Path: utzoo!attcan!uunet!samsung!zaphod.mps.ohio-state.edu!uwm.edu!marque!lakesys!mikes
From: mikes@lakesys.lakesys.com (Mike Shawaluk)
Newsgroups: comp.sys.amiga
Subject: Re: Self Extracting Archives
Message-ID: <1759@lakesys.lakesys.com>
Date: 19 Mar 90 13:02:42 GMT
References: <55.25f441b5@uoft02.utoledo.edu> <195@sai.UUCP> <2675@leah.Albany.Edu> <79.25f87ef0@uoft02.utoledo.edu> <10102@cbmvax.commodore.com> <90.25fc5fd6@uoft02.utoledo.edu> <492d7755.1a5bf@moth.engin.umich.edu> <104.25fdcdd4@uoft02.utoledo.edu> <270@estinc.UUCP>
Reply-To: mikes@lakesys.UUCP (Mike Shawaluk)
Distribution: na
Organization: Lake Systems - Milwaukee, Wisconsin
Lines: 50

In article <270@estinc.UUCP> fnf@estinc.UUCP (Fred Fish) writes:
>I would have stayed out of this discussion except for the fact that I find
>SXA's particularly distasteful.  Anyway, the above statement is clearly false
>because there exists a set of arguments against SXA's that cannot be applied
>against executables inside a normal archive for the simple reason that they
>are not applicable.  For example:
>
>	1.	There is no simple way to get a listing of the contents
>		of an SXA.  There is a simple way for non-SXA's.
> . . .
>	3.	SXA's are machine dependent and can only be run/used/extracted
>		on the target machine for which they were built unless a
>		separate archiver exists that supports both SXA's and
>		non-SXA's.  A non-SXA can be unpacked and examined on any
>		machine for which the required archiver is available.

I beg to differ.  In the MS-DOS world, there are several SXA methods in use;
three that I can think of are for .ZIP, .ZOO, and .LZH files.  In each case,
a short program is prepended to the appropriate archive file (for the .ZIP
file, there is some internal tweeking, but read on).  When this is done, the
file is STILL readable by the "normal" utility (i.e., PKZIP and PKUNZIP can
be used to read the contents of and/or extract from a self-extracting ZIP
file; the same is true of ZOO and LHARC).  The only requirement for the
utility that does the extraction or viewing is that is look past the initial
self-extraction code & "lock on" to the signature that's at the start of the
archive part of the file, and decode it from there.  In a ZIP file, there are
also pointers with the file headers and central directory, which are modified
when a file is made into a self-extracting one, such that it's not required
to seek over this code; one can seek to the end of the file, back into the
centra directory there, and go directly to whatever file or files are
required.  I know that, in the case of PK*ZIP, that Phil specifically took
steps to assure that it would be possible for people to access files in
self-extracting ZIP files using the "trusted" tools, for exactly the reasons
you and others have mentioned.

Again, please do NOT take the intent of this message as one of advocating
SXA's (btw, I like that acronym; saves a lot of typing; thanks Fred :-).
Rather, I am trying to shed a little light on this whole discussion (I would
have done so earlier, but our host here hasn't been allowing postings
lately, due to disk shortages.)  I would agree with Fred & others regarding a
self-extracting format like PAK that is unique to one platform (the Amiga),
and that isn't a very flexible tool (and, as a separate aside, doesn't
compress all that well either, but then, it wasn't designed to...)

  - Mike
-- 
   - Mike Shawaluk             
"Rarely have we seen a mailer  ->  DOMAIN: mikes@lakesys.lakesys.com 
 fail which has thoroughly     ->  UUCP:   ...!uunet!marque!lakesys!mikes 
 followed these paths."        ->  BITNET: 7117SHAWALUK@MUCSD