The collected works of robespierre - Page 9

"Binaries" could mean two different things in Lisp systems in general, not just Lisp Machines.
The first is what are called fasl files, or compiled files. These are a compiled representation of a lisp program file. What this means is that it contains the same objects (symbols, functions, and constants) as the source in a compiled form. Usually there is a map of the symbols that are used, and instructions to recreate the constants and functions when the fasl is loaded. Whether this is really a binary format is kind of irrelevant since all that matters is recreating the objects. You could just print out the compiled functions in octal, it would still be fasl.
The second is what is called a world or a band. That's a large file that contains an entire virtual memory image. For acceptable performance this does need to be a bit-for-bit copy of the memory in the machine at the time the world is saved. Some systems copy the world eagerly into memory and others use a lazy copy-on-write scheme. You can also build worlds that are incremental on other worlds, although only Symbolics seems to have done this.

If the above isn't clear enough, there is no relocation or linking involved in compiled files. Functions are visible by being bound in the environment, whether interpreted or compiled makes no difference.

The MacIvory wasn't the first Lisp Machine processor that fit inside a Macintosh, that would be the Texas Instruments MicroExplorer from 5 months earlier. It was actually faster, due to its onboard memory, and was even sold through Apple sales channels (with a Mac II chassis customized to say "Microexplorer"). Today they seem especially rare, although the only other person I know who has one got it for the same as I paid, $5. The problem is that the software from TI is not nearly as interesting.

mia wrote: I saw on the opengenera cdrom a few languages (fortran, c, basic, etc.) certainly those do not produce lisp as output.

That is how F2L and Zeta-C worked.

These notes are really great.
http://blakemasters.com/peter-thiels-cs183-startup

from https://design.osu.edu/carlson/history/ ... tware.html :

In 1993 Alias started the development of a new entertainment software, later known as Maya which would become the industries most important animation tool. Steven Spielberg chose Industrial Light & Magic to provide the visual effects in 1993's Jurassic Park. In turn the animators at ILM picked PowerAnimator as the software of choice to model the huge prehistoric beasts. They delivered the very real looking dinosaurs with PowerAnimator and reaped the Oscar for Best Visual Effects.

One unusual brochure that sold recently called the 4D/85 the "Omni IRIS", never heard that one before.
http://www.ebay.com/itm/Silicon-Graphic ... 1672985292

Also there was the "IRIS 4D/70 Hardware Specifications", which has descriptions of the G graphics system (identical to IRIS 3130 graphics), and other details.

http://www.ebay.com/itm/SGI-Silicon-Gra ... 1354397945
http://www.ebay.com/itm/SGI-Silicon-Gra ... 1340550605

well at least three pages of that TR are readable enough to transcribe from the photos ;)
I used to have a G, it's remarkably strange. 17 GE1.5s...

there were several HP-PA SBCs on VME (or even something newer like TCA??) boards. ISTR they used wacky micro-D-sub connectors to fit on the narrow panel.

You can use a MacIvory III in any of the Macintosh II series, and (if it is revision "i" or above) in the non-AV Quadras, or a '040 WGS. You have a little more choice than with the MacIvory II because separate slots for the NS 8•16 memory board aren't needed.

The high-resolution option required a Sony DDM-2802C display, which looks like an ATC monitor. There are LCD replacements for them today (search "eizo raptor")

that is not a good evaluation of security vulnerabilities... there are many attacks that do not rely on open ports or malicious local users. Heartbleed being only the most well-known.

foetz wrote:

robespierre wrote: that is not a good evaluation of security vulnerabilities... there are many attacks that do not rely on open ports or malicious local users. Heartbleed being only the most well-known.

please do list those that affect irix. i'm sure that'd be very interesting for everybody running irix.
heartbleed for example is not a problem.

Do you use glibc?
http://www.openwall.com/lists/oss-security/2015/01/27/9

Is it an 800K floppy? I'd be interested to know whether the NeXTStation can read that.
Are there any raw-floppy-tools specifically for the NeXT slab?

That always happens, the space bars are molded from different batches of plastic.

block 0 *is* the beginning of the disk, but the disk label is cached in a separate area of memory while you're in fx.
so to clear the label you need to deal with it explicitly. this is for your own good.

SoftWindows95 can assign the virtual COM1-COM4 to either of the physical serial ports ttyd1, ttyd2, etc. You can grab and release the ports while running.
It can only emulate a serial port up to 9600 bps and some modem signals are not passed through.
I don't know about "sharing", however: if more than one process tries to open the same serial port, then 1) if they try to use the port in different modes, the subsequent opens will fail, and 2) if they use the same mode, then they will both send and receive data to the port. there is no port shadowing.

Check the price of the Altium Nanoboard... lol

The Presenter cards for the Indy XL and XZ graphics (they are different cards) require special connectors that are only found on those graphics adapters. So no, they cannot work in an Indigo. There are different cards for Indy Presenter (1024x768) and Presenter 1280. The situation is similar on Indigo2, there are different boards for XL/XZ/Extreme and for Impact, and the versions for Indy Presenter will not work with Presenter 1280. In addition there is an issue with the Presenter 1280 that can permanently damage the screen if the system is not patched to the right level.

The situation on O2 is much better, there is a single adapter for all Presenter screens and they didn't have the compatibility problem. They also don't require the kludgey audio jumper that feeds audio to the Presenter. It must be routed internally.

jsloan wrote: are there any other pieces I am missing here ?

Yes. The whole system architecture.
The northbridge on these systems, supplied by Intel, contains a separate processor called the Managability Engine that is totally invisible to the OS and bypasses all its access controls. It can be accessed remotely over the Internet to gain full, undetectable, control of the machine and cannot be turned off.
This on top of the 386SL System Management Mode, which executes invisibly to the OS and is still present in every x86 sold.

jsloan wrote: How would you avoid this ?

I don't know whether the chipsets from AMD, VIA etc have something equally nasty, at least they don't advertise that they do.

looks a lot like a hole saw.

the thermal switches on top of the card cage will still trip if the blowers fail. whether that's fast enough is another question.
the other problems you're having: those old Wren drives really suffer from stiction, try all the old methods like putting them in the freezer and tapping them on the side with a screwdriver handle. The GM1 failure could be one of the sockets: those LGA sockets kind of suck and can get misaligned. Fortunately there's a debug port right on the GM1.

to make a function in C++ that uses C calling conventions you normally use the extern "C" {} syntax

The keyboard and mouse are both rs-232, like Sun used, but they are not directly compatible. The pinout is different for starters, and the Indigo connector is mini-din-6. The voltages and baud rates are also different, as are the key codes.

None of the scsi-to-ide adapters are what I would consider affordable. They cost from $150-250 generally. What you can get are cheap adapters that allow using 68-pin or 80-pin scsi disks in an 8-bit scsi machine like the Indigo. This is a much preferred solution because the computer will actually talk directly to the drive. ultra160 scsi disks are still available NOS for reasonable prices.

Your Apple CDROM will only work if it is able to switch to 512-byte-per-sector mode. I don't know if that's the case. The types of readers that are typically used on these machines are Toshiba/Plextor, which all (IIRC) do have that capability.

Edit: looks like you may be in luck:
http://www.pimpworks.org/sun/suncds.html

Another thing, you didn't specify if your have an R3000 or an R4000 indigo. The R4000 Indigo can run up to Irix 6.5.22, but the R3000 Indigo isn't supported after Irix 5.3

The only thing that comes to mind is that the boot disk is not where the installer expects to see it. The disk bays in an Octane are "dks0d1" or bus 0, device 1, in the bottom bay, then device 2 above that, then device 3 on top. The SCA drives don't need jumpers, they are set by the socket they plug into. If the lowest bay is empty, the system can still work, but it could confuse the installer.

That or the partitions are arranged differently. I don't think you can boot from an XLV (RAID), so that's not it.

It's also possible that the 6.4 media (the miniroot on the 6.4 CDROM) doesn't understand the version of the filesystem that's on the boot disk. If you use new enough install media it should understand it fine.

you can use the "shell" option from the install program to get a prompt, and then "vi /root/etc/passwd" to edit the passwords.
You said you were familiar with OSX command line stuff, but I don't know if you also know vi; you can move the cursor to the letter after the first : in "root:...:0:0:..." and keep hitting x until you clear that field. (so it reads "root::0:0:..."
the cursor keys are hjkl
then type :wq to save and quit

However, if the first field after root is literally "x", then you need to edit /etc/shadow instead

With the root password cleared, you can boot "single" and then change the other accounts passwords with "passwd username"

'Street' is not a simple calculation: they were only available through authorized resellers, and the amount of discount you got depended a lot on who you were and which system you ordered.

in 2005, an SGI reseller would have listed these prices for refurbished Octane2s:

$10,995 WT5-2P600V12-536 - Ref SGI Octane2 V12, R14K 2x600MHz, 512MB, 36GB
$3,495 WT5-1P400V12-536 - Ref SGI Octane2 V12, R12K 400MHz, 512MB, 36GB
$1,995 WT5-1P360V6-518 - Ref SGI Octane2 V6, R12K 360MHz, 512MB, 18GB
$1,295 WT5-1P300SE18 - Ref SGI Octane SE, R12K 300MHz, 512MB, 18GB
$400 WT5-BASE-004 - Ref SGI Octane2 Base System

But the system was discontinued at the end of 2004, so these were stock-clearing prices.

Prolog is a very interesting approach, but it suffers in certain practical aspects. A lot of the things you want to do are not "reversible" and so Prolog feels forced. But I'm sure that once you "get it" your mind opens in some small but significant way. It could also be that by being forced to use Horn clauses, some classes of logic bugs are eliminated, but I don't have any evidence of that.
The development of Prolog was much more influential on academic computer science than on the software development world. There is a whole swath of logical programming languages that take their cue from Prolog, including some Lisp packages.

it's hard not to see such questions as a cry for help.
most people can figure out themselves enough ways to flip the switch.
but if you're determined to use dangerous drugs, I suggest you find some Ibogaine.

Sure, you'll burn them out with write amplification.

The ThinkPad 800 used the less-capable PPC603 processor, which aside from its slower clock, also has smaller L1 caches than the PPC603e in the 850.

better check what the graphics actually is, 9-pin could be MGA, CGA, EGA, or Hercules, all different and seldom compatible.
If there is also a composite RCA port (for a television) it's CGA. CGA also has a cool light pen port.
Conversion of CGA/EGA to regular VGA-style video is simple and there are very cheap adapters out there, but not all monitors can sync at the lower frequencies used

ivelegacy:
That was just a recitation of the common video cards for a PC XT :)
There are of course other kinds of video that could be found with 9 pin connectors. The relevant factors are
1. video signal level, VGA (and most contemporary "analog RGB") uses 1Vp-p, but CGA and EGA along with most earlier standards use TTL levels (between 1.4 and 5 Vp-p).
2. Sync level and polarity
3. horizontal scan frequency and #lines
4. vertical scan frequency and refresh rate
5. for accurate color, you need to know the color data format. Digital RGB may or may not have Intensity control (RGBI), EGA uses weird 2-signals per primary color, and some systems use YUV to make life difficult.
When you're especially lucky, the monitor may be designed for multiple formats and all you need to do is flip a switch.

In reality, in person trading has been extinct since the '90s. Today the random trends are produced by HFT algorithms fighting each other in something akin to Corewars.

commodorejohn wrote: (5:4 is pretty alright too, but I doubt anybody's ever done that on a laptop.)

There were several. The Thinkpad 770Z was one of the first, I think.

also: A wider screen just means there is more room for palettes.

ClassicHasClass wrote: Macs used to face the user also, and then Steve-O somewhere around the G3 era realized it was bad marketing.

The logo on the Pismo faces the user, and it started facing the other way on the Powerbook G4.
The iBook also changed direction from the G3 to the G4.

It looks like the CZ-611C has 128KB of IPL and 768KB of character ROM. I would expect those to be soldered, but the IPL might be socketed.

uunix wrote: Has anyone plugged a PS/2 Mouse into the keyboard of these machines? I know it's stated you must use the Indigo Keyboard.. but mouse?

A PS2 mouse would be fried by the high voltages that are present on an Indigo. It might also damage the keyboard or the computer.

hamei wrote: I've never had much luck with those ps/2-usb converters or I'd risk the jewels on an Octane test

They don't convert anything. It's just a passive plug that tells the mouse to switch to ps2 mode.