struts-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chappell, Simon P" <Simon.Chapp...@landsend.com>
Subject RE: [OT] Estimating a Struts-based project
Date Tue, 25 Mar 2003 15:48:00 GMT

>-----Original Message-----
>From: Andrew Hill [mailto:andrew.david.hill@gridnode.com]
>Sent: Tuesday, March 25, 2003 9:00 AM
>To: Struts Users Mailing List
>Subject: RE: [OT] Estimating a Struts-based project
>
>
>Well given such constraints, your most accurate estimates are 
>likely to come
>from an astrologer - or even an economist - nah! thats going a 
>bit far. Ill
>stick with the astrologer.

:-)

>Failing that you could do what I do when subjected to the 
>gnashing of teeth
>routine.
>These estimates arent worth [one of Mark's this's] but the 
>proportions are
>correct - just not the magnitude:
>
>First you estimate the times using the method you think would 
>actually come
>up with something thats relatively accurate. This will of 
>course give you
>estimates that the client won't accept.

I'm very good at coming up with estimates that make them spit blood.

>The client of course already has their own 'estimate' (they call it a
>deadline).

Yes. Every project starts with an acronym and a deadline. Well known fact. Actually, that's
the only well known thing at the start of most projects.

>You see how many hours there are until that 
>deadline, and then
>allocate your estimates to add up to that based on the 
>proportions in the
>'real' estimate.

So, don't bother creating an estimate. Or create one that you are happy to have ignored. This
is why I want to algorithmise (not a real word) the process so that I can easily and with
little real effort create a quasi-realistic estimate that management can then safely ignore,
but that I could actually work from the 1% of the time that they supprise me and accept the
estimate.

>Then think of something that sounds quite unlikely to occur 
>(but is in fact
>100% certain (an unexpected windows bug perhaps?)) and make a 
>small note on
>the quote that if this happens it just might possibily in the 
>worst case
>increase the estimates by a factor of x (where of course x is 
>the value that
>the clients estimate of total work needs to be multiplied with 
>to get the
>real estimate of total work...)

This is too much like CYA and I never do that. If a project fails after my estimate was ignored,
I refuse to feel bad about it.

Simon

>-----Original Message-----
>From: Chappell, Simon P [mailto:Simon.Chappell@landsend.com]
>Sent: Tuesday, 25 March 2003 22:38
>To: Struts Users Mailing List
>Subject: RE: [OT] Estimating a Struts-based project
>
>
>These have all been good suggestions, but I am constrained 
>here by a number
>of factors.
>
>All our estimates have to be in hours.
>There is screaming and wailing and gnashing of teeth if the 
>estimate is "too
>big".
>We are not a software house, so the business gets to set schedules.
>We are a HIGHLY seasonal business, this forces many scheduling 
>decisions.
>etc. etc.
>
>Thanks to everyone.
>
>Simon
>
>-----------------------------------------------------------------
>Simon P. Chappell                     simon.chappell@landsend.com
>Java Programming Specialist                      www.landsend.com
>Lands' End, Inc.                                   (608) 935-4526
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: struts-user-unsubscribe@jakarta.apache.org
>For additional commands, e-mail: struts-user-help@jakarta.apache.org
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: struts-user-unsubscribe@jakarta.apache.org
>For additional commands, e-mail: struts-user-help@jakarta.apache.org
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: struts-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: struts-user-help@jakarta.apache.org


Mime
View raw message