Re: [NTLK] MP2100 Offset of Recognition

From: Stefan Thorsteinson <>
Date: Wed Aug 26 2009 - 00:13:27 EDT

Hi Frank,

Interesting. I did note at one point, while doing a couple of hard resets,
that Newt generated the calibration screen about three times before it
accepted where I tapped. And yes, I had my glasses on. :) I hadn't
considered cheating the recognition at that point. ;)

The "Jaggies" by hardcore users may not be what I'm seeing. I can't say that
I saw the generation of squiggly lines as opposed to a straight line drawn.
I don't do graphics very often. It started as seeing a tap to the left of an
actual tap. I also experienced an insertion of an underscore where I was
trying to select a point to insert.

When writing, I would see the character I was writing show up to the right
of where I wrote it. It's usually been to the left. If it were a four or
five character word, the last character would look nothing like what I drew.
Hence my thoughts toward Jaggies.

Perhaps my experience is somewhat unique. Regardless, I will download the
test and report whatever results I encounter. What gets me is that I opened
Newt right up for a good bath and it seems all problems went away. As
previously stated, time will tell. For the balance of today I have accessed
Newt for contacts or info with no issue.

I will report back to Andy on what I find out. I mention this particularly
because I thought it interesting that I normally work off battery power and
rarely with AC power and, like Andy, the latest incident occurred while I
was plugged into AC and adding a contact. (Coincidence?) Things went
sideways while trying to input. Did I experience a combinations of things at
once? Not sure. I intend to play a bit and report back.

I feel bad to offer more curiosities than assurances. Frank & Andy - I will

Almost forgot: Frank: More testing required but, particular quadrants of
screen are being more sensitive than others has merit. Lower right quad in
portrait seems a little unresponsive. More testing required...

Too late to relate....

Stefan Thorsteinson

-----Original Message-----
[] On Behalf Of Frank Gruendel
Sent: Tuesday, August 25, 2009 5:24 PM
Subject: [NTLK] MP2100 Offset of Recognition

> I'm noticing that my 2100, for no apparent reason that I can trace, will
> start offsetting the point of recognition of a tap on the screen.

Note that this is NOT (as has been posted) an incarnation of what Newton
users lovingly call "the
jaggies". The latter results in all kinds of weird lines being drawn when
writing on the screen
because the touchscreen is misled by pressure where no pressure should be.

Alignment problems are a different cup of tea altogether. For one, they
usually do not change with
pressure on the bezel. And, unlike the jaggies, they have a tendency to be
stronger at certain parts
of the screen, which you can easily test for yourself if you download
NewtTest from the software
section of the site in the signature and use the subtest where the screen is
filled with small
buttons that you need to tap to disappear. If you tap a button and keep the
pen down, you will see a
dot where the Newton thinks you tapped. You'll notice in almost every case
that the offset differs
depending on where on the screen the particular button is.

The good news is that within certain limits you can resolve this hardware
problem by cheating during
the alignment routine. All you need to do is click with a deliberate offet.
I always forget in which
direction the offset must be, but if you want to compensate e. g. a vertical
offset in the upper
part of the screen, you need to click either above or below the topmost "X".
Note that if the offset is too large, the Newton will not accept your tap
and show the "X" again.
But you can increase the offset by using the alignment routine multiple
times with smaller offsets.

Hope it helps


-- Newton software and hardware at

The NewtonTalk Mailing List -
The Official Newton FAQ -
The Newton Glossary -
WikiWikiNewt -

The NewtonTalk Mailing List -
The Official Newton FAQ -
The Newton Glossary -
WikiWikiNewt -
Received on Wed Aug 26 00:16:08 2009

This archive was generated by hypermail 2.1.8 : Wed Aug 26 2009 - 00:30:00 EDT