Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!rpi!dali.cs.montana.edu!ogicse!hsdndev!husc6!purdue!haven!wam!dmb
From: dmb@wam.umd.edu (David M. Baggett)
Newsgroups: comp.sys.atari.st
Subject: Max of 113 files in a subdirectory: fact or fiction?
Message-ID: <1991Jan10.084308.13614@wam.umd.edu>
Date: 10 Jan 91 08:43:08 GMT
References: <10472@lanl.gov> <1991Jan08.061722.26635@convex.com> <1991Jan8.153539.19118@wam.umd.edu> <1991Jan09.010526.15973@convex.com> <1991Jan9.090428.29529@wam.umd.edu> <1991Jan09.215647.5890@convex.com>
Sender: usenet@wam.umd.edu (USENET Posting)
Reply-To: dmb%wam.umd.edu@uunet.uu.net (David M. Baggett)
Organization: University of Maryland at College Park
Lines: 39

In article <1991Jan09.215647.5890@convex.com> rosenkra@convex.com (William Rosencranz) writes:
>
>In article <1991Jan9.090428.29529@wam.umd.edu> dmb%wam.umd.edu@uunet.uu.net (David M. Baggett) writes:
>>It's DOS (and therefore, for compatibility, TOS) that has the file 
>>limit restriction.  Only 113 files in a FAT.  Or is it 112?  In any
>>case, don't ask me why.

>i thought this restriction was only for the root directory of a disk,
>not subdirectories. but my memory, as usual, is fuzzy, and i have no
>way right now of verifying this. i know u can have more than 113 files
>in a hd partition, though i can't be sure if u can have more than
>113 files in total on a floppy, even if they are in directories. easy
>enuf to test, however.

Well, I seem to recall running into the 113 file limit when archiving
(once again) HacMan II -- I couldn't have all the files in a single
directory at once because there were more than 113.  I tried to copy
the whole mess into d:\usr\games\hacman2\dist\ (i.e., not a root
directory) and got an error code (I forget the number) when gulam hit
the 113th file or so.

I guess it could be gulam, or maybe I did something odd.  I don't think
any of the archivers themselves impose a restriction on # of files; I
just meant that since you can't have more than 113 files in a single
directory (under TOS), a potential standard archiver must be able to
handle subdirs easily, otherwise we won't be able to archive things 
consisting of more than 112 files on all machines.

>                         . i distinctly remember lharc taking 6 min on
>a file that zoo did in around 2 min. still, it was a long time ago, so
>this is more of an impression than fact.

That time comparison sounds about right.  There's simply no denying
that LZH encoding takes considerably more CPU time than whatever
methods Arc and Zoo employ.  I don't know why this is since I
haven't looked at the algorithms.

Dave Baggett
dmb%wam.umd.edu@uunet.uu.net