's a pointer to the information about the OpenVMS Hobbyist program which can be used as a source of licenses for just about any available OpenVMS software provided you are strictly a hobbyist user (no commercial use).

I'm currently running OpenVMS under SimH on an Athlon XP 1600 using hobbyist licenses. This gives me performance approximately equivalent to a MicroVAX II processor but I/O is quick since disk data is normally found in the linux disk cache.

Re:Software License barriers

[Anonymous Coward]

Well, it all depends on what you want to run. OS/MVS 3.8 is freely available, and that was the current OS version back when I was a mainframe systems programmer, even if I can't get free copies of CICS or RACF, at least it's from after IBM stopped charging for TSO.

Of course, I think my Athlon actually emulates a mainframe at a faster speed than the Amdahl 470/V6 we used to use!

16MB seems so small anymore...

Re:Software License barriers

[kcbrown]

Unfortunatly, the massive cost of liscencing the MVS (OS/390, zOS) operating systems means there is no way that a normal user can run a PC based mainframe. IBM employees can do it, of course.

You can't run the current version of the OS without paying big money (or having the right connections and agreements), but you can run older versions, fortunately.

For instance, you can get MVS 3.8J for free. There's even a prebuilt distribution of it for use with Hercules available. The main page is here, and the ISO image can be downloaded from Jay Maynard's PD software page (the turnkey site's links to downloadable images no longer work).

There are other operating systems, e.g. VM/370, which are also freely available.

HP1000

[johannesg]

I once considered writing an emulator for the HP1000 - until I realized it is totally impossible to find documentation on the hardware. I guess that is a major problem for most of these old systems.

Re:HP1000

[seh1]

Impossible? Not quite. I just pulled a copy of the HP 1000 M/E/F-Series Computers Technical Reference Handbook off of the shelf. For some reason I have two copies ot the HP 21MX E-Series handbook as well.

These were the first systems I ever admin'd and I just couldn't get rid of the docs. I even have one of the 14 inch 2.5 MB disk packs and a the bootloader paper tape. I do wish I still had one of the 16 Kb core memory boards out of the old 1000. Now that would be cool to have framed and mounted.

Somebody stop me before I get all mushy and dribble into my keyboard.

Re:HP1000

[johannesg]

So _YOU_ have it!

If you could just post your address and exact instructions as to which shelf I may be paying you a quiet visit later on...

Actually the last few years of my life were spent retiring two of these systems. They have been used until about three years ago to perform thermal vacuum tests on spacecraft by the European Space Agency. Since they were getting a bit long in the tooth, I was contracted to write a replacement system on modern hardware.

We discussed emulation at an early stage, but we realized quickly it would never be a solution since you cannot possibly emulate the incredible amount of IO channels the things have (dozens of serial ports, GPIB ports, etc.).

Let's find out!

[WarPresident]

So how many PDP-11's can you run on a Pentium 4 anyhow?

You could shell out some bucks for Ersatz11 and find out. It runs under Linux, and it runs fast. You can even attach Q-Bus and Unibus hardware with an adapter.

But how do they emulate...

[Anonymous Coward]

...the nipples?

Until that is resolved, I'll stick to nursing my old hardware.

You might be a vax geek if...

[schnits0r]

TOP TEN SIGNS THAT YOU'RE A VAX GEEK

Key traits identifying individuals tendencies towards abnormal preoccupation with VAX computer systems

9. When talking about building software you make reference to
compilation times in weeks and days instead of minutes and seconds.

8. You stopped purchasing new furniture when you realized that
your computers work just as well.

7. Your electricity bill is more than your monthly rent payment.

6. You've been hospitalized with muscle strain injuries after
performing some routine hardware maintenance on your computer.

5. You don't have an SO, but it's okay because your computer keeps
you warm at night.

4. While doing laundry, you occassionaly have a mental lapse and try to
wash your socks and underwear in your 11/750.

3. Friends who visit you want to know why there are old-time movie reels
stuck on your refridgerator(s).

2. Your house is pleasantly warm in the dead of winter, even with the air
conditioning turned all the way up.

1. The lights in your home dim or flicker when you reboot.

0. It doesn't matter to you if someone else's computer is faster because
you know your system could smash theirs flat if it fell over on it.

Re:You might be a vax geek if...

[Jahf]

Come around my place for the holidays and you'll hear my father and uncle talk about one you missed: the wonders of VMS lexicals.

Humbug to the holidays, I say.

Re:You might be a vax geek if...

[Rorschach1]

Funny? That should be modded 'informative.' I've never been hospitalized, but I did wind up with a wrist brace for a week or two thanks to some RA-92 hard drives.

Seriously, if anyone wants a free VAX 6000-510, let me know. I need the garage space back. I'm on the central coast of California. I'll even throw in a MicroVAX II or two if you want. They make good end tables.

Re:You might be a vax geek if...

[awx]

Yes!

Shame I live in the UK, really :(

Get over to classiccmp.org and post about it on the Classiccmp mailing list and you'll find quite a few people who'll help you get that space back ;)

Re:You might be a vax geek if...

[derrith]

I'm interested, I live in the bay area and I have a vehicle I can pick them up with. My old high school computer repair class would love to have them.

e-mail me logansaltsman(at)sbcglobal(dot)net

Thanks

Re:You might be a vax geek if...

[Anonymous Coward]

You might as well have the high school shop class make those end tables for you instead. At least someone will learn something useful.

Re:You might be a vax geek if...

[otomo_1001]

3. Friends who visit you want to know why there are old-time movie reels
stuck on your refridgerator(s).

Yeah right, a vax geek has friends? When pigs fly. :)

j/k I wasn't even born yet when these beasts of burden were unleashed.

Re:You might be a vax geek if...

[Anonymous Coward]

Shame I live in the UK, really

Ya, that must suck.

Re:You might be a vax geek if...

[Graabein]

> 0. It doesn't matter to you if someone else's computer is faster because
> you know your system could smash theirs flat if it fell over on it.

Which reminds me: I'll always regret not saving a Cyber 180 at a former employer before it (the Cyber) went into a landfill back in '89 or '90. It had a MAP and all.

Now that thing would easily crush a VAX 11/750 if it fell over on it. And a PC? Pfft! Roadkill!

*sniffle*

Nice idea, but good luck

[garver]

Regression testing. Emulation's nice, but it ain't the same as the original hardware. Which means, people will need to regression test. Trick is, the people that know what that old PDP-ll is actually doing retired a long time ago. So who's going to write the test cases?

SIMH URL

[Gropo]

SIMH Website.

Been wanting to buy an old 11/780 shell for a while. Not for a bar, but to mount both my Mac and Gaming PC innards in. This'd be a real trip to run as an emulator during parties. Now to interface the VT-120... Hack the shell I suppose. Run everything USB. >:D

The problem with this...

[freeze128]

The main problem with this is that a lot of the emulator software doesn't get upgraded. Where I work, we have a 5250 terminal emulator with a specific macro language that will only run on Windows 95/98/NT 4.0. We are already emulating a terminal. Who wants to emulate a 5250 terminal on an emulated windows 95 machine?

Changing WOW factor.

[Damon C. Richardson]

Problem is... The WOW factor on these servers has gone from WOW look what it can do. To WOW your still using that.

storms

[the_Bionic_lemming]

Nice - until Storms Blow thru and knock out the air conditioning, which short out the backup airconditioning, and having the server room heat to over 120 degrees, causing the machine doing all the emulation to lose three of the drives on the raid.

Which is what happened to us overnight in Illinois. Makes me wish that at least some of our stuff was in the old server room on old machines serving up stuff faithfully.

:(

Re:storms

[wintermute740]

" Nice - until Storms Blow thru and knock out the air conditioning, which short out the backup airconditioning, and having the server room heat to over 120 degrees, causing the machine doing all the emulation to lose three of the drives on the raid."

In central Ohio here... We had the AC in our server room go out in the middle of February one year. Akamai called because one of their colocated servers quit responding to pings... I opened the server room door, and thought I'd been hit by the backdraft from a fire... It was over 150 degrees in the server room, and the only machine effected, the Akamai machine, was restored to service with a reboot... But does anyone know how hard it is to get an AC repair shop to believe it's not a prank call when it's below 0 outside?!?!

My First

[PingPongBoy]

The first laptop I had was a pdp-11

Re:My First

[jdunn14]

So sorry to hear about your legs.

Re:My First

[TWX]

"The first laptop I had was a pdp-11"

An LED-screen-based terminal emulator with a 300 baud modem dialing up to the PDP 11 shouldn't really count...

So many pitfalls!

[ikegami]

The article does well by pointing out a great list of problems that can be encountered when emulating a machine.

Some of the projects on which I work are for nuclear power plants, many of which here in Canada use computers from 1972 -- I was born in 1976 -- to control the plant. While spare parts are dwindling, the prospect of having to retest all of the code is daunting, not to mention the costs of making a program as complex as an emulator in the first place.

I've seen (the assembler equivalent to) the following code used in embeded processors to perform a sleep():
counter = 500; while (counter--) { /* nothing */ }
Imaginine executing that on an emulator that didn't pay any attention to timing?

Re:So many pitfalls!

[Anonymous Coward]

I think it is far better off emulating the hardware in hardware ie. Putting the CPU core in a FPGA. You can go down to clock cycle accurate that way very easily.

Re:So many pitfalls!

[Anonymous Coward]

Many emulated cpu cores implement cycle counters. It doesn't take much to have the host processor use that to keep the timing correct.

Re:So many pitfalls!

[rreyelts]

Imaginine executing that on an emulator that didn't pay any attention to timing?

Any half-decent emulator will pay attention to cycle counts. It's one of the few things that distinguishes an emulator from a virtual machine. Take MAME for example - all the CPU emulation in there tracks cycles.

Re:So many pitfalls!

[Anonymous Coward]

that's what the turbo button is for =P

Re:So many pitfalls!

[SnarfQuest]

counter = 500; while (counter--) { /* nothing */ }

And, if you compile this code with a decent optimizing compiler, it will be reduced to

counter = 0;

This type of delay loop is awful for even moving around to a different CPU from the same family: say from an PDP-11/24 to a PDP-11/83.

This code is just not portable, and never has been.

Re:So many pitfalls!

[babbage]

counter = 500; while (counter--) { /* nothing */ }

My first programming class in high school -- in which we were writing QuickBasic on "fancy" Mac LC 520s (the ones with shiny new tray loading CD-ROM drives and color screens) -- this was how we were taught to write brief pauses into our programs. The numbers were bigger -- loop for 10000, loop for 30000, etc -- but it was the exact same idea.

We were told even then that this was inaccurate & unreliable -- we could all see how much longer the Mac II would pause compared to the LC 520s -- but it was basically the only option available to use in QuickBasic. Even the manual suggested this idiom IIRC.

But that was the 80s, and people were all crazy back then :-)

You Can't Emulate Hardware...

[fuzzybunny]

...if you actually need the physical computer bits.

Maybe you can run a virtual machine on a Linux box that lets you have a little software VAX on your PC, but try keeping your beers cold in it

Wow - dedication to a task

[acvh]

This kind of thing amazes me; not that it can be done, but that someone would do it.

"Attempts to bring up the PDP-15 ADSS (Advanced System Software) were stymied by lack of the proper paper-tape bootstrap. Eventually, a PDP-9 bootstrap turned up in France, but no paper-tape reader was available to transcribe it. The collector in France scanned the tape in sections on a flatbed scanner and then wrote a program to recognize and transcribe the holes and splice the transcribed sections together. The PDP-15 simulator writer then discovered, through debugging the boot process of ADSS, what changes had occurred between the PDP-9 and PDP-15."

Re:Wow - dedication to a task

[ClippyHater]

That's simply amazing. Made me realize that I'm not even close to being the geek I thought I was--I would've bought a six-pack and called it a day.

Microsoft's initiative? My foot….

[nurb432]

Umm if you haven't noticed they bought someone else's product.. its not THEIR initiative..

And have you heard of VMWare? its been a viable product for some time now.. ( this is perhape the reason they bought conectix in the first place )

( other, even older, free sandbox applications not withstanding )

Bob Supnik!

[ackthpt]

Man, I haven't seen that name in 20 years!

I have such warm, fuzzy memories of hacking a PDP 11 and rabidly tearing away the wrapping from each DEC Professional magazine that graced my mailbox...

Yeah, emulation sounds more reasonable than what some nut did, he got the schools old PDP 11/50, with 1 TU16 and 2 RP04 drives and had his house (I sh!t you not) raised 12 inches so he could set it up in the basement. No idea what's happened since.

Re:Bob Supnik!

[SnoBall]

Funny, I misread his name as "Bob Sputnik".

Just a guess ...

[Anonymous Coward]

So how many PDP-11's can you run on a Pentium 4 anyhow?

All of them? ;-)

Re:Just a guess ...

[Cragen]

Probably not far from right. Considering, in 1984, we had a PDP-11/44 that ran on 4 Mb RAM (1 Mb per double wide board), with a 25-MB removable "platter". (a disk slightly bigger than a LP record. Remember them?) I don't remember what the DEC RM102 h/d had or even if I have the nomenclature right. The 11/44 was cool cuz there were no more switches on the front of it for inputting troubleshooting instructions (in octal). We actually got to use a keyboard for that! The good old days -- not.

Re:Just a guess ...

[earthforce_1]

Actually, it would be fun to see how many pentium-4's you could fit inside a PDP-11 case.

I miss those 8 inch 100K? floppy drives though - those were the days when floppy lived up to its nam, and the drives made interesting squeaking noises while they were being read. I still have a couple of them downstairs, I was part of the last class that used PDPs before the university retired the old 11/70.

The PDP OS had a weird (and annoying habit) of automatically making numbered backups every time you saved - in theory sounds like a good idea, except with an entire class editing and saving assignments, the main drive ran out of space and the whole system froze every 15 minutes, and we had to hunt down the TA to reboot it. "Delete your files, delete your files" was a cry heard every 5-10 minutes in the lab, lest the whole system hang and die yet again.

Back in the 1970's, that astronomically expensive PDP 11/03 in the Heathkit catalog was my dream machine, as it was the only true 16 bit PC on the market. I even bought the paper tape software and manuals for it!

efficient versus speed

[www.sorehands.com]

If you keep the old system, it is running at .8mhz, 5mhz, or some other slow speed. It still runs faster emulated, though the emulation is not efficient. It also takes care of problems of repair and replacement of parts.

Wow, now I can play Adventure...

[Anonymous Coward]

... as it was MEANT to be played - on a VAX!

Re:Wow, now I can play Adventure...

[TimMann]

Heck no, it was meant to be played on a PDP-10! :-)

Oh my god

[Conspiracy_Of_Doves]

This is just so very, very wrong.

Don't get me wrong, a program that lets linux programs run in windows isn't a bad idea, but that picture... it just should not be.

When napster went legit, I had images of the napter cat, once a symbol of freedom, as a borgified drone of the RIAA, that picture sends the exact same shivers down my spine.

Re:Oh my god

[Performaman]

"industry leading Windows 2000 stability and performance"
That's up there with "I did not have sexual relations with that woman." in terms of truth.

Why bother

[CarrionBird]

There's nothing to stop HP from making replacement parts on an ongoing basis if they wanted to. As long as there is a market for the repair services, why landfill working hardware?

Re:Why bother

[antispam_ben]

why landfill working hardware?

You're asuming that the only alternative to keeping old hardware in service is to put it in landfills. Admittedly, too much decommisioned harware will be landfilled, but it doesn't have to be that way. Old hardware should firstly be offered to computer museums and collectors (perhaps have them sign something saying THEY won't landfill 'em), and if no takers, dismantled and recycled. There's lots of good steel in these old things - 19-inch racks can be reused as-is. The electronics (made of lead-bearing solder-laden printed circuit boards) should definitely be disposed of following HAZMAT regulations rather than just tossing it in the trash.

There is only one possible answer

[O0o0Oblubb!O0o0O]

that is, of course, 42

My MITS Altair emulation project

[AltairMan]

I've used Bob's emulator a bit, playing with the PDP-11 emulation when I had an 11 in my basement that was failing. I now use the VAX emulator running BSD. I've also used SIMH as inspiration for my own emulation project for emulating a MITS Altair 8800 (with the front panel).

The next version is done and will be released within the next week or so after I update the docs to synchronize with the changes made.

Anyway, the project page is here:

http://highgate.comm.sfu.ca/~rcini/classiccmp/Alta ir32.htm

Please explain again....

[greymond]

Why you wouldn't just replace the old (insert ancient) servers?

Emulators never seem to work as well as the original, or better than the modern replacement.

Re:Please explain again....

[Anonymous Coward]

The reason you wouldn't replace the old machines with new stuff is because you lost your source code, or it's in a language no one knows any more, the last guy who understood the system retired 10 yeasr ago, you've been planning for years to shut down this system anyway but you still have a few customers, etc. etc. etc. Unless your old application has decades of life still left in it, it's almost always a better deal to shell out a few grand for a PC + emulator, than to spend tens or hundreds of thousands of dollars rewriting it.

Re:Please explain again....

[Anonymous Coward]

Most likely they used proprietary software that they can't port to the new system.

Usermode Linux

[Brian Blessed]

I've recently done this with a small webserver to keep it running. Some sort of deposit had appeared on top of the electrolytic capacitors on the motherboard, and the machine became unbearably unstable.

I took out one of it's mirrored drives and connected it in a different (larger) machine and then booted it using Usermode Linux.
I found it was best to be running 2.6.7 on both the host and the uml and it is bridged onto the host's network, so it appears exactly as before.

- Brian.

Re:Usermode Linux

[Anonymous Coward]

Some sort of deposit had appeared on top of the electrolytic capacitors on the motherboard, and the machine became unbearably unstable.

Sounds more like they leaked...

VMS-on-Linux with SIMH

[Phil Wherry]

For those interested in trying VMS on Linux using the SIMH emulator (entertaining if you once used it), I've written a set of installation instructions that might be of some use.

Phil

Re:VMS-on-Linux with SIMH

[awx]

My god, that's you?

*shakes your hand*

Thanks, i've referenced countless people to your page, it's really well-done.

Re:VMS-on-Linux with SIMH

[OlRickDawson]

I'm using VMS running on Linux, under SIMH. For some ancient modelling software that is still needed, the performance gain of running it on a Pentium 4, 2Ghz is about 3x the original speed of a microvax 2.

Re:VMS-on-Linux with SIMH

[Phil Wherry]

Thanks; I really appreciate your kind words!

Phil

Emulation

[PingPongBoy]

I've written emulators for devices. It depends on what you want to achieve. Emulating the interface may be relatively easy. If you used an old server to provide specific data or perhaps some kind of interface to another service, well you don't need to duplicate the entire server, you just have to implement a "jumper" system to provide a different path for the information flow.

Emulating an device comprehensively just to simplify servicing it could be futile or infeasible when you need to know the fine details of the device's characteristics. The manufacturer of a device might supply an emulator but I wonder just how many PDP-11's or machines lacking backwards compatibility still provide a vital nonupgradable function.

One may point to certain programs that used to run in DOS or in my case Win95 that don't run in XP. I want to speed up these programs on new hardware without having to buy the latest version. This is the downside of using Windows - if backwards compatibility is broken, a faster processor may force an expensive upgrade. Then again, all this backwards compatibility could be slowing Windows down.

stability

[vasqzr]

How stable is, say a Dell/IBM/HP server, running an emulator?

I wonder if the guys at DEC ever thought, one day we'll be able to put 50 of these computers inside one box.

Mod parent up? Has Slashdot discussed this?

[gptelemann]

That press release sounds like a joke, or the most arrogant thing I've ever heard from Microsoft. Inclusion of the Microsoft 2000 code eliminates the freezes and crashes Linux users have reported? Or am I missing a joke? -_-

Hardware stability is a problem

[Anonymous Coward]

One of the key reasons to switch to emulators from old hardware is simply replacement parts. The "bought-and-paid-for hardware is cheap" argument falls apart when you just can't get replacement parts or trained people to install them. If you think it's hard to hang a 5-1/4-inch floppy on a new box, just try to find a working drive. If the floppy example doesn't work for you, think about a working RP03 or RM05 -- with a supply of replacement heads and platters. Finally, the emulator really can be cheaper to run. Consider that you can run the emulator in a virtual machine partition so that you're only using, say, 1/10th of a $1000 multi-gigahertz server (plus the cost of the virtualization software, I admit) - and it will probably still run faster in real-time than the original system did.

On a digression, it was great working for Bob Supnick. He's a nice, bright guy and I'm glad to see he's still keeping things stirred up.

Re:Eww...

[Anonymous Coward]

Come on man, just re-load with the 'it.' part stripped out.

PDP-8/E runs great on OS X

[caveat]

the PDP-8/e Simulator for Macintosh is a LOADED system (up to 32K words of memory, KE8-E Extended Arithmetic Element, ASR 33 Console Teletype, ASR 33 Auxiliary Teletype, PC8-E High Speed Paper Tape Reader and Punch, RK8-E Disk Cartridge System, LP8-E Line Printer, and a KC8-EA Programmer's Console) that runs a quite a bit faster than the original - fastest benchmark is a G4/450 at about 22x; my 2x1.25 runs the tests well under 1sec. If you need to support an -8 legacy, this seems like the logical way to go.

Re:PDP-8/E runs great on OS X

[Anne Thwacks]

runs a quite a bit faster than the original - fastest benchmark is a G4/450 at about 22x Actually, if the app is still running, there is a good chance speed is not a big issue. If speed was an issue, they would have upgraded 20 years ago.

and people will never learn how to really build hardware if all they ever see is Dell kit.

Simulation/Emulation vs Conversion

[mykepredko]

About fifteen years ago, I was involved in the retirement of a number of older computing systems (specifically IBM "Series/1", "System/7" and "1130"s) used in manufacturing. At the time, these systems were critical in supporting older products (most notably FAA radar displays) but had been withdrawn from regular IBM support and parts were only available from returned equipment.

I could appreciate the article's comments about engineering detective work; we had some source code on paper, some source and binary code archived on disc and some binary code saved on cassette tape (seriously). Product, tester and controller documentation was spotty to say the least. For the most part, we had enough understanding of what was happening to be able to recreate the test specifications for all the products.

The big problem was understanding actual timings and electrical parameters; few of the part numbers were built from standard TTL ("VTL" in IBM parlance) and most were built using IBM "SLT" technology implementing RTL and DTL logic.

After collating all the data we had, we decided we could: we could simulate the controller operations in a PC. In many cases, we could emulate the operation of the controller/tester hardware with basic digital I/O cards connected to a PC. Finally, in quite a few cases we were completely on our own due to unusual (for today) electrical requirements.

Due to the large number of part numbers (1500), we wanted to come up with a single solution that made the most sense and, ideally, worked for all the different part numbers. We looked at simulating the controllers with PCs and passing the I/O to the old tester hardware, emulating the tester using a PC with I/O cards or converting the tests to run on a standard InCircuit Tester (ICT).

In virtually all the cases, it made the most sense to convert the tests to run on a standard ICT tester (GenRad (new Teredyne) 228x was chosen) rather than simulate or emulate the hardware. The conversion applications generally converted the binary code into digital I/O operations (or GPIB instrument I/O) rather than come up with compilers for the original source code (although we did do this in one case). This was still a rather large job, but it was completed before parts sources for the old controlling computers completely dried up.

I suspect that from the lack of hardware interfacing information in the article, the author has run into similar problems. Despite that, having a simulator could be very useful in understanding how an old computer system operated and what is required to properly emulate/convert it into more modern hardware.

myke

Re:Simulation/Emulation vs Conversion

[CharlieG]

Wow, another old hand who's done GPIB!! That's actually how I got my start as a professional programmer - we were running a bunch of gear that used the GPIB bus, either off of some HP controller (which ran HP Basic) or some old Tectronics terminals

We used some CEC corp IEEE-488 (aka GPIB) cards and an emulator to emulate that old HP basic. Then, as the programs were in BASIC, the natural progression was to the Microsoft BasComp compiler, then to the PDS. Then windows and VB came along, and us old hands at HP basic (a DEAD language) and the Old Microsoft Basics became hot properties, and some of us caught the front end of that wave. It's been a hell of a ride, where I've seen a lot of really bad code, and seen some real nice stuff. I still miss controling BIG machines with my little computer

++Interesting!!

[Anonymous Coward]

. . ..

Why not port? Why bother?

[Mr. Roadkill]

If the DEC VAX systems are so antiquated, wouldn't it make more sense for businesses using these systems to simply upgrade to newer/better technology? I mean even if they have incredibly stability and "wow" factor, wouldn't it be easier than both solutions just to upgrade the software to newer systems.

Say you've got an existing system performing a particular task - maybe your inventory or payroll system, and it's coping with the load and has sufficient flexibility to allow you to make changes in response to foreseeable business and legislative changes. You are faced with a bill of a couple of thousand dollars every time something breaks, and that's happening more frequently. You know you want to get newer software in the future, but you're not entirely sure what you want and want time to look but are concerned the old system will not last until then.

In this situation, you could spend a couple of thousand dollars on a new machine and run your old software on it using a free emulator. Hell, the machine you just ordered for your secretary would probably out-perform the old server. The new machine will be one twentieth the size of the old, use one twentieth of the electricity and won't have twenty years of accumulated wear and tear.

If you've got a custom application written in a dead language, emulation may be the best way to continue fulfilling your requirements too. Sure, a new app in the shiny new language du jour might be nice, but if you've got proven 30 year old code and performance on the old hardware is adequate then it makes sense to keep things ticking on a virtual vax or pdp. And think of the kind of server room consolidation you could perform!

Other posters have commented on the proven behaviour of hardware v. emulators - how the latter could have bugs that aren't apparant, and may thus be unsuitable for users like NASA etc. Surely it's easier to produce a bug-for-bug compatible emulator than it is to re-write possibly millions of lines of legacy code in a bug-free manner. Sure, it's nice in the long run to re-write software with greater portability in mind, but while that's happening wouldn't it be worth making sure you can continue to run your existing programs without having to worry what will happen if some obscure I.C. elects a new Pope?

Re:Why not port? Why bother?

[Doctor Faustus]

Sure, it's nice in the long run to re-write software with greater portability in mind

If you have an emulator that runs basically anywhere, I think I'd have to say that old PDP-7 assembler code has become portable.

Re:PDP-11s *still* in use! If it ain't broke...

[Anonymous Coward]

This just goes to show you that if it ain't broke, don't fix it!

I wish that someone would *fix* M$!!!

What's even awesomer...

[Asprin]

...is that through resources such as Ebay, half.com, your local public library, garage sales and thrift stores, you can still get manuals for such things -- AND CHEAP! I found a PDP-11 technical programming textbook at my local Goodwill for a buck. Heh, heh, heh. All I gotta do is crank club977 on Shoutcast and it's 1987 all over again! W00t!

Re:What's even awesomer...

[antispam_ben]

...is that through resources such as Ebay, half.com, your local public library, garage sales and thrift stores, you can still get manuals for such things -- AND CHEAP! I found a PDP-11 technical programming textbook at my local Goodwill for a buck.

The Net is awesome for finding any used or out-of-print books or manuals of any kind - it used to take months to find something unusual, now it can be in your hands in a couple of days. Here are two valuable resources:

http://www.bookfinder.com/
http://used.addall.com/

You can do your own shopping at thrift stores and yard sales, finding the occasional RCA Receiving Tube Manual and such (I've done that a lot and now have about 10k books, including 20 tube manuals) but this is hit-or-miss for something specific. For a few more bucks per book, you can often find exactly what you want at one of the metasearch sites above or (if it has an ISBN number) used on amazon.com.

If you still don't find it, you can subscribe to this list:
http://www.bibliophilegroup.com/ ($30 per year subscription, two week free trial)
and send a WTB: (Want To Buy) post, where hundreds of used book dealers have large portions of their inventories they've yet to enter online, but may have it for you.

Computer manuals before 1970 or so are actually in demand and worth something (maybe $10-$25).

Worked on a PDP years ago

[MemoryDragon]

During a high school internship. Fascinating machine. Back then, I think it was around 84 or 85, this machine served around 5 people on plain text terminals, and as soon as you started an integrated pascal compiler everybody complained that the machine was slow as a dog. Needless to say, that the pascal compiler was slower than Turbo pascal on a normal PC (given the recursive descension nature of turbo pascal no wonder, although the pc was myriads slower) But one thing I never saw with this machine, it never fell, and the multi language binaries, as far as I can remember where you could hook different languages easily together were really nifty. No wonder the company used this machine mainly for development and holding customer data, this beast was as solid as a machine could get.

5.25 inch drive

[Kozar_The_Malignant]

>How are you going to emulate a 5.25 inch drive to read old disks?

Write an emulator for your 8 inch drive.

use it or lose it

[Doc Ruby]

At least as big a problem of losing these legacy processors is losing the I/O devices that process their data, especially the storage units that read their media. Dead media is a problem catching up with our mediated society, as we data predators starve when our media prey go extinct. Copying legacy data to new media formats before their original formats become obsolete ought to be a required task in the archive process. And IT budgets should reflect that. The cost differential between hosting all old emulated system/data complexes, which is very cheap per datum, and rebuilding obsolete hardware later to recover only the data required in hindsight, favors hosting everything. And the steep increases in newer storage technologies mean that the old data is not only very cheap to store, but can be stored economically and conveniently in distributed copies, ensuring full recovery.

The networked storage itself will make the archiving, hosting, and retrieval process extremely economical. And while older data devalues for a while, beyond the threshold of "living memory" it become increasingly valuable, defining the past. Consider the expense, and value, of recovering even tiny records from as little as 800 years ago. By keeping old data entirely in the actually digital, virtual realm, where it can be be used, rather than in the doomed physical simulations of digital media, we're investing in our future by more accurately remembering our past.

I want to emulate my old Vic-20...

[Anonymous Coward]

Choplifter withdrawels...

Re:I want to emulate my old Vic-20...

[antispam_ben]

Chances are you can find your emulator for most any 8-bit computer through http://www.google.com/.

VDM sound for dos sound under newer windows

[Mustang Matt]

For anyone with old dos games, don't bother dual booting or trying to use dos sound drivers.

Just emulate the hardware!

http://www.ece.mcgill.ca/~vromas/vdmsound/

Emulating PDP-11's on Pentium 4's...

[Anonymous Coward]

Pentium 4 eh.. let's see, that depends on how many Pentium 4's I can emulate on my dual opteron...

Gotta spend money to earn money, right?

[maynard]

They haven't upgraded a critical piece of equipment in 30 years? No wonder US manufacters are losing out in the global marketplace.

It's not like the guy's exporting to the international market. The guy makes custom sheet metal ducts for heating and A/C systems and services a local market. It's only one part of his business, but that one purchase (originally bought by his father) has earned its original investment many times over. If his market position was declining due to lack of modern tooling I'm sure he'd re-invest. I guess. Not that he clues me in on the particulars of his finances and business. Would you buy something you didn't need on credit just to "win in the global marketplace"? --M

You, sir, are an imbecile.

[Anonymous Coward]

"Duhh, huh huh huh, 42, duh, 42, 42, 42, duhhhh, 42! HUH HUH HUH ME CLEVAR CLEVER!"

What sort of lesional psychosis drives people like you to think that typing "42" a) was ever funny to begin with, and b) is STILL funny, decades later, after millions of mindless repetitions?

I'd put you down like a sick dog if I could reach you. It's the only merciful thing to do.

Only For The Most Critical Situations

[blueZhift]

While I think building all sorts of emulators would be a lot of fun, I can only see this as making sense in the most critical situations where porting just isn't practical or timely enough. This is often the case encountered when trying to preserve some old arcade game for example. You've got the code but it's 500K lines of assembly for some proprietary board. I can see the government using emulation to delay or smooth the transition to new systems when the services can't afford to be unavailable for even a few days.

We've had one running for quite a while

[Graemee]

We replaced a VAX 330 with an emulated system. The first comments from the field were much better response time was. However the old system ever had downtime in the last three years, and that was due to a UPS failure. Stability is not quite the same, but it reboots quicker.

not willing

[dfj225]

You know, until a few months ago, I never realized how man businesses there are that are not willing to give up their old hardware/software. I came to this realization when I went for an interview with a company who sells services to companies looking for solutions. One of their products that they were working on takes output from old COBOL and other programs and creates a mondern interface using XML that can be presented through a browser or whatever. When the user gives this modern interface input, this input is then given back to the old program to allow it to process it. I never even knew there was such a large market for software of this type. So, I'm guessing that taking the old hardware and emulating it would probably be the logical next step.

Bob Supnik ..

[Anonymous Coward]

was also the dude responsible for 'translating' the "dungeon" game (aka Zork) from Muddle to Fortran so it would run on PDP11's.

Maybe he did some work for Infocom as well.

I'm reminded of the book Ol'Yeller...

[LifesABeach]

the ending in the book Ol'Yeller made me sad. But the solution used was the only one that was humane.

VAX is a good machine. i've gone to VAX School for VMS 0.1, ya 0.1; it dates me. many of CPM, Apple, MS-DOS, and UNIX commands could be found on a VAX.

how does this apply? when its time to say good bye to a friend for the last time. it hurts. but the pain passes.

we should place a PDP-8, PDP-11, VAX, DEC-10, DEC-20, and DEC Alpha in the Smithsonian. it will remind us from where our computing origins came from.

Those who make a living maintaining software from machines 20 years obsolete cannot support a bench mark analysis of desktop pc's to their equipment. for me, anyone that maintains software on an UNIVAC 1100/8, IBM 1402, IBM 370, IBM 4200, IBM AS-400, DEC-20, et.al.; these people are not being completely honest to their supervisors.

Re:I'm reminded of the book Ol'Yeller...

[corellen]

There is a PDP-11 in the Smithsonian. Its in the Amercian History Building. In the exibit about the digital revolution, I forget the actual name of the exibit. There is a few other ancient iron there too.

Using VMware to emulate DOS machines (for games)

[Esion Modnar]

I've gotten that to work fine (using FreeDOS), except for a couple of problems. No support for sound, and with some games, especially on recent hardware (Athlons, P3's, P4's), they run really fast.

I wanted to fire up Return to Zork again, but it went through the intro in about 5 seconds, and silently.

And of course, there's no way to emulate physical media readers (i.e. 5.25" drives) if you want to get at your very old data.

(Oh, and if anybody knows where I can still download old DOS soundblaster drivers, I'd appreciate it. The Creative site has cut out much of their old stuff.)

Reminds me of one of my favorite jokes...

[cblguy]

(old school guys will get it, anyway...)

How can you tell the DEC repair man on the side of the road?

He's the one changing out the tires, trying to figure out which one's flat...

No vaxen here, just Alphas...

Three

[Anonymous Coward]

So how many PDP-11's can you run on a Pentium 4 anyhow?

Mr Owl: Let find out. One, two, three, (crunch).

Mr Owl: Three.

Emalaters are grate...

[theskipper]

As a mater of fakt I uze the sam dikshunary emalater as /. so I dont missspel werds lik microprossesor.

What about the Alpha?

[hak hak]

On the HP VAX site, they recommend to move from VAX to the AlphaServer. What happened to their plans to phase the AlphaServer out?

Re:What about the Alpha?

[vidarh]

See, they probably won't phase out the AlphaServer until they've moved all their VAX customers over to it. That way they get to sell their customers new hardware twice.

Re:What about the Alpha?

[Anonymous Coward]

On the HP VAX site, they recommend to move from VAX to the AlphaServer. What happened to their plans to phase the AlphaServer out?

They are still going to sell Alphaservers until about 2010 but without any major speed bumps. VMS on Itanium has had a developer's edition out for some time, and will be having an initial release later this year.

If you're still running on VAX hardware, you're hardly an early adopter, so Alpha would seem the logical choice for a few years yet.

What About the Durability and Reliability

[JoeCommodore]

The old Minis were built like tanks, modern computers aren't always as robust. Beyond emulating the CPU itself you will also have to emulate all the related software specific peripherals ("select next tape reel image now.")

Also how would you emulate those special peripherals that you may use with your old iron system?

On the other hand, it would be a lot of work, and it may buy a company a few years but probably it will lead to a re-write of the system. The good part about that is you then get the emulation team to properly document the data formats, etc. so the upgrade/conversion team can convert over all the old data.

Y2K called a lot of attention to this

[Fencepost]

In a lot of cases there's old software that performs mission-critical functions, but for which the source code was lost long ago. There's a lot less of it now since quite a bit was replaced during all the Y2K buildup, but there's still some out there. I suspect that there's also a disturbing amount of it that now has pre- and post-processing of data to massage dates.

Screen scrapers have also been around for quite a while just providing a GUI on top of a terminal-based software that it's not worth replacing yet.

While we're talking emulators

[leob]

I'd like to point people's attention to Fabrice Bellard's free portable retargetable emulator QEMU. The technique employed in QEMU allows to enjoy most benefits of JIT (except dynamic register allocation and cross-basic block optimization) without actually writing any code generator for the host processor.

Implementing PDP-11 and VAX in QEMU will achieve unprecedented (in a free emulator) speeds.

If a SID falls over in the forest, can you hear it

[NotZed]

All well and good, but not all hardware can be emulated.

A 2Ghz cpu strugges to emulate a 1mhz 8 bit cpu with a couple of tiny audio/visual chips. It would be more effective just emulating the COBOL they ran if they are old business systems, but add any funky hardware and you're up shit creek without the proverbial.

Some chips, like the MOS SID cannot be emulated fully, as each chip has its own electrical and audio electrical characteristics making them individually unique and unemulatable.

Re:Mod parent up? Has Slashdot discussed this?

[operagost]

Yes.

Come on, man - it mentions Bob twice! And Packard Bell was absorbed into NEC years ago!

This is not new.

[Anonymous Coward]

There are lots of old machines that run under emulation. For example, the Wang 2200 series machine is emulated by a couple of vendors on PCs. I'm sure people run DOS apps under VMWare as well.

Series 1 Tivo!

[cat_jesus]

That is the answer for our dying series one Tivos. Emulate the hardware! I really like the software at its current software version and would like my Tivos to last forever, especially since I've paid for lifetime subs. Done properly I could emulate both Tivos on one box -- with dual tuners of course.

I agree!

[NoMoreNicksLeft]

(Warning: Sarcastic or not, this is not a troll. I own a PDP-11/04 that I'm trying to get running, not to mention a bunch of other vintage machines)

Just think of the possibilities! Why try to preserve the Mona Lisa, when we can just photocopy it?

David the statue? Laser scan it, and upload the mesh triangles to sourceforge!

There is nothing that this strategy can't be used on for outrageous savings. We don't even have to manufacture new CPUs at all, just emulate the Pentium5 on your PII!

Emulation is for those that go "Gee, wasn't that nifty", once in their lifetimes. The true enthusiast wants the real thing. If someone restores old cars, they're an auto enthusiast, and people honk their horns at the things on the road, in admiration. If it's home furnishings, they're antique collectors, and magazines do photoshoots of the treasures. But if it's a computer, you were supposed to throw that out after 6 months, to buy another. It makes no sense.

Re:I agree!

[Junta]

This is not for 'enthusiasts', this is for businesses. The cost of maintaining old, failing hardware grows exponentially as both hardware problems become more frequent and availability of 'vintage' parts decreases. A businesses bottom line is hurt dramatically if they have applications that only run on the old hardware and by extension, feel a need to run that old hardware just because their core business requires the apps.

Of course, someone actually has to *write* an emulator, and if no third party steps up to do so, that would be a harder task than porting your old applications. If the antiquated system, in retrospect, turned out to be pretty mainstream and therefore modern emulators were written with the platform in mind, you are in luck. However, if a business picked an ultimate loser in the marketplace no one cared about, the chances of an emulator existing, and indeed being worthwhile to produce at all, are slim.

Re:I agree!

[NoMoreNicksLeft]

Then it's an even worse idea. How so? They think running a critical app on hard-to-repair hardware is tough? Then they need to wait until the emulator bugs out on something subtle.

They end up with still failing "virtual" hardware, and the only consolation is that if they persist long enough, they may eventually fix it completely. Oh, at least until they need to port the emulator to Windows 2009 Gold Pro edition on the Pentium 9, then it bugs out again.

Start porting the damn apps, or rewrite them. And even as you're doing this, plan for the next changeover in 10-15 years.

Re:I agree!

[TimMann]

There is more than one kind of enthusiast. Some folks are enthusiastic about writing emulators. It's fun to get to know the old machines well enough to emulate them, and to see the old software run again on them.

I could argue (overstating my case) that emulator writers are the true enthusiasts for the old machines. Those who merely (;-) collect old hardware and perhaps get it running may have little understanding of how it works inside. Those who emulate it know it intimately.

KEEP THE OLD STUFF

[Anonymous Coward]

You should nurse old hardware, and not emulate it, for the same reason you should read keep old editions of books and not download 300k e-book versions to their hard drives.

Emulators aren't perfect

[poot_rootbeer]

But would a software-based emulator accurately reproduce the behavior of the infamous "More Magic" switch?

That's a serious question, by the way. How can it be proven that an emulated system will perform exactly the same way that the original system would?

Consider that even among the most popular emulators, those for videogame consoles and handhelds, you won't find many claimed by their authors to have more than 99% compatbility.

Yes, gaming hardware may possibly be more difficult to emulate than well-documented business hardware due to the number of custom chips that effectively have to be reverse-engineered, but do you want to migrate your mission-critical systems from physical hardware to emulated hardware only to find that they depend on the 1% of functionality that's not accurately emulated?

PDP-8, running on PDP-11, running on...

[notthepainter]

This was way too long ago, but I had a college job at MIT doing programming for particle tracking from colliders (doubt they were even super colliders back then!)

Anyway, it was PDP-8 work, but we didn't have any 8s around, but there was an emulator for the PDP-11, and of course we didn't have any of those around.

But we did have a PDP-11 emulator for our hardware, wish I could remember what it was, DEC for sure.

Ahh, the gold old days.... This would have been '79 or '80, I love being able to tell people that my first email addres did have an @ but that hadn't put the . in yet...

Re:PDP-8, running on PDP-11, running on...

[javiercero]

That could have been a VAX (probably an 11-780 for the vintage you described). The VAX was able to run PDP-11 code, in fact the name of the VAX was VAX-11 as in a virtual address extension of the PDP-11.

You might read Slashdot too much . . .

[TheLetterPsy]

If you read the first line as:

Top Ten SIGS That Make You A VAX Geek

HOW DO YOU EMULATE THE UNDEFINED?

[blair1q]

They may think they know the specs of those machines, but really the science of requirements capture wasn't matured at that point.

They'll be chasing bugs from now to infinity.

Doubly redundant?

[Anonymous Coward]

The coolant system on our Sperry (Dopey) is triply redundant. If your data and throughput is important; you scale the necessary redundancy linearly when you move to a server environment.

Can't emulate dem blinkenlights!

[lemongrass]

What good is a simulator if you don't have any blinking lights!

Re:Can't emulate dem blinkenlights!

[Anonymous Coward]

It's been done, several times. My own PDP-11 emulator drives a little LED board through the LPT port. But no one will shell out the $14.68 it cost me to have blank PCBs made, so I guess they don't care as much as they say they do. There's a project afoot to make something much more complicated for SIMH, it'll be interesting to see how that develops.

Re:Eww...

[Anonymous Coward]

Come on man, just re-load with the 'it.' part stripped out.

Whatever, tool. That makes about as much sense as when Microsoft recommended that we "manually type URLs into the address bar" in response to that browser bug involving null bytes in URLs.

What a wonderful way to use the net. Hover over a link, memorize it and then type it manually into the address bar. Talk about STUPID.

You'd be surprised...

[Anonymous Coward]

As an SE for a large hardware company I was on-site in a customer's datacenter last year. They still had a fully-operation VAX 11/780 for simulation work. This was 2003! I asked why they hadn't migrated off of it, or why they hadn't looked at emulation.

EDS runs the shop that the customer outsourced to. They were making a mountain of cash off the monthly service fee for the old big iron. Also, because it was a government contractor they couldn't easily move off of the 11/780 to using open-source emulation software because it had to be verified and tested.

You're right again, President Bush!

[Anonymous Coward]

Sure, spending money with absolutely no return on investment whatsoever is the path to true competitiveness!! It's in the Bible, somewhere, I bet.

Now, about those tax cuts...

Why: Coolness and Bootstrap Education

[seawall]

There are several reasons to have emulators.

A big one (not the one that funds them) is they are cool.

Useful sometimes. E.g. PDP-11 on a PCI board with a PDP-11 hardware interface is buyable. It's used, among other places, in the postal system to run hardware that needs a PDP-11. Interestingly, it used to use the PDP-11 on a chip but last I checked used an Intel CPU. XEROX had allegedly bought all the remaining PDP-11 on a chip machines for their use in copiers.

The article was pushing the "where we came from" aspects. I KNOW how hard it is to keep PDP-10 hardware running. It's rather handy in defeating patents to come up with prior art...from, say, 1964. The thing here is it's use it or lose it time. You write an emulator and you understand the machine.

Bootstrap Education. A young person can understand a CP/M machine on a level that just isn't going to happen, say, my iBook G4 and use that understanding to bootstrap up to the next level of complexity.

MVS vs VMS and "five 9's"

[xtermin8]

There's also the power of redundancy at issue here. Remember that you could buy 1000 linux boxes for a fraction of the price of short-term maintainance of a z/390, and virtualizing the hardware would multiply that number again- Result: disposable systems. Paper towels are indeed less reliable than rags, do you keep rags around the office? Also, somewhat off topic, but certainly related: I have heard that VMS was the only system software that had certified "5 nines" reliability. True?

Re:MVS vs VMS and "five 9's"

[javiercero]

I believe Tandem's Non-stop OS and Stratus' FTX had the five 9s thing going for them. Although I am unaware that there was an actual certification process.

Not very fair...

[MooseByte]

"The truth is, Slashdot is basically a Gamer site."

Oh come on now, that's not fair. We also talk about pr0n.

emulation.net

[network23]

There's a lot of really good emulators for MacOS X on http://emulation.net/.

Including Edsac. IBM Series 1, PDP-8/E, VAX, CP/M, Sinclair QL, Windows PC, Oric and other obsolete stuff. Lots of fun!

two

[Anonymous Coward]

Actually, I run 1 PDP10 and one PDP20...
The PDP20 is running KLH10, though, sorry Bob.

PDP-11s

[giminy]

So how many PDP-11's can you run on a Pentium 4 anyhow?

How about a beowulf cluster of PDP-11s?

Commercial Support? Unix lost? OpenBSD?

[emil]

Originally intended as an educational project, it is increasingly being used in long-lived production environments as a substitute for real systems.

Our VAX administrator is considering (what I consider to be) an excessive amount of cash for the commercial Charon VAX emulator. Will anyone support SIMH in a production role?

For example, Unix versions 1-4 appear to be irrevocably lost.

I see that Caldera/SCO's UNIX versions 5-7 are available. What happened to 1-4? How were 5-7 rescued?

As of this writing, there is still a timing problem in the MSCP driver for OpenBSD/VAX, although not for VMS, Ultrix, or BSD 4.3.

Have the OpenBSD/VAX developers been approached to patch their source? I see a kernel patch and a preinstalled disk image - hopefully this would be easy to do.

Re:Commercial Support? Unix lost? OpenBSD?

[trb]

I see that Caldera/SCO's UNIX versions 5-7 are available. What happened to 1-4? How were 5-7 rescued?

The UNIX v1-4 that are being discussed as lost use a different numbering scheme from the UNIX System 5 and SCO/Caldera numbering scheme.

For a chronology of early UNIX systems, see my UNIX timeline from 1984. The v1-v4 are in the 1970-74 timeframe, and System V was from around 1982/83, and that's what latter-day UNIX System 5 is based on.

Case Statements?

[rrowv]

Instruction execution models the behavior of the real system, with a fetch phase, an address decode phase, and instruction execution. Often, the instruction breakout is simply a large case statement. This is fast and models the structure of microcode, but can be bulky and difficult to read. Has nobody ever heard of jump tables??

ECAP -- layers of emulation

[jms]

As of a few years ago, I was a system administrator at a university. One of the more esoteric things I did was to bring ECAP over to VM/CMS.

ECAP (Engineering Circuit Analysis Program) was written around 1965 by IBM. I never used it (other than to port it and test it), but I believe that it basically solved simultaneous differential equations.

The interesting thing was that all that we had was a binary originally written for and compiled on an IBM 1620 computer. At some point, a 1620 emulator was written for OS/360, allowing the original binary to run on the IBM 360 architecture. It was running as an MVS loadlib. I moved the binary to VM/CMS, and it ran perfectly fine in OS emulation mode.

So, we have the 1620 binary (in the form of a virtual punch card deck)
running under a 1620 emulator
running under an MVS emulation library
running in VM/CMS on an IBM 3090
in a virtual machine, emulating an IBM 370.

All for an old professor whose tenure probably predated the 1620 computer that the software originally ran on.

Specifics of the patent

[tepples]

Nintendo patented emulators where the emulator turns I/O speed hacks on or off based on a hash of the binary. Speed hacks are useful for an emulator designed for one or a few titles, but most of the GBA emulators out now are designed for accuracy in order to run unknown programs, including freshly compiled programs.

In other words....

[Stephen Samuel]

Just imagine:
You can have a Beowulf Cluster of Those(tm) running on a single CPU!

MESS?

[Anonymous Coward]

Can't believe that noone has mentioned the MESS project http://mess.org/.

Yes, its primary goal is to emulate all of the home computers, but a lot of the systems it covers were used in businesses as well.

...on the head of a pin

[tsm_sf]

So how many PDP-11's can you run on a Pentium 4 anyhow?

All of them?

May have to connect peripheral to modern system

[AHumbleOpinion]

You have some odd storage devices (that store a tiny fraction of modern HDDs, thus you can emulate them with an image file). You have printers (emulatable with... printers!). Perhaps really ancient input devices such as a cardreader (scanner -> conversion tool -> file). No doubt other exotic peripherals exist, but you can somehow emulate them all.

There is a chicken and egg problem. Many government agencies are interested in old hardware because they want to read old data before it is lost. Consider magnetic tapes. You need a tape reader to recover the ageing data. Emulating such devices only works once that data has been read and converted to a modern form. Now you don't necessarily need the old computer, perhaps you only need to jury rig an interface from the legacy peripheral to a modern system that is emulating the old computer.

Au contraire

[hey!]

You don't just emulate the hardware, you also emulate the physical access to the hardware via a GUI.

Open the 11/05 emulator window, grab the wirewrap tool from tool pallette and go to town! Just remember to hit Alt-S to toggle your anti-static wrist band on, or right click on the radiator to emulate putting your bare feet on it (if the system clock says it summer).

RX50s aren't hard sectored

[TimMann]

Nothing will read hard sectored RX50 disks, because there's no such thing. They are soft sectored, and PC hardware can read/write them. Please see the link another poster gave: http://home.iae.nl/users/pb0aia/vax/rx50.html

Just imagine where this is going...

[Chemisor]

I can just imagine a Palm pilot running a PC emulator running a PDP-11 emulator running a program in an ENIAC emulator. Life just doesn't get better...

Somewhat amazing...

[c0p0n]

Imagine a Beowulf cluster running on top of that 1000 emulated machines...

bugs

[firewood]

The difficulty of developing a historically accurate emulation is not only in developing relatively a bug free software emulator, but in emulating all the bugs of the original historical hardware. All the hardware bugs may not only include known ones, but ones yet to be discovered or characterized.

Perhaps some historic item of software will be discovered at some point in the future which actually requires one of these undiscovered bugs to in order to function "properly". An interesting problem if the hardware (or at least a physical component accurate simulation) is no longer is available for examination.

Re:Using VMware to emulate DOS machines (for games

[TimMann]

DOS itself doesn't have sound drivers. DOS games drive the sound hardware directly. (Well, I guess some newer sound cards have initialization software that you have to run to set them up and/or turn on their emulation of older sound cards for DOS programs to use, but that's still not exactly a driver...)

In a VMware VM, the virtual sound card can either be one that's compatible with the old Soundblaster 16, or one that's compatible with the newer Soundblaster Ensoniq AudioPCI (aka es1371). The newer card is used by default in new VMs. You can get the older card if you need it by setting sound.virtualDev = sb16 in your .vmx file.

I've heard the sb16 emulation isn't quite complete enough for some utilities that autodetect the card to find it (sorry, I don't have the details), but it's there if you can tell the software to just go ahead and use it.

How's my swing?

[BorgCopyeditor]

How do you emulate an 8 inch drive?

Let's see ... use a putter and tap the ball very lightly?

Emulate the emulators ?

[arjay-tea]

What happens when the emulating system becomes obsolete? Do you emulate it in order to run the first emulator?

Re:Emulate the emulators ?

[Anonymous Coward]

Why not? That's probably the easiest way :-). Or, if it is open-source, why not just recompile the stuff? *g*

Mac OS X Emulates 35 Computers

[Slur]

Check out http://emulation.net/ which provides a one-stop resource for emulation on Mac OS and Mac OS X. They list 35 different computer systems for which you can get emulators. Most of them appear to be free.

Acorn Atom, Acorn BBC Micro, Amstrad CPC, Amstrad PCW, Apple I, Apple II, Apple ///, Atari 800, Atari ST, Baby (SSEM), Commodore Amiga, Commodore 64, Commodore VIC-20, CP/M, Edsac, IBM Series/1, Macintosh 68000, Memotech MTX512, MIPS R2000, MO5, MSX, Oric, PC-9801, PDP-8/E, SAM Coupé, Sinclair QL, Sinclair ZX81, Sinclair ZX-Spectrum, TI/99, TO8, TRS-80, TRS-80 Color Computer, VAX, Windows PC, X68000

If you prefer game consoles you have 13 to choose from. ROMs are hard to come by but if you look hard enough you can find them. And there is a huge selection available.

Atari 2600, Atari 5200, Colecovision, Intellivision, Nintendo, Nintendo 64, Odyssey^2, PC-Engine/TurboGrafx-16, Sega Master System, Sega Genesis, Sony Playstation, Super Nintendo, Virtual Boy

If handheld units are more your speed, there are Mac-based emulators for 11 different varieties. I haven't tried any of these, but if MacMAME and the other console emulators are any guide, these should run at full speed and beyond.

Atari Lynx, Dreamcast VMU, Gameboy, Gameboy Advance, HP-48 Calculator, Magic Cap, Neo Geo Pocket, Palm, Sega Game Gear, TI Calculator, Wonderswan

This has been the goal in digital audio production

[Anonymous Coward]

for about the last 8 years.

As an interesting comparison.
I would never buy a vintage board over an emulation.
What a hassle.

IBM HACMP

[kupci]

Hmm. That sounds roughly similar to IBM's HACMP, High Availability something or other. Basically if there is a hardware failure it has to shift processing to the other box (RS-6000 I think, on the system I'm familiar with, no expense was spared). I just recall this as I had to work on a shutdown process for my component. But as I recall it was very fast, the idea was that processing (from a mainframe) was uninterrupted.

Good one!

[kupci]

I recall the community school I took calculus at while in high school had a Dec PDP-7 (or 11? but I'm guessing, 32K?). Anyway, it was about the size of a refrigerator, and one of my friends spent most of the semester trying to boot that sucker, as you say with _tape_! And flipping the switches and the whole bit. Although he claimed various successes, This guy eventually got a Masters at Stanford and works at MSFT, graphics/sound (DirectX I believe) so he was no slouch.

WRONG!

[kupci]

I believe Patterson wrote the original version of CP/M on an emulator, simply because it was too expensive or difficult to run on native hardware. And it worked. I may be off on the details, it's all in Robert X. Cringley's excellent book about how Silicon Valley Geeks are billionaires but still can't get a date. Greate book.

It'll never be the same...

[The Master Control P]

I've got a Tandy M100, along with an Sanyo MBC-4000. I don't care what kind of emulator you try to offer me for the M100: It doesn't have the same keyboard with a wired switch for every key. It doesn't have the same screen. I can't take it with me. And I severely doubt that my emulator will be running on the original hardware after more than twenty years of use.

The MBC-4000 (If anyone has ANY disks for this thing, please E-Mail me!) is another unique thing. I've never met anything with a similar keyboard.

My point is, you can't duplicate the feeling and the (for lack of a better word) aura of the original hardware... So what if I can get an Altair 8800 emulator? That doesn't mean I have a piece of computer history. I don't have a computer that's proven it's durability for more than 30 years.

As a side note, I am an active M100 user... Most reliable goddamn computer I've ever had. Never crashes, never BSODs. Just does what I program it to.

Re:Using VMware to emulate DOS machines (for games

[Anonymous Coward]

Try dosbox.
http://dosbox.sourceforge.net/news.php?show_news=1 /