eBay flooded with IBM PS/2s

[PeterNY]

Is it me or does it look like eBay is flooded with IBM PS/2s this month? Some have been there for months already but there are a lot of new listings.

Model 25: http://cgi.ebay.com/200618265015
Model 25: http://cgi.ebay.com/320748350310
Model 25: http://cgi.ebay.com/320748352865
Model 30: http://cgi.ebay.com/140595940856
Model 30: http://cgi.ebay.com/330582747978
Model 30: http://cgi.ebay.com/370522987879
Model 50: http://cgi.ebay.com/180715145294
Model 53: http://cgi.ebay.com/130565337771
Model 55: http://cgi.ebay.com/330604415418
Model 55: http://cgi.ebay.com/370522988006
Model 55: http://cgi.ebay.com/150652781370
Model 56: http://cgi.ebay.com/150642517008
Model 56: http://cgi.ebay.com/290603360046
Model 56: http://cgi.ebay.com/150642517008
Model 56: http://cgi.ebay.com/190558373224
Model 80: http://cgi.ebay.com/190558373268
Model P70: http://cgi.ebay.com/250800081241
Model P73: http://cgi.ebay.com/320723009820

 

[DOS lives on!!]

There's more than I've seen in a while. That Model 25 looks nice.

 

[Chuck(G)]

Unless you have an unusual MCA-dependent card, what practical good is a PS/2? I could never wrap my head around that one.

In the current IEEE Computer, the issue is dedicated to the IBM PC. there's an interview with Mark Dean, CTO IBM, where he admits that the PS/2 was a huge mistake for IBM and a complete misreading of the market. The guy in places is clearly full of hot air (he talks about the "Zilog 6502" at one point), but he has some valid points.

BTW, if you're interested in 5150 history, it's a great issue to have; there's a couple of photos of the original IBM 5150 prototype (wire-wrapped PCBs).

 

[IBMMuseum]

Unless you have an unusual MCA-dependent card, what practical good is a PS/2? I could never wrap my head around that one...

Well, the models numbers under 50 were not microchannel, with the 25 and 30 specifically set to be replacements of the PC, XT, and AT. Unless you had an unusual adapter that couldn't be replaced by a comparable MCA version, what was to hold you back (other than the price) for the Model 50 and above? At the time, some microchannel cards were relatively innovative, and didn't have much competition for their features outside of the platform.

 

[Chuck(G)]

The huge difference was that the 5150,60 and 70 systems were open-architecture, with pretty much full disclosure. I submit that this is what made them successful. To quote Dean:

Now if you asked me if we should have done our product plan differently, I'd have to admit that we lost sight of why the PC had become successful when we went to the PS/2. To enable continued growth, we should have continued with the model of building it so that other people could play. That would have allowed us to stay in control of the market. When we did the PS/2, we lost control.

(On the evolution of the ISA bus) ...but there were others who said, "No,we're going to shift everything over to the microchannel." We shouldn't have done that; it wasn't the right approach.

I recall that when I purchased my 5150, in addition to PCDOS ($40) and MASM 1.0 ($99), I purchased the tech ref manual ($100), that laid out the 5150 design in great detail, such that it wasn't long before I was buying Vector prototype cards (much cheaper than the IBM version) and cutting them down to fit a 62-pin bus connector and building my own experiments.

I didn't see any of the same level of openness with the PS/2 line, particularly with the MCA stuff.

That isn't to say that IBM didn't make a pile of money from the PS/2 gear. They have a superb marketing department that could sell cockroaches to apartment dwellers. But they lost control of the market in the process. Before PS/2, people paid a lot of attention to what IBM was doing. Afterwards, not so much.

 

[Unknown_K]

I kind of like the easy MCA plug and play method and the machines seem well built. IBMs issue wasn't MCA, it was the very expensive liscence to make MCA cards that screwed them up. If anyone could make MCA cards for a token amount then it would have caught on. EISA which was the PnP setup by the non IBM power sellers never caught on that well either (MCA and EISA were for servers and workstations mostly anyway, home users could not afford either). Even if MCA was used by all the PC makers cheap ISA machines would still have been what the home market wanted.

IBM was losing the PC wars because they were slow to get out the new chips compared to others (Compaq was first with the 386) and because of price. If you looked at Compaq prices back then they were just as expensive as IBM.

 

[krebizfan]

I kind of like the easy MCA plug and play method and the machines seem well built. IBMs issue wasn't MCA, it was the very expensive liscence to make MCA cards that screwed them up. If anyone could make MCA cards for a token amount then it would have caught on. EISA which was the PnP setup by the non IBM power sellers never caught on that well either (MCA and EISA were for servers and workstations mostly anyway, home users could not afford either). Even if MCA was used by all the PC makers cheap ISA machines would still have been what the home market wanted.

IBM was losing the PC wars because they were slow to get out the new chips compared to others (Compaq was first with the 386) and because of price. If you looked at Compaq prices back then they were just as expensive as IBM.

The higher end PS/2 cases were very sturdy, close to being literally bulletproof. But the motherboards frequently had problems with memory slots that didn't work and some MCA cards would not work with some PS/2 models. Slow systems with high prices and with many flaws was not a long term recipe for success. IBM put too much focus on not competing with their workstation and minicomputer lines and didn't bother trying to compete with the higher end PC manufacturers.

I think both MCA and EISA had made a mistake by requiring the configuration be stored on disk. PROMs were cheap enough back then to avoid the problem of misplacing the configuration disk.

 

[Ole Juul]

There's more than I've seen in a while. That Model 25 looks nice.

But the price! Is that even close these days? Especially, as the ad says, "without remote" hehe.

 

[kishy]

But the price! Is that even close these days? Especially, as the ad says, "without remote" hehe.

I sure hope that isn't market value...something unrealistic about that.

Consider the perhaps 75% chance the CRT will be utterly destroyed during shipping, the 50% chance that either the CRT or a power supply board inside will fail in some way upon being powered on...I'm pretty content with the ~$2 I paid for mine.

OK, maybe I'm exaggerating, but really...there is an enormous risk of it not surviving shipping.

And yeah, that is quite a lot of PS/2s on eBay. As for Chuck, I'd have to say the appeal of (ISA) PS/2s is that they pretty much are the same thing as PCs etc...with the additional bonus of things like high density floppy drives, realtime clocks, etc. Some are in more manageable form factors, even if they lose expandability because of it. PS/2 ports too...

 

[njroadfan]

$700 for a PS/2 Model 80? I lugged many of them off the street corner on garbage night (ok 2 of them actually). I even got monitors and Model Ms with them. The most exotic PS/2 I ever picked up was a Model 50Z with a AOX Micromaster 386 card, thats just about it. I tried to obtain a rare Leading Edge Model D3/MC from work, but it was tossed before I could get at it, did get the Adaptec AHA-1640 card out of it though.

 

[Lutiana]

I'd be curious to see what the Model 30 8086 go for. I have 2 of those that I want to get rid of once I verify that the HDDs in them work.

 

[IBMMuseum]

...it wasn't long before I was buying Vector prototype cards (much cheaper than the IBM version) and cutting them down to fit a 62-pin bus connector and building my own experiments...

I heard the fine pitch to the adapter bus, as well as the more odd dimensions (relatively more narrow for most MCA adapters) were to prevent low-end manufacturing of the time from making cards. The MCA bus improvements (more grounding pins, and slot formats being different if the functionality needed it) were good. It was still very much an evolutionary product, as much as revolutionary, and took later designs to take some of the nuttiness out.

 

[Great Hierophant]

The problem I have had with IBM PS/2s is that they are simply not enough fun. The Microchannel PS/2s all suffer from the same issue, lack of decent sound options. Sound cards for the MCA bus are extremely rare, expensive and not particularly reliable. Roland and Creative's offerings seem just quickie ISA ports, which simply won't cut the mustard for gaming.

The ISA PS/2s suffer from a couple of problems. The biggest issues is that they just aren't fast enough to be exciting. The IBM PS/2s Models 35 & 40slc have a 386SLC/20, which simply is too slow for many a VGA game. That is such a shame because they well-featured machines, with true IBM VGA on the motherboard, serial & parallel ports, ps/2 mouse and keyboard ports. They are also very well built machines.

There are other issues too that tend to vary by model. Nonstandard power supplies, power-in-drive floppy pinouts, piezo tweeters, no space for a 5.25" drive, few slots, no hardware slowdown, proprietary memory expansion, and Dallas RTCs. The 8-bit machines are especially weird with MCGA, nonstandard mouse and bidirectional parallel ports, ST-506/412 hard drive interfaces, and a dead-end upgrade path with 8086s.

 

[DOS lives on!!]

The ISA PS/2s suffer from a couple of problems. The biggest issues is that they just aren't fast enough to be exciting. The IBM PS/2s Models 35 & 40slc have a 386SLC/20, which simply is too slow for many a VGA game. That is such a shame because they well-featured machines, with true IBM VGA on the motherboard, serial & parallel ports, ps/2 mouse and keyboard ports. They are also very well built machines.

There are other issues too that tend to vary by model. Nonstandard power supplies, power-in-drive floppy pinouts, piezo tweeters, no space for a 5.25" drive, few slots, no hardware slowdown, proprietary memory expansion, and Dallas RTCs. The 8-bit machines are especially weird with MCGA, nonstandard mouse and bidirectional parallel ports, ST-506/412 hard drive interfaces, and a dead-end upgrade path with 8086s.

My Model 30-286 is a solid performer in DOS, but once I get into Windows 3.1, 1mb of RAM doesn't cut it. Everything takes a long time to load, especially Microsoft Word (which I'm suprised someone could fit it on a 30mb hard drive). They're very well built machines, just not on the RAM part.

My Model 50's HDD died. BIG ISSUE!!! None of my twenty some hard drives will fit it, and none of my 5.25 drives either. No turbo button. But, for slowing DOS down, I use MO'SLO. And we must thank Compaq and IBM for definetely pushing the Dallas RTCs.

 

[PeterNY]

I'd be curious to see what the Model 30 8086 go for. I have 2 of those that I want to get rid of once I verify that the HDDs in them work.

May not sell. I tried selling 6 IBM PCs recently and none sold.

The problem I have had with IBM PS/2s is that they are simply not enough fun. The Microchannel PS/2s all suffer from the same issue, lack of decent sound options. Sound cards for the MCA bus are extremely rare, expensive and not particularly reliable. Roland and Creative's offerings seem just quickie ISA ports, which simply won't cut the mustard for gaming.

The ISA PS/2s suffer from a couple of problems. The biggest issues is that they just aren't fast enough to be exciting. The IBM PS/2s Models 35 & 40slc have a 386SLC/20, which simply is too slow for many a VGA game. That is such a shame because they well-featured machines, with true IBM VGA on the motherboard, serial & parallel ports, ps/2 mouse and keyboard ports. They are also very well built machines.

There are other issues too that tend to vary by model. Nonstandard power supplies, power-in-drive floppy pinouts, piezo tweeters, no space for a 5.25" drive, few slots, no hardware slowdown, proprietary memory expansion, and Dallas RTCs. The 8-bit machines are especially weird with MCGA, nonstandard mouse and bidirectional parallel ports, ST-506/412 hard drive interfaces, and a dead-end upgrade path with 8086s.

Very true unfortunately.

 

[Chuck(G)]

Would it be fair to say that IBM caught the "Mac" disease from Apple? Closed-architecture, proprietary components?

 

[Unknown_K]

Would it be fair to say that IBM caught the "Mac" disease from Apple? Closed-architecture, proprietary components?

Only if they shipped all the PS/2 machines with OS/2. PS/2 ports and 72 pin SIMMs did become standard with the clones as well. If you want an "Apple" type IBM machine thats closed you need to get one of the MCA derivatives that only ran IBM UNIX (RS/6000 machines) and the 800 series PPC thinkpads.

 

[PeterNY]

Would it be fair to say that IBM caught the "Mac" disease from Apple? Closed-architecture, proprietary components?

IBM PS/2s were never intended primarily towards the customer market I believe, rather more for corporations who did not have any problems spending lots of money on machines. :D Apple never managed to get a foothold in the corporate market aside from graphical design firms, at least in Europe.

 

[tezza]

IBM PS/2s were never intended primarily towards the customer market I believe, rather more for corporations who did not have any problems spending lots of money on machines. :D Apple never managed to get a foothold in the corporate market aside from graphical design firms, at least in Europe.

Yes, indeed.

The funny thing is that if I'm ever going to any vintage BASIC or MS-DOS/PC-DOS programming I always drag out the old PS/2 30-286. I think it's because it's got a smaller footprint that other PC-type machines of the same era. Out of all my vintage machines it probably gets used the most.

(Yes I know I could also do similar coding in Virtual PC on my laptop but I do like the Model M keyboards).

Tez

 

[njroadfan]

I heard the fine pitch to the adapter bus, as well as the more odd dimensions (relatively more narrow for most MCA adapters) were to prevent low-end manufacturing of the time from making cards. The MCA bus improvements (more grounding pins, and slot formats being different if the functionality needed it) were good. It was still very much an evolutionary product, as much as revolutionary, and took later designs to take some of the nuttiness out.

...and that edge connector got a lot of use in the end. I guess the economy of scale was high enough with IBM's production that it was a cheap option for re-use as VL-Bus and 33Mhz 32-bit PCI slots.