TGIF

[altivo]

Today's achievements, some may be dubious:

• Cataloged a pile of books, which may sound ordinary but these days it's anything but.
• Checked and found that the VPN tunnel is still live.
• Found that my suggestion that we use IRC for interlibrary quick communication has been taken seriously. A test IRC server is now active, though it needs configuration.
• Managed to dodge a line of severe thunderstorms by closing the library a few minutes early and running like heck.
• Installed the WATFIV language compiler on my IBM mainframe simulation early this morning.
• Resisted the temptation to send scathing and contemptuous letters to various high profile individuals at Six Apart (this last one was difficult.)

The IRC thing perhaps deserves some background details. The evil managers of the Big Messed Up Consortium had decreed that all of us had to install and run DBabble to permit quick announcements and communication between institutions. DBabble only runs on Windows. Obviously, I was unimpressed, and got the usual runaround when I complained: "No one else has complained." So when I was asked for suggestions, I pointed out that IRC can be obtained and installed for free, is supported by clients that run in just about any environment, and can be made more than secure enough for our small needs. In this very different group, my words were taken seriously it seems.

Comments

[doco.livejournal.com]

As long as it doesn't end up like

<@tivo> lol sum n00b wantz to rent harry p0thead 7
<larry_> tell him dumbledore dies on page 677
<+brAyn> wtf dats hp 5 u n00b
<@tivo> stfu brain kthx, effin' spoiler
+++ tivo has kicked brAyn from #library (gtfo now)

you will be fine, I guess ;)

[altivo.livejournal.com]

I don't think we have to worry. We're talking about library staff here, not the users. For an example of why I think that makes a difference just see this comic strip.

[dragon-punk.livejournal.com]

You work Library stuff? That's pretty interesting. My cousin's on the staff of OPAC/Evergreen. Supposedly other library systems are picking up on his software.

http://www.gapines.org/

LOLcomic

[altivo.livejournal.com]

While we were considering whether or not we could successfully get out of the bad contract with BMUC, I did take a look at GA Pines/Evergreen. It has a lot of promise, but isn't featureful enough to replace what we are using now. I'll be watching to see what happens with it, certainly.

Yes, I'm a public librarian, and before that I was a college librarian. Before that, I worked for a library software house (that no longer really exists.)

Hu-freaking-zzah!

[vakkotaur.livejournal.com]

Someplace might actually use the right tool for the job and not try, yet again, to pick the ideal wrench to hammer a screw in.

Re: Hu-freaking-zzah!

[altivo.livejournal.com]

Heh. Actually, for the purpose they want here, they didn't really need to waste server space and cycles. I told them that. They could just as well have set up a group account with any of the existing free IM services, or, for that matter, used e-mail, the telephone, or the fax machine. The BMUC has oveer 100 member libraries. Our group has only eight, which does tend to make things much simpler. IRC is overkill, but at least it's easy to use, or can be made so.

[jmaynard.livejournal.com]

What OS are you running these days, and wheredja get WATFIV?

[altivo.livejournal.com]

At the moment, mostly MVS 3.8j, based on Volker Bandke's turnkey distribution #3 but I've been starting to make some major modifications in it to suit myself. MVS is most familiar to me, though I'll probably take advantage of the opportunity to play sandbox with VM eventually (I never worked with VM at all.) I have to say, I'm very impressed with the reliability and performance of the Hercules code. It makes me snicker every time I "mount" a tape and MVS accepts it with perfect equanimity.

The WATFIV came from following a link on your site, in fact. Once the JCL was properly adjusted it installed without a hitch. The test job got lots of errors, but they all appear to be related to some version discrepancy between the compiler and the testbed code. The compiler/preprocessor itself is running as expected. I went after WATFIV because I couldn't find a working RATFOR implementation. While I prefer BAL for serious, long lasting, high performance code, you just can't beat Fortran for engineering, science, and statistics in my opinion.

[jmaynard.livejournal.com]

We try to keep Hercules at a level that would enable it to be used as a production system, although no DP manager in his right mind would do that: there's no paid support available (dammit). That's a pretty lofty goal, and one I'm proud to say we meet pretty well.

I went looking...and found WATFIV on Jay Moseley's page. I didn't know it was there. Whee.

[altivo.livejournal.com]

Maybe the link was on Volker's "New User Cookbook" site then. I've spent a lot of time reading both that and the Hercules pages.

My mate here is asking for IITRAN. (Teaching language, sort of a primitive PL/I if I understand correctly, that was developed and used at Illinois Institute of Technology in the 60s and 70s.) We haven't found a trace of it, but he'd be delighted should someone unearth it. He actually used PL/I for years, and is happy to have access even to the MVT version of it, but IITRAN has nostalgia value I guess.

[jmaynard.livejournal.com]

Toss out a query on the hercules-390 list and see if you get anything more than a "huh?"...

[altivo.livejournal.com]

I'll probably do that eventually. Just friended you, BTW, so your comments won't be screened any more. That was pointless.

[jmaynard.livejournal.com]

Oh, and if you do find a working RATFOR, you'd make at least one guy on the turnkey-mvs list happy.

Oh, wait, that's you. I didn't make the connection. What kind of ham radio stuff do you have in it?

I asked in another forum, and apparently the best way to get a working RATFOR implementation is to work through the exercises in Software Tools...

[altivo.livejournal.com]

Heh. Yes, that was me.

The code I wrote in RATFOR (and ran through the Misosys RATFOR processor, first to Microsoft F80 on the TRS80 and later the MSDOS version of the same processor, dumping into Lahey Fortran 77 on MSDOS) deals largely with propagation predictions on HF frequencies and antenna modeling. I could do it again using some other platform, since it's just a translation of the theories and equations laid out in innumerable handbooks, but my version of it was easy for me to use and ran fast. Why reinvent it if I can recover the original? ;D

I'm looking at SIMH for the VAX too, thanks to your suggestion. Unfortunately, winding your way through that maze of Encompass/DECUS and the hobbyist group in order to get license keys turns out to move at a snail's pace. I have the emulator up and running, but it may yet be weeks before I can actually install VMS on it.

VMS would be an easier environment for engineering programs, I think. But I've always liked IBM JCL. How's that for a perversion?

[jmaynard.livejournal.com]

I've got no problem with MVS JCL at all. Many folks don't realize that it, not Unix, was the first system to provide device-independent I/O, and the complexity of JCL is due in large part to achieving that goal.

I need to dig through my old libraries; somewhere in there, I've got a generalized repeater site intermod analyzer in PL/I (the Optimizer, though it could probably be made to run under PLIF fairly easily). The neat thing about that one is that I needed a way to loop through a variable number of array elements at a time, and wound up writing a function to do that, called ODOMETER.

If you do get RATFOR running, I'm sure the community would love to use it...

[altivo.livejournal.com]

If I get RATFOR going from somewhere, I'd certainly share it with anyone who wanted it. We did have the working source for it on a nine track tape sitting in a closet at our old house, along with several other tapes containing my own code salvaged from Time-Life, Federal Reserve Bank, and NOTIS Systems. Of course when we moved to the farm in 1998, I said "There's no room for these old tapes and we could never read them again anyway," and chucked them into the trash. A backup of my entire Panvalet library! Sigh.

[jmaynard.livejournal.com]

Ack!...but it's a common story.

You won't feel any better hearing that I have both 9-track and 3480/IDRC tape drives, I'll bet.

[altivo.livejournal.com]

You're right. I don't, though we did know of service bureaus at the time that would have read the tapes and put them onto other media for a fee. I'll keep that in mind if I manage to unearth any treasures in tape format.

[altivo.livejournal.com]

Oh, the host for Hercules, if that's what you meant, is Linux. Slackware 10.2 to be precise.

[jmaynard.livejournal.com]

No, I meant the guest OS.

I keep meaning to sit down and build up an MVS system my way, dammit. (Was there ever an MVS systems type who didn't have firm, and firmly stated, opinions about how he thought his system should go together?)

[altivo.livejournal.com]

Well, I confess that a lot of my preconceived notions are actually based on the way the systems at Time-Life in Chicago were laid out back in the 80s. Though I hasten to add that there were elements of that job that I really don't miss, like the two years I spent shepherding an IBM mass storage system, complete with aging 3330 drives and about 900 of those irritating bullet cartridges. Now there's an emulation challenge for someone. Except you'd have to put in lots of code to make it break down and misbehave about once a week, like the real thing did.

[jmaynard.livejournal.com]

Haw. Considered to the 2821 noodle snatcher, the 3850 was a model of reliability.

There's been a singular lack of demand for a 3850 emulation. (Much less the 2821.) I never saw one in action, and from all of the stories I've heard, am just as glad.

[altivo.livejournal.com]

I never saw a 2821 "in the flesh" as it were, though my fellow specialists in data administration used to reminisce about them a lot. Time never had one, but they had both worked for CNA, who did.

Actually, you missed something by not seeing a working 3850. For its time, it was a pretty amazing piece of robotics. Let me tell you, though, you did NOT stick your hand (or any other appendage) into the box when it was powered up. Those accessors moved like steel cheetahs, and probably could have amputated a hand in a flash. Time had the second largest production 3850 in Chicago, I think. There was a smaller one over in the IBM building that they used for training both customers and CEs, of course. Standard Oil /Amoco had the largest one, though. It was one of those tandem units with two full cabinets and was stuffed full of geological survey data.

My mind boggles at the realization that today people carry around that much storage on key fob flash drives and think nothing of it. Or that the entire Time-Life production system as it stood in 1983 can reside in such a tiny percentage of my $500 desktop PC's memory and disk space. It's like Heinlein's foldboxes or Dr Who's Tardis...

[jmaynard.livejournal.com]

Yeah, I occasionally have to stop and boggle at just how far computing has come in terms of processor power and storage space, and how thoroughly we've wasted all that capability. My main Hercules box at home (a dual-processor Opteron 275) can run faster than most 3090s, and it's not a particular speed demon. When you compare it to the first machine I ever worked on, a 370/158AP with 6 MB of memory (huge!)...

The real shame is that that 158AP served a hundred or so users (mostly CICS, a few TSO) with quite acceptable response times, and the average desktop can't even run Vista well. It's not that the computers are slow; it's that nobody pays attention to writing small, fast code any more.

[altivo.livejournal.com]

nobody pays attention to writing small, fast code any more.

Amen to that. Of course we're starting to sound like a couple of real geezers now. But even Linux has deteriorated to a bloated monstrosity, especially when you look at the so-called user-friendly distributions like Kubuntu. I have a strong dislike for Windows, and that's one reason. It takes two or three times longer to boot up a single user Windows XP than it took your S370/158 to IPL cold. When I started at Time-Life in 1980, a similar 370 (may have been a 16x version) was supporting their training department and systems testing. The "production" system was a new, shiny 3033 just like the one used in Volker's setup. It had about 700 end users on CICS and Intercomm application screens and performed much better than Windows does, just as you say. At the same time it was providing massive batch services for the SAMI division, generating tons of printouts of market analysis and statistics for the supermarket industry. They still had their prior S360 system "up on blocks" but it went away when they brought in dual 3084s just a couple of years later.

[dragon-punk.livejournal.com]

Huzzah IRC~!

(~'.')~

[cabcat.livejournal.com]

Kitty suddenly feels very IT inept hearing all this talk XD

[altivo.livejournal.com]

You'll get there if you stay in the field. It's just accumulated experience.

Note also that most of the web sites related to the Hercules mainframe emulator feature artwork or logos of dinosaurs. I particularly like the front page for Volker Bandke's Hercules site.