View Source: UglySolutionsNeedLoveToo

Note: This page has been locked and cannot be edited.
!!! Ugly Solutions Need Love Too

Phil Hollenback%%%
[philiph | http://www.twitter.com/philiph ]

First, consider this picture:%%%
[https://farm8.staticflickr.com/7015/6784578285_acfccbb78e_d.jpg]%%%

That looks pretty horrible, right?  It gets even better: PG&E did this
on purpose when they replaced the power pole!

When I first noticed this hanging extra piece of pole, I
[made fun of it|http://twitter.com/philiph/status/163726516939792384].
I'm sure many of you would do the same.  That looks like an incredibly
horrible, sloppy job.

But then I got to thinking.  What problem was PG&E trying to solve?
I thought back to when they replaced that pole.  It actually took
two attempts.  The first attempt at replacing the pole resulted in
disaster - the workers hit the underground water main and flooded the
street.  I can only assume that the water main was buried directly in
front of the old pole.

PG&E gave up that day and put the old pole back.  Then, they came back
again a few weeks later and replaced the pole again, leading to the
curious setup you see in the picture above, with the hanging piece of
the old pole still in place.

Now if you know the story about the broken water main, this whole
thing starts to make more sense.  PG&E couldn't put the new pole in
exactly the same location as the old one, because that would be
sitting right on top of the water main.  That could lead to damaging
the pipe again in the future, and they can't use their automated power
tools to dig the hole.  Thus, they decided to move
the new pole back from it's old location.

However, as you can see, the lower cables (TV or telephone I assume) on
the pole are already stretched tight.  That means PG&E couldn't move
the pole back any further without a lot of additional work and
coordination with different companies.  Thus you get to a classic case
of having to support competing goals: don't hit the water main, and
als you can't move the pole back from the road any further.

As a result of this, PG&E decided to leave a section of the old pole
in place, and anchor it on the new pole.  While this certainly does
look ridiculous, it's not an unreasonable solution.  Sure, they
probably could have gotten the water company to move their pipe.
Maybe they could have gotten the phone company to splice in a longer
cable so they didn't have to leave that piece of the old pole hanging
there.

In the end, PG&E chose a solution that they could implement easily
(leave a piece of the old pole in place) instead of a solution that
would require more complicated interactions with other companies
(moving the water main or other cables).  That might not be the
optimal answer, but it gets the job done.  To be fair, note the rope
at the bottom of the old piece of pole - that might indicate a
temporay solution.  Maybe PG&E will come back and solve this all for
good at a later date - I don't know.

While I was working on this blog post, I read
[@evilchili|http://www.twitter.com/evilchili]'s piece
[Good Enough Isn't|http://amokti.me/2012/02/14/good-enough-isnt/].
That got me thinking: was the PG&E solution an example of a 'good
enough' answer that's just a bandaid?  Perhaps.  However given the
real-world constraints involved I don't think I can fault them for the
answer they ultimately came up with.

My point with all of this is that you should never be too quick to
judge someone's solution to a particular problem.  Maybe they did just
do the minimal amount of work necessary to get the job done.  Maybe
there will be more failures in the future because the comprehensive
solution wasn't implemented.  Ultimately you can't make that decision
until you understand all the variables.  Take the time to consider the
options, and weigh the pros and cons of available solutions.  Some
times the quick hack is the right answer, and sometimes a more
comprehensive solution is needed.

CategoryGeekStuff

CategoryDevops

CategoryBlog
View Source: UglySolutionsNeedLoveToo

Main Categories

Search

Toolbox

RecentChanges

Favorite Categories

Views
