The collected works of ClassicHasClass - Page 13

Continuing the story of IBM hate:

So today the HMC arrived. The vendor had kindly updated the HMC software on it for me, so it was ready to go. Fired it up. It found the server. It prompted me for the processor activation code. I typed it in. There was a pregnant pause, and then

HSCL9017 HSCL0500 A CUoD processor activation code to permanently activate 2 processors has been entered. The sequence number of the CoD code indicates that this code has been used before. Obtain a new CoD code and try again.

F$%k me. So, my second main core is deactivated because your damn little service processor lost its mind, but NOT ENOUGH OF ITS MIND to accept the code that originally activated it, and I'm going to have to ask IBM, with whom I do not have a service contract, to GIVE ME A NEW CODE FOR THE PROCESSOR I ALREADY BOUGHT.

Hopefully my vendor can do the regeneration, or else I guess I have a "free spare core."

At least now I know why ASMI wouldn't take the code, though Error 254 Could not complete request is a less than useless error message.

I really hope Tyan makes a POWER8 box so I can upgrade to that instead of waging war against another IBM box the next time this happens.

Kira wrote:

ClassicHasClass wrote: I really hope Tyan makes a POWER8 box so I can upgrade to that instead of waging war against another IBM box the next time this happens.

They will - I just wouldn't expect it to run AIX.

Given my experiences with AIX, I'd view that as a net positive - but opinions differ on that point.

(iSeries, on the other hand, is a legitimate loss.)

No, I'm quite sure it won't run AIX, but if that's the price I have to pay for not having to screw around with Big Blue corporate ...

josehill wrote: I consider OS X 10.6.8 to be the "reference release" of OS X. Everything after that seems to have valued form over function. Too bad it looks like we've seen our last 10.6 security update. It's nice to see a little bit of color get reintroduced to Finder icons in Yosemite, though. I never liked the "let's make everything gray/invisible" approach that has been in place since 10.5.

Tiger forever.

(I'm typing this in 10.4.11.)

For nostalgia purposes, it would be the Indy, hands down, and in fact an Indy was my only SGI initially because of nostalgia (I used one at the Salk Institute as an undergrad, and I remember lusting over it in Wired).

For practical purposes, I like the Fuel. It's quiet, easy to work on, does a lot, not hard to find. But, as others have pointed out, it can be pricey, though that can apply to any higher-end SGI (example: Tezro).

That might work. Thanks!

I have an old Perl 4 binary installed on it now but it's not suitable for most of the tasks I need it for.

PPC4xx is definitely too underpowered for my typical workload. I haven't had much experience with Freescale designs but it's worth a look; thanks for the recommendation.

It's not AIX. It's the hardware itself. AIX's licensing is actually pretty simple (if you have the hardware, you are entitled to run AIX). The problem is that the hardware prevented me from getting the features turned on I already paid for, but that's at a far lower level than the operating system.

Similarly, it's not PowerPC that's the problem, it's the IBM big POWER servers. That's why I have high hopes for OpenPOWER, though they've been dashed before.

IBM eventually had to generate me a new code.

That is very encouraging. I think I'll save my pennies for one.

No. Firefox is slow because JavaScript is an interpreted language and the interpreter is slow. XUL is actually pretty efficient as far as markup goes, even without a true native widget toolkit.

I'd rather have NetSurf than Origyn, but that comes from my long-standing anti-WebKit bias, which is abating a bit now that Google has Blink to take over the world with instead.

I sympathize greatly, and I know the desire to run a RISC workstation, but I would strongly recommend against this. First, POWER:

- IBM boxes run two things well: AIX and Linux (and then really only RHEL, though CRUX PPC should be okay). The *BSDs on ofppc have issues and don't support all the hardware. AIX has not been a good workstation OS since 5L, and not arguably even then. And then you still have to deal with endian problems in the huge body of Linux software written as if only one architecture exists, though this is coming to an end with ppc64le and POWER8.
- IBM POWER hardware runs hot and heavy. This 2-way 4.2GHz POWER6, throttled down, still draws around 300W and it will heat a room just as well as the G5. In fact, it's worse than the G5, because there's no liquid cooling, and even with the baffle on a deskside unit it's loud. The newest POWER7 machines are better, but good luck trying to find one at a reasonable price from a reseller, and you can forget buying from IBM directly. My POWER6, which is 2008-era, would still sell for several thousand in 2014 (I bought it for $10K in 2010). When it blew its system backplane, it was $1700 to fix and a whole lot of cussing at IBM's CUoD hardware lockout (see my thread on why IBM sucks).

It's a fast number cruncher and I love AIX jackboots, which is why I have it. But it would be horrible as a workstation.

I don't know much about modern SPARC hardware, but I can't imagine the situation with SPARC boxes and Solaris is substantially different.

As much as it pains me to say it, the last great, useable, powerful RISC workstation was the quad G5. And then, only in OS X. I really wish there were a better option too.

commodorejohn wrote:

TeamBlackFox wrote: Why the anti-webkit bias? I don't care for Safari myself, but I think its a little unfair to judge a rendering engine.

Because monocultures are inherently inbred and are exceptionally vulnerable to plague or unfavorable recessive traits. Diversity equals resilience!

You took the words right out of my mouth. I despise monocultures. We cursed Trident for this back in the day, remember.

TeamBlackFox wrote: I see what you mean then I guess. I love Gecko, but making a non-GTK port of Firefox would entail a lot of work, I don't know if we have anyone here willing to take it on. And I'd much prefer a newer Firefox, slow as it will be, to Netsurf because of licencing, which we won't get into.

You'd have to port widget/, at least, and then you'd need some way of handling native controls. It's doable, but as you say, non-trivial.

What's wrong with NetSurf's license? It's GPLv2, so no worse than GPL ordinarily ...

Look, I get your feelings; I've had this fight on another board that shall remain nameless. And for a time the PowerPC workstation form factor was still doing quite well compared to x86 workstations. Unfortunately, that time was around 2008. My 2007 C2D Mac mini gets walloped by the 2005 quad G5 on some tasks, but easily cruises by it in others, and by 2008 there wasn't much question. I despise the x86 ISA as well, but Intel has had a lot of money and time to invest in making it run well despite its warts, and in fairness to Intel they've tried to kill it at least three times (iAPX432, i860/960, Itanium) and the market wouldn't let them. So I can't really blame them anymore though I used to.

This isn't to say that a RISC workstation is useless in this day and age (my quad is still my daily driver, and does all my tasks except TurboTax and Android development), but all the current RISC hardware you will buy will be server oriented, so you'll be going a few years back for something useful. Unfortunately, the most powerful and still useful RISC workstation that does, you know, client workstation applications, was the one you sold me last week.

An SGI with Irix will still be a decent machine as a workstation, but for web browsing you might want to keep a Linux or BSD box in the closet to run Firefox/whatever and use the SGI as an Xterm for that purpose. You could even run the assisting box on ARM if you like. The Fuel should do the trick just fine. I like the Tez, but it's a loud and hot machine, and there are faster loud and hot machines for general purpose usage. Or you could install *BSD on the SGI, but that takes all the fun out of it!

The $1000 p520s are stripper units, and none of them are deskside, which means they're going to be even noisier (you can't install a baffle on them and their fans run faster). If you decide to go lower than that, say a POWER5 or even a POWER4, remember that the 970 is not only a POWER4 derivative, it has AltiVec and they don't! IBM didn't add VMX/AltiVec to "big POWER" until POWER6.

I just don't think POWER is going to do the job for you here, and that's coming from someone who's been a pro-PowerPC bigot for a very long time.

Kira wrote:

ClassicHasClass wrote: I despise the x86 ISA as well, but Intel has had a lot of money and time to invest in making it run well despite its warts, and in fairness to Intel they've tried to kill it at least three times (iAPX432, i860/960, Itanium) and the market wouldn't let them. So I can't really blame them anymore though I used to.

Offtopic historical note: 860 and 960 were completely distinct designs. 960 was a combination of a Berkeley RISC with some concepts from the iAPX 432 - notably, in the beginning, tagged memory; on the other hand, 860 was an odd proto-VLIW chip with no relation whatsoever to the 960 and an emphasis on HPC and graphics.

Editorial point conceded; I didn't mean to conflate those in my post.

hamei wrote: Can you run Rhapsody on a G5 ? That might be decent.

No. I did get it running on this Wallstreet G3 and, well, it's NeXTStep with Platinum but no app compatibility. Also, the compiler is a freaked-out gcc 2.7.2.1. I like it in theory, but when I use it, I end up spending all my time in the Blue Box.

surrealdeal wrote: could be worth the $200 in gas:
viewtopic.php?f=4&t=16728769

Mmm, no. That's a POWER3-II. Interesting as a collector's item, can do some basic tasks, might make a fun server to mess around with, but not a daily driver by any means.

Wow, that sucks. I hope you can find new employment. I'm sorry to hear it.

Fuel vs Octane comes down to a matter of choice. I don't disagree the Fuel is "less special" with respect to design than the Octane, but I do love Big Red.

PA-RISC does have the advantage of cheap, but mostly because no one loves them anymore, including HP. They're pretty zippy systems, but my top-spec PA-8900 C8000 does use a lot of juice, and Linux is best described as a "work in progress" on it. At one stage I was thinking of porting Firefox 3.6 to it since it already does run Firefox 3.5, but I never got around to it and the Fuel replaced it in the KVM slot. It runs 11i v2 TCOE.

Like POWER, I have a soft spot for PA-RISC because my first job was on a K250 and later I did contract work with a C3750. It's a very clean RISC and HP at least crammed incredible amounts of L2 in it, something Apple could have learned from with their criminally undercached designs. I came to hate HP-sUX and yet become fluent in it. Now I have two HP-sUX machines (a 9000/350 in the huge tank-like minirack and the C8000), plus a "425t" that I need to figure out if it even still works and what the hell is in it.

If you can live with the C8000's limitations, and they're significant, it's certainly not a bad system.

No, no, don't get me wrong, it's "interesting." The fact I have it on a Wallstreet laptop means I don't have to devote a KVM seat to it, I can just toss it on the project table when I want to play. One of these days I might make it do something practical, but the problem is that there weren't many NeXTStep applications then or now, and it doesn't have Carbon, so most Mac source code is out. OmniWeb was cool in the day but I'm better off running Classilla in Blue Box!

No, I know what a 425t is -- I put it in quotes for a reason. Someone slapped a 715t/33 sticker on the front, but the backplate says 425t, and it does have the Domain/OS keyboard port suggesting the original motherboard. I just never get around to dragging it out of storage since my big 9000/350 satisfies my HP jones.

libxul isn't just xul, though. It's virtually *everything* -- NSPR, XPCOM, layout, content, DOM, the compositor, NPAPI, JavaScript, docshell, appshell, you name it. The actual portion of layout/content that handles XUL is a relatively small portion. Only a few glue libraries live outside of libxul mostly so that the JS shell doesn't have to be linked against the whole damn thing.

You could make the argument that XUL is not much good without everything else, but there's nothing preventing someone from making a competing implementation; nobody bothered. I think it's more instructive to look at things like Camino and K-Meleon, which were essentially Gecko without XUL, and the delta in code size compared to the contemporary 3.6 was not large (even considering that they were swapping a JavaScript-based browser front end for a Objective-C/C++ one).

I guess it boils down to how you view it, but I'd still call that an L2 cache even if it's not as good as it could be. HP certainly did in all the spec sheets and it serves the same role.

EPIC/VLIW certainly needs huge cache for those instruction word sizes.

commodorejohn wrote: OpenBSD has been, in my experience, the free Unix that least makes me want to kill myself.

And people say you're too sparing with praise.

Hey, man, computers are much more fun as a hobby -- well, the old ones -- than they ever were as a career.

But then I'm collecting as a hobby the computers I used to work on when it was a career.

Now that I think about it, maybe you're right.

I can attest that mopar5150 doesn't bite.

I'm actually one of those nuts who runs their own mail server. I'll never go back.

But I've heard good stuff from people on Fastmail.

I seem to remember a similar problem with an iMic on my Fuel. I gave up and put the M-Audio board in.

Depends. If the machine came pre-activated it should -- and the operative word is should -- stay that way as long as the VPD module is not tampered with. However, in my case, trying to reset the system backplane upset the apple cart even with a known good code. IBM eventually did spit out a new one, thank goodness, but if you experience a hardware failure that requires replacement of the entire backplane as I did, you might wind up in the same situation. Essentially you should not take more than a single core and 16GB of RAM for granted; more cores or more RAM requires activation codes. At least POWER6 has SMT.

Unfortunately I'm quite sure that HP and Snoracle are the same way about their proprietary hardware, which is why I'm looking forward to getting an OpenPOWER machine to play with Real Soon Now.

Yeah, this is where having a home T1 and an SLA is a Damn Good Thing, plus a friend in the Bay Area I can mailpeer with. My power's been out for longer stretches than my network has (typically repaired same day, even on weekends).

I've thought about an Amazon instance as a backup cloud mail transfer point, but I just don't care enough yet.

Exactly why I question the effectiveness of encrypting mail in the general sense. Unless you know it's going point-to-point where it's not going to get tampered with, someone is going to read it, somewhere (or store it to try to read it later -- which the NSA is already documented to have done).

robespierre wrote: there are schemes called "forward secrecy" (or, redundantly, "perfect forward secrecy") that are designed to prevent that type of prospective ciphertext storage. the idea is that a conversation takes place in or near real-time, and after each step, the secret that authenticates the previous chain of steps is published. so once the recipient has received and opened a message, and used that step's secret to verify it, the protocol makes future authentication of the message impossible.

Yes, but that still doesn't deal with the insecure endpoint problem, and as long as people use $FREE mail services that don't care about such things, there's going to be lots of insecure endpoints.

Big sucker, isn't it?

As the resident Mozilla Tier 3 porter, I think you'll find most of the same problems porting SeaMonkey that you would with Firefox. While there are some widget differences, they are ultimately both XUL-based. Ditto for Songbird or T-bird.

Nice setup. Always like the Indys.

What did you use for file serving services? Was it the Mylex RAID card?

My ANS 500 cost me a "summer of work" (actually, I was paid, and they threw the server in as part of the contract, since it was of no use to anyone else).

My ANS 700 was bought as a set of two, one with a blown logic board and one that was stripped, so I combined the two and scrapped the bad one. I think I paid $100 in 2004?, but we went and picked it up (which with these units is no mean feat).

The ANSes are wonderful machines whose quirks are more endearing than annoying. Their biggest drawback is their (by modern standards) weak CPU; otherwise, they have great I/O, even for their age.

Wait, AWS or ANS? The ANS is not part of the Workgroup Server line.

On my ANS, I used uShare until it drove me crazy, and then migrated to Samba.

Unfortunately, both 10.4 and 10.5 retail packages are incredibly expensive. I was fortunate to buy my Leopard discs just after Snow Leopard came out, so they weren't that expensive. Tiger, on the other hand, can be a couple hundred itself, especially the 10.4.6 retail version which will boot any Tiger-compatible Power Mac.

osx tiger is to old.

I hate to break it to you, but 10.5 is no spring chicken either.

10.5 runs better on G5 systems than 10.4 does, but 10.4 has Classic, and I still run Classic apps from my long investment in Mac stuff. So Tiger forever, which is why TenFourFox runs on 10.4.

USB booting on Power Macs is very hit-and-miss. On the same machine, it worked once, and then I used a different device, and it didn't work. So it probably did work for them, but you can't duplicate the conditions. Stick to FireWire.

UPGRADED!!! See end

I'm making a new thread since this really is a "new" bigred -- new machine, new motherboard, new PIMM, new RAM, thanks to mopar5150. The only things carried over were the old hard disk, the SCSI card, the FireWire card and the DAT; I also installed the front bezel from the first bigred because it's a little nicer. I even installed a Pioneer 305S since it's a 40X CD-ROM drive (besides DVDs).

Just to see if the bump in cache speed makes any difference, I did a little testing in dhrystone, which is not a great benchmark but is easy to configure and run. When it was 700MHz, I got

...

At 800MHz,



I ran each one about three times and the numbers were stable. That's a 14% increase ... as expected from the simple ratio of clock speeds. Oh well. 14% is still better! And it runs very well.

It looks just like it did before except for the new drive, so refer to those pictures . Internally the M-Audio was replaced by the Audigy that mopar5150 provided, and he also threw in a USB card that works well. I'll let him comment on those, since I don't know if they're in the Fuel aggregator yet.

New hinv, gfxinfo, etc., with scsicontrol list of SCSI devices:

And after I posted that, it occurs to me that there's no way dhrystone could have tested the cache speed difference anyway, because the entire thing fits into cache and then some. Any benefit from 350MHz to 400MHz would have been well below the noise floor.

I should post a picture with the door open. The black Pioneer 305S I have really looks sharp in there, like it was made for it.