The collected works of SAQ - Page 13

InfiniteReality4 was based on the same basic chips as other IRs, just with some speed-bumped and with more memory.

Looks like Bali was to be to IR what ODYSSEY was to IMPACT.

The POK(x) errors are board-level, so try pulling boards to solve that. The system controller will shut off the system on these errors, but you should still be getting system controller power and the +48V lights, so look at your OLS module(s).

Check this out if you haven't yet: http://www.futuretech.blinkenlights.nl/chalonyxdiag/

I have never had my OLSes open, so I"m not sure where the ACOK tap is.

The only reason that the US has two breakers is because the center tap of the distribution transformer means that you can get 120V on either side relative to neutral/ground (and therefore current), so both sides need to be protected. If you have 240V without the center tap you only need one breaker because the other side would be the "ground/neutral" (similar to the way the US does 110V).

His response could be read as not wanting to package and ship at all (hence the reference to SFBA). Properly packing a computer takes some work (finding a strong box, proper antistatic and packing materials, etc.) and it can actually be a good indication - anyone can drop something in an envelope or an old shoebox, but it wouldn't get to you in very good condition.

I wouldn't be surprised if the "303"/"305" boards had them too - but those are also likely rare (IR graphics). I certainly don't have any with my CHALLENGE.

My elementary school had Apple ][-series boxes, and programming was a part of the computer curriculum, starting at least in 4th grade. Nothing special - we used LOGO, then BASIC - but a lot more in-depth than what they're doing even at the Jr. High level in the school district where I work.

There are implementations of these simple languages still around, but programming is not considered educationally important enough to devote the resources to. The high schools offer Java as an elective, but that's about it (they've still got a decent electronics program at at least one of them).

I have a IP25 thowing POK errors - I can give it a try with my 100W+ stained glass iron.

I'm not sure that I would - if you're not careful with those you can pop the board layers apart. I had a KSU controller board that started spitting molten solder as the layers delaminated - not fun.

No - that would really cause damage. If I were using a torch it would be one of those small ones used in brass instrument repair. Hot flame, small enough to be directed at the solder joint without burning the board.

How are you on binary patching and debug/writing of partially undocumented firmware?

Do a search on "Indy RM9000" on the fora.

Good advice, but it does sound like it belongs in some sort of techno cross-talk act: "you backed up before trying to back up, right?..."

I don't have a FireGL, but for Visualize-FX pretty much everything you need is on the MCOE.

They didn't spend the time building up the backend that SGI did - the Visualize-FX hardware was OK, but HP didn't spend the money and time putting the support stuff in HP-UX that SGI put in IRIX.

I think I heard rumors of a bulldog for one of the PDP OSes. Perhaps in reaction to Albert.

Anyone know of somewhere online where they have the logos in bigger files than the openvmshobbyist.org does?

If that's their console then it isn't "interesting" so much as a good exhibition of many of the things that are wrong with UI design now.

s/VMS/DCL

You might want to look at small servers.

Itanium workstation hardware is rare, as are late Alpha workstations. In addition, many Alpha workstations had WinNT-only parts and/or are very expensive. OpenVMS really doesn't gain much from having a local framebuffer, anyway.

Hmm...Interesting... I hope that the concept of building UIs around the concepts shown in movies doesn't go any farther. Especially not to the GNOME/KDE developers.

rx1600/1620 is pretty much your 1U option for Integrity, and as with pretty much all other OpenVMS HW the "inexpensive" option involves finding someone nearby getting rid of one. The reseller market is pretty strong, so prices are high.

Provided that you're running the (single) copy of Windows XP that was licensed on the mainboard on a VM running on that mainboard I think you could make a strong case that it is within the license (since you are running a single copy of the licensed software on the licensed machine). The sad part is that you might really need to make a case in court, as MS' approach seems to be a blanket "you have to buy another license" in most cases, and they're very creative sometimes.

I would be a bit surprised if "IRIX 6" really means "IRIX 6". It more likely means IRIX 6.2 or later (IRIX 6.0.1 and 6.1 were based on IRIX 5.2/5.3).

VMS being labeled "obsolete" surprised me, but then I remembered that it has EDIT/TPU and so doesn't need EMACS. "Obsolete" must be their way of saying that the grapes must be sour.

Have you checked and verified proper PSU voltage? The IIci PSUs do tend to have problems after this many years, though it's often with the +5VSB line. The other bit I'd look at is the memory and SIMM sockets. The rest of the IIci is very well built.

False scarcity has been a recognized marketing maneuver for a while.

My big beef with Apple is that they're capricious. Almost all other manufacturers commit to supporting a product for a set amount of time, whereas for Apple it gets dropped when it isn't cool any more (and OS updates seem to stop after one year, which is well behind the accepted standards (usually 5 years)). I understand that IBM probably required royalty payments for Rosetta/QuickTransit, but why not keep it around as an extra cost option for people who need to run PPC apps? MS kept around XP mode (as an "extra cost option" so to speak).

Also - MS may be a scumbag as far as business goes, but they do have a pretty good in-kind donation program for nonprofits.

Let's qualify that - for many of us Apple gets a lot of things wrong. There are plenty of people who buy technology as an appliance, don't mind getting a new one every 3 or so years, and don't mind getting new applications every time. If there weren't then Apple would be in trouble.

Compare that with the occasional "hacker friendly" company that puts together an interesting architecture with openish, easy to expand specs and dies after a few years. Apple's attitute is just "follow the big bucks and to hell with the rest of the world."

The EV7z only shipped in the big systems, though you might be able to use some boards in smaller systems provided you could find them.

The 21164 was the last of the Alphas where DEC was seriously courting third-party builders, but you should still be able to find something on the 21264. The 364 came after Compaq bought DEC, so I wouldn't expect as much there.

Multia will run v7.3-2 with a bit of hacking, and when run vertically and with a jimmied fan it will be cool enough. The downside is that most don't come with 3.5" internal drives (so you have to add in an external drive box), and memory is tight.

VAXstation 4000/VLC is a thin pizzabox, but only 5 VUPS. VAXstation/MicroVAX 2000 is a lunchbox sized machine, but even slower and harder to keep running (MFM/RLL drive).

DS10L is a 1Uer, but can be noisy.

I'd say a 3000/300 series Alpha would be smaller than the PWS - the PWS is a midtower PC size machine.

I didn't say I recommended it, but a Multia was my first VMS system and it is possible to get it working, and they are small and cheap if those are big considerations, though I was very happy to move up to a DEC 3000/300X that I swapped a Sun for. My Multia came with the stand and I "hotwired" the fan to run all the time. Note that I wasn't running it the summer when it was hot enough to lock up the Indy. At the time I wrote the first post I forgot the most irritating part of the Multia - the SRM console needed for OpenVMS is taken from a different machine (but similar) and it has a bug on the Multia - when the battery dies, the console is trashed until you flash the Multia SRM back on (blind) and then flash back to the OpenVMS capable firmware.

The downside with most DEC systems is that you're either in an area with bunches of them that are given away if you know people, or you're in an area where they're hard to come by and a low-middle range 21164 system will run you three figures.

I also didn't really recommend the 2000, just noted that it was an option. For actual use, even a SOC VAX system such as the VLC can feel a bit slow, especially with DECwindows (I have a 4000/200 32MB). If the OP wants a workstation with graphics and can afford it I'd recommend a 21264 or better, since those will work with the Radeon 7000/7500 cards and have the '264 extensions.

I guess the problem is that we don't know what the OP's definition of large is. If the shipping cost is the big factor then perhaps a PC164LX provided he can source a S3 Trio64, QLogic (or other supported SCSI card), and a DE500/other supported network card assembled in a found case. Again, it would require a bit of work, but there are good "how do I make this run OpenVMS" howtos out there.

If you and enough people like you shelled out $20k+ for a RISC workstation 4-ish years ago they'd still be around. If enough people bought an Indy or Alpha when they were pushing them to the power users they'd still be around. The inexpensive software wasn't there, true, but then Intel and AMD caught up. Take a look at the PCs we have now - they make an Ultra-80 look so-so for about $500-$1000 (multicore somewhat-crossbarred, 2/3-3/4-assed switched I/O, 2/3rds assed drive interfaces that will kind of do reordering - yeah, doesn't sound so fantastic that way, but compare it to where PCs were 10 years ago). The fancy, really nice RAS-type stuff generally wasn't in the workstations, anyway, unless it was the ones that really leveraged server parts, and there are still companies that make very nice servers with non-Intel stuff.

That said, I do wish that PCs came with decent firmware. EFI is a step up from PC-BIOS, but still not in the same league as SRM or OpenBoot.



BSD? Windows is the way it is because MS marketing likes gegaws. Linux distros are the way they are because it's "cool" and everyone feels the need to do their own thing. Solaris is slow because... well, I don't know why Solaris is so slow - but ZFS is neat. The BSDs tend to be pretty speedy, so try one of those.

Something like that - I can't remember exactly right now. It was not my favorite machine. Nice form factor, but too slow, too few SIMMS, occasional VMS weirdness, and the external drive bay negated the size advantage. It was, however, only $20 and local - at the time the best Alpha deal I had ever seen. Of course, shortly after that the 3000 and then a 1000A showed up local and cheap.

I know that I've heard 150MHz R10k had video capture issues, and I have a lingering suspicion that 175 might have as well.

Depends what the OP is going to use it for. The VLC is a nice mini-pizzabox size, and provided that 5VUPS is enough it works well. 4000/60 and 4000/9x are nice machines, but if the OP wants to learn VMS all VAXes are stuck at 7.3, and while VMS is a mature OS and HP fortunately doesn't feel the need to rearrange the UI and add a bunch of SDVs (stuff of dubious value) every couple of years to drive sales there are some differences between 7.3 and 8.3/8.4.

VAX peripherals and add-ins are also harder to come by.

I don't know if I'd classify the VS2000 and 4000/VLC in the same class as Multia. They were both designed as usable standalone machines in their own right, with sufficient memory/disk expansion and computational resources. Keep in mind the 2000, while slow by today's standards, was a mid '80s machine and in the low end of DEC's lineup. Compare it to what you got with a 8088 or 80286.

Maybe. If you're lucky.

The "Enterprise systems" sales teams sometimes can't be bothered if you're not buying at least 20 of whatever it is they're selling.

IBM seems to be better, but I've still heard of people trying to get stuff (i and AIX) and been brushed off because they only wanted one.

Never said they didn't add new legitimate features, just said that they didn't make gratuitous changes. If you fire up a WinNT4 or Win2k system there will be noticeable differences in how it is administered. Same with Macintosh and many other systems - but you can learn basic OVMS admin on an old version and most of it carries over because DEC/CPQ/HP didn't gratuitously change things.

Sounds like finances are probably an issue for the OP, which is the case for a number of people, and unfortunately Alphas hold their value more than many other computers. Perhaps he should look at SIMH or another emulator - while you don't get the feel of the hardware you can run OVMS fairly easily and get a feel for it.

As far as HP-UX equipment goes there isn't much that's small, except for really old stuff that runs 10.20 or earlier. HP-UX is a pretty straight SysV anyway, so I wouldn't worry too much about it (unless there's a definite job attached).

I wish that OVMS ran on a DS20L, though that would probably boost prices on that one substantially. 1U duallie.

Then SGIs are out - Shake.

Better definition goes back to the Xeroxes, Apollos and Suns that "defined" the genre. A high-end desktop machine equipped with a graphics head and networking (though this does not necessarily have to be used), solid and stable hardware/software and purchased for a task-driven reason, and used most often by a single user (though most legitimate workstations offered a second concurrent user license for remote admin - something that MS still hasn't picked up on).

Back when the hobbyist program was being run by volunteers it could sometimes take a day for the Encompass data to propagate, and a couple of weeks for the disks to be sent out. All things considered that really wasn't too bad, but now it's even faster.