Flutterby™! : _x_ time 2007-11-19 18:08:02.931396+00

_x_ time

2007-11-19 18:08:02.931396+00 by meuon 9 comments

All projects take X time, even the supposed 30 minute ones. Actually, 30 minute projects take x² while month long projects only take X *2. X is a constant and is equal to more time than you have to do a good job at the project, feature request, bug fix.. etc..

X also has additional multipliers including the Good Deed Factor, the Non-Profit Org Factor, the For A Loved One Factor, and the ClueLess DotCom Style Startup Factor.

We probably all need to take the pledge:

I hereby promise to not start another "small" project without a free month of immediate time and $5k up front.

#Comment Re: made: 2007-11-19 18:22:42.580971+00 by: ziffle [edit history]

I once had a programmer say "it will only take ten minutes" - I looked at him and said nothing takes ten minutes. His feelings were hurt. He was attending UTC and they had him convinced he was gods gift to software.

I inherited a number of websites. In several of the cases, the owners called up later, claiming the site was not functioning, and further it used to work. Looking at the code, there were large gaps - as in empty - as in no code. In other words it could not have ever worked- EVER! But since they were 'good christians' and would never lie, and I wanted to mollify them and settle them down I fixed the code for free.

Usually followed by 6 hour searches looking for a period where a comma should be.

There are some major flaws in programming syntaxology, periods and commas are one of them, colons and semi-colons in CSS, using ' in sql.. () and {).. I need to create a font for X-term that makes all of these VERY noticeable.

was the first major programming project. It set the minimum acceptable time overrun standard factor of 2.7. If any project takes less time, then it is a) not really done, b) grossly overbid or c)grossly underspeced.

JT, I find that on Windows it's 10 minutes to solve the problem, followed by 6 hour searches to make it work when the user has done something completely unrelated, like change their desktop background, wherein you discover that in fact the API you used is now 5 generations out of date, the current one was thoroughly covered at some Microsoft event you didn't attend, and it'll take two weeks to implement to anyway.

On the Mac it's 10 minutes to solve the problem, followed by 6 hours of tearing your hair out while you try to figure out why the open source libraries that ship with the machine are from 5 year old source compiled with utterly bizarre config options.

On Linux it's 10 minutes to solve the problem, followed by 6 hours of trying to convince the customer that, really, it is that simple.

In Java they tell you it's 10 minutes to solve the problem, but six hours later you'll have managed to do nothing but generate big coredumps and huge CPU loads.

In Perl it's 10 minutes to solve the problem, but next time you solve the problem you'll just have to solve it from scratch all over again.

In C++ it's 10 minutes to write the necessary template structure, and then 6 hours to figure out why some implicit constructor side effect is actually solving the problem.

Tomcat, spend 10 minutes installing the software under Tomcat, 6 hours trying to tweak Tomcat to actually run, and then two weeks trying to figure out why there are edge cases that work fine in one Tomcat install but not in another one.

#Comment Re: made: 2008-01-03 11:22:11.0703+00 by: meuon [edit history]

#Comment Re: made: 2008-05-16 11:25:00.009755+00 by: meuon [edit history]

Proof. Shortly after posting this in November, Dan referred me to a small project by nice people, and while, yet again, it soounded like a days work. It came online in May. Broker Risk Management was a joy to work with, and it took so long because Lesie cleaned up hundreds of documents by hand. It took days to get their local IT Guru and I synced up with hosting and DNS, and now it's done. Most of the real work is not visible, but there is some neat stuff once logged in as a client. With the way things are going, it may be my last small web project for a while. And yes, I charged way under $5k, and in retrospect (and this original post) should not have. But I'm also glad they like it and are happy. Sometimes that's a good part of a project.