Share commercial Unixperiences

I remember typing "make -j && beep" into a terminal to build xemacs on some new toy the compute center on the university had obtained. Turned around to get off the seat to get me some coffee, but there was this damn beep. This was my first contact with a Cray T3E.
 
vmisev said:
Please do tell us more 🙏 Which software run on Sun servers and SGI WS? Did SGI maybe run Linotype-Hell DaVinci? AFAIR SGI was huge improvement and much smaller and cheaper than previously used Linotype-Hell LinoServer beasts (their size was comparable with AS/400 B20-B40 models)
Click to expand...
Honestly I don't remember the specifics of the software, but they did not directly drive any printing presses, they were for prepress production to make plates for large printing presses. The SGI workstation was used for color/color separation work, as I recall. Most other workstations were Macs, and around that time macOS X was born, and we were experimenting with that on one machine to see where the future lay.
 
vmisev said:
Just remembered something, did any of NT’s in your company had Apple YellowBox running? YellowBox was superset of NeXT’s OpenStep (which had Adobe's Display PostScript), combined with other Apple software tech and Java.

I had it at $DAYJOB running on PC with NT 4, serving as a base for Heidelberg Signa Station imposition. AFAIR, it was rock stable in that role.
Click to expand...
No, I don't recall any of the NTs running YellowBox.
 
JWJones said:
Honestly I don't remember the specifics of the software, but they did not directly drive any printing presses, they were for prepress production to make plates for large printing presses. The SGI workstation was used for color/color separation work, as I recall. Most other workstations were Macs, and around that time macOS X was born, and we were experimenting with that on one machine to see where the future lay.
Click to expand...
Thanks, so SGI vas probably running DaVinci, that fits your description of the role perfectly, but Sun was obviously running some RIP, and it’s driving me crazy now that I can’t think off the top of my head of any older RIP that run on Sparc, grrr 🤬 (mad at myself, not you!)
 
JWJones said:
No, I don't recall any of the NTs running YellowBox.
Click to expand...
Heidelberg was only company that I know of that licensed and shipped it with their products – AFAIR it costed an arm and a leg, not only because it was Apple thing, but they had to pay sublicense for Adobe's Display PostScript as well.
 
I just learned something about Sun and endianness IDK before! Looks like that we (FreeBSD), at least x86_64 version can't mount Sun .iso UFS images if they were made on Sparc (big-endian). Surprisingly, Debian (tested on 12 x86_64 QEMU VM) is mounting them without any trouble with simple
mount -t auto –o ro disc_name.iso /media/cdrom/

I was frustrated with myself because I couldn't remember which RIPs were used on Sun Workstations, so I googled and googled and found on Internet Archive "Sun NeWSprint v2.5 for Solaris 2.x" There is a PDF brochure, but not much info about it in it. I knew that this is hardly for image/platesetters, probably for Sun rebranded Lasers and Inkjets so I downloaded newsprint-2.5.iso hoping that some more info can be on the disc, but whatever I tried, I couldn't mount it on FreeBSD.

So, I googled some more and found somewhere in the mailing lists from the time of FreeBSD 10 message that Debian can mount them okay.

I rebooted into Alt Sisyphus - there is where my other Linux's VMs are - started virt-manager and Debian 12, scped newsprint-2.5.iso to it and mount succeeded on the first try; to repeat myself, with:
sudo mount -t auto –o ro newsprint-2.5.iso /media/cdrom/

How that endianness magic works, and what Debian have that FreeBSD lacks?

P.S. edit: s/UDF/UFS/g # thanks covacat for pointing out this typo

newsprint-2.5.jpg
 
file says it's ufs not udf or cd9660
newsprint-2.5.iso: Unix Fast File system [v1] (big-endian), last mounted on /CD_Image/2.5NEWSP_M2, last written at Wed Jun 23 19:36:37 1993, clean flag 1, number of blocks 133263, number of data blocks 130558, number of cylinder groups 4, block size 8192, fragment size 1024, minimum percentage of free blocks 0, rotational delay 0ms, disk rotational speed 90rps, SPACE optimization

freebsd wont do cross endian ufs but netbsd will
 
covacat said:
file says it's ufs not udf or cd9660
newsprint-2.5.iso: Unix Fast File system [v1] (big-endian), last mounted on /CD_Image/2.5NEWSP_M2, last written at Wed Jun 23 19:36:37 1993, clean flag 1, number of blocks 133263, number of data blocks 130558, number of cylinder groups 4, block size 8192, fragment size 1024, minimum percentage of free blocks 0, rotational delay 0ms, disk rotational speed 90rps, SPACE optimization

freebsd wont do cross endian ufs but netbsd will
Click to expand...
UFS, of course, that is what I meant! 🤦‍♂️ UDF was typo, thanks for pointing it out; edited and fixed! Man, if you only knew in how many ways I can misspell 'search' 😵‍💫

Another question: will mount of this iso work in FreeBSD/PowerPC in QEMU?
 
covacat said:
freebsd wont do cross endian ufs but netbsd will
Click to expand...
This explains an awful lot. I still have disk images from netbsd/68k and for the heck of it could not mount them.

As for normal .iso files, tar (as from base) can read them fine.
 
vmisev said:
started virt-manager and Debian 12, scped newsprint-2.5.iso to it and mount succeeded on the first try; to repeat myself, with:
sudo mount -t auto –o ro newsprint-2.5.iso /media/cdrom/

How that endianness magic works, and what Debian have that FreeBSD lacks?
Click to expand...
Turns out, there is nothing special about Debian in this regard, any Linux can mount cross-endian UFS .iso images, Alt (6.12.44) can do it, even old Fedora 25 (4.13.16) can:
Screenshot_20250906_103831.png
 
ralphbsz said:
Interesting tidbit: The first HTTP web server outside of CERN, the first one in the US, and the second web server in the world, ran on a NeXT cube, in an office about 50 feet down the hallway from me.
Click to expand...

That's neat :-). Unix/*BSD/Linux/etc gets a lot of credit for powering the Internet. Even today ! (As I stare down at my iPhone..)

I then spent two years employed by HP, with a HP-UX machine on my desk, and a Windows laptop. The HP-UX machine was also completely boring and perfect; I used Motif under X for a desktop (mail and web browser), and otherwise mostly shell work. Lots of development in languages such as C++, Tcl and others. Given that I was inside HP (matter-of-fact, in HP's building 1, the same building where Mr. Hewlett's and Mr. Packard's offices were), there as no system administration to be done, we had a competent set of professional admins. HP's compiler was extremely good; some colleagues tried to use gcc, and were always amazed at what a complete piece of crap gcc was.
Click to expand...

Ya HP-UX and the HP-9000 series were good machines. In the shops I was employed at there were a lot of HP-UX running and producing production work loads on HP-9000 -- (10.10/10.20/11.x/etc). I actually had an HP/UX 9000 Workstation on my cubical desk at a few of these shops (mostly model 715 or related). Excellent machine ! I was able to get a lot of C/C++ programming done on it. And "yes!" the X/Motif interface on HP/UX worked very well ! Nice work by HP.

I had colleagues that used Sun machines, with both Solaris and SunOS.
Click to expand...

THAT was a real delima when Sun switched from SunOS to Solaris. I was not a huge fan of mixing BSDisms with AT&T SVRxisms to create the (BSDthing that is ALSO the AT&T SVRx thing -- or the thing that no one can understand. But it's "cool", right?). I understood that the industry was desperately in search of an "open standard" (call it POSIX, Open Desktop, (anyone remember OpenDOC? :-), etc) but (I personally think) that was just industry smoke and mirrors to say "we are OPEN but we are really closed". Linux busted the door wide open on that front and that ended the "Clone Wars"... I mean the the "Unix Wars". At that time I was pulling for BSD side of the house because I really liked the innovations that were coming out of the west coast labs - I guess they call the *BSD folks the "long hairs" now.

I still enjoy in the year 2025 that the following still works:

Linux Shell> ps -ef (gives you all the running processes on the sytem - but DOES NOT on *BSD)
*BSD shell> ps -aux (gives you all the running processes on the sytem - but DOES NOT on Linux/AT&T SVRx/etc)

I guess now days it's:

*BSD shell> ifconfig (whatever args) -- something works !
Linux shell> ifconfig (whatever args) --- COMMAND NOT FOUND

Who came up with THAT idea? I don't remember the 'vote' for that.
 
In my search of which RIP (for image/plate setters) run on Sun Sparc Workstations, I learned something IDK before:

Scitex Brisque was running on IBM AIX on RS/6000 boxen.

Back in th'90s I was hearing only praises about Scitex, but I never had hands-on experience with one – I was working only with Heidelberg and Screen separation systems (scanners and *setters) and bunch of other software (Harlequin/Xitron, Fiery, Kodak, etc.), but they all worked on usual suspects - PC/Mac WS.

So, if someone is looking for RS/6000 and AIX, best place to look will be on used print/prepress equipment sites.
 
CShell said:
I guess now days it's:

*BSD shell> ifconfig (whatever args) -- something works !
Linux shell> ifconfig (whatever args) --- COMMAND NOT FOUND

Who came up with THAT idea? I don't remember the 'vote' for that.
Click to expand...
It happened to me yesterday (again 🤦‍♂️). I started Debian 12 VM and wanted to scp from the host (pls. see #32), so needed to check guest IP (I can't remember $hit), typed (muscle memory) ifconfig and dang!
Only then I remembered that I'm in a Deb world now and that proper way to ask for own address is ip a.
Triple 🤦‍♂️🤦‍♂️🤦‍♂️
 
It was not for the desktop.
It was the operating system by itself which was rock stable. It was a time of Windows Blue Screens of Death.
Also no memory leaks.
Our customers providing applications here in Belgium where making Miljons of Euro's. I was providing a basic block for a stable solution.
Which domain was it. Text to speach , speach recognition, plugin-boards with FPGA's and real-time-operating systems and SS7 protocol.
Used by telecom companies to setup simultaniously miljons of phone calls together.
I looked at current prices of SCO-unix, its between 150 Euro to 2500 Euros.
 
Alain De Vos said:
It was not for the desktop.
It was the operating system by itself which was rock stable. It was a time of Windows Blue Screens of Death.
Also no memory leaks.
Our customers providing applications here in Belgium where making Miljons of Euro's. I was providing a basic block for a stable solution.
Which domain was it. Text to speach , speach recognition, plugin-boards with FPGA's and real-time-operating systems and SS7 protocol.
Used by telecom companies to setup simultaniously miljons of phone calls together.
I looked at current prices of SCO-unix, its between 150 Euro to 2500 Euros.
Click to expand...
I see much, much higher prices on buysehi
 
vmisev said:
Thanks, so SGI vas probably running DaVinci, that fits your description of the role perfectly, but Sun was obviously running some RIP, and it’s driving me crazy now that I can’t think off the top of my head of any older RIP that run on Sparc, grrr 🤬 (mad at myself, not you!)
Click to expand...
I don't recall DaVinci, but it might have been. I personally didn't work much on the SGI machine. I have worked with all the other RIP systems you mentioned: Harlequin/Xitron, Fiery, Kodak. The move to the Win NT machines coincided with installing Prinergy (formerly Creo, now Kodak). I also worked with Fuji's RIP system, XMF, at a couple of different jobs (also run on Win servers). At this point, I think I have forgotten more about prepress stuff than some new entrants into the field may know, haha.
 
JWJones said:
I don't recall DaVinci, but it might have been. I personally didn't work much on the SGI machine. I have worked with all the other RIP systems you mentioned: Harlequin/Xitron, Fiery, Kodak. The move to the Win NT machines coincided with installing Prinergy (formerly Creo, now Kodak). I also worked with Fuji's RIP system, XMF, at a couple of different jobs (also run on Win servers). At this point, I think I have forgotten more about prepress stuff than some new entrants into the field may know, haha.
Click to expand...
Do you maybe remember brand of plateseseters that Sun was running? That might help me with the search.

As for Kodak, I used only Preps as standalone app but I liked it far more than Signa Station. Although, as quality of output on film/plates goes, no RIP matches Heidelberg Delta Towers, their screening algos are superb (as much as I don't like one-vendor closed workflows, I have to give them credit for that)
 
You must log in or register to reply here.
Back
Top