[Mozile] Problem with path for rng

[James A. Overton]

Yesterday I did some testing on this issue, and discovered that the  
the loading system was working differently when using a local file  
system than it was when using a server. This was breaking the server  
paths in various cases. So the source of your problems is a bug that  
slipped in somewhere along the line (probably when I was fixing  
Reto's problem), and which I hadn't noticed.

The good news is that I fixed it. I've also changed the schema  
loading system to start at the mozile.root, like all the other file  
paths in Mozile function calls. I'll commit the changes to our CVS  
repository later today, once I've done a few more tests.

To answer your questions:

The mozile.js file is a "compiled" version, which is meant for  
deployment but not for development. It's built using Apache Ant: we  
concatenate a list of JavaScript files from the "src" directory, then  
we strip the whitespace and comments. The idea is that a single  
compressed file loads faster than a dozen uncompressed one. But  
changes you make to the source code inside "src" aren't reflected in  
mozile.js until you recompile it.

There's more general documentation here:
http://mozile.mozdev.org/0.8/doc/html/index.html

When you're developing, you want to load "src/core.js" and the  
modules from "src" instead of "mozile.js", as explained here:
http://mozile.mozdev.org/0.8/doc/html/ch02s03.html

To build a new mozile.js file, see the instructions here:
http://mozile.mozdev.org/0.8/doc/html/ch03s07.html

James


On 2006-Nov-06, at 7:16 PM, Gemstar wrote:

>
> Ah, I just went to look at it again.  Something else I see is that  
> basic.html
> is including mozile.js which has both parse(), load() and  
> mozile.xml.load()
> functions.  I don't see it using rng.js or xml.js.
>
> When debugging, I see that it seems to get to parse() successfully,  
> but when
> it goes to load(), it seems like it's trying to load the path of  
> the file as
> a string rather than loading it actually as a file.
>
> Let me know if this sounds likely.
>
> Thanks so much,
> Meg
>