Nekonomicon - jan-jaap

jan-jaap
Donor
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Challenge L, 6*R4x00, 1GB, HIPPI at Feb. 7, 2010, 11:59 a.m...

Indyboy wrote: Thanks for your time and devotion! I'm eagerly waiting for your next observations or solutions for this problem. It'd to be nice to build a 16 proc system

Not possible, at least not in a deskside. Only 5 E-bus slots, and you need at least one for memory, and one for I/O. So max is 3 CPU boards. I'll have to settle for 12 * R10K 8-)

But they will be the 2MB L2 model.

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Donor
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Challenge L, 6*R4x00, 1GB, HIPPI at Feb. 10, 2010, 2:06 a.m...

Indyboy wrote: According to lunatic it's possible so I'm really excited how it could be done.

Me too. It's just a curiosity thing; I have both the L and the DM sitting here so this is the perfect occasion.

I could simply install the R10K boards in the Challenge L and call it quits but that wouldn't be satisfying. I need to know how things work and how to hack them

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2689 notes

Wrote on the subject of 4D/240GTX troubleshooting at Feb. 22, 2010, 9:50 a.m...

zuluchas wrote:

She's ALIVE!

Good work! And PM me if you need a newer rev. MC2.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Donor
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of O300 drive sleds at March 3, 2010, 5:39 a.m...

recondas wrote:

hamei wrote: Now just needs some quieter fans.

Looks like jj's got the quiet 0300 covered: viewtopic.php?f=3&t=16718040&

Two things:
* Change all of them. It is a pain to take the PSU apart, but if you don't you might as well forget about it entirely.
* The fan I used has been discontinued. There's the S-FLEX SFF80D (2,800 rpm) but at 32dB it louder than the NMB Minebea I used (28dB), and I believe the bearing had a shorter MTBF as well.

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of eBay: SGI O2+ R12000 400mhz 512mb at March 5, 2010, 1:34 a.m...

Hi Ian!

mapesdhs wrote: > Btw, what is a T-cut? ...

See:

http://www.carbasics.co.uk/what_is_t-cut.htm

It's designed for car paintwork but does wonders on other surfaces aswell.

I bought a 150g tube from B&Q , was about 5 UKP. Practice on some plastics you don't much care about first, get
used to it. Use it too much btw and it will discolour the surface.

[...]

PS. Some of the above solvents are toxic and damaging to skin, lungs, etc. Be careful if you use any of
these things. Well ventilated area essential.

I use a cleaning agent made for cleaning stainless steel in kitchens. It's an abrasive, fine powder ('white clay'), not toxic, doesn't dissolve anything. I got mine from one of those "eco" / "bio" shops.

It works best to remove scruffs: http://vdheijden-messerli.net/sgistuff/ ... .24-skins/

Just for kicks I give it a try on a pair of totally scratched O2s. Here's a photo after I treated the worst of the two: viewtopic.php?f=3&t=16721530&p=7305524#p7305524

As with any scraping agent: apply too much or too long and you will have no texture left. Also not really suitable to remove scratches deeper than the texture on the skins.

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Missing IP numbers.. at March 5, 2010, 2:05 a.m...

Several of the IP numbers you mention did exist: http://old.sgistuff.net/hardware/tables.php#ip

The rest probably never made it to a real product. Which makes you wonder what happened between IP35 (Fuel) and IP53 (Tezro).

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of eBay: SGI O2+ R12000 400mhz 512mb at March 8, 2010, 12:22 p.m...

mapesdhs wrote: (does the abrasive powder you mention have a typical brand name or product name?)

It's called First Stone: http://www.nettoyant-argile.com/

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx2 deskside IR2E, 2xR14K 500Mhz, 2xR12K 400Mhz, 8GB, DIVO at March 10, 2010, 2:46 p.m...

eMGee wrote:

bplaa.yai wrote:

This is the kind of machines that is really more impressive/fun for real, but pretty boring in a video...

A computer that is powered up with an ignition key and a gigantic graphics pipe? I tend to disagree!

I still have one that's looking for a new home

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of SGI systems power usage at March 14, 2010, 9:58 a.m...

zuluchas wrote: This is interesting because the power supply is only rated at 1050W. Perhaps the VGX transplant is too much for this little guy? It certainly makes a 16-gauge cable warm (so I've switched to 12).

The 1050W number is the combined output power (U*I) of all power modules of the PSU. You have to add the power lost in the PSU to calculate the power drawn from the net. I believe they are quite efficient especially considering their age, but even (let's say) 80% efficiency means it generates 200W heat inside the PSU. There's a fan in it for a good reason!

There is another rating (1500W) for the max input power the that the primary stage can handle.

These two numbers for the same PSU, combined with the fact that you have to disassemble quite a few parts before you can even read the label, has led to endless discussion threads in comp.sys.sgi about "RealityEngine capable" PSUs and which system configurations would need those. In the real world, my 4D/440VGX consumes significantly more power than my Crimson RE.

At least, that's how it is for a single tower PowerSeries. Here's the label for the PSU of one:

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of name the next zombie to rise from the dead at March 19, 2010, 9:35 a.m...

Attachment:

File comment: The new SGI O3!

new-sgi-o3.jpg [ 13.07 KiB | Viewed 333 times ]

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Personal IRIS 4D/35, R3K, 72MB at March 22, 2010, 8:18 a.m...

Looking good! I see you've got a PCI cage, but it doesn't show in the hinv (XTALK_PCI)? A hacked 3c996b-T card would be a nice addition to get gigabit ethernet on the cheap.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Personal IRIS 4D/35, R3K, 72MB at March 22, 2010, 9:21 a.m...

bigD wrote:

Thanks all! I'm pretty excited about this. Interestingly, the seller went for my price on the RM10s, but not the GE-16. Doing some searching, it appears that I'll be able to mix the RM10s and my older GE14-4 for the time being - is that correct?

Yeah, I ran my previous Onyx2 with a GE14-2 and two RM10's for some time. Ridiculous combination but it works fine.

bigD wrote:

Could it be because I don't actually have any PCI cards installed? I'll look into it, and cheap gigabit sounds good to me.

No, it should always print the XTALK_PCI near the top of the hinv. Try to reseat it, that might be enough. Otherwise, inspect the backplane for excessive dust where the compression connector mates, or scavenge a compression connector from another XIO card.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Personal IRIS 4D/35, R3K, 72MB at March 22, 2010, 9:51 a.m...

Yes, that's the one. My fault, XTALK_PCI is a PCI shoehorn

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of name the next zombie to rise from the dead at March 23, 2010, 2:58 a.m...

We have a winner! Cyclone

Back in the nineties a fancy Indigo, Indy or O2 with a PostScript RIP, now fully buzzword compliant:

Quote:

Cyclone™ is the world's first large-scale, on-demand cloud computing service specifically dedicated to technical applications. Cyclone enables you to benefit from two high-performance service models: Software as a Service (SaaS) and Infrastructure as a Service (IaaS):

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of name the next zombie to rise from the dead at March 23, 2010, 4:01 a.m...

skywriter wrote:

jeepers! already?? i missed the press release...

Had it in the mail this morning

http://www.sgi.com/products/hpc_cloud/cyclone/

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Anyone for Ramen (compositor software)? at March 23, 2010, 5:29 a.m...

Dependencies:
OpenGL version 2.0 or greater.

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Missing IP numbers.. at March 24, 2010, 12:47 p.m...

Oskar45 wrote:

My above comment regarding Indigo's hardware IP seems to be still valid, though...

Had a look at an R3000 mainboard and it says "HP1 assy xxx-xxxx-xxx". But I know for a fact IRIX calls it IP12.

The codename for the Indigo was 'Hollywood', maybe that has something to do with it?

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Fuel V10 R14000 600 MHz 2GB at March 24, 2010, 4:05 p.m...

Looks very clean! I wouldn't worry about the V10, I've got one at work and it works fine even @ 1920x1200

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Irix Pascal at March 26, 2010, 8:03 a.m...

That compiler is for ISO/ANSI Pascal. It won't compile Turbo Pascal code.

Otherwise it should work.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of SGI systems power usage at March 31, 2010, 4:52 a.m...

bri3d wrote:

my AlphaServer4/266 tends to leak over 80(!!!!!)W when off (somehow).

Wow that's bad.

An O2K/Onyx2 module uses ~ 25W when powered down, but the breaker in the back engaged.

It also has a hideous power factor of 0.25 or so when powered down, when running it's 0.97 or so.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Houdini v4.0.4 glitches? at April 2, 2010, 12:48 a.m...

SAQ wrote:

Express graphics (XS/XZ/Elan/Extreme) are OpenGL native - everything from RealityEngine on had a native OpenGL implementation.

Nope, the first OpenGL implementations were Indigo2 IMPACT, and Onyx InfiniteReality respectively.

The best reference I could find right now:

SGI graphics FAQ wrote:

Subject: -75- Why does my GL application run slower on newer SGI hardware than it did on older SGI hardware?
Date: 8 Mar 1997 00:00:01 CST

One probable explanation is that your program is using IrisGL (sometimes referred to as just "GL") rather than OpenGL. Starting with Impact graphics, SGI graphics hardware is optimized for native OpenGL. IrisGL calls are executed through an emulation layer known as IGLOO, or "IrisGL On OpenGL." This layer of emulation reduces performance.

The best solution is to port your program to OpenGL

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Houdini v4.0.4 glitches? at April 2, 2010, 8:32 a.m...

Yes, I was talking about hardware.

Here's a paper about the design of the RealityEngine by Kurt Akeley of SGI. Even though the original RE was introduced in the IRIX 4.0x days, this paper talks about OpenGL. Then again, I assume they didn't design OpenGL on a Friday afternoon

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2689 notes

Wrote on the subject of WTD: RM6-64 for Onyx1 Infinite Reality at April 9, 2010, 5:43 a.m...

An Onyx1 IR found it's way to my place recently. I managed to work around the passwords and dead NVRAM, but it's IR set was a mess: The RM was in the wrong slot, and when installed into the right slot, it fails irsaudit. It boots into a login screen though.

It's got a couple of other issues, so before I decide what to do with it, I'd like to know if there's anybody out there willing to sell/trade me an RM. I've got *lots* of SGI bits and boards, so if you're looking for something let me know.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Talk at GCC about obsoleting older IRIXes at April 12, 2010, 3:47 a.m...

SAQ wrote:

We should be able to scrape up equipment for test builds here - I personally have a Challenge R10k that has a 6.2 system disk option.

I've been doing this for while, but you must not underestimate building GCC. First of all you need a dozen or so support packages in place, and a fairly complete set of GNU tools. A full bootstrap build + regression test takes the best part of a day on my Origin 300. Older IRIX means older hardware means longer build times.

Rainer Orth, who has put by far the most effort into GCC on IRIX says building an IRIX 5.3 GCC takes nearly a week for him. I cheat a bit and can do it in an IRIX 5.3 sandbox on my Origin 300 (yes: IP35-irix5.3 ) in around 7 hours I think. You cannot sandbox IRIX 6.2 inside 6.5 because nothing depending on pthreads will work.

You cannot expect "GNU" to support systems indefinitely, having been EOLed by their manufacturer for 10+ years.

So I've got the O300 in the rack at work and it does automatic regression builds. It need a bit of hand-holding to keep up with dependencies and the occasional breakage. But that's not the problem. You need to analyze what comes out -- compare results, find the source of breakage, provide patches, and be responsive on the mailing lists. In reality, I don't have the time for this. I've got a busy job, wife, baby boy etc. etc. Instead I maintain a private patch set which fixes the worst breakage and build compilers when I have the time, not on a schedule or when upstream decides it's the time.

I've got a new 12x R10k Challenge L which is almost ready to roar. I can do the occasional GCC builds on that too, and I plan to help Tom with TGCware, but again, it's not realistic that I'm going to keep up with GCC SVN HEAD.

tgc99 wrote:

Donating cycles to making the very best gcc 4.4.x (or 4.5.x) release possible for IRIX < 6.5 would be nice.

This is exactly what is feasible in the limited time I have available.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx2 deskside IR2E, 2xR14K 500Mhz, 2xR12K 400Mhz, 8GB, DIVO at April 12, 2010, 6:10 a.m...

sybrfreq wrote:

the camera I use most often doesn't have a microphone.

Who needs a camera, when the Onyx2 has goodies like DIVO

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Unisys machine - can they run linux? at April 12, 2010, 12:07 p.m...

Pontus wrote:

if you feel like owning the biggest backup on the street, I might be able to hook you up

You could mod it into a sauna cabin

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Octane SSE/SE R12000 400MHz 1.5GB Presenter at April 12, 2010, 12:48 p.m...

Gerhard.Lenerz wrote:

Not so nice is that the system almost immediately turns fast fan on as soon as it has passed POST. I don't believe that the little Presenter card makes such a difference, I guess it's probably that a third slot is equipped with a board.

Try to arrange your XIO cards such that the lower right corner XIO slot is unused. It more or less guarantees fastfan mode which I personally find unbearable

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of small firefox 3.5 / Seamonkey 2.0 update at April 13, 2010, 1:20 a.m...

dexter1 wrote:

Anyway, i will be getting a pressie next friday which consists of a former Cave machine here at the TU, an Onyx2!

Which means finally i can try out Firefox/Seamonkey builds without having to rely on remote X connections.

Haha, until you realize how much heat an IR pipe puts out doing nothing while the machine is busy compiling

Good catch otherwise!

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Talk at GCC about obsoleting older IRIXes at April 13, 2010, 7:25 a.m...

SAQ wrote:

The newer GCC probably doesn't even run well on the older Personal IRISes.

GCC 4.x currently doesn't run at all on any machine with an R3000 CPU, and neither does any code built with it.
I have the feeling some mips2 instructions have sneaked into the crt code.

Both GCC 3.x and 4.x can consume more memory than you can install on many of the early systems, especially when compiling larger C++ files.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of readline-6.1 compiled? at April 14, 2010, 1:07 a.m...

You need to link the C library

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of readline-6.1 compiled? at April 14, 2010, 3 a.m...

ritchan wrote:

How is that done?

Add '-lc' somewhere near the end of the 'ld -shared ....' invocation.

ritchan wrote:

Why isn't it automatically done?

Normally you link by invoking the C compiler (cc), not ld directly. cc links in libc automatically.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of OS/2 redux! at April 15, 2010, 1:03 a.m...

porter wrote:

other than that it's all good.

(c) If your app thrashes the stack, it crashes the system, not the app.

I did s/w development back in the OS/2 days. We were so happy when Windows NT appeared. Windows NT (3.5.x) was slower, but more robust.

FWIW, it seems this redux thing is more about creating some sort of OS/2 personality / desktop on top of Linux.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Freebie arrived in my office: a naked Onyx2 at April 16, 2010, 3:22 a.m...

Nice freebie

Assuming all panels are installed, you cannot really fry it as long as you put it on blocks so it can blow the hot air out the bottom?

Was that an "official" rackmount or a deskside stripped naked? I still have some of the Onyx2 rackmounted deskside skins / rails, if you want to properly mount it in a rack. It's not as pretty as a deskside, but the metal front has a no-nonsense "don't mess with me this is serious business" feel to it, and it solves airflow problems.

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Freebie arrived in my office: a naked Onyx2 at April 16, 2010, 7:38 a.m...

dexter1 wrote: Is there any trick in removing the nodeboards btw? I'm testing removal on my old Origin2000 and it feels like i have to apply an unreasonable amount of force. I removed the blanking plate next to the nodeboard, but i cannot find any extra screws of latches holding the boards in place.

You need to release both the two screws and the two 7/64" hex bolts.
It can take quite a bit of force to release a nodeboard. Use a foot to keep the chassis in place and pull with both hands

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
From Wijchen, The Netherlands
Who joined June 17, 2004, 11:35 a.m.

Wrote on the subject of Freebie arrived in my office: a naked Onyx2 at April 16, 2010, 8:13 a.m...

dexter1 wrote: UPDATE: Aha! the Red cyllindrical bolts in the middle.

Orange, but yeah, those are the ones.

Beware that using European, metric tools won't work (or you're going to damage something).

I ordered a set of 'inch' hex keys online, but apparently Harley Davidson shops have them as well

Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum : almost every MIPS/IRIX system.
Wanted : GM1 board for Professional Series GT graphics (030-0076-003, 030-0076-004)

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx InfiniteReality: 4*250MHz, 256MB at April 19, 2010, 2:43 p.m...

Not too specacular, but the fact that it works at all is a small victory

hinv

Code:

  IRIS 1# hinv -mv
  
  CPU Board at Slot 2: (Enabled)
  
  Processor 0 at Slot 2/Slice 0: 250 Mhz R4400 with 4 MB secondary cache (Enabled)
  
  Processor 1 at Slot 2/Slice 1: 250 Mhz R4400 with 4 MB secondary cache (Enabled)
  
  Processor 2 at Slot 2/Slice 2: 250 Mhz R4400 with 4 MB secondary cache (Enabled)
  
  Processor 3 at Slot 2/Slice 3: 250 Mhz R4400 with 4 MB secondary cache (Enabled)
  
  CPU: MIPS R4400 Processor Chip Revision: 6.0
  
  FPU: MIPS R4000 Floating Point Coprocessor Revision: 0.0
  
  Main memory size: 256 Mbytes, 2-way interleaved
  
  MC3 Memory Board at Slot 1: 256 MB of memory (Enabled)
  
  Bank A contains 16 MB SIMMS (Enabled)
  
  Bank B contains 16 MB SIMMS (Enabled)
  
  Bank C contains 16 MB SIMMS (Enabled)
  
  Bank D contains 16 MB SIMMS (Enabled)
  
  Instruction cache size: 16 Kbytes
  
  Data cache size: 16 Kbytes
  
  Secondary unified instruction/data cache size: 4 Mbytes
  
  Integral SCSI controller 0: Version WD33C95A, differential, revision 0
  
  Integral SCSI controller 1: Version WD33C95A, differential, revision 0
  
  Disk drive: unit 1 on SCSI controller 1 (unit 1)
  
  Integral EPC serial ports: 4
  
  Integral EPC parallel port: Ebus slot 3
  
  Graphics board: InfiniteReality
  
  Integral Ethernet controller: et0, Ebus slot 3
  
  I/O board, Ebus slot 3: IO4 revision 1
  
  VME bus: adapter 13
  
  VME bus: adapter 0 mapped to adapter 13
  
  EPC external interrupts

gfxinfo

Code:

  IRIS 2# /usr/gfx/gfxinfo -v
  
  Graphics board 0 is "KONA" graphics.
  
  Managed (":0.0") 1280x1024
  
  Display has 2 channels
  
  4 GEs (of 4), occmask = 0x0f
  
  4MB external BEF ram, 32bit path
  
  1 RM6 board (of 1) 1/0/0/0
  
  Texture Memory: 64MB/-/-/-
  
  Medium pixel depth
  
  32K cmap
  
  brd: 80160004 2020c/-/-/- f9390404
  
  ge: 3 14831057 14731057 14231057
  
  rm0: 15032057 15431057
  
  4631057 1/1/1/1
  
  4d31057 0/0/0/0/0/1/0/0
  
  4938057 3/3/3/3/3/3/3/3/3/3/3/3/3/3/3/3/3/3/3/3
  
  dg: 05532057
  
  5838057 1/1/1/1
  
  5631057 0/0
  
  GE:   NIC #:      0000.0005.096e (family: 0b)
  
  Serial #:   CFY884
  
  Part #:     030-0681-003
  
  RM0:  NIC #:      0000.0004.b80e (family: 0b)
  
  Serial #:   BYR571
  
  Part #:     030-0684-004
  
  RM1:  No NIC serial number available.
  
  RM2:  No NIC serial number available.
  
  RM3:  No NIC serial number available.
  
  DG:   NIC #:      0000.0002.a6b9 (family: 0b)
  
  Serial #:   CFT316
  
  Part #:     030-0686-003
  
  DGOPT:No NIC serial number available.
  
  Input Sync: Voltage - Video Level; Source - Internal; Genlocked - False
  
  Channel 0:
  
  Origin = (0,0)
  
  Video Output: 1280 pixels, 1024 lines, 60.00Hz (1280x1024_60.vfo)
  
  Video Format Flags:  (none)
  
  Sync Output(s):
  
  Composite sync on Green
  
  Composite TTL sync on Aux 0
  
  Using Gamma Map 0
  
  Channel 1:
  
  Channel's Source Size: 1280 pixels, 1024 lines; Origin = (0,0)
  
  Video Output: 640 pixels, 486 lines, 29.97Hz (640x486_30i.vfo)
  
  Video Format Flags:  (none)
  
  Sync Output(s):
  
  Composite sync on Green
  
  Horizontal TTL sync on Aux 0
  
  Using Gamma Map 0
  
  IRIS 3#

Unfortunately, the RM-64 fails irsaudit. It doesn't seem to result in visible artifacts, though.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx InfiniteReality: 4*250MHz, 256MB at April 20, 2010, 2:20 a.m...

jan-jaap wrote:

... the fact that it works at all is a small victory

Here's what I had to do to get this far:

* The IO4 board had a Dallas with an empty battery. This makes the machine forget it's configuration and it causes IRIX to print a continuous stream of warning messages on the console -- not very handy. I replaced the entire IO4 for the time being.

* The machine was password protected, and it's security policy was rather strict. Futile once I get physical access, of course :twisted:

* The RM6 was in the wrong slot (RM2 slot rather than RM0 slot). Graphics doesn't work like that.

* The InfiniteReality FP1 front plate had been engaged rather cruelly which had damaged the connectors on DG4 and RM6. It had also crushed a handful of wires on the back side into a solid mess. I undid that, but it seems the RM6 is toast anyway. In fact, I think someone 'borrowed' an RM from this system to fix up another Onyx, and left this one in a mess.

* The GE12-4 would make 'irsaudit' crash. Fortunately, after a fresh IRIX install, GE flash and clearing the IR-EEPROM that solved itself :oops:

* All SCSI on this box is high voltage differential. I had to mod it to attach a CDROM.

* Two out of five original harddisks were dead. Bummer, but all these 2GB and 4GB disks don't add up to much anyway.

Tasks left:

* I'm going to mod the IO4 with a CR2032 coin cell holder to solve the Dallas problem once and for all. Pics will follow.

* The system controller is giving me a handful of '???' when powering down. Might be another dead Dallas. This type is still available, but I think I'll mod it anyway. I'm tired of these things.

* The hinge of the front door broke, it has been repaired but in such a way that it doesn't close anymore. I'll have to fix that.

* The system was loaded with Alias|Wavefront, ProEngineer, etc. I'll have to see if there are any licenses and other interesting bits left.

* Upgrades?

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx InfiniteReality: 4*250MHz, 256MB at April 20, 2010, 7:10 a.m...

Ryan Fox wrote:

A nice find...

It found me! I'm innocent, honest!

Ryan Fox wrote:

What condition is the machine in cosmetically ? Pictures?

Attachment:

ONYX_01.jpg [ 594.87 KiB | Viewed 208 times ]

Bah, even with a professional desoldering station it seems to be impossible to remove a Dallas from an IO4. I'm going to have to think of a plan B

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Finally got myself an Onyx2 Deskside at April 21, 2010, 3:23 a.m...

This batch, I assume ? Congratulations

Every unit is supposed to have a PSU and a fan tray, but they are amongst the parts that fail most often. Looks like someone gutted one or two systems to keep the rest going.

7/64 inch hex keys can be ordered online, or from shops that deal with machinery from the US (like a Harley Davidson shop).

If you search the forum you'll find some recipes to recover scruffed skins. Often an amazing recovery is possible. At least you've got plenty of parts to pick and choose, and you can recover part of your cost by selling of the remains.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

jan-jaap
Old Salt
Who joined June 17, 2004, 11:35 a.m.
and authored 2688 notes

Wrote on the subject of Onyx InfiniteReality: 4*250MHz, 256MB at April 28, 2010, 12:20 p.m...

I've tried it all. Even tried a professional desoldering station. It won't give in. I'm not going to push it too far and burn the PCB.

I've seen pentium's work:

Attachment:

P1270273.jpg [ 78.03 KiB | Viewed 125 times ]

I've reworked a Dallas before. This is the result (from a SUN SparcClassic):

Attachment:

IMG_2747.med.jpg [ 124.39 KiB | Viewed 125 times ]

Find the differences :lol:

Usually Dallas chips are socketed. This one isn't. Since I cannot remove it from the PCB, I'll have to rework it while it's still installed. That, in turn, means the support next to it will have to be removed.

I intend to use an SMD style CR20232 holder again like I did with the SparcClassic. There's room for one on the IO4 next to the Dallas, putting it on top of the Dallas might cause problems with IO4 mezz cards.

_________________
Now this is a deep dark secret, so everybody keep it quiet

It turns out that when reset, the WD33C93 defaults to a SCSI ID of 0, and it was simpler to leave it that way... -- Dave Olson, in comp.sys.sgi
Currently in commercial service:

(2x)

In the museum: almost every MIPS/IRIX system.

The collected works of jan-jaap - Page 6