home *** CD-ROM | disk | FTP | other *** search
- Path: soap.news.pipex.net!pipex!usenet
- From: m.hendry@dial.pipex.com (Mathew Hendry)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: Want to boot FAAAST?
- Date: Fri, 22 Mar 96 22:40:29
- Organization: Private node.
- Distribution: world
- Message-ID: <19960322.3F8A50.1446D@ai051.du.pipex.com>
- References: <68771929@0humpty.tomate.tng.oche.de> <1280.6645T903T2008@mailbox.swipnet.se> <4ipc7i$6n@sun1000.pwr.wroc.pl> <4irl0n$1dng@columba.udac.uu.se>
- NNTP-Posting-Host: ai051.du.pipex.com
- X-Newsreader: TIN [AMIGA 1.3 950726BETA PL0]
-
- Erik Trulsson (t92etr@sabik.tdb.uu.se) wrote:
- : Robert Wojcik (wojcikr@alpha.ok.ae.wroc.pl) wrote:
- :
- : : This is strange wondering, regarding one block of Ram Disk takes 1024 bytes.
- : : Anyway another 1024 for file header ...and it sucks like hell.
- :
- : That is a very common (and natural) misconception.
- : The 1024 bytes block size reported bt RAM: is a lie from the filesystem.
- : The ram disk doesn't use fixed blocks to hold the files, it only uses the
- : amount of memory that is actually needed (plus some small overhead most likely)
- [...]
-
- The overhead is about 130 bytes or so per file (a bit more, I suppose, if the
- entire file cannot be contained within a single memory chunk).
-
- Try this script:
- --- CUT ---
- .key dummy
- .bra {
- .ket }
-
- set counter 1
- set endtotal `avail total`
- set starttotal `avail total`
-
- echo "Testing..."
- lab copyloop
- echo >ram:test{$$}_$counter "t" noline
- set counter `eval $counter + 1`
- if val $counter not gt 100
- skip back copyloop
- endif
-
- set endtotal `avail total`
- delete quiet ram:test{$$}_#?
- echo "RAM used to store 100 1 byte files = `eval $starttotal - $endtotal`"
- --- CUT ---
-
- which will write 100 1 byte files to RAM: and report on the amount of RAM
- used. On my machine this produces:
- Testing...
- RAM used to store 100 1 byte files = 13600
-
- or 136 bytes per file, amounting to an overhead of 135 bytes per file.
-
- -- Mat.
-