Path: utzoo!utgpu!water!watmath!clyde!rutgers!ucsd!ucsdhub!hp-sdd!ncr-sd!crash!pnet01!haitex
From: haitex@pnet01.cts.com (Wade Bickel)
Newsgroups: comp.sys.amiga
Subject: Re: Speeding up directory access time
Message-ID: <2471@crash.cts.com>
Date: 3 Feb 88 11:37:36 GMT
Sender: news@crash.cts.com
Organization: People-Net [pnet01], El Cajon CA
Lines: 61

mp1u+@andrew.cmu.edu (Michael Portuesi) writes:
>haitex@pnet01.cts.com (Wade Bickel) writes:
>
>>          When using various File Requester systems I notice a common flaw.
>>        If I "accept" a selection, the Requester closes.  Often I am loading
>>        a series of files into buffers or some such thing, and immeadiately
>>        need to repeat the process.  The problem is that the Requester must
>>        read the relevant directory each and every time I call it up, which
>>        when using a floppy is often a bit of a wait.
>>
>>        It seems to me that maintaining a directory of the floppies currently
>>        in the drive(s) in RAM would solve this.   Since (I assume) the system
>>        is aware of human disk manipulations this should not be too difficult
>>        to implement.
>
>Of course, you could simply buy a copy of FaccII and save yourself
>the trouble...

          Yes, I could buy FACC and it would speed up my floppy accesses, but
        this does not help users' of code I'm writing.

          I can retain the data myself, but was hoping C= might include this
        feature in 1.3 or 1.4.  Seems the algorithm should go something like
        this:


                        1)  Set up some structures to support retaining
                              of directories.

                        2)  Add code to the system so that directory listings
                                are routed through new/custom routines.

                        3)  When an access is attempted the drive's status
                              is checked.

                                a) If the disk's directory has allready been
                                     read, return it.

                                b) If the disk's directory has not been read 
                                     read and return it, and retain it in the
                                     structures (dynamic or quasi-dynamic
                                     allocations, or course).

          It would also be nice to retain directories of disks which have
        been removed, as an option.


          In order to avoid interferance from unknown disk accesses, I assume
        the datestamp field should be checked to determine if the directory
        has been read?


                                                                Thanks,

                                                                      
                                                                     Wade.


UUCP: {cbosgd, hplabs!hp-sdd, sdcsvax, nosc}!crash!pnet01!haitex
ARPA: crash!pnet01!haitex@nosc.mil
INET: haitex@pnet01.CTS.COM