The collected works of edefault - Page 2

edefault wrote: afaik there once came a screen saver with IRIX (named "CPU eater") which was quite similar.

Anyone any idea where to this one vanished?

regarding IRIX 6.5.22 on my ancient Indigo2 R10k you´re rigtht,
but no luck on IRIX 6.5.28 on my Fuel - there´s no CPU eater,
and the backround stays light blue regardless of what I choose.

stuart wrote: I have a 300MHz R5200, but I've no idea whether there's a blue wire or not... is this a deal-breaker?

Have a look at it: the voltage regulator is housed in a TO-220 package (See below).

Maybe it works with other PCB versions, but two modules without the blue patch wire I had reworked years ago showed up not to boot into IRIX afterwards though all tests from the PROM monitor ran through error-free. I guess it might have been mere coincidence, but I don´t want to find out that it is the other way and mess up another rare CPU module.

Please look here: http://www.in4tec-mbh.com/O2@600/gallery.html

To anyone with real deep (SGI) hardware knowledge and adjacent test equipment:
I´ll gladly send you one of these modules for free to find out what´s the matter

About the time frame: three modules get reworked next week, one of which is sold already, another one i wanted to keep myself.
But I´m as well willing to ship it - on the condition that I get the trade-in RM5271 module later (otherwise the chain gets broken).

Ah, now for the vermouth drop (does one use that phrase in english at all?): Material, Soldering, X-Ray and Shipment still totals to 250 Euro.
It is unlikely that this will change because tere are no hordes of O2-owners who trade in their modules by the dozen.
If still interested please pm me and I´ll gladly tell you my PayPal accredited email address.

Cheers

did a little googlin´ ... CPU eater is a wallpaper and once came with demos.sw, dunno in which IRIX version.

Hello,

is anyone about to provide a binary of seamonkey_1.1.18?
With 1.1.17, ssh can eventually becom compromised:
http://www.mozilla.org/security/announce/2009/mfsa2009-42.html

Oskar45 wrote: Anyway, for fun I did a <time ./testrun.sh> [with the default sizes/iterations] and get on my Fuel 700Mhz:

Code: Select all

real    0m58.05s
user    0m46.82s
sys     0m0.19s

quite fast, compared to my 600MHz R14000 IP35 Fuel


or have I missed something here eventually?

dexter1 wrote: Rich and i are still busy with Seamonkey 2.0 / Firefox 3.5 port. Hopefully i can get it done soon

any progress?
Every now and then FF/SM alert about available updates ... while we IRIXers are still stuck with FF2.0/SM1.1.
I wish you´d succeed in doing your port before the "amigafication" process of IRIX progresses too far 8-(

Nice situation ... I've been there at least three times: LED red means the system is not running at all.

Most likely, the main board is toasted - it is quite fragile, pulling it with power cord plugged kills it for sure,
even when the power is off. Just recently I had to find out that a somehow dead CPU module can also kill it.

There' s a small chance to revive it through the "jumper" trick ... never did it for me, though.
Go search for a thread about it anywhere around here.

If you want a spare contact Ian Mapleson - maybe he's got some left he's willing to sell.

ajerimez wrote: When seating the CPU, you really need to squeeze it against the motherboard to ensure that it firmly grabs the signal risers.

Same for the main board: just pushing the plastic lever may not be enough.
Press firmly against both main board and front of O2 housing (give it a hug).

a little late, now, but ... just for the records:

it is possible to make a VGA monitor work with the Indigo 2 machine,
even if that monitor is _not_ capable to handle SOG.

One has to carefully read through the tech specs to find out that HSync/VSync
are already put out to some pins on the SGI's 13W3 monitor jack on the indigo2.
Then figure out how to connect these to a standard HDSUB15 - et voila.
Soldering expertise required, though.

I did let go all of my Indigo 2 beauties (had a dozen or so) on eBay long ago,
but still own the two adaptor cables I once built.

It is also possible to implement a hardware hack that makes a SGI GDM20D11
accept standard VGA signals - the input stage hardware can process H/Vsync,
one only has to connect the adjacent pins and add one R and C component.

For some reason unknown, SGI preferred the SOG over HSync/VSync so much,
they simply did not connect the latter sometimes even if were available.

Those eBay SUN/SGI adapters are SUN adapters. They will only work with SOG monitors on SGI as I described above . If you wish to confirm, use a multimeter and compare the pinout to those for your workstation in the owner's manual on techpubs.sgi.com. Pay particular attention to pins 3, 4, 5 if IIRC on the 13W3. You will find they don't line up with the pinout of a standard VGA 15 pin connector.

As mentioned above, this is true in general, but: SGI Indigo2 machines *do* put out Hsync/Vsync as well as RGB/SOG,
so it is possible to connect them to non-SOG monitors without any additional hardware, solely by correct wiring.
I don't know if there were variants in the Indigo2 family, mine all were of the R10k Impact type.
These worked fine on Eizo FlexScan L985EX monitors, which are - AFAIK - not capable to synchronize on green.

I got two working 13W3-to-HDSUB25 cables here and am willing to let go for shipment costs if someone wants one.

[...] it doesn't don't do the SGI/IRIX user community any favors in the long run.

yep, agreed.

Best practice for Indigo2 owners:
First read tech specs, esp. pinouts, then practice some soldering.
That makes buying adaptors, regardless of what flavor, obsolete.

Both Indigo2 machines and GDM20D11 monitors have dormant options
already built in which allow for interfacing w/o need for sync-on-green.

After leaving these dormant for years, I'm planning to revive a couple of SCSI disks from an XVM array that was originally attached to my Origin file server long ago ... I vaguely remember all the fun I had back then with the logical volume manager GUI, obviously programmed in java , esp. there was a "steal foreign disk" function that would come in handy now.

However, I can't find the lvmgr on the Fuel I now plugged the external disks to.
It once was invoked from somewhere in the toolchest ... yes, I love GUIs - at least, this is silly cone gfx

A "hinv" shows the external disks as "XVM Foreign Disk". Hardware ok, then. About the related system software:
eoe.sw.xlv, eoe.sw.xlvplex and eoe.sw.xvm are installed, as "inst eoe.sw." clearly shows.

The command "xvmgr", however, is "not found", nor is "/usr/bin/xvmgr" ... and I don't remember where else it could be.

What am I missing here eventually?
TIA

installed are *both* eoe.sw.xlv and eoe.sw.xvm ... but where is the starter button?
Or should I reinstall eoe.sw.xvm on spec eventually? Which of the CDs does it come from?

thanks a lot ...
sysadm_xvm is already installed as well, still, I have no clue how to invoke the GUI.

Anyway, I got my task done through the xvm CLI. JFTR (and in case I later need this again):
from the "hinv" output I knew that there were disk 4 and 8 on controller 2.
xvm started from a shell, I issued

xvm:local> show -v dks2d4

that gave me a hint what the former machine name was ("RUTH", here)

xvm:local> steal RUTH dks2d4
phys/dks1d4

whatever that means ... same thing with dks2d8.
then I had to forcefully remove the xvm label through

xvm:local> unlabel -force dks2d4

and voila! The two raw disk devices appeared in Disk Manager.
Now I can make use of it again

Case solved! Though not through the java GUI