home *** CD-ROM | disk | FTP | other *** search
- Path: hearst.acc.Virginia.EDU!adastra!mbs
- Newsgroups: comp.sys.amiga.programmer
- From: mbs@adastra.cvl.va.us (Michael B. Smith)
- Subject: Re: YASL (was: AsynchIO (was: fastest file read method ??))
- References: <3116c08e@lls.se>
- X-NewsReader: GRn 3.0a February 1, 1996
- MIME-Version: 1.0
- Content-Type: text/plain; charset=iso-8859-1
- Content-Transfer-Encoding: 8bit
- Message-ID: <mbs.4892@adastra.cvl.va.us>
- Date: Wed, 7 Feb 96 07:06:57 EDT
- Organization: Only if you insist...
-
- In article <3116c08e@lls.se> cmh@lls.se (Magnus Holmgren) writes:
- > In a message of 29 Jan 96, Michael wrote to :
- >
- > <AsyncIO>
- >
- > MBS> Actually, SeekAsync() is still badly buggy, in every version I've
- > MBS> yet seen.
- >
- > Even release 7? I have fixed a couple of bugs, and I don't know of any
- > holes at the moment, but that doesn't mean that there aren't any. ;)
- >
- > If you could report them (preferably with a compilable example in C), I'd
- > have a chance in fixing them. ;)
-
- I've not tried a version of yours since 4 (which I had already said in an
- article early in this thread).
-
- The base asyncio code (i.e., Martin's) has a real problem with early buffer
- spills. Up through v4 of your stuff, while it didn't appear to be reading
- extra blocks, the data point was seriously offbase.
-
- My "example" is a mailreader, and is several thousand lines of source. To
- prove it wasn't my code, I re-did it using dos.library buffered I/O at one
- point. That worked. The original seekasync() worked (slowly). Yours gave
- me the wrong data.
-
- I'll try to find time to look at v7, but it'll be a week or so.
- --
- // Michael B. Smith
- \X/ mbs@adastra.cvl.va.us
-