shulman@topaz.RUTGERS.EDU (Jeff Shulman) (11/05/85)
Delphi Digest Monday, 1 Nov 1985 Volume 1 : Issue 3
Today's Topics:
Power Line Questions
TML Pascal compiler
RE: TML Pascal compiler (Re: Msg 3003)
RE: Usenet Digest Volume 1 Issue 3 (Re: Msg 3004)
----------------------------------------------------------------------
From: ELL
Subject: Power Line Questions
Date: 30-OCT 23:59 Hardware & Peripherals
If you were starting from scratch in adding a branch circuit for computers and
other electronic devices, how would you set it up?
Would it be possible, or even desirable, to put line filters/surge suppressors/
spark gap arrestors/UPS's/etc. between the circuit breaker and wall outlets? (
Also, what are the merits of using or not using these devices)
What do you like/have/use? What should be avoided? Any reference to articles or
particular devices would help too. Thanks.
------------------------------
From: WINTERS
Subject: TML Pascal compiler
Date: 1-NOV-02:44: Tools for Development
Has anyone tritried out TML's new Pascal compiler? I'd like to hear
about it . Thanks Steve Winters
------------------------------
From: RONB
Subject: RE: TML Pascal compiler (Re: Msg 3003)
Date: 1-NOV-18:26: Tools for Development
I received versoin 0.9 of TML's Pascal a few days ago. I think this is
the current version. I've only briefly looked at it, but it looks
pretty good so far. It compiles to machine code or assembler, with the
option of using the MDS system if you have it. Comes with same
editor,and RMaker as MDS (modified, I think). Problems we've run into
so far are errors may be pointed out without line #'s or procedure
reference so an error in some statements ( such as an ELSE sitting by
itself) can be hard to find, also Units, EXIT, and SIZEOF are missing,
also the documentation is not yet completed but promised "at TML's
expense". It comes with examples and complete interfaces to toolbox.
Let me know if you need more info. Ron (RONB)
------------------------------
From: PEABO
Subject: RE: Usenet Digest Volume 1 Issue 3 (Re: Msg 3004)
Date: 1-NOV-19:25: Mousing Around
Concerning the serial drivers and flow control --
The usual method of flow control is to issue XOFF (^S) when you want the
other end of the link to pause, and issue XON (^Q) when you are prepared to
receive. There is a bit in the status which controls whether this is
enabled by the driver or not. You can set this bit, but my understanding
is that there is a bug in the ROM version of the driver which prevents
the flow control from working.
There is another driver for the serial ports, called the RAM serial driver,
which does perform flow control properly, but which must be packaged with
an application, and opened explicitly. It is a resource called SERD in the
application resource fork.
The RAM serial driver is part of the Apple Software Supplement, and is
available from a number of sources.
You can issue XON-XOFF from within your application as an alternative, but
that can be quite awkward, since you have to anticipate when you are going
to lose control, issue the ^S before that happens. Of course, using a
bigger buffer helps.
Also, a device typically does not respond to flow control right away ...
quite a few characters might appear after the ^S, depending on what kind of
device is on the other end. You have to wait for things to settled down
after the ^S to be sure that it has taken effect. This is one great
advantage of using the RAM serial drive ... it responds at interrupt level
even if the application is busy.
On the other hand, I don't know what strategy is used to prevent chattering
of flow control. Ideally, you want some hysteresis, so that flow control
is exercised when the buffer is say, 80% full, and not released until the
buffer is less than 70% full. Otherwise you could be sending a *lot* of ^S
/^Q combinations while the buffer utilization hovers around 80%.
Beware of sending multiple ^S. Some devices could get confused if you send
^S several times without sending ^Q in between.
Jeff, take a look in the database under the topic Macintosh in Fact ... you
should be able to find the RAM serial driver there, along with
documentation on how to use it.
[ I will upload the doc file if someone wants me to but I think the RAM
driver itself is part of the software supplement and can't be made
"public" - JSS ]
Concerning Imagewriter loss of data --
This also sounds like a flow control problem. The Imagewriter has a DIP
switch setting to control whether it uses XON-XOFF or DSR (control lead)
busy signalling. As an experiment, you can try flipping the switch and see
if the problem clears up.
------------------------------
End of Delphi Digest
********************