Some Mail No Longer Fully Displaying


Environment: Mail 2.0.1 on OS X 10.4.1

Some mail, which displayed completely with Mail 1.X under OS X 10.3.X, no longer displays completely on Mail 2.0.1 on OX X 10.

The mail in question is generated by Mailman software in digest mode. Some parts of the mail - generally, the text of the messages comprising the digest, do not display: i.e., the message content appears empty.

The raw source of one of these messages is appended below. In this message it is the material between matching "--===============1313329100==" that does not display.

Thanks for your help,



From Sat Jun 25 12:00:11 2005
Return-Path: <>
Received: from ( [])
by (Postfix) with ESMTP id DC43A9A93C
for <>; Sat, 25 Jun 2005 12:00:11 -0400 (EDT)
Received: from ( [])
by (8.13.2/8.13.2) with ESMTP id j5PG08Zo765986;
Sat, 25 Jun 2005 12:00:08 -0400
Received: from ( [])
by (Postfix) with ESMTP
id 898F99A96B; Sat, 25 Jun 2005 12:00:01 -0400 (EDT)
Content-Type: multipart/mixed; boundary="===============1406747642=="
MIME-Version: 1.0
Subject: DataWeekly Digest, Vol 20, Issue 4
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Distribution for Data Analysis Group Weekly Reports
List-Unsubscribe: <>,
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>,
Message-Id: <>
Date: Sat, 25 Jun 2005 12:00:01 -0400 (EDT)
X-Virus-Scanned: amavisd-sophos
X-PSU-Spam-Flag: NO
X-PSU-Spam-Hits: -0.869
X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on
X-Spam-Status: No, hits=-4.5 required=5.0 tests=AWL,BAYES_00,NO_REAL_NAME
autolearn=no version=2.63

Content-Transfer-Encoding: 7bit
Content-Type: text/plain;

Send DataWeekly mailing list submissions to

To subscribe or unsubscribe via the World Wide Web, visit
or, via email, send a message with subject or body 'help' to

You can reach the person managing the list at

When replying, please edit your Subject line so it is more specific
than "Re: Contents of DataWeekly digest..."

Content-Transfer-Encoding: 7bit
Content-Type: text/plain;

Today's Topics:

1. Tiffany - Jun 24, 2005 (TIFFANY SUMMERSCALES)
2. Daily Report (John McNabb)
3. Keith - Weekly Report 06/24/2005 (Keith Thorne)

Content-Type: multipart/digest;

Content-Transfer-Encoding: 7bit
Content-Type: message/rfc822

This week I worked on reviving and automating the Poisson test. I now have a
driver function and scripts that will allow me to run over and produce Poisson
test results on the data specified in a segwizard list. The driver function
uses data conditioning and can produce results for any IFO and band. I also
updated my confidence interval finding function so that given the Poisson test
results and an estimate of the model parameter, k, it will produce a confidence
interval for theta (the strength of the correlations in time) for each test
Currently, I am working on finding optimum thresholds and number of bins for
differnt data lengths. Once that is done, I plan on running the test over all
of S3.
Next, I plan on writing a function to take all of the results and produce
summary webpages.



Content-Transfer-Encoding: 7bit
Content-Type: message/rfc822

My Weekly Report can be found at:

"In theory, theory and practice are the same,
but in practice they're different."
John W. C. McNabb

"In theory, theory and practice are the same,
but in practice they're different."
John W. C. McNabb

Content-Transfer-Encoding: 7bit
Content-Type: message/rfc822

From: Keith Thorne <>
Precedence: list
Subject: [DataWeekly] Keith - Weekly Report 06/24/2005
Date: Sat, 25 Jun 2005 09:52:19 -0400 (EDT)
To: DataWeekly@Gravity.PSU.Edu
In-Reply-To: <>
Message-ID: <>
Content-Type: TEXT/PLAIN; charset=US-ASCII
MIME-Version: 1.0
Message: 3

Weekly Progress Report - June 24, 2005
Keith Thorne

**Tasks Accomplished This Week**
1) r-statistic results on white-noise MDC
I completed r-statistic analysis on MDC set WNB2_BN_S4 of
20-ms white-noise bursts. I presented this together with
a completed non-MDC background analysis at the Tuesday

The MDC data also shows a tail at large Gamma (as
did the background). However, even a modest cut
at Beta_3 = 3 ( which keeps only 1% of the background)
only keeps 30% of the MDC signal. This is worrisome
for trying to use the r-statistic to remove background.

2) R-statistics comparisons with CorrPower
I worked with Laura Cadonati to see if the tail at Gamma>5
in S4 background was real or a bug. Laura both analyzed
my BNETG event sample with her updated CorrPower code, and
started on the S4 WaveBurst studies. She reported that she
also see a tail at Gamma>5 in the WaveBurst triggers, but can
remove most of it using a H1/H2 energy-matching cut. (This cut
can not be used in the present BNETG sample as the old V2 calibration
with bad H2 alpha-betas was used). She also fed back a list
of the outliers at large Gamma.

3) BNETG CPU usage vs segment duration
I conducted an investigation of the quadratic scaling of BNETG CPU time
vs. segment duration was coming from. I determined that while the
'nithro2' scaled linearly, the 'refine' code scaled quadratically.
The culprit in the 'refine' routine was found to be the code that remade
the CP list after removing a Change-point. This was replaced by
functionally-equivalent code which marked points to be removed, than
removed them all at the end. This change removed the quadratic scaling.
The study also revealed that 'nitrho2' was producing change-points at
a rate 20 times greater than output from 'refine'.

4) Update to S4 spectral line lists
Following last week's DetChar I updated my web pages on the S4 spectral
line features. I also added tentative violin mode identifications from
Gregg Harry

5) New Burst MDCs with V3 calibration data
I made new Burst MDC sets (SG18_S4, SG19_S4) which used the V3 calibration
constants. This was to allow the testing of the H1/H2 energy ratio
cut on MDC data (as V2 had bad alpha-betas for H2).

6) New White-Noise MDCs for PSURG analyses
I also made new Gaussian-envelope white-noise burst MDC sets for
our use at PSU. These used the V3 calibration constants. They
also attempt to show a distribution in amplitudes by using
sets which have a 1/h^2 relative population (128 at 1e-19,
64 at 1.414e-19 down to 1 at 11.314e-19) Set WNB3_BN_S4 is like
WNB2_BN_S4 (20-ms envelope), while WNB4_BN_S4 has 1 sec envelope
for external-trigger studies. A web page was created to track them

7) Re-clustering, analysis with V3 calibrations
I re-submitted the S4_Pass1 pipeline at the clustering step
to see the effect of the V3 calibrations (and to be able to test
the H1/H2 ratio cut). I also added an updated results page to that
pipeline to automate things. I also started analysis of the
new white-noise MDC (now generated with V3 constants) to also study
that cut. These are currently plowing through the grid computer

**Tasks Planned Next Week**
1) Results on re-analysis with V3 calibrations
2) Study of H1/H2 ratio cut on white-noise MDC
3) Start tuning BNETG for 1sec white-noise bursts
4) Check performance on white-noise MDCs
- time reconstuction, r-statistic, etc.


Content-Transfer-Encoding: 7bit
Content-Type: text/plain;

DataWeekly mailing list