Analysis, design, and testing always take longer than programming. Even
with a code change of one line you may take a lot of time analyzing the
problem and the best way to fix it.

Jerry Banker


> -----Original Message-----
> From: John Rodgers [mailto:[EMAIL PROTECTED]
> Sent: Friday, March 07, 2008 4:01 PM
> To: [email protected]
> Subject: [U2] Estimating Timelines
> 
> Snipped from Louie's contribution on the Include Wierdness
> 
> -     I have noticed that some programmers inflate their estimates by
> 10
>       times, just so they will look good finishing early.  We have to
> be    accurate, without lying.
> 
> 
> Speaking for myself, this is usually necessary because today's
> applications are so complex that all the dependencies are impossible
to
> forecast.
> 
> It is not a question of lying - it is a question of trying to
> anticipate
> the worst case.
> The first estimate you give is the only one that "management" (your
> customers)  will ever remember. You can explain the changed
> circumstances until you are blue in the face - it makes no difference.
> The customer cannot see the issues. You are late in delivering.
> 
> 
> A good example just occurred today.
> 
> Set up a control flag for particular function.
> 
> Easy enough - we already have a screen for these things - just add one
> field
> Should be about an hour. Tops. No big deal.
> 
> Add the field and test it.
> SB+ spits it back at me with a "matrix out of range" error.
> 
> These are linked screens - we are on the fifth screen.
> I smell a big problem.
> 
> I create a sixth screen and put the field on that additional screen.
> Now I get a different (more informative) message. "Too many fields on
> linked screens".
> 
> As I suspected - we have hit the limit of this arrangement. We are on
> an
> old version.
> 
> Now I have to totally re-design this set of screens.
> With Page forward and back keys
> Don't forget the Inquiry mode - How do I make those use the same
screen
> defns?
> Make sure it functions "exactly" as it did before - it does not want
to
> because the sub-screens are a different herd of cats.
> 
> Now, how much time should I have estimated for this task?
> And this is not unusual.
> 
> Does anyone out there have any thoughts on THIS issue?
> I think it (estimating timelines) is one of the more difficult tasks
we
> face. Especially when interfacing to an existing mature application.
> Coding is the easy part of this job.
> 
> 
> Cheers
> 
> 
> JR
> 
> 
> 
> 
> 
> 
> 
> 
> John Rodgers
> 
> Masonite International
> 
> Tel:  (813) 2612396 ext 3036
> -------
> u2-users mailing list
> [email protected]
> To unsubscribe please visit http://listserver.u2ug.org/
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to