[MetaCRS] Proj4js 1.0.0 release candidate 1

Mike Adair madair at dmsolutions.ca
Tue Oct 7 13:52:27 EDT 2008


Richard Greenwood wrote:
> * The dynamic look up code should be in a separate module so that it
> can be excluded, or a different service lookup service used.
>   
You may not have a requirement for dynamic loading, but I do (and I 
suspect other people do as well).  I think the implementation I came up 
with is a reasonable compromise because:
- if you don't need dynamic loading, just include the required defs 
through a script tag and the dynamic loading code is never used.
- The code to handle dynamic loading amounts to 5k uncompressed and 2k 
compressed so the difference is not a large hit in terms of download time.
- I initially had that code separated out, but doing that complicates 
the overall library un-necessarily for the end-user. 
- there isn't an dependency on any external libraries now so Proj4js is 
a self-contained package.

> * The NZGD49 to WGS should fail. It's a grid shift datum transform.
>   
It would be good to get Stephen Irons input on this but the test cases 
submitted with that contribution use a datum code of nzgd49 (New Zealand 
Geodetic Datum 1949) which is a 7 parameter datum and in theory should 
work with proj4js.

> * Why did you remove my bi-directional transform form my test page?
>   
This is still in SVN.  I can add a reverse transform button to the index 
page if you felt that was important.
> * The object model is overly complex, adds nothing to the clarity or
> usability, and distances the library from its C language origins.
>   
This is a JavaScript library, not a command line C program, and as such 
I think it is a good thing to use current JS best-practices to implement 
it.  The object model is actually very simple.

Mike



More information about the MetaCRS mailing list