yahoo groups mangles some syntax?

[Daniel]

I started a new topic (should have done this before),
because this has diverged from N / N+1 topic. You can
see previous posts under N / N+1 topic. I know this
"mangling" is not directly about sed. But I think it
matters that everyone sees syntax as a poster intended.

Thanks for sending screenshot of your MUA. I repeated
the test from a test yahoo group I made. Email sent to
Thunderbird email client displays the syntax correctly.
So not a total failure. MUA works. But here's the
apparent score so far, for three environments:

A) yahoo groups archive mangles some syntax.
B) yahoo email reader mangles some syntax.
C) MUA (Thunderbird and at least one other) works.

Two out of three tested environments seem subject
to failure. Kind of a problem... Especially because
A) should be the archive, the common reference.

The basic cause seems to be yahoo "knows about"
HTML tags and entities, is not a text-only thing.
But our discussion are really text-only (I think).

I have no idea how often this arises in practice.
Certainly caused me confusion in viewing (on web
site) N / N+1 replies. The replies seemed way off.
Then I looked back at mine, saw the mangling...

A partial workaround might be "don't use LT and GT
as word delimiters when posting", use \b instead.

Here's some tests I devised. If anyone wants, you
can see what gets mangled in your environment. In
my tests, 2, 4, 6, 7, 16, 17, 18, 19 don't display
properly for environments A) and B). Dan

#1 - Using \bfoo\b syntax:

s:(.*)\bfoo\b(.*)\n(.*)\bbar\b(.*):\1FOO\2\n\3BAR\4:

#2 - Using \< \> (BackSlash-LT BackSlash-GT) syntax:

s:(.*)\<foo\>(.*)\n(.*)\<bar\>(.*):\1FOO\2\n\3BAR\4:

#3 Using HTML entities for GT and LT:

s:(.*)\&lt;foo\&gt;(.*)\n(.*)\&lt;bar\&gt;(.*):\1FOO\2\n\3BAR\4:

#4 \<foo\> is foo surrounded by BackSlash-LT and BackSlash-GT.
#5 \&lt;foo\&gt; is same as #4, with HTML entities for LT and GT.

#6 &bull; is HTML bull entity. Displays as bullet or as literal?
#7 &nbsp; is HTML nbsp entity. Shows space or literal?

#8 < is LT all by it's lonesome.
#9 &lt; is also LT, but as HTML entity.

#10 > is GT all by it's lonesome.
#11 &gt; is also GT, but as HTML entity.

#12 \< is BackSlash-LT all by it's lonesome.
#13 \&lt; is also BackSlash-LT, but LT as HTML entity.

#14 \> is BackSlash-GT all by it's lonesome.
#15 \&gt; is also BackSlash-GT, but GT as HTML entity.

#16 <strong>Hi</strong> is Hi surrounded by "strong" tags.
#17 <i>Hello</i> is Hello surrounded by "italic" tags.

#18 <br> is HTML br tag. Shows as literal?
#19 <p> is HTML p tag. Shows as literal?

Well, I viewed it with a MUA, not using webmail.
Screenshot here: Loading Image...
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Hope this helps.
--
D.

[Daniel]

Here are more pictures showing the problems I saw
last month. I finally got back around to this.

To make it more clear, I've marked (encircled) the
problems. In the three cases showing the mangling (the
glitch), you will see "foo" drop out, or you will see
HTML interpreted, instead of literally displayed.

1) Reading post in yahoo mail:
Loading Image....html

2) Reading post on sed-users website, chrome browser:
Loading Image....html

3) Reading post on sed-users website, firefox browser:
Loading Image....html

These problems are NOT currently occurring. In other
words, yahoo groups is now displaying the literal text
characters, instead of trying to interpret characters
as HTML. I'm glad it's working now.

This was not some local problem with my computer or
browser. As I mentioned before, I tested with three
separate devices: windows xp computer, windows 7
laptop, ipod touch. On each device, I tested with
several browsers. In February, the mangling occurred
on every device/browser combo. Now no problems occur.

http://help.yahoo.com/kb/index?page=content&id=SLN4391&actp=search&viewlocale=en_US&searchid=1364167824501&locale=en_US&y=PROD_GRPS
is a cryptic yahoo "help" message which suggests the HTML
display may somehow depend on the egroup attachment setting.

However, the same problems were occurring in a test group
I created, and the problems also went away on the test group.
And I didn't change any settings. So I don't think the
glitches were caused by any setting on sed-users group.
Or maybe our group moderator figured something out and
fixed the problem...

Anyway, nothing to be done at this point. My best guess
is somebody at yahoo screwed something up, it got reported,
it got fixed up. I just wanted to document this in case
it ever comes up again, because it was so confusing and
frustrating when sed syntax was being incorrectly displayed.

Post by Daniel
I started a new topic (should have done this before),
because this has diverged from N / N+1 topic. You can
see previous posts under N / N+1 topic. I know this
"mangling" is not directly about sed. But I think it
matters that everyone sees syntax as a poster intended.
Thanks for sending screenshot of your MUA. I repeated
the test from a test yahoo group I made. Email sent to
Thunderbird email client displays the syntax correctly.
So not a total failure. MUA works. But here's the
A) yahoo groups archive mangles some syntax.
B) yahoo email reader mangles some syntax.
C) MUA (Thunderbird and at least one other) works.
Two out of three tested environments seem subject
to failure. Kind of a problem... Especially because
A) should be the archive, the common reference.
The basic cause seems to be yahoo "knows about"
HTML tags and entities, is not a text-only thing.
But our discussion are really text-only (I think).
I have no idea how often this arises in practice.
Certainly caused me confusion in viewing (on web
site) N / N+1 replies. The replies seemed way off.
Then I looked back at mine, saw the mangling...
A partial workaround might be "don't use LT and GT
as word delimiters when posting", use \b instead.
Here's some tests I devised. If anyone wants, you
can see what gets mangled in your environment. In
my tests, 2, 4, 6, 7, 16, 17, 18, 19 don't display
properly for environments A) and B). Dan
#4 \<foo\> is foo surrounded by BackSlash-LT and BackSlash-GT.
#6 &bull; is HTML bull entity. Displays as bullet or as literal?
#7 &nbsp; is HTML nbsp entity. Shows space or literal?
#8 < is LT all by it's lonesome.
#9 &lt; is also LT, but as HTML entity.
#10 > is GT all by it's lonesome.
#12 \< is BackSlash-LT all by it's lonesome.
#13 \&lt; is also BackSlash-LT, but LT as HTML entity.
#14 \> is BackSlash-GT all by it's lonesome.
#16 <strong>Hi</strong> is Hi surrounded by "strong" tags.
#17 <i>Hello</i> is Hello surrounded by "italic" tags.
#18 <br> is HTML br tag. Shows as literal?
#19 <p> is HTML p tag. Shows as literal?

Well, I viewed it with a MUA, not using webmail.
Screenshot here: http://img248.imageshack.us/img248/6656/email2y.png
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Hope this helps.
--
D.