Re: Yarn: DOS vs. Win95
From: A.R. & F.L. Scott-Thoennes (firstname.lastname@example.org)
Date: Tue, 25 Feb 1997 01:02:10 -0800
In article <0QjEzcGUbV7A092yn@gate.net>,
email@example.com (John A. Stanley) wrote:
>Ever since my system crash in the DOS import.exe due to a junk email
>with a 100+kb To: line I've been using import95, and I'm happy to
>report that it is far more robust. I used it to import that same
>crash-packet, and it simply ignored the offending email and filtered
>the rest of the email into the appropriate pseudo-newsgroups.
>However, I noticed that import95 doesn't import packets nearly as
>quickly as the DOS executable, so I've been fiddling around with Yarn95
>and comparing it to yarnx.exe. The bottom line, is that Yarn95 takes
>a big hit in the speed department compared to yarnx.exe (specifically
>in terms of text searches.) Isn't 32bit software supposed to achieve
>_higher_ performance rather than lower performance?
This is a myth. 16bit software is often as fast as a 32bit equivalent.
The advantages of 32 bits lay more in the posibility of expanded
functionality than in efficiency of size or speed. However, yarnx.exe
_is_ 32bit software, just a completely different type.
Something you should be aware of is that import runs the filter1.exe
program once for each piece of mail (not news) that it imports. Could
you have had import95 running the dos version of filter1.exe? It seems
to me this might be slower that dos import running dos filter1, but
I don't know enough about win95 to say for sure.
>And one more thing: is it at all feasible to make URL's in articles
>"clickable" so that a running web browser would go there with a
>simple click instead of a cut'n'paste into the Open location window?
Good idea, but Chin would have to add mouse support first :)