[Talk-us] Fwd: Fixing TIGER street name abbreviations

[Anthony]

On Sat, May 12, 2012 at 7:08 AM, Dale Puch <dale.puch at gmail.com> wrote:
> He is only editing ways originally imported by tiger, based on the tiger
> prefix, suffix, road type, and base name.  If all were rev 1 unedited
> imports it would work darn well.  The testing is mostly for bad tiger data
> and subsequent edits that confuse things.

Surely anything less than "darn well" is unacceptable.

> Your just saying it cant/shouldn't be done, and he is figuring out a way to
> make it work correctly.

He just said in the other thread that he can't get it to work
correctly, and that the focus should be on fixing the mistakes.

> Let him do his tests and provide results, and then
> you can try and find the faults in that rather than telling him to not try.

I don't think it's a good idea to find the faults.  He'll just fix
them and come out with a new version.  And then, evenutally, we'll
stop finding faults.  He'll assume that means the faults aren't there.
 But what it really means is that the faults are ones that are harder
to find.

> He isn't running this on the live DB so why not encourage/help him produce
> better results.

Because if he produces better results it's more likely people are
going to allow this to be run on the live DB.

> If there are things you know that will cause problems,
> provide real examples of it so he can edit and test the script to handle
> them or gracefully skip those.

See above.  Manually finding the errors in the script is equivalent to
manually editing the names in the first place.