InfoMagic Internet Tools 1995 April/Internet Tools.iso/dos_win/winsock/maillist/94-05.Z/94-05/text0110.txt

Text File | 1994-05-31 | 7.4 KB | 153 lines

> In article <[e-mail]> [e-mail] (Matt Pierce) writes:
> >From: [e-mail] (Matt Pierce)
> >Subject: Re: Win Mosaic alpha 4 (my fix)
> >Date: Fri, 6 May 1994 21:25:06 GMT
> >Keywords: winmosaic,win32s
>
> >>Everything about my system says overkill (which is why it was named
> >>Goliath). However, the only thing I *can't* get it to do is 32-bit disk
> >>access. I think it has to do with the disk system being SCSI based but I
> >>don't know for sure. I was able to get 32-bit on my old 386 which had an
> >>IDE 200MB before I went with my current system?
>
> >>Anyone got a clue on this one? Thanks for any information.
>
> Does your drive have more than 1024 cylinders? I've heard drives with more
> than 1024 cylinders will not work in the 32 bit mode.

  Here's the message I read on a local BBS about large drives and 32 bit disk access:

    Only nuisance was having to trim down the size of the drive to 504 megs
to get the WFW and WIN 3.1x drivers to handle the 32-bit-disk-access=ON.
The 504 Megs is from 1024 cylinders by 16 heads by 63 sectors per track by
512 bytes per sector then divide by 1048576 (which is 1024 x 1024).

    So it was a trade-off:  To get 32bitdiskaccess=ON, I had to give up 12
megs of formatted capacity (the WD caviar 2540 formats out at 516 Megs when
used with 1048 cylinders by 16 heads by 63 sectors per track).   I figured
it was worth the 12 megs.

[e-mail]
From [e-mail] Sat May  7 20:52:38 1994
Received: from bigblue.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.07) with SMTP
          id AA25787; Sun, 8 May 1994 11:14:52 -0400
Received: by bigblue.oit.unc.edu (AIX 3.2/UCB 5.64/4.03)
          id AA22486; Sun, 8 May 1994 11:04:47 -0400
Received: from GATEWAY by bigblue with netnews
    for [e-mail] ([e-mail])
To: [e-mail]
Date: Sat, 7 May 1994 20:52:38 GMT
From: [e-mail] (David Arnstein)
Message-Id: <[e-mail]>
Organization: My Own Internet Account!
Sender: ses
References: <[e-mail]>, <2pv14a$[e-mail]>, <[e-mail]>
Reply-To: [e-mail]
Subject: Re: Win Mosaic alpha 4 (my fix)

In article <[e-mail]>,
Matt Pierce <[e-mail]> wrote:
>>Everything about my system says overkill (which is why it was named
>>Goliath). However, the only thing I *can't* get it to do is 32-bit disk
>>access. I think it has to do with the disk system being SCSI based but I
>>don't know for sure. I was able to get 32-bit on my old 386 which had an
>>IDE 200MB before I went with my current system?
>
>>Anyone got a clue on this one? Thanks for any information.
>
>You have hit the problem right on the head.  32-bit disk access turns on the
>the 32-bit block access mode that is, for some strange reason, turned off by
>default on IDE hard drives.  I still don't know a whole lot about this, but
>the way I understand it, just about all IDE hard drives have the capability to
>transfer data in 32-bit chunks.  However, they default to 16-bit because of
>the controller cards.  I don't know much more beyond that.  I do know that
>there are several ways to turn on the 32-bit disk access outside of Windows.
>Some of the newer BIOS manufactures (AMI for instance) have the capability of
>turning on 32-bit access from the CMOS setup.  I've also seen device drivers
>(Drive Rocket and VLIDE.SYS) that do the same thing, I think???
>
>SCSI drives do not have the same problem.  32-bit disk access is the norm for
>those with controllers that can handle it.  That is why the option is greyed
>out in Windows.  You can, however, still enable 32-bit file access if you're
>using Windows for Workgroups.  Believe me, it makes a BIG difference in speed.
>I'd love to know exactly what is going on, as I am, by no means, an expert on
>this.  I have used two SCSI drives in two different systems, though, and
>32-bit disk access was not an option on either one.

Everything above is FALSE.  Microsoft Windows 32-bit disk access has absolutely
nothing to do with block transfer mode.  In fact, 32-bit disk access is a very
misleading name (blame Microsoft).  It is simply a Windows device driver that
operates in protected mode that takes the place of some code which would
otherwise execute out of your BIOS (in real mode).  Executing BIOS code is
slow, as is switching back and forth between real mode and protected mode.
That's why 32-bit disk access improves your performance.

Now as to the issue with SCSI drives, consider that 32-bit disk access uses an
(internal) Windows device driver called *wdctrl.  This device driver knows how
to access IDE disks (Western Digital design) but does not know about SCSI.
Therefore, to use 32-bit disk access with a SCSI drive, you need a Windows
device driver which is specific to your SCSI host adapter.  As far as I know,
there is only one SCSI manufacturer that offers such a device driver:  Future
Domain.
--
David Arnstein       |          What do you mean, "get a life"?
[e-mail]  |          This *is* my life!
From [e-mail] Tue May  6 18:19:38 1994
Received: from bigblue.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.07) with SMTP
          id AA03268; Sun, 8 May 1994 12:14:44 -0400
Received: by bigblue.oit.unc.edu (AIX 3.2/UCB 5.64/4.03)
          id AA22312; Sun, 8 May 1994 11:55:16 -0400
Received: from GATEWAY by bigblue with netnews
    for [e-mail] ([e-mail])
To: [e-mail]
Date: 6 May 1994 18:19:38 GMT
From: [e-mail] (Hans Peter Kluender)
Message-Id: <2qe1nq$[e-mail]>
Organization: iXOS Software GmbH, Munich, Germany
Sender: ses
References: <2qacce$[e-mail]>
Reply-To: [e-mail]
Subject: Re: Does recv() receive what send() sends ?

P. Eivind Jenssen ([e-mail]) wrote:
: I'm programming with TCP-sockets, and I wonder:

: When I use the send() function, it returns the number of bytes that was
: sent out on the net.
: Can i be sure that the recv() function will receive all these datas
: that was sent, or do I have to tell the 'send-machine' how many bytes
: I received from it so that it can send the rest of the datas?

: I'd be very grateful if anyone could answer this.
: Thanks anyway!

send() and recv() use UDP, not TCP; both are based on IP.
Anyway: data sent by send() arrives complete or not at all.

 _  _  __ __   __     ______________________________________________
|_|\ \/ //  \ / /
 _  \  // /\ \\ \      Hans Peter Kluender     iXOS Software GmbH
| | /  \\ \ \ \\ \     email [e-mail]       Bretonischer Ring 12
| |/ /\ \\ \/ /_\ \    Tel [phone]     D-85630 Grasbrunn
|___/  \_\\__/|___/    Fax [phone]     Germany
From [e-mail] Sat May  7 22:21:13 1994
Received: from bigblue.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.07) with SMTP
          id AB06671; Sun, 8 May 1994 12:44:44 -0400
Received: by bigblue.oit.unc.edu (AIX 3.2/UCB 5.64/4.03)
          id AA23953; Sun, 8 May 1994 12:18:04 -0400
Received: from GATEWAY by bigblue with netnews
    for [e-mail] ([e-mail])
To: [e-mail]
Date: Sat, 7 May 1994 22:21:13 GMT
From: [e-mail] (Kenneth Kasajian)
Message-Id: <[e-mail]>
Organization: NETCOM On-line Communication Services (408 241-9760 guest)
Sender: ses
Subject: NSLOOKUP FOR WINSOCK???


Anyone got NSLOOKUP for winsock?