The mention of WWWMAN is misleading; the "tests" done aren't valid. From
the results given by some users, it is obvious the problem is *NOT* in
Arachne, or WWWman ...
This is the message received when you enter WWWMAN at the DOS prompt:
WWWman, Arachne cache and directory maintenance tool
(G)1997 New Wave Microtechnology Ltd. Portions (G)1998,1999 Arachne Labs
Portions (G)1999 Bernhard "Bernie" Eriksson
WWWman -c <cache_index> converts cache index to HTML
WWWman -d file:fullpathname DOS directory, table, icons from WWWman.cfg
WWWman -l file:fullpathname DOS directory, text only
Program writes to STDOUT; syntax is WWWMAN [options] > file
-------------------------
The line below
> cd \arachne
> wwwman -l file:@:*.* >\temp\yourpc.htm
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
when entered at the DOS prompt, will of course cause problems!
The "@" is an instruction to Arachne to use maximum memory. I'm pretty
darned certain that is filtered out before the "command line" gets to
WWWman... but wwwman is capable of ignoring it if it *is* placed on the
command line.
I know that because wwwman will write the information, using the
yourpc.htm code (if it resides in the same directory as wwwman.exe) as
a framework, to screen if the output to file ">" is omitted ( which,
of course, means the subsequent path to any location is ignored).
This type of test, using internal codes as command line items, isn't
valid. What *appears* to be a path statement -- like the above --
*inside* Arachne is *NOT* ... that "temp" is a VARIABLE containing the
actual path to the ARACHNE.TMP directory.
So unless you have a \temp directory set up within the Arachne
directory, the command *cannot* be properly processed from CLI ...
period.
The result is a "file creation error" .... which doesn't bother DOS but
*COULD* bother any Win9x installation, because Windows is notorious for
being unable to handle most error messages.
> at DOS prompt, and you'll get the lockup on MS-DOS 7.10 as well).
>> a) MS-DOS 6.0 on different machines: works
>> b) MS-DOS 7.10 (AKA the DOS from W98) on different machines: blocks
FYI, MS-DOS 7.x isn't a complete "DOS" and AFAIK does not handle error
messages nearly as well as the old, "outdated" versions from 3.0 thru'
6.2x ... [I know I darn sure wouldn't want to try running Arachne in a
"command line mode" on W2K !! I don't even try to run my DOS games
there. That's what the A: drive is for on the Toshiba Tecra. <G>]
> It looks like WWWMAN has trouble with interpreting available drive
> letters under plain MS-DOS 7.10;
Not the job of WWWman ... WWWman just passes the commands thru to
get the requested data; the failure is in the OS itself.
Since Arachne is designed for "low end" systems, those which operate on
Linux, or MS-DOS 5.x or lower, I think it is counterproductive for people
to be "testing" it on Win9x or W2K or WinMe or WinXP.
Those Windows systems are *not* the machines/systems Arachne is designed
for, and to expect Arachne to be redesigned to run on those systems --
to take into consideration all the shortcomings of the truncated "DOS"
that Win9x and higher use -- is ludicrous! That would require Arachne
to bloat up damn near as big, and hog resources damn near as much, as
Windozes does!
> <snip> Some internal system structures are much different
> in DOS 6.x and 7.10, and there are some bugs in MS-DOS 7.10 that
> "disappear" when Windows GUI (or rather its virtual drivers) takes
> over much of the DOS functionality.
DURRRRRRRRRRRRR, GEORGE!
Arachne is not a Windows program!!!!!!!! If you want to run Arachne,
then use an OS that is fully functional ... don't blame it on the
software design of Arachne.
It would appear that the only "disclaimer" that needs to be added to
Arachne release is:
WARNING! Fully functional OS is required.
Attempts to run this program using any of the crippled MS-DOS
programs integrated into Win9x or higher can cause lockups and
system crashes!
And if Billy Gates don't like it, he can kiss my ASCII
l.d.
