[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Date Index][Thread Index][Author Index]

OT Re: the electrix repeater/estimating software jobs



As a professional software developer currently involved in very 
complex video effects software, I can sympathize.

Here is my partly tongue-in-cheek algorithm for estimating a software job.

1.      Estimate the amount of time the job would take if nothing went 
wrong.
2.      Allow for the fact that things will surely go wrong, as they 
have in the past, and double it.
3.      Allow for the fact that, in addition to the things that have 
gone wrong in the past, new things will go wrong. These will be even 
harder to fix, so double the estimate again.
4.      One of the things most likely to go wrong is the initial 
estimate i step 1. Allow for this by doubling your estimate again.

If you work hard, and are reasonably lucky, you won't exceed this 
estimate by more than 100%.

Although the above analysis was  conceived of as a joke, I suspect 
that it would  be reasonably accurate at estimating software jobs.

Back to tweaking loops on a Nord Modular so I can twist the input of 
s sax player friend who is coming by tomorrow.-:)

At 9:42 PM -0700 8/3/01, Mark Hamburg wrote:
>Here's my guess as to what's happening at Electrix. Software is hard. Big
>software is really hard. People who've written small pieces of software 
>and
>had them work don't always realize that it doesn't all just scale 
>linearly.
>
>Add to that the fact that there are probably a lot of time-dependent and
>parallel operations going on in the unit and you've got a debugging
>nightmare. Things go wrong but you can't make them go wrong again.
>
>So, if you'd be prepared to only do the things they demo'd in the sequence
>and with the timing they demo'd them, you'd probably be fine. Some tester
>probably didn't stick to the script and had something bad happen, but they
>are having a horrendous time reproducing it reliably.
>
>My first job in the commercial software industry was working for Ann Arbor
>Softworks. At the time A2S was famous for FullWrite Professional being
>vaporware (though in total it was less than a year late). I joined at a
>point where we spent several months operating under the lines: (a) we'll 
>be
>done in two weeks and (b) if we don't ship in two weeks we're going to go
>out of business. A few years later I was talking to one of the lead
>engineers on the team and he confessed that he'd realized afterward that 
>he
>and the other lead engineer could hold about two weeks worth of work in
>their heads. Hence when asked how long it would take to finish, they'd
>reflect for a while and come up with the answer "Two weeks".
>
>My question is: Has anyone at Electrix read _The Mythical Man-Month_?
>
>Mark
>
>on 8/1/01 6:04 PM, Mark Sottilaro at sine@zerocrossing.net wrote:
>
>>  I'd love to here an
>>  uncensored version of what's going on at Electrix from an 
>engineer/software
>  > engineer.  I'd bet we'd all have a nice chuckle.
>  >
>  > <nelson> Ha ha </nelson>

-- 

"The reasonable man adapts himself to the world. The unreasonable man 
persists in adapting the world to himself. Therefore, all progress 
depends on the unreasonable man.

--  George Bernard Shaw

                Emile Tobenfeld, Ph. D.
Video Producer                  Image Processing Specialist
Video for your HEAD!                    Boris FX
http://www.foryourhead.com              http://www.borisfx.com