• Julia Hayward (google)

    TRWTF is not trying things out in a sandbox before bringing a highly flaky production down. Or maybe TRWTF is expecting management not to throw you under the bus to protect their golfing privileges. Or maybe TRWTF is not running like hell from the interview, or at least as soon as you realise there's no budget for sandboxes...

  • Dude (unregistered)

    Am I the only one who expected it to end "the network (maybe, barely) is running, but there are no backups"?

  • (nodebb) in reply to Dude

    Backups? What the ... are backups?

  • Quite (unregistered)

    TRWTF is a) not wearing a wire to that crucial first meeting where the agreement to go ahead was forged, and b) not lawyering up. At that point, a mild "that's not what happened ..." (plays tape) "... and you didn't ring once ..." (shows telephone records) "... so Bob's being somewhat unfair here."

    It may actually take a court subpoena to get those telephone records, but at least it will result in some money coming his way from the weekend he worked.

  • (nodebb)

    Btw, https://en.wikipedia.org/wiki/File:Project-triangle.svg (AKA Iron Triangle)

    It can be clearly seen that CHEAP is the base of the entire project. Any PHB will tell you if this isn't obvious nothing ever will be.

  • LCrawford (unregistered)

    His plan to wipe, reload and reprogram an entire small company over a 3 day weekend was entirely un-realistic. No testing in a development environment first? I'm surprised that they even had a system where they could wipe and reload without it being a totally manual operation. I'll bet employees still lost some personal information or settings they had chosen

  • poniponiponi (unregistered)

    So the new guy bullies his boss into letting him completely redo the company's broken-as-fuck-but-apparently-still-somewhat-functional system from scratch in three days with no backups or fallback plan, fucks it up royally, and got shitcanned.

    I fail to see the problem here.

  • Starlokk (unregistered)

    Gregg is TRWTF. As soon as I read his proposal I couldn't believe how stupid it was. It might the riskiest possible plan, I can't think of much worse unless he was also re-coding the system from scratch.

    Manager-guy is a douche for not accepting any responsibility and lying, sure, but Gregg set everyone up to fail.

  • Bitter Like Quinine (unregistered)

    The real WTF is not knowing WTF "sign-off" means, as in...

    "In order to execute this plan, he would need signoff from Bob, EduLoans' Vice President"

    If you don't have written clearance, you don't have clearance.

    Managers lying to my face, or in my presence to senior managers, is a common enough occurrence to be a "feature" of working in IT. The work-around is to "have it in writing".

  • BoogieDaddy (unregistered)

    Senior management refusing to take responsibility.. there's a new one. Hah. I've had a similar situation, on a re-deployment project. I told management I needed four days to do it correctly, mostly over a long weekend, moving systems incrementally to reduce overall effects to a few hours at most for each impacted system. Not like I hadn't done the same, successfully, oh several dozen times previously. Oh, no, the boss decided some flashy advertisement for deployment software we had never used was the legit stuff; it would give us the edge. We could do it in hours, and have it all done; oh and while we're at it do it during the week so we can "pull the switch" in real time! Whoo!

    Needless to say it failed - miserably - and several days were spent fixing the mess. Natch I was responsible, because reasons. Bosses love having a scapegoat to take the blame, especially when it's their decisions that lead to the problem, and are happy to keep the praise if there's praise to be had.

    Getting it in writing isn't always practical, but email and voicemail still works for internal purposes. Never stand up in court, I guess, but at least the boss' boss has some evidence to back you up.

  • (nodebb)

    Netware 4.11 is quite capable of emulating Netware 3.12, so there was no need to change any of the workstations until the new server was up and running.

  • getwiththetimes (unregistered)

    Nothing like saving the riskiest part until last!

  • LH (unregistered)

    I got scolded (and the aftermath eventually led me out of the company) because I not only refused to sign off (I was the "expert") a technical decision but I went against my immediate manager and raised my concerns in public, in writing, in an email chain. Schadenfreude-wise it was ok because shit happened just as I anticipated, but having to leave a job because you want to do your job is not very nice.

    But in this case I fail to see how the tech guy is not to blame. The road to hell is paved with good intentions.

  • Vietcongster (unregistered) in reply to BoogieDaddy

    Doesn't email count as a legal proof? I've even heard of Facebook and WhatsApp messages being accounted for in some cases.

  • Meh (unregistered)

    For me, TRWTF was Greg thinking "I'll be a hero," and then not immediately thinking "no, that never ends well, this is a horrible idea."

  • (nodebb)

    Usually the daily WTF is some obscure coding horror which I can't see unless someone explains it. But this time it sounds familiar. In a previous life I was a Netware (2.15, 3.12, 3.2, 4.x, 6.x) admin for years, in a no-budget environment. I've done upgrades and maintenance on production servers because, eh, no money for a test lab. But in that case you make small, careful changes. So as I read Gregg's plan my face slowly turned white. Holy cow what a recipe for disaster. I read this as a "look how stupid I was when I started working" piece. He deserved the kick, but he learned a lot about IT in one hour!

  • (nodebb)

    Usually the daily WTF is some obscure coding horror which I can't see unless someone explains it. But this time it sounds familiar. In a previous life I was a Netware (2.15, 3.12, 3.2, 4.x, 6.x) admin for years, in a no-budget environment. I've done upgrades and maintenance on production servers because, eh, no money for a test lab. But in that case you make small, careful changes. So as I read Gregg's plan my face slowly turned white. Holy cow what a recipe for disaster. I read this as a "look how stupid I was when I started working" piece. He deserved the kick, but he learned a lot about IT in one hour!

    Addendum 2016-06-21 10:22: Edit: sorry for double post, it seems I can't delete it.

  • operagost (unregistered) in reply to urkerab

    Indeed, urkerab. This could have been done incrementally. First, a new server should have been procured to run Netware 4.11. The way this reads, it's unlikely the existing server was anything special and could have been replaced for peanuts. Second, a scratch PC could have been set up to talk to the new server. Once it was working there, THEN the entire lot of workstations could have been migrated. And as you said, the old 3.12 client could have talked to Netware via bindery emulation, so they actually could have just had their scripts reconfigured.

    It bothers me that I even remember this much about Netware.

  • Bill T (unregistered)

    "Greg unhatched a bold plan". Unhatched, the opposite of hatched, i.e. it went back into the egg where it should have stayed...

    Wiping all the workstations and restructuring the way the organically-grown system worked over a weekend? There's no way that could possibly have problems...

  • DrPepper (unregistered)

    Wiping all the workstations... I would have stopped right there. You NEVER do that. Ever. Ever. Buy new machines, put them side-by-side for a while, until the user feels like they don't need the old machine anymore. Then backup the old machine onto the new machine (or external storage) and put it in the warehouse for a while. Then maybe a year on, you can finally get rid of it.

    I still have backups of my previous 2 computers stored away on an external drive, in case I need files off them (haven't yet, but you never know!)

  • b.a. freeman (unregistered) in reply to Bitter Like Quinine

    ... and if the boss won't sign the document granting clearance, then U send him an email spelling out in detail the agreed-upon strategy, and copy his boss on it. for legal reasons, it might also be a Good Idea to BCC oneself.

  • JG (unregistered) in reply to DrPepper

    Buy new machines, put them side-by-side for a while

    Er, Budget?

  • Developer Dude (unregistered)

    Good deeds never go unpunished

  • Ron Fox (google)

    The road to hell is paved with good but untested intentions.

  • TenshiNo (unregistered)

    I agree that Greg's plan was ridiculous. You never try and replace an existing network like that. You setup a new one, get it running solid, then migrate the clients over. Preferably, in batches, so that any issues you run into don't bring down the whole company at once. I worked for a school district many years ago where we did exactly this, upgrade from Novell 3 to 4, and switching the network to pure IP in the process. Hardest part of that was when we finally had to update the Novell Client on 4,000+ PC's (running Windows 95) by hand.

    As for managers throwing people under the bus: at a previous job, the VP of the company walked up to me one day and said "We want to implement Salesforce, so shop a contractor that can help us get it setup, interview a couple, and let me know what you're recommendation is." At the time, Salesforce was (relatively) new, and a contract bid to get it setup was not cheap I found 3 local companies, did phone interviews, got prices, etc. Suggested to the VP that we go with the mid-price bid, and he told me to call and set it up. Come to find out, the day that the contractor came out and started doing training, that the VP had brought it up with the companies management board and had been shot down. I got ripped a new one for spending thousands of company dollars "without authorization". The whole time, I'm just sitting there like "the #2 person in this whole company told me to do it. How is that not authorization?" I didn't get fired, but I did get told, in no uncertain terms, that I was not getting my standard bonus that year because of the incident.

  • dhasenan (unregistered)

    The signoff process involves documenting in detail what you plan to do, finding who needs to read that document to see if it's a reasonable plan, getting them to review it, and documenting their approval. This Gregg didn't do any of that.

    Loretta was a hero for being able to manage (even poorly) a corporate network with only a two week training course.

  • (nodebb) in reply to operagost

    NetWare was pretty cool technology, especially for its time. Even today, in some aspects NDS (or eDirectory as it was called later) would blow Active Directory out of the water.

    It's a shame that Novell never managed to properly counter the Microsoft marketing machine by the time Windows NT4/2000 Server came around, which included AD for free.

  • LK (unregistered)

    I love the way people are talking about this like it happened this year.
    Seriously, they're using Novell and batch files, it probably predates Windows for crying out loud.

    Most people couldn't afford or wouldn't justify "wasting" one computer as a server, let alone having a second one to test changes with.

  • (nodebb) in reply to Ron Fox

    To be fair, even testing intentions could fail.

    Afterall, we've heard a lot about how the testing environment has subtle difference to the production one.

    Addendum 2016-06-21 22:14: testing -> tested in the first line.

  • hsp (unregistered)

    I find it completely fine that he got fired. This is not a WTF for coders, this is a WTF for managers.

    I would totally fire anybody who event suggested this freaking plan. Small steps at a time.

  • (nodebb) in reply to LK

    "Most people couldn't afford or wouldn't justify "wasting" one computer as a server" This. I had the hardest time convincing management that our most powerful computer (a 386) should sit under a desk doing nothing. The year was 1991, not 1981.

  • Nate Scherer (google) in reply to The_Dark_Lord

    Sit under a desk doing nothing, or sit under a desk doing server things?

  • Jistuce (unregistered) in reply to Nate Scherer

    Obviously it is doing nothing. I mean, if it was doing something, the screen would be on. This "waiter" system doesn't even HAVE a screen, how COULD it do things?

  • Lerch98 (unregistered)

    No budget for new equipment. No budget to get paid. A place like doesn't pay ####. So WTF, get fired, get better job...lesson learned.

  • AstorLights (google)

    "Combined with the Netware disarray was the crappy loan processing software EduLoans ran its entire business through."

    I wish I could write sentences like this one.

  • (nodebb)

    I wouldn't have touched that situation with a ten foot pole and a golden parachute.

  • Guntank17 (unregistered)

    Indeed. Anyone who comes up with a solution this extreme as the first and only option from the very start, especially over 3 days, should, like the traders at Enron, have their motives for even suggesting something like that put under very, very close scrutiny.

  • Endaar (unregistered) in reply to DrPepper

    That's absurd. Done right the end-users workstations should be effectively interchangeable. Sure, e-mail might need to be reconfigured, etc. and they'll lose their wallpaper if you don't take care (or simply don't care) to migrate their profile, but really, reimaging a workstation should be a non-issue.

  • Jasper (unregistered) in reply to Endaar

    Remember, Novell 3 to 4 migration. This is in the early nineties. There was mostly no such thing as a roaming profile or an interchangeable client PC — the software had barely started supporting it and was doing it badly.

Leave a comment on “Putting the "No" in "Novell"”

Log In or post as a guest

Replying to comment #:

« Return to Article