Bob here...developer of RFinder...
> Have anyone here looked at repeaterbook.com at
all?

>>Heh, free, a sane web interface, and they have an iOS app too.
 I wonder 

>>how comprehensive the database is.  It also makes me wonder
what the 

>>RFinder value proposition will be :)

Honelstly I never saw it before...it isn't found easily
on search engines...my first impression is that it is similar paradigm
to the ARRL Repeater Dir...his sitename is self descriptive...our db is
geolocation enbabled...I did not look deeply into his data but at first
glance he is not geolo enabled as when I drilled into his data there were
no lat/Lon's...mapping seems approximate and based on city.
RFinder for iOS code cuts this summer.

> It also has some areas that need work.  The guy who wrote the
site

> is a self-taught php developer.  He's been pretty open to the

> suggestions/changes that I've proposed.

>>It returns queries semi-instantly, so at least he has a grasp on
the

>>backend. There is some information in there on "paper repeaters"
in my

>>area that are coordinated, but never got put up. Looks like some
of the

>>data was imported from the local coordination body (gotta get it

>>somehow, I suppose).
His queries are quick...that is our weakness and we
are addressing that.  We will change our DB this summer...the web
services will be functionally the same so there will be no front end programming
changes (I think, LOL)....we prototyped this but did not expect the popularity
we have achieved in 6 weeks...600 users already!!! We were trying to get
out the gate...bottom line we know there is a performance issue and it
is being addressed.

> There are a couple of export formats currently supported. I think
we

> have a reasonable chance of getting chirp csv export format added.

> (Chirp native xml is in the range of possibilities too of course)

>>That would be great. Maybe you can bring it up if you already have
a

>>relationship with him? It'd be even nicer to have a web services

>>interface to the query, but that would cut down on his ad revenue.

While I love a good competition, our scope is different
from his...we are creating the first Worldwide Repeater Directory...I don't
belittle anyone's previous attempts at creating online repeater databases,
but we are not regional.  The data is user maintained (on the android
platform you can request changes to an entry, and hopefully by the weekend
add a new repeater...this will migrate out to interactive web and web service).
 If you want something on RFinder's back end, just ask.  We are
already a web service.  We started there...interactive web is not
our primary interface although it exists from and admin perspective right
now and will be end after we change DB's.  I don't want to sound like
a megalomaniac, but....One Master Database.  Please support this...for
obvious reasons. 'Nuff said.

> There are lots of repeater databases out there, but most seem to

> have put little to no thought into the data maintenance process.

>>Yeah, I've just never been one to preload my radio like that before

>>traveling, so I'm usually not on the lookout for stuff like this.
 I

>>guess some people are, though.

We put end user maintenance in within a week of publishing
to the market.  By user request. Again, we will allow adding interactively.
 We have adding by batch already although it is an email process currently.
 We have been asked by reps of Chile, Brazil, Israel, Switzerland,
South Africa and Sweden so far to work with them on improving the data.
 We have provided their data in our spreadsheet format for them to
edit and return...all are done except SA and Sweden.

> All the effort seems to go into the GUI and the database/site quickly

> looses relevance as the data slowly rots.

 >

> Thoughts?  Comments?

>>Yeah, the RFinder guys seem to have no consistency checks in the
data

>>they put in (or imported, or whatever). Tone values with transposed

>>digits and what-not. Does the repeater book site handle that better?

>>Requiring a human editor is a nice idea, but weeding out invalid
tones 

>>and things of that sort is much better done in code, I'd say.

We have been working over the last couple of weeks
to fix inconsistencies in the data.  We have an algorithm worked out
to fix the PL's.
Our data is sourced from dozens of places....once
Dan pointed out the inconsistencies with the PL's our new mode is to validate
on import.  We have to retroactively fix what's wrong though.  Work
in progress.

>>Unless nobody finds it interesting/useful, I think the integrated

>>RFinder client functionality in chirp is still worthwhile.  Would
you 

>>agree, or not?