🚀 go-pugleaf

In article <19788@agate.BERKELEY.EDU>, steve@violet.berkeley.edu (Steve Goldfield) writes:
> 
> Below are listed discrepancies between what I got from BMUG
> and the posted list (the dates listed are those of last modification
> or the latest date listed in GetInfo). . .
> 
> Name                  Version              YourSize   MySize
> 
> GateKeeper            1.0 (1/30/89)        42905      42804
> Interferon            3.10 (5/10/88)       26244      26218
> RWatcher              1.0 (11/14/88)        1450       1448
> Virus Detective       2.0 (1/27/89)        41686      35587
> Virus Rx              1.4a1 (1/3/89)       62286      62276
> VirusWarning (10/3/88)                       963       1344
> Warning 1.0d2 (12/20/88)                    3052       3412

This brings up a question I've wondered about:  How many bytes represent
a significant difference between two theoretically equivalent files?  I'm
an unrepentant archivist, and have lately been building a library of all
of the official Apple system releases, including fonts, DAs, CDEVs, etc.
My question stems from the discrepancies in the sizes of various copies
of the same version of some of the DAs, particularly the Control Panel
and Chooser.  Sometimes these differences are on the order of 1K or even
more.  Why?  I haven't mucked around with ResEdit to see whether any of
them have resources their smaller brethren don't, but I assume that this
must be the case.  Again, why?  Do these DAs build their own resources
based (in the case of the CP, for example) on the number of CDEVs they
have ever displayed?  How different constitues *different*?  It's been
driving me crazy...

In general, then, to what extent can the number of bytes occupied by a 
file be used to determine whether it is truly different from another?  
Is there a better way?

Thanks, folks.

Steven E. Bruun
AT&T Bell Labs
Naperville, IL
att!ihlpa!sebruun

"Is a dream a lie if it don't come true, or is it something worse?"