Programming with
a dose of satire.
Soapbox
|
| ||||||||
 |
 |
|||||||
|
Soapbox |
||||||||||||||||||||||||||||||||||
15 Commandments to Curb Bad Programmer HabitsBy Mark Collins-Cope :)
Message Forum Agree/disagree with the advice listed here - or got some "commandments" of your own to add? Why not leave a message... Message Index:
The Messages: You cannot just command accurate estimationThese "commandments" seem quite reasonable, except for #4. The reality of software is that it is an adventure through territory with the potential for an unexpected surprise at any turn. Before we start, we cannot know what we will need to know that we do not know. We aren't just building the same old toaster time and time again. Commanding accurate estimation is just not realistic. Demanding compliance to estimates will just result in inflated estimates. Realistically, open communication with the customer, iterative development in order of business priorities, and always having the latest working and tested system available is the best we can do to give the customer the most value by any date (whether this date is predetermined by us, or by the customer, or by changing market conditions). Steven Gordon sagordon@asu.edu http://sf.asu.edu/, usa Sun Oct 03 12:17:16 BST 2004 Good Point ... Even the omnipotent can make mistakes... Yes, thou art correct in that thou aren't just building the same old shit so estimating can be difficult... In these circumstances a contingency factor (which may be over the top) is the only way of guaranteeing that the software can be delivered to the agreed estimate... But perhaps that is better than being late... :) Mark Collins-Cope mcc@ratio.co.uk London Mon Oct 04 13:19:34 BST 2004 Alternative wording ... 4. Thou shalt strive to learn how to estimate work and deliver within agreed timescales ... Comments? Mark Collins-Cope mcc@ratio.co.uk London Mon Oct 04 13:56:41 BST 2004 Accurate Estimates and Schedule Pressures G'day, Maybe add numbers 4a, 4b and 4c to reflect the more usual way that a request for an estimate is given: 4a) Thou shall not respond immediately to a request for an estimate no matter how much pressure to do so is applied. 4b) Thou shall not agree to deliver to schedules thrust upon thee without first having time to evaluate whether the schedule is realistic. 4c) Thou shall refuse to adjust only thy schedule in the face of management pressure. Schedule djustments are acceptable when accompanied by adjustments to scope of delivered features. The number of times where I have been asked to give "corridor estimates" or estimates in meetings off the top of my head. Having a schedule dropped on to you at a meeting is also a good one. The other one that is good is collapsing under pressure applied by management to accept a tighter schedule. I used to work for someone who did this regularly. Only one person in the group actively resisted his pressure. A couple of times she directly said, "Well you can write down that date, but it won't be finished then." Later on when I was out having a drink with this guy he told me how this girl was the only team member who was always on schedule! This was where I learnt my lesson concerning this. cheers, Rob Rob WELLS robwells57@btinternet.com Maidenhead, UK Tue Oct 05 11:24:47 BST 2004 Estimate boundary Statistical analysis is possible only on large base sample. Programming do not usually have large base sample to give fairly accurate estimate. All estimates in this world not only programming are by nature inaccurate. That is why they are estimates. Some ideas: - ask your marketing to tell you when EXACTLY they are going hit one milion users for our product and EXACLY how much money they will spend to reach this marketing goal. Try to evaluate accuracy of these predictions. - Ask your sales force EXACTLY when they believe that bangs and whistles for a peanut price which they promissed to the just caught client will be delivered regardles of the production line? What is wrong here is to ask any production to give their estimates of the production output and what is more wrong to ask them to give that estimates accurrately. This shows incompetence in business proces - it is not their job to make management estimates - that is management job. For #4: thou shalt find better project manager that does not ask stupid questions about estimates but does his job of monitoring and managing the production process. lj lj011@yahoo.com Wed Nov 03 06:38:34 GMT 2004 RE: 15 Commandments The obvious superficial observation: God only needed ten. Sean skip_203@hotmail.com USA Wed Feb 09 21:00:01 GMT 2005
|
||||||||||||||||||||||||||||||||||
All trademarks and copyrights on this page are owned by their respective owners.
Stories and articles are owned by the original author.
All the rest Copyright © 1998-2008 Matt Stephens. ALL RIGHTS RESERVED.