Difference between revisions of "Amira Quirks"

Revision as of 15:44, 8 May 2009

Root Directory Saving Problem

You SHOULD NOT save any network under the root directory of a drive.

For example, if your USB drive is located at E:\ do not save at E:\ directly but in a folder beneath E:\ like E:\Amira\ or any other name.

Why is this happening?

You need to understand Amira network structure first.

...so what happen if you do save it there?

Let's say you have a network called "test.hx" that loads three files a.am, b.am, c.am that are located in the data directory of the same network ("test-files" folder)

Well, when you save at NON-ROOT directory, you get correct relative referencing of the .hx network by loading the data modules at the proper RELATIVE locations. Eg. "test.hx" will load "test-files\a.am" "test-files\b.am" "test-files\c.am" with REFERENCE relative to the .hx file location PROPERLY.

However, if you save at the ROOT directory, you will only get PRECISE definition of the .hx network. Eg. "test.hx" will now attempt to load "C:\test-files\a.am", "C:\test-files\b.am", "C:\test-files\c.am" with ABSOLUTE path definition and result in file problem once you copy the networks around since it will always look for the file in those SPECIFIC locations (C:\test-files\).

Save Network

Another key issue you need to be careful is how you save your network. ALWAYS save as "Pack and Go".

Why is this happening?

All data modules are saved seperately. However, if you loaded the data module from elsewhere (eg. recycling an existing .surf or .am file from other networks), you need to save your OWN copies of it instead of relying on the reference to the source.

...so what happen if you don't save it as pack and go?

Let's say you have a network called "test.hx" that loads three files a.am, b.am, c.am that are located in the data directory somewhere else in the master database in server (eg. Z:\Database\).

Well, when you save without using "pack and go", every time you load up the network, you are still attempting to load the data from Z:\Database. This is even worse when you might accidentally mess up the source data by overwriting them (hopefully, i should already placed READO NLY there). So when you move your network around, you will get error and no data loaded if the new location does not have access to Z:\Database. Eg. "test.hx" will load "test-files\a.am" "test-files\b.am" "test-files\c.am" with REFERENCE relative to the .hx file location PROPERLY.

However, if you save using "pack and go", all externally loaded data file will be saved at "test-files" folder and hence not absolutely loading from any other places. This ensures you have proper portability of the file and will not run into problem when loading it on the presentation machine (Shuttle, or cube)