[X Newbies] jpegs, etc.

[Vincent Cayenne]

At 1:07p -0400 2003.05.20, Florin Alexander Neumann wrote:
>This goes back to a basic problem. How does the computer know what 
>kind a particular file is? One way is to include this information in 
>the file itself. For instance, PNG files always begin with "âPNG". 
>Another is to include it in the file name -- e.g., "My Picture.png". 
>This is what Windows (still) does.

The file metadata is data about the data in the file so both of the 
above methods describe two locations for _placement_ of metadata as 
well as two different ways of stating the metadata.

>  Yet another is to include this information neither in the file 
>itself, nor in the file name, but in the file metadata, i.e., in the 
>information about the file.

The third option that was used by Apple-preference pre-OS X was to 
store the metadata in a resource fork separate from the data fork of 
the file. Again this both a difference in location and in the way the 
metadata is presented.

>  This information is present in a database maintained by the operating system.

The metadata was held within each file (otherwise it isn't really 
metadata?). The database in question was where looking up the file 
metadata (type and creator) would give the associated application and 
other applications capable of dealing with the type of data within 
the file.

>  Each method has its pros and cons.
>
>Macintosh used the latter option. For each mounted volumes, Finder 
>maintained a database with information about each file

No, the information remained in the individual files.

>, including a four-character Type code, e.g. "PNGf". This code was 
>invisible to the user, who could only access it with utilities such 
>as ResEdit or File Buddy.

The type and creator wasn't particularly hidden and could be seen by 
the simple expedient of using the "Get Info" in the Finder. Setting 
the type and creator outside of the associated application/s was 
often accomplished by various utilities.

-- 
'tis as said. [Reality is defined by being described]