• Anonymous (unregistered)

    Ignorance and stupidity this way

    | | | | V

  • n/a (unregistered)

    HAI GUYS. IM UR NEW MANEFRAME CODERz.

  • KC (unregistered)

    haha, I can writez codez??

  • Anonymous (unregistered)

    /END IGNORANCE AND STUPIDITY

    Meaningful discussion and commentary this way:

    | | | | | V

  • JackD (unregistered)

    Not likely. Masterfully written. A beautiful tale of intrigue. Bravo.

    captcha: jugis... jugis inDEED

  • bob (unregistered)

    And twtf is what? The kid screwed up, big deal. I want to know what The Sorcerer did to fix the problem.

  • Wizard's apprentice (unregistered)

    It is widely know that when summoning demons to bind them to your command, even a little mistake in the summoning pentacle can cost you your life and soul.

    So beware, dabblers of dark arts....

  • Gordon (unregistered)

    Not really a WTF (more like an honest, easy-to-make mistake), but an excellent story. I can appreciate that sinking feeling you get when you realize things have just gone from bad to worse.

  • return of the spelling nazi (unregistered)
    At that point, the basement was full of mops and buckets

    Comparing the core dumps to the mops in Fantasia was a great idea. I loved this story.

    Also, it had nothing to do with cleaning ladies unplugging servers. Hooray!

  • diaphanein (unregistered) in reply to bob
    bob:
    I want to know what The Sorcerer did to fix the problem.
    Fired the kid?
  • JamesQMurphy (unregistered)

    If the Sorceror really was a professional, he would have shown poor Rod what he did to fix it. Here was a guy who was willing to do the work, he just made an honest mistake. I wish I had more people like that around me; here, most folks just bitch and whine for someone else to fix it. Besides, what happens when (if?) the Sorceror goes on vacation? What if he leaves? Then the company would be really screwed.

    It's appropriate that he's called a Sorceror. Rather than relying on reason, they rely on luck.

    Captha: damnum. My sentiments exactly!

  • (cs) in reply to JamesQMurphy
    JamesQMurphy:
    What happens when (if?) the Sorceror goes on vacation? What if he leaves?

    Or dies? Unlike the COBOL they manage, the sorcerers will eventually die!

  • Franz Kafka (unregistered) in reply to diaphanein
    diaphanein:
    bob:
    I want to know what The Sorcerer did to fix the problem.
    Fired the kid?

    Teach the kid (in the morning).

  • Steve (unregistered)

    I'd've just gone to the operator (remember those?) and asked to have the job killed from the console. Easy-peasy.

    By the way, I find it difficult to believe that an S106 would be all that difficult to find -- it just means that a program in the library is unreadable, probably because of a compilation or linkage editor error. There are several subcodes but all of them are similarly obvious. S0Cxs were generally more problematic. Trashing General Register 13 or the return register vector (I forget the IBMese for that) was the usual culprit.

  • noname (unregistered) in reply to Steve

    my condolences to you ;)

  • Zap Brannigan (unregistered) in reply to JamesQMurphy
    JamesQMurphy:
    If the Sorceror really was a professional, he would have shown poor Rod what he did to fix it. Here was a guy who was willing to do the work, he just made an honest mistake. I wish I had more people like that around me; here, most folks just bitch and whine for someone else to fix it. Besides, what happens when (if?) the Sorceror goes on vacation? What if he leaves? Then the company would be really screwed.

    It's appropriate that he's called a Sorceror. Rather than relying on reason, they rely on luck.

    Captha: damnum. My sentiments exactly!

    I'm guessing you never worked in a mainframe environment, that's not how things are done in big iron world. Every mainframe data center has one or two guys that have had the training so they know the cryptic commands to make it go. If they shared that information they wouldn't be able to lord it over the code monkeys.

  • (cs)

    Aaaaaahhhhhh. I knew what ABEND was before reading this ... and TSO. I must have accidentally become a student of the dark arts of the mainframe.

  • anonymous sissy (unregistered) in reply to Steve
    Steve:
    I'd've just gone to the operator (remember those?) and asked to have the job killed from the console. Easy-peasy.

    By the way, I find it difficult to believe that an S106 would be all that difficult to find -- it just means that a program in the library is unreadable, probably because of a compilation or linkage editor error. There are several subcodes but all of them are similarly obvious. S0Cxs were generally more problematic. Trashing General Register 13 or the return register vector (I forget the IBMese for that) was the usual culprit.

    NERD ALERT!!!!!!!!!

  • (cs) in reply to mxsscott
    mxsscott:
    Aaaaaahhhhhh. I knew what ABEND was before reading this ... and TSO. I must have accidentally become a student of the dark arts of the mainframe.
    When this came out, I was technically a student of the dark arts of the mainframe, having been given TSO access. Most stuff I did was on the ISPF screens, however. But hey, I learned some basic RACF commands as well!

    ALU WTFUSR RESUME ALU WTFUSR PASS(omgwtf) NOEXPIRE

    Oh no. 8 months since I last touched one of those big irons and I still remember the commands.

    Addendum (2008-09-03 15:33): By the way, for those who haven't enjoyed the 3270 terminal experience, in this one it does matter where you type your commands. So raging alerts will make it impossible to execute a command.

    By the way, one popular command on ISPF to avoid session timeout was what we dubbed the "dipping duck" command:

    BOTTOM&30

    which sends the command "BOTTOM" every 30 seconds, effectively working as if I had a "dipping duck" pressing Ctrl every 30 seconds.

  • (cs) in reply to danixdefcon5
    danixdefcon5:
    By the way, one popular command on ISPF to avoid session timeout was what we dubbed the "dipping duck" command::

    BOTTOM&30

    which sends the command "BOTTOM" every 30 seconds, effectively working as if I had a "dipping duck" pressing Ctrl every 30 seconds.

    If the sea was whiskey And I was a dipping duck I would dive to the BOTTOM Drink my way back up

  • diaphanein (unregistered) in reply to Franz Kafka
    Franz Kafka:
    diaphanein:
    bob:
    I want to know what The Sorcerer did to fix the problem.
    Fired the kid?

    Teach the kid (in the morning).

    He's never going to learn if you baby him like that.

  • CoyneT (unregistered) in reply to bob
    bob:
    And twtf is what? The kid screwed up, big deal. I want to know what The Sorcerer did to fix the problem.

    I suspect the correct answer is "IPL" ... Initial Program Load ... which is the fancy mainframe name for the equivalent of ctrl-alt-del.

    NOT something you want to have happen at random times due to noobs ... though I've even seasoned Sorcerers can do things like this accidentally (one even admitted it to me).

    ...and I've done it too. I was Sysadmin on an HP/3000 MPE (now HPE) system and typo'd a device type. Many hours of bit work later, I had seriously learned a lesson about ENSURING that I had a usable/bootable backup of the accounting structure before changing device configuration.

    See, we'd just switched from the HP/3000 Series III to the Series 40. The Customer Engineers had done a system boot tape all right, but there was this tiny gotcha that if you don't back up at least one file on that tape (and they didn't) then the user accounting structure is not on it.

    So I had tapes with the user accounting structure that would only boot on the Series III (which was out the door) -- and a Series 40 boot tape that had no accounting structure. I spent the next 15 hours or so bit-pushing to build something that would read the accounting structure off of the Series III boot tape.

    ... and that O/S is about as fool-proof as you can get.

  • Jaded (unregistered)

    such WTFery

    Because the comments continually go down hill. Because people who have never had the pleasure of being awakened to a system alert that requires input at 3AM when a cleaning crew unplugged equipment don't believe it could happen. Because the same people think others should hand all of their knowledge out so that they don't have to learn from experience. Because I don't even care what I think.

  • (cs)

    I remember submitting COBOL jobs to Chicago in class via JCL and having to wait 3 hours to find out you had a syntax error in line 1398. Good times. NOT.

  • Steve (unregistered) in reply to anonymous sissy
    anonymous sissy:
    Steve:
    . . . Trashing General Register 13 or the return register vector (I forget the IBMese for that) was the usual culprit.
    NERD ALERT!!!!!!!!!
    You betchum.

    All those years slaving over an IBM Model 028 and 029 (bonus nerd points for knowing what one is without looking it up) and 3277 terminal have to count for something.

    Not very much, mind you, but something. . .

  • (cs)

    Why is now that all I can think of is..

    Hey Mickey, you're so fine. You're so fine you blow my mind. Hey Mickey! <clap,clap> Hey Mickey!

  • Peter (unregistered) in reply to Steve

    I know the 029 well, and have used the 026 when the 029 wasn't available.

    The 082 and its newer cousin the 083 were designed to bisect and fanfold cards lengthwise.

    And always remember: "Face down, 9 edge first!"

  • Witch hunter (unregistered)

    I see MATCHLIM not MAXLIM, so spell was cast.

    CAPTCHA: nulla

  • (cs)

    , END

  • diaphanein (unregistered)

    I can't believe some of you claim to work in IT and don't know what an abend is. What is this world coming to? HTML experience does not a programmer make.

    And overnight phone calls for failed jobs are not uncommon, nor are they a rite of passage. They're part of the job. I was just pissed this week when I received a 5:30 am wake up call on a holiday for DEV one night, and the next, I get no call for a 12:30 am job fail due to a database server deciding to reboot itself. Op center just replied to the automated alert with "job abended.". No shit. I can ead your email, but not the one you replied-all to. Nevermind the alert contains the runbool entry that explicitly said to call me.

  • Unbelievable (unregistered) in reply to anonymous sissy

    Uh - you're calling Nerd Alert in the daily WTF? You mean there's someone reading this who isn't a nerd?

  • PSWQ (unregistered)

    Lot of fairly standard non-mainframe programmer type responses to this one, predictably. First, an Abend S106 is trivial to track and fix. Certainly doesn't need a SLIP trap. Second, there's no need to go into the machine room to "the terminal" to enter a SLIP trap. Third, as soon as the SYS1.DUMPxx datasets (files) fill, it stops taking dumps - nowhere to put them. Fourth, anyone who has ever issued any Unix command has no place criticizing z/OS command formats. Fifth, even if you are dumb enough to use the operator console to issue a SLIP, 3270 screens and emulators are not like Unix TTY crap. Messages issued by the system do not clear what you have typed in to the command area of the screen.

    And we haven't even got into the supposed syntax of the SLIP command here...

    Off topic, I heard a neat description of C the other month: "C and its derivatives are the world's first write-only programming languages, because nobody can read them".

  • (cs) in reply to PSWQ
    PSWQ:
    Lot of fairly standard non-mainframe programmer type responses to this one, predictably. First, an Abend S106 is trivial to track and fix. Certainly doesn't need a SLIP trap. Second, there's no need to go into the machine room to "the terminal" to enter a SLIP trap. Third, as soon as the SYS1.DUMPxx datasets (files) fill, it stops taking dumps - nowhere to put them. Fourth, anyone who has ever issued any Unix command has no place criticizing z/OS command formats. Fifth, even if you are dumb enough to use the operator console to issue a SLIP, 3270 screens and emulators are not like Unix TTY crap. Messages issued by the system do not clear what you have typed in to the command area of the screen.

    And we haven't even got into the supposed syntax of the SLIP command here...

    Off topic, I heard a neat description of C the other month: "C and its derivatives are the world's first write-only programming languages, because nobody can read them".

    Yeah, I smell the dirty fingers of a clueless editor here. Too many things that are just wrong. Particularly the command that's the focus of the article.

    SLIP is relatively modern command. The ancient incantations are done with 1 letter commands: C, D, E, F, H, K, M, P, R, S, T, U, V, Z. It's been a long, long, time since I worked at a console, but I think those are the cancel, display, reset, modify, hold, control, mount, stop, reply, start, set, unload, vary, and halt commands. I've probably forgotten 2.

    I could look them up, but what's the fun of that?

    And, uh..., I know what dozens of system abend codes mean. You would too, if you worked on MVS for 25<< a long damn time.

    btw, I also know a couple dozen eunuchs commands, and can even use vi without having to restart every time.

  • Rand (unregistered) in reply to PSWQ
    PSWQ:
    Lot of fairly standard non-mainframe programmer type responses to this one, predictably. -snip- Off topic, I heard a neat description of C the other month: "C and its derivatives are the world's first write-only programming languages, because nobody can read them".

    The mainframe programmer strikes back? COBOL predates C, so C certainly can't be the "world's first." I've seen some COBOL, and though I'm not trained to read it, it's a mess. It takes much more code to accomplish the same tasks as it does in C, and severely handicaps the readability of the code.

    There's nothing inherently wrong with C, and one could argue that "it's derivatives" comprise the bulk of modern languages today (C, C++, Java, C#). Of those, I would say that C itself does tend to be the least readable, but it's not the worse language in terms of readability. Perl gets that award.

  • Drahcir (unregistered) in reply to Rand
    Rand:
    PSWQ:
    Lot of fairly standard non-mainframe programmer type responses to this one, predictably. -snip- Off topic, I heard a neat description of C the other month: "C and its derivatives are the world's first write-only programming languages, because nobody can read them".

    The mainframe programmer strikes back? COBOL predates C, so C certainly can't be the "world's first." I've seen some COBOL, and though I'm not trained to read it, it's a mess. It takes much more code to accomplish the same tasks as it does in C, and severely handicaps the readability of the code.

    There's nothing inherently wrong with C, and one could argue that "it's derivatives" comprise the bulk of modern languages today (C, C++, Java, C#). Of those, I would say that C itself does tend to be the least readable, but it's not the worse language in terms of readability. Perl gets that award.

    No, Lisp gets that award. Brackets as far as the eye can see, and as plentiful as your sperm count. I'd hate to program in Lisp and lose count of how many brackets I have. That would be a nightmare.
  • Steve (unregistered) in reply to Drahcir
    Drahcir:
    No, Lisp gets that award. Brackets as far as the eye can see, and as plentiful as your sperm count. I'd hate to program in Lisp and lose count of how many brackets I have. That would be a nightmare.
    That's why you use the One True Editor, emacs. Lisp mode will provide all the tools you need.

    BTW, shame on me for forgetting the Model 026 keypunch. I have no idea what an 028 might be -- I never saw one and must have made it up.

  • EPE (unregistered)

    I am here, at the office, reading this story about ancient mainframes; and I read these comments, about how bad things were a far time ago...

    Then, I stand up, look around, and see one or two hundred of my coworkers (*), not-so-furiously banging their keyboards, in front of their screens full of green text over a black background; they talk about PL/I, JCL, IMS, ABENDs, ...

    (*) No kidding, there are about four hundred people in this single office, six hundred total in the building.

  • (cs)

    Marvellous. PIC X(20), 88-levels, and the endless full-stops. I remember the bad old days. I actually wrote a system to replace an IBM punch-card machine...in COBOL. Wonder if it's still running. Perhaps it's still compiling...

  • Anonymous Cow-Herd (unregistered) in reply to sas
    sas:
    And, uh..., I know what dozens of system abend codes mean. You would too, if you worked on MVS for 25<< a long damn time.

    25-to-life? That's harsh.

  • Bitter Like Quinine (unregistered)

    Long long ago, when starting my first real job, I was introduced to the most ancient computer I had ever seen (core memory, punched-tape, etc). In fact I'm fairly sure that Babbage copied bits of the design in his difference engines.

    It was big, hot and noisy, and needed frequent rebooting, hence I was told the proper sequence of actions to perform a clean restart.

    "Press this, flip this then this then this, then press and hold this switch until the lights start flashing and you're done." said my new co-worker.

    "Seems simple en..." I began.

    "And whatever you do, do not flip the switches in the wrong order!" She added.

    "Er, why? what happens?"

    "Dunno. I've never made that mistake!"

    "So how do we know not to do it?" I asked.

    She looked at me pityingly, "That's what I was told when I started here."

    Sure enough, the day came when 'Old Faithful' needed another restart, and I was put to the ultimate test. Could I handle the pressure of pressing five buttons and switches in the right order? Would I crack under the strain of a front-line outage? Was I up to the physical and mental demands of the job?

    Walking up to the 'beast' I looked at the switches I was supposed to press. The first was a halt/reset switch, the middle three switches toggled bits in one of the machine's registers (set by a dial to the right, which had obviously never been changed since the damn thing was first assembled), and a proceed button that initiated the boot. It was obvious within the first few seconds that it mattered not one whit which order the middle switches were tripped in, regardless of stern warnings to the contrary.

    By way of a practical demonstration to myself, I stopped the machine, toggled a few bits at random a dozen times to the tune of the 1812 overture, toggled bits until just the three flags required were lit and pressed the start button. It worked perfectly.

    In nearly ten years there, the only other person I saw not reboot the machine with exactly the sequence I was told was the hardware engineer who occasionally came to fix it.

    So, yes, sometimes the mighty sorcerers of old knew the real secrets of magic. And sometimes it's just a bunch of mumbo-jumbo hoodoo.

  • YAGrammarNazi (unregistered) in reply to Rand
    Rand:
    PSWQ:
    Lot of fairly standard non-mainframe programmer type responses to this one, predictably. -snip- Off topic, I heard a neat description of C the other month: "C and its derivatives are the world's first write-only programming languages, because nobody can read them".

    The mainframe programmer strikes back? COBOL predates C, so C certainly can't be the "world's first." I've seen some COBOL, and though I'm not trained to read it, it's a mess. It takes much more code to accomplish the same tasks as it does in C, and severely handicaps the readability of the code.

    There's nothing inherently wrong with C, and one could argue that "it's derivatives" comprise the bulk of modern languages today (C, C++, Java, C#). Of those, I would say that C itself does tend to be the least readable, but it's not the worse language in terms of readability. Perl gets that award.

    If you're going to quote someone's words, why not copy and paste the text? Then you wouldn't make a fool of yourself by changing his correct "its" to your illiterate "it's".

  • (cs) in reply to Steve
    Steve:
    I'd've just gone to the operator (remember those?) and asked to have the job killed from the console. Easy-peasy.
    TopCod3r? Is that you?
  • Dave G. (unregistered) in reply to diaphanein
    diaphanein:
    I can't believe some of you claim to work in IT and don't know what an abend is. What is this world coming to? HTML experience does not a programmer make.

    What are you talking about? So people either work on mainframes or they only have "HTML experience"? Anyone who doesn't know what an arbitrary error code on an arbitrarily chosen system means is relegated to HTML markup duties because they are just toooo stupid for anything else?

    Get over yourself pops. You are not special. You are not important. That's why you're the nerd who knows all about error codes, because other people are too busy worrying about shit that actually matters. You and your knowledge of random error codes aren't the slightest bit impressive, and nobody here gazes in awe at how old school and elite you think you are.

    What the hell do you talk about at parties? Do you share all the error codes you remember with the women there? Are they impressed?

    My condolences. You must have a dreadful sex life.

  • Everything (unregistered)

    Well, if the command trapped Everything, the output should have been easy enough to read through:

    42

    (Now understanding all the implications of what this might mean, that's an entirely different thing, of course...)

  • (cs) in reply to Dave G.

    What's a sex life?

  • Sorcerer (unregistered) in reply to bob
    bob:
    And twtf is what? The kid screwed up, big deal. I want to know what The Sorcerer did to fix the problem.
    Reboot.
  • Bo, the ancient mainframer (unregistered)

    S106s are easy. The reason code, which describes the type of error in detail, is in register 15.

    Say, f'instance S106, look up content of register 15, find the value 0C and you know that you ran out of memory space to load the module. Try increasing REGION and rerun the job. Common knowledge.

    Damn, I'm scaring myself here. I actually KNEW that.... Why is it that I can remember this but can't for the life of me remember what I was supposed to pick up from the grocers' on the way home? Getting old is hell.

    Bo

  • (cs) in reply to Dave G.
    Dave G.:
    […] nobody here gazes in awe at how old school and elite you think you are.
    I do!
  • (cs) in reply to Dave G.
    Dave G.:
    What are you talking about? So people either work on mainframes or they only have "HTML experience"? Anyone who doesn't know what an arbitrary error code on an arbitrarily chosen system means is relegated to HTML markup duties because they are just toooo stupid for anything else?

    Get over yourself pops. You are not special. You are not important. That's why you're the nerd who knows all about error codes, because other people are too busy worrying about shit that actually matters. You and your knowledge of random error codes aren't the slightest bit impressive, and nobody here gazes in awe at how old school and elite you think you are.

    What the hell do you talk about at parties? Do you share all the error codes you remember with the women there? Are they impressed?

    My condolences. You must have a dreadful sex life.

    Ummm...Dave? Dave? Your doctor just called. He really thinks you should start taking your medications again. I mean, he really insists on it. I think you should listen. We certainly don't want a repeat of last time, when the men in white jackets came for you, do we?

  • (cs) in reply to bob
    bob:
    And twtf is what? The kid screwed up, big deal. I want to know what The Sorcerer did to fix the problem.

    Hah! Yeah right! A sorcerer is going to tell someone what he did to fix the problem? Pssssh! They keep those cards up against their chest. Glued to their chest even. That way they can stay the martyr... the sole saver of souls as well.

Leave a comment on “Classic WTF: The Sorcerer's Apprentice”

Log In or post as a guest

Replying to comment #:

« Return to Article