• cb (unregistered)

    I'll post the first comment in six weeks

  • nexekho (cs)

    So what did he have to say on the phone? DON'T LEAVE ME HANGING LIKE THIS FFFFUUUUUU

  • @Deprecated (cs)

    I think in the suit meeting, Gary should have bluffed.

    "My estimate was 5 months! Whatever gave you the idea it would be done in six weeks?"

  • Wilhelm Svenselius (unregistered)
    Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.
    TRWTF is that he did not quit right there and then.
  • JamesQMurphy (cs) in reply to @Deprecated
    @Deprecated:
    I think in the suit meeting, Gary should have bluffed.

    "My estimate was 5 months! Whatever gave you the idea it would be done in six weeks?"

    ... and produced his copy of the estimate as proof.

  • OzPeter (cs) in reply to Wilhelm Svenselius
    Wilhelm Svenselius:
    Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.
    TRWTF is that he did not quit right there and then.
    Or when put in front of management.
  • Anon (unregistered) in reply to Wilhelm Svenselius
    Wilhelm Svenselius:
    Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.
    TRWTF is that he did not quit right there and then.

    If you quit your job every time something marginally shitty happens, I doubt you'd be in any job for more than about....six weeks.

  • temptar (unregistered) in reply to Wilhelm Svenselius
    Wilhelm Svenselius:
    Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.
    TRWTF is that he did not quit right there and then.

    Agreed. I would have been gone. 8 months to 6 weeks is crazed pruning of time.

  • Not Steve (unregistered)

    Two lessons to take home from this:

    1. Cover your ass: where did the original estimates go? Are there no emails, paper trail, etc?

    If there's a daily meeting discussing this project, I would personally amend everything I said with "ok, we're two weeks in, only 7 months and 2 weeks to go". Passive agressive, yes, the point here is that the developers and anyone involved needs to be fully aware of the actual time estimate, and don't pay lip service to the 6 week estimate.

    1. Say no to micromanagers: unless a manager is writing code, why are you listening to his suggestions to write everything from scratch. I've been in this situation before, and I've flatly rejected requirements for projects containing implementation details (i.e. database schemas, suggestions to implement something from scratch, suggestions to store XML in a database) -- because 99% of the time, implementation details written by business analysts are marketing no-nothings are WTFs in themselves.

    Worst case scenario, say "sure, I'll write my own ORM from scratch, wink wink"

  • Hasteur (cs)

    Worked in a shop where there was constant threats by our customers to leave the system because they weren't getting the features they wanted in a reasonable time, sales people who would promise anything as long as they got the client's signature on a piece of paper, customer advocates that would put up token resistance to ideas that the clients have so that the developers can be stuck holding the "No" card.

    I parted ways with the shop over a year ago, and I've found out from friends who were in the department afterwords, that the entire structure we had in place fell apart and the department was gutted for "Failing to meet commitment dates"

    If you don't take the advice of people who are doing technical work underneath you for estimates, you're always going to end up going back to the customer and digging yourself out of a hole you never wanted in the first place.

  • Bogolese (unregistered) in reply to Anon

    From 5 months to 6 weeks is "marginally shitty?"

  • frits (cs)

    Unfortunately, this is a common conflict between marketing/sales guys and engineers. At an old job, the VP of Marketing asked my boss how long it would take to develop certain type of electronic test equipment (hardware and software/firmware). After analysis, my boss replied that it would take about two years. The VP replied that it had to be done in one year due to market conditions or not at all.

    If we were releasing something that was an emerging technology, the VP may have had an argument. However, in this case we were just releasing a competing product in an established market that had added capabilities over our competitors as well as a better user experience. After more discussion, my boss realized the VP of Marketing was not going to budge. The VP offered to hire a new developer (me) in order to meet the deadline.

    My boss knew that his original estimate was still more accurate that the aggressive one year deadline, but agreed to the deadline anyway. He also knew that after a year, if we were behind schedule, management would still want a return on their investment, and would not scrap the project.

    Long story short, it took a year and ten months to release the initial version of the product. Additionally, our team members got to stay employed and the company got a nice product to sell.

  • powerlord (cs)

    "After three months of work, well passed the 6-week deadline"

    I feel like nitpicking: It's well past the 6-week deadline.

  • My Name Is Missing (unregistered)

    What a slacker for a boss. The industry standard is two weeks.

  • Anon (unregistered) in reply to Bogolese
    Bogolese:
    From 5 months to 6 weeks is "marginally shitty?"

    In the grand scheme of all things that happen at work? Yes.

  • CNinja (unregistered)

    I don't see a WTF... this is SOP here.

    At least they started paying OT this time around. Probably because half the team quit last time.

  • stillinbeta (cs)

    How could anyone think that writing your own version of a library would take LESS TIME? Maybe I haven't worked in industry enough...

  • Adriano (unregistered) in reply to Bogolese
    Bogolese:
    From 5 months to 6 weeks is "marginally shitty?"

    Well, it is a shitty margin.

  • airdrik (unregistered)

    Kind of reminds me of what people in one part of Brazil that I'd traveled would say when asked about such things as how long they've lived there. The usual answer was "dois dias", literally: two days, but in usage it meant: some indeterminate amount of time; I don't keep track.

    So, how long will the project take? Six weeks (and by "Six weeks", I mean however long it takes to complete the project)

  • SysKoll (unregistered)

    TRWTF is that someone managing a development team thinks, in spite of overwhelming evidence, that it's faster to develop an in-house basic module than to buy a product or to adopt an OSS solution.

    We are in a very agressively scheduled project and we are bringing in OSS modules as much as possible to limit our own coding effort.

  • boog (cs)

    What may have helped Gary out is if they followed a policy of outlining smaller project phases with individual estimates. For example, if the herbal remedy (or "HR") database needed eight months, I'm sure Gary could easily break that down to a 20 or 30 smaller chunks, each with its own estimate.

    If Steve wanted to change the estimate to 6 weeks, he'd need to provide estimates for all of those chunks. Since each phase would be estimated at roughly a day of work, it'd become obvious very quickly that 6 weeks is nowhere near enough time.

  • IamNotLegend (unregistered)

    Sounds like a similar experience; where the Architect spent weeks analyzing and coming up with a plan in which he said it would take 1yr to roll out, management decides that it should only take 3months; "after all it's only programming". LMFAO.

    He quit, in came the next architect and before he could even review the requirements, he was told, again "we see this rolling out in Jan 10" and this was back in Oct09. No points for guessing what he did afterwards. LOL.

    Fast forward to Jan 2011, a few overpaid consultants later, management still think it will be out in 3 months.

  • boog (cs) in reply to SysKoll
    SysKoll:
    We are in a very agressively scheduled project and we are bringing in OSS modules as much as possible to limit our own coding effort.
    I hope you meant to say "necessary" instead of "possible".
  • Steve A (unregistered)

    I thought this was a good article.

    When you come up with an estimate, you have to be objective and not let anyone else's expectations or wishes to interfere, and have the guts to stick by it.

  • piskvorr (cs) in reply to Anon
    Steve had altered Gary’s letter
    something marginally shitty happens
    Whoa. I wouldn't want to be on your team (I smell "middle management" from your attitude - "outright fraud is okay, as long as it's not me getting screwed"). Indeed TRWTF is Gary not objecting - it's not like such an abusive relationship will magically get better.
  • JB (unregistered)

    Managers like this can destroy a business, you have to go over their heads or walk away. However some of the projects mentioned sound quite feasible in 6 weeks, in fact I've written similar functionality to the drug/herb DB in 2-3 weeks solo with additional features you do not describe, so maybe you should consider more solid (potentially commercial) frameworks too.

    The manager was obviously barking when thinking of building new drivers, etc. but a hand-written 10,000+ line data-layer in ~10 days solo shows that some people are just slow coders! The drug database as a system would easily have been achievable, you just have to make it clear the data is a separate task. Sounds like a lack of knowledge on one side, and a lack of drive and ability on the other!

    Managers need to listen to their senior developers, but the developers also need to say "This isn't feasible" or just do it for the business to work and be able to employ them!

  • airdrik (unregistered) in reply to stillinbeta
    stillinbeta:
    How could anyone think that writing your own version of a library would take LESS TIME? Maybe I haven't worked in industry enough...
    Oh, that's easy! When you write your own library, you know exactly how it works and you can even write in your own custom hacks so that the code on top of it will be more efficient. When the library is more efficient then your code will be more efficient and you can get things done faster. You also don't have to spend any precious time reading documentation or setting up and configuring the library to make sure that it will work in the context of the current project.

    Of course you can't trust open source anyway - it is constantly changing, so we would need to be constantly updating our software around the library, and when the source code is open then any one can examine it and find holes to exploit and we don't want that.

    Besides, who reuses anything these days when it is so easy to just write things from scratch. On that matter, you are going to be writing your own custom programming language designed around this project, which of course will require its own compiler (which can be used to optimize things saving even more time), as well as a new IDE built around the new language and compiler. We will also be using our own custom protocols for data transmission, and we need our own custom database engine to optimize our custom schema and data processing needs.

    With all of the time that I estimate that we are saving by writing custom, optimized versions of everything (and about 100 more developers), we should easily be able to trim the whole project down to six weeks!

  • Severity One (cs) in reply to stillinbeta
    stillinbeta:
    How could anyone think that writing your own version of a library would take LESS TIME? Maybe I haven't worked in industry enough...
    Why, have you worked for less than six weeks?

    Reminds me of a manager at work, who stated that implementing a certain feature would take "half a page of code". After our, um, comments, he doesn't use that particular turn of phrase any longer. We, however, do. :)

  • Steve A (unregistered) in reply to boog
    boog:
    What may have helped Gary out is if they followed a policy of outlining smaller project phases with individual estimates.

    Yes. Also, breaking the project into "Phases", and only giving an estimate for one phase at a time.

    Marketing and Sales will promise the customer the moon and all of the tea in China too. It is better to split it up into an initial phase addressing the most basic core functionality, and put all of the bells and whistles in later phases.

    By the time the initial phase is complete, the scope of the project will change to an extent that anything beyond that would not have been applicable anyway.

  • frits (unregistered) in reply to boog
    boog:
    If Steve wanted to change the estimate to 6 weeks, I'm pretty sure I would have strangled him before he had a chance to change it.
    I'm pretty sure you would find yourself in jail in addition to being unemployed.

    I bet you don't think your so smart now, do you?

  • Bushea (unregistered)

    I used to work for a company with a slightly different take on things. The product was high profile, since it related to compliance management; but the management used.. interesting tactics to get things done.

    First project I was on, we all got called in, told we had six months to complete, otherwise the department would need to be "pruned". We worked weekends and late nights (three of us becoming dads for the first time while working on a laptop in the delivery room).. and just scraped the product out. This was just around the time of the 9/11 attacks in the US. Did we get a small bonus for the 12-month project we delivered in 6 months? No.. hell no. We all got 12% knocked off our salary because "times were hard", but obviously not too hard for the directors to get new company mercedes that year..

    Fast forward to next big project. We all got called in, told we had six months to complete otherwise the department would need to be "pruned". We worked weekends and late nights, through birthdays, anniversaries and divorces. We managed. This time we got a major bonus: 6% of our 12% lost salary in vouchers for the cheapest shop in town. Super.

    Fast forward to next big project. We all got called in, told we had six months.. I think you get the picture.

    Companies do this, and they get away with it for a long time; most often because they prey on environmental situations: is there a recession on? Time to cut the salaries. Is there major instability in country xyz? Time to drop bonuses. Employees typically get entrenched out of fear in these companies.

  • boog (cs) in reply to JamesQMurphy
    JamesQMurphy:
    @Deprecated:
    I think in the suit meeting, Gary should have bluffed.

    "My estimate was 5 months! Whatever gave you the idea it would be done in six weeks?"

    ... and produced his copy of the estimate as proof.

    I've worked on projects where the guy in Gary's position has to sign off on the estimate before development could begin. Sadly, if that happened, Gary's original copy of the estimate wouldn't matter to the suits at all.
  • Beta (unregistered) in reply to JamesQMurphy
    Comment held for moderation.
  • TopCoder (unregistered) in reply to airdrik
    airdrik:
    stillinbeta:
    How could anyone think that writing your own version of a library would take LESS TIME? Maybe I haven't worked in industry enough...
    Oh, that's easy! When you write your own library, you know exactly how it works and you can even write in your own custom hacks so that the code on top of it will be more efficient. When the library is more efficient then your code will be more efficient and you can get things done faster. You also don't have to spend any precious time reading documentation or setting up and configuring the library to make sure that it will work in the context of the current project.

    Of course you can't trust open source anyway - it is constantly changing, so we would need to be constantly updating our software around the library, and when the source code is open then any one can examine it and find holes to exploit and we don't want that.

    Besides, who reuses anything these days when it is so easy to just write things from scratch. On that matter, you are going to be writing your own custom programming language designed around this project, which of course will require its own compiler (which can be used to optimize things saving even more time), as well as a new IDE built around the new language and compiler. We will also be using our own custom protocols for data transmission, and we need our own custom database engine to optimize our custom schema and data processing needs.

    With all of the time that I estimate that we are saving by writing custom, optimized versions of everything (and about 100 more developers), we should easily be able to trim the whole project down to six weeks!

    that's very true.

  • cod3_complete (cs)
    A few days later Gary was holding a signed letter from the client agreeing to the proposal. Gary reviewed the document and discovered everything was exactly as he had written it, except for the time estimate. Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.

    This is a key anti-pattern at work here. The wrong people made the estimate and at the wrong time. How the hell does a moron like Steve keep his job????? It baffles the mind. Not only that but why did Gary stay or at least raise hell with upper management?? If upper management knew a story besides Steve's at least one light-bulb might come on that something is amiss.

  • DaSteve (unregistered)

    This comment will be funny in six weeks.

  • Ken B. (unregistered)

    Reminds me of a boss I had years back...

    We would have meetings, where he would discuss what he wanted, and and the dev team for an estimate. If, for example, we estimated 6 months, he would ask "what would it take to get it in 3 months?" We would respond, "if we changed X, Y, and Z, and were given resources A, B, and C, we could probably do it in 3."

    Well, he would go off with the "3 months" estimate, totally ignoring the X, Y, Z, A, B, and C modifiers, and wonder why the project was always behind schedule.

  • Steve's a wuss (unregistered) in reply to My Name Is Missing
    My Name Is Missing:
    The industry standard is two weeks.
    Agreed.
  • Shawn (unregistered)
    Comment held for moderation.
  • Jeff (unregistered) in reply to JB
    JB:
    Managers like this can destroy a business
    ... if they're allowed to. That means, no bailouts, no too big to fail.

    When you have multiple small businesses pursuing the same market, the bad ones get weeded out. When the government grants patents and other types of monopolies to one big corp, complete with immunity from lawsuits, customers have no choice and the big nightmare stumbles on forever.

  • Dr.Evil (unregistered) in reply to frits
    frits:
    boog:
    If Steve wanted to change the estimate to 6 weeks, I'm pretty sure I would have strangled him before he had a chance to change it.
    I'm pretty sure you would find yourself in jail in addition to being unemployed.

    I bet you don't think your so smart now, do you?

    Hahaha unregistered frits, I know you're trolling, but you know we just can't let this sort of thing go here on this site.

  • davee123 (unregistered)

    Underestimating is a key ingredient in getting the contract. The people making the decision NEVER KNOW what's realistic, all they see is numbers on paper, and they'll pick whatever is smallest.

    Almost 2 years ago, our company had a new web service to develop. Our department bid 5 months, using our already-established framework. IE, we'd use our own user system, our own already-in-place-and-tested servers, our delivery and monitoring systems, the whole works.

    Another department was also bidding. They had NO framework in place. No plan for staging, no user system, nothing. But they bid 4 months, and were armed with a SECRET WEAPON: pretty screen mock-ups, which management LOVES.

    Guess who won the bid?

    6 months after starting, their project manager abandoned ship. A new one was assigned to the project, and the whole thing was renamed. 8 months after THAT, the NEW project manager ALSO jumped ship. And 2 months after THAT, the project was AGAIN renamed. It's now 18 months into the project. And I believe their target date is about 2 months away. We're not holding our breath.

    DaveE

  • Alex (not that one) (unregistered)

    There are basically two possibilities:

    1. It will take as long as it will take.

    2. The amount of time it will take can be decreed by management, and reality will conveniently jump into line.

    If you really believed #2, why wouldn't you just wave your magic wand and declare everything will take just one day? No, the fact that they "negotiate" down to six weeks implicitly admits that they know it will take some nonzero amount of time. So what this really means is that they are hoping by applying a lot of pressure they can get the lazy slobs who just play video games six hours a day to buckle down and actually work. In short, it is an extremely disrespectful and unprofessional assumption, and they should be maneuvered into stating it explicitly.

    I worked for one corp where the pressure trick was they spent $BIGBUCKS on an ad campaign that was irrevocably scheduled to "drop" on $DATE. So, they thought, the developers won't dare run the project even one day late.

    The problem was, we weren't bluffing. There was just no possible way to meet their date. So we began frantically negotiating scope. We ended up launching a radically scaled-down solution the day before the ad campaign... and by radically scaled-down I mean it was a single web page that said "coming soon".

  • piskvorr (cs) in reply to Ken B.
    Ken B.:
    Well, he would go off with the "3 months" estimate, totally ignoring the X, Y, Z, A, B, and C modifiers, and wonder why the project was always behind schedule.
    See, that's just Management Physiology 101: selective hearing. This is an example of a conversation, as heard by the manager:
    "Hello $developer, could we deliver $product two months earlier than your estimate?" "Well, *a minute of white noise*, it is possible, b-*another minute of white noise*." "Great, I'll tell the higher-ups!"
    It is a well-known byproduct of the partial strangulation by a tie.
  • valerion (cs)

    My shitty deadline story is a project from a couple of years ago. I gave an estimate of 6 months. I was told this was ridiculous and I could have 3 months. I protested but got nowhere. It would be due in June, regardless.

    Then, management delayed the start of the project by 2 months due to some financial wrangles with a vendor who would be co-developing, but wouldn't accept the completion-date moving back. So now it was due in 1 month.

    Unsuprisingly it wasn't ready in time(!) and then I ended up in hospital for a few days and unable to work for almost 3 weeks. I went back to work a week earlier than the doctors recommended and had the pleasure of the CEO coming over to see me. He wasn't impressed I was in tracksuit bottoms and not a suit. He said to me "So it's all going to be ready by August?" to which I said "We'll try our best" and then he looked at me and said "Well, it WILL be ready." and walked off.

    I was due to go on annual leave for 2 weeks towards the end of August - a holiday I had booked 3 months before this project was even conceived. It was indicated down to me that if it wasn't done my leave would be cancelled. I indicated back up that they could kiss my backside.

    It all worked out in the end - we released when it was ready towards the end of September. There was no rush at all for it anyway.

    That CEO has gone now, and been replaced with somebody slightly more reasonable. But not much.

  • Gary (unregistered)

    It's like the intranet implementation at my old job, which was always 18 months off. True in 2000 when I joined, true in 2007 when I left, and still true today I believe.

    Steve's not lying when he says the job will take six weeks. It's just that you don't decrement the time remaining as real time passes - so in four months there are still six weeks of project. You only get to countdown when there are REALLY only six weeks left.

    if (realTimeRemaining > sixWeeks) {
       reportedTimeRemaining = sixWeeks
    }
    else {
       reportedTimeRemaining = realTimeRemaining
    }
    
  • Mike (unregistered) in reply to Anon
    If you quit your job every time something marginally shitty happens, I doubt you'd be in any job for more than about....six weeks.

    Let me guess, you're in sales or in management?

    Repeatedly lying to your clients and disregarding the recommendations/input of the people you hired (because you supposedly trusted them and valued them) isn't "marginally shitty". Perhaps compared to losing your life or limb in a blue-collar job it could be "marginally shitty", but in context of technical work it's a slap in the face over and over again.

    It's called "ethics" and "integrity", and to some people those are important values instead of taking part in manipulative, ignorant, & shortsighted schemes.

  • hoodaticus (cs) in reply to Wilhelm Svenselius
    Wilhelm Svenselius:
    Steve had altered Gary’s letter to indicate that it would take six weeks to finish the project.
    TRWTF is that he did not quit right there and then.
    In the corporate world, silently going along with that == "you owe me a favor of my choice in the future".
  • blah (unregistered)

    Come on! No Total Recall jokes?

  • YR (unregistered)

    I wonder if I would last if I rise into lead/analysis/manager positions.

    I doubt I could hold my temper and not: (a) show the middle finger... (b) shout "F YOU!"... (c) or slam my original estimate on his desk... ... and turn my back on Steve, no matter the expensive suit guys.

Leave a comment on “Six Weeks Steve”

Log In or post as a guest

Replying to comment #:

« Return to Article