DEC Alpha and Digital Equipment Corp. If you love the terminal on Linux or similar thank them.

[Uttamattamakin]

This is a really interesting video on computer history.  The DEC Alpha is interesting as it was an architecture that competed at the high end.  One thing they did that has had really staying power was create the VT-100 terminal which is based farther back on the old teletypes that were used to interface with mainframes in the past.  Your command line, your "terminal" app are all trying to emulate the functions of their hardware.   In fact the first way I accessed the internet was via dialup and a terminal app in Tandy Deskmate.   So many great computer types that are just gone now. 

 

 

[Rex Hite]

I had Dec Alpha workstations with Windows NT4 in my shop in the 90s. When Microsoft ended support for the Alpha I had them wiped and replaced the OS with a version of Red Hat Linux for the Alpha and sold them off. The emulator program FX!32 which enabled 32 bit x86 code to run was essentially witchcraft. It re-wrote x86 dlls and then used the new optimised ones next time the software was run.

 

[Uttamattamakin]

On 12/17/2022 at 3:55 PM, Rex Hite said:

I had Dec Alpha workstations with Windows NT4 in my shop in the 90s. When Microsoft ended support for the Alpha I had them wiped and replaced the OS with a version of Red Hat Linux for the Alpha and sold them off. The emulator program FX!32 which enabled 32 bit x86 code to run was essentially witchcraft. It re-wrote x86 dlls and then used the new optimised ones next time the software was run.

 

Then did it better than a native X86 32 processor.  Intel killed so much innovation thank God AMD's X64 was just too good for them to strangle it.   We could've had processors that were so much more efficient, and open (Unlike how ARM has been treated almost always sold with a locked platform for an appliance like experience).   

 

[Radium_Angel]

On 12/17/2022 at 2:39 PM, Uttamattamakin said:

 So many great computer types that are just gone now. 

The enemy of "great" is "good"

x86 was "good enough" and more importantly "cheap enough" that all the cool tech (SGI, Sun, DEC, etc) died off 

[MinnesotaSteve]

Back in 70s/80s a lot of computers used terminals.   The ANSI escape codes which are used in Linux and such didn't really come from the VT100, but rather the VT100 was the first to implement the ANSI spec as they were trying to standardize these controls at the time.

 

I had a DEC Alpha on my desk at the university in 1994, it was one of the first ones we had.  They were a fine computer.   I later worked on a Alphastation 400 which was running Windows NT.   But by the time Alpha had come around, DEC was in trouble.   So I don't think they were ever able to really market or take advantage of it.   The quality of the hardware had been declining.   Those Alphastations that had been purchased, were sent back and replaced with Compaq Proliant 5000's as the DEC hardware wasn't reliable at the time.   This would have been 1995 or so.   DEC was bought out in 1998 by Compaq.
 

 

 

 

[jasonvp]

On 12/20/2022 at 6:06 PM, MinnesotaSteve said:

Back in 70s/80s a lot of computers used terminals.   The ANSI escape codes which are used in Linux and such didn't really come from the VT100, but rather the VT100 was the first to implement the ANSI spec as they were trying to standardize these controls at the time.

 

The OP is sort of conflating what happened.  I think it's safe to say that the terminal folks use on any *NIX system got its genesis from X11.  Sure, VT100 ANSI-based terminals existed far longer than X11 did, but what we're used to playing on and with is a graphical representation of said.  And that came from X11.  Each *NIX OS had its own X11 server, and DEC was certainly no stranger to it.  But that far back, it was their Ultrix OS, not the Alpha and OSF/1 (original name for the UNIX that ran on the Alpha).  Ultrix existed way, way before the Alpha did, but it still had ye ol' X11 server and the xterm application.

 

Make no mistake: the Alpha was awesome for its time.  I attended Clarkson University from '91-95 and the school was saturated in IBM RS/6000 workstations and servers.  They were pretty fast; certainly trounced the snot out of the Sun workstations and servers we had.  The last year I was there, we finally got an Alpha and it was eye opening.  One of the guys I worked with there had written a Fortran 77 program that computed pi out to some ungodly number of digits.  The quickest time he could get out of the fastest RS/6000 we had was almost 21 minutes.  The Alpha did it in under 7.

 

A couple of years after I graduated I found myself as a fledgling network engineer at a little ISP no one's ever heard of: AOL.  We had fleets of Alphas used as web proxies for our dial-up users.  The I/O throughput on them was insane and the 64-bit memory addressing helped keep all that stuff in a RAM cache for quick retrieval.  I even had one under my desk as my workstation. 

 

[wseaton]

Ran NT 3.51 on an Alpha.

 

Holy Jesus that platform was fast.

[Spindel]

On 12/19/2022 at 5:20 AM, Uttamattamakin said:

Then did it better than a native X86 32 processor.  Intel killed so much innovation thank God AMD's X64 was just too good for them to strangle it.   We could've had processors that were so much more efficient, and open (Unlike how ARM has been treated almost always sold with a locked platform for an appliance like experience).   

 

OMG stop with the ”open” argument. All  the CPU archs have been more or less the same in terms of openness be it the Z80, 6502, 68000 or x86. 
 

People bought in to brand recognition and That’s what killed most of the competition. 

[Uttamattamakin]

On 12/31/2022 at 10:10 AM, jasonvp said:

 

The OP is sort of conflating what happened. 

 

I'm not conflating those things.  

When you run a DOS emmulator it is not a conflation to say you are emulating DOS.  Your Xterm program is itself an emulator of the VT-100 terminal.  https://tldp.org/HOWTO/Text-Terminal-HOWTO-10.html  

The rest of what you said is true.  The things the VT-100 did had precedents going back to the Teletypes (TTY) used to interface with Unix systems way back, built on that.  Then X11 came and one of the first things they wanted was a way to interact with all the old programs that expected a VT-100 terminal.   

 

 

Quote

 

 The last year I was there, we finally got an Alpha and it was eye opening.  One of the guys I worked with there had written a Fortran 77 program that computed pi out to some ungodly number of digits.  The quickest time he could get out of the fastest RS/6000 we had was almost 21 minutes.  The Alpha did it in under 7.

 

A couple of years after I graduated I found myself as a fledgling network engineer at a little ISP no one's ever heard of: AOL.  We had fleets of Alphas used as web proxies for our dial-up users.  The I/O throughput on them was insane and the 64-bit memory addressing helped keep all that stuff in a RAM cache for quick retrieval.  I even had one under my desk as my workstation.  

 

WOW.  I only got to read about DEC Alpha in PC magazine and that sounds like just what I would've expected.  Meanwhile there I was perhaps putting together a barebone PC with an AMD K5 or K6.  

 

On 12/31/2022 at 12:11 PM, Spindel said:

OMG stop with the ”open” argument. All  the CPU archs have been more or less the same in terms of openness be it the Z80, 6502, 68000 or x86. 
 

People bought in to brand recognition and That’s what killed most of the competition. 

 

REally consider this then read the rest. Apple Silicon with undocumented instructions so you can't just run any OS on it. 

 

https://news.ycombinator.com/item?id=25559145  

 

The "Open" argument comes from this fact.  Can you painlessly install another OS on your ARM PC (aka your phone or even ARM Mac)?  Not really.  So far only x86 and RISCV are really open to build a PC onto which a full desktop can be installed without hassles and it can then be used to run Crysis or whatever else you want and have the code for. ARM SBC's are just not there yet.  

[Spindel]

17 hours ago, Uttamattamakin said:

REally consider this then read the rest. Apple Silicon with undocumented instructions so you can't just run any OS on it. 

 

https://news.ycombinator.com/item?id=25559145  

 

The "Open" argument comes from this fact.  Can you painlessly install another OS on your ARM PC (aka your phone or even ARM Mac)?  Not really.  So far only x86 and RISCV are really open to build a PC onto which a full desktop can be installed without hassles and it can then be used to run Crysis or whatever else you want and have the code for. ARM SBC's are just not there yet.  

You really wan’t to go this route?

 

Well for starters I replied to this: ”Intel killed so much innovation thank God AMD's X64 was just too good for them to strangle it.   We could've had processors that were so much more efficient, and open”

 

I also gave several examples to contemporary archs that all are/where basically the same in terms of openness as x86 and thus pointing out the ”openness” fallacy of your argument.

 

ARM is als just as ”open” as x86 and you can build your own system with an ARM CPU. It probably won’t be as versitile of a system but throw in a Linux dist and a realtek, annapurna or snapdragon CPU and you have a computer. Not being able to install windows or play your game of choice on that computer has nothing to do with the openness (or closedness) of the arch it has only to do with software support of the arch. If someone wants to write the software they are free to do it.

 

Apple Silicon is just one flavour of ARM, the SoC are not sold separatley but sold as part of a system, so here it’s OK to talk about a bit less of openness. But again Ashai Linux project showed that at it’s base it is ARM and they had their PoC dist running fairly quickly and working. What Ashai project struggles with are the non ARM parts of AS like the GPU and special function blocks. 
 

I’m guessing you don’t understand the cincept of ”open”, just like the user on this forum that a while back claimed that Windows was more open than MacOS (hint it’s exactly as closed). 
 

You confuse ”open” with ”broad software support”.

[Uttamattamakin]

On 1/6/2023 at 2:47 AM, Spindel said:

You really wan’t to go this route?

 

Well for starters I replied to this: ”Intel killed so much innovation thank God AMD's X64 was just too good for them to strangle it.   We could've had processors that were so much more efficient, and open”

 

I also gave several examples to contemporary archs that all are/where basically the same in terms of openness as x86 and thus pointing out the ”openness” fallacy of your argument.

 

ARM is als just as ”open” as x86 and you can build your own system with an ARM CPU. It probably won’t be as versitile of a system but throw in a Linux dist and a realtek, annapurna or snapdragon CPU and you have a computer. Not being able to install windows or play your game of choice on that computer has nothing to do with the openness (or closedness) of the arch it has only to do with software support of the arch. If someone wants to write the software they are free to do it.

My point exactly. 

 

On 1/6/2023 at 2:47 AM, Spindel said:

Apple Silicon is just one flavour of ARM, the SoC are not sold separatley but sold as part of a system, so here it’s OK to talk about a bit less of openness. But again Ashai Linux project showed that at it’s base it is ARM and they had their PoC dist running fairly quickly and working. What Ashai project struggles with are the non ARM parts of AS like the GPU and special function blocks. 
 

I’m guessing you don’t understand the cincept of ”open”, just like the user on this forum that a while back claimed that Windows was more open than MacOS (hint it’s exactly as closed). 
 

You confuse ”open” with ”broad software support”.

Which misses the point you just stated you got.  Let me lay it down for you.  For normal people if it does not run Windows and is not a Mac then it's not a computer.  For people who are just a bit smarter they can use Linux.  Sure one could run Crysis on a Cray Super Computer, one could run SCO Unix on a X386.  You know what you cant do with those? 

You couldn't go to a store and buy a box of software, back in the day, or now download random software and just run it.  Given that the instruction set is not fully documented for most ARM SOC's one can't even really write a proper compiler for it.   So yeah you got me a software engineer with an EE background can make a toaster run Doom. 

[Spindel]

13 hours ago, Uttamattamakin said:

My point exactly. 

 

Which misses the point you just stated you got.  Let me lay it down for you.  For normal people if it does not run Windows and is not a Mac then it's not a computer.  For people who are just a bit smarter they can use Linux.  Sure one could run Crysis on a Cray Super Computer, one could run SCO Unix on a X386.  You know what you cant do with those? 

You couldn't go to a store and buy a box of software, back in the day, or now download random software and just run it.  Given that the instruction set is not fully documented for most ARM SOC's one can't even really write a proper compiler for it.   So yeah you got me a software engineer with an EE background can make a toaster run Doom. 

But running Windows or MacOS has nothing to do with openness as per your initial statement. It’s a software support issue.