[TAG] (forw) Re: [lg-announce] Linux Gazette #126 is out!

[Rick Moen]

Quoting Benjamin A. Okopnik (ben at linuxgazette.net):

> Because it would have added to the workload instead of decreasing it.
> Here, let me run through it again:
> 
> 1) According to your own description, it takes about six hours for you
> to do all the processing. This does not include the time (~1/2 - 3
> hours) that I've had to spend on it.
> 
> 2) Kat did it her own way in about five hours.
> 
> 3) Therefore, if she undertook to learn the old system, it would have
> taken her _at least_ 6 hours plus the learning curve - and would still
> have required me to spend the X hours fixing it.
> 
> What would you estimate - 12 hours? 15? Whatever it is, it was far
> greater than the five or so that it did take.
> 
> *THIS* is why it was the case.
[...]
> If learning a new process is so horrible, then why are you insisting
> that Kat should do it?


Just popping into the conversation briefly to confirm that the above
seems to make a great deal of sense, and is a compelling argument.
Kudos again to Kat for stepping in and finding a way to produce good
work efficiently and in a timely fashion.

I've been in similar situations, where I've made a judgement call to
switch to a possibly cruder mechanism of my own devising that I can
understand, rather than working longer to master a possibly better one
that seems to pose greater difficulties.  This is especially important
under time pressure -- and your initiative, Ben, to insist that
deadlines really matter is likewise laudable.

Really, it's a matter of whatever works best for the people who end up
doing the work.  Results matter.  Timely results _really_ matter.