View Single Post
Old 12-21-2007   #11 (permalink)
Robert Aldwinckle


 
 

Re: multiple Inboxes

"Ildhund" <jnllb@xxxxxx> wrote in message
news:eHq6dlyQIHA.5288@xxxxxx
Quote:

> "Robert Aldwinckle" <robald@xxxxxx> wrote in message
Quote:

> news:#5DKD9xQIHA.1164@xxxxxx

What a coincidence! WLM obviously needs to escape that #.

Quote:
Quote:

>> "Ildhund" <jnllb@xxxxxx> wrote in message
>> news:O1EdzjvQIHA.3940@xxxxxx
Quote:

>>> It's been established that Quick views is fragile. When it breaks, the
>>> message count doesn't work and it will only show (mail) messages received
>>> after the accident. The treatment has been posted many times: delete
>>> HKCU\Software\Microsoft\Windows Live Mail\SearchFolderVersion, and the
>>> prophylactic: don't delete any folder under Storage folders.
>
Quote:

>> Yet the latter is what Ron Sommer seems to be suggesting as the first step
>> in removing unwanted messages from the ("Quick View" of) All News <eg>
>
> There's a subtle difference. What breaks the Quick views is deleting, say,
> an empty Recovered items folder showing under Storage folders in the Folder
> pane - i.e. using the UI. Ron very bravely used Explorer to delete the
> newsgroup folder found in the Store folder, but found that a subsequent
> compaction - reindexing of the MessageStore - brought everything into line
> again. If this had caused a(nother) malfunction in Quick views, I think he'd
> have said so.

Yes, of course. Did you notice my smirk? ; )

Quote:

>
Quote:

>> news:%23xW9xtpLIHA.3940@xxxxxx
>> Subject: Re: Unwanted newsgroup postings
>> Date: Sat, 24 Nov 2007 07:10:49 -0600
>>
>> BTW I would have sworn that I had posted that news link correctly
>> yesterday
>> but somehow in Sent Items I found the %23 converted back to a #
>> (which won't work if you try to use it from the IE Address bar.)
>> In any case I am sending it today with %23. If it changes again
>> there is a bug somewhere.

So there *is* a bug in WLM about this. See the above Message-ID that you replied to. ; )
I have just gone back to repeat the Ctrl-Shift-F Find and Ctrl-g on the one from Ron that started
this little digression and confirmed that that would be how the error occurred in my original post.
Now that I am aware of the bug I will have to remember to do manual conversion of that character
even when I use Ctrl-g to generate a news: link (which rather defeats the purpose of going
to the trouble of doing a Ctrl-g in some cases anyway.)

Quote:

>
> That one works fine for me with its %23. Of the two in the preamble to this
> post, yours (which has a #) doesn't work, but mine (with neither # nor %23)
> does. What's your explanation for that?

# is one of the characters which must be escaped to be used as a URL.
Your Message-ID didn't contain any characters which required escaping.
Mine did.

Here specifically is the source of the problem:

http://www.rfc-archive.org/getrfc.php?rfc=3977

(Live Search for
RFC news NNTP "Message-ID"
)

<quotes>
9.8. General Non-terminals

message-id = "<" 1*248A-NOTGT ">"

A-NOTGT = %x21-3D / %x3F-7E ; exclude ">"

</quotes>

Elsewhere in the document that specification was interpreted as:

<quote>
For the purposes of this specification, message-ids are opaque strings
that MUST meet the following requirements:

o A message-id MUST begin with "<", end with ">", and MUST NOT
contain the latter except at the end.

o A message-id MUST be between 3 and 250 octets in length.

o A message-id MUST NOT contain octets other than printable US-ASCII
characters.
</quote>


Now you have to compare that spec. with this one:


http://rfc.sunsite.dk/rfc/rfc1738.html
(Live Search for
RFC news NNTP "Uniform Resource Locator"
)


<quotes>
url = httpurl | ftpurl | newsurl |
nntpurl | telneturl | gopherurl |
waisurl | mailtourl | fileurl |
prosperourl | otherurl


newsurl = "news:" grouppart
grouppart = "*" | group | article
group = alpha *[ alpha | digit | "-" | "." | "+" | "_" ]
article = 1*[ uchar | ";" | "/" | "?" | ":" | "&" | "=" ] "@" host

alpha = lowalpha | hialpha
digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
"8" | "9"
safe = "$" | "-" | "_" | "." | "+"
extra = "!" | "*" | "'" | "(" | ")" | ","

escape = "%" hex hex

unreserved = alpha | digit | safe | extra
uchar = unreserved | escape
</quotes>


E.g. notice that the NNTP spec permits # in its definition of Message-ID
but the URL spec does not. Hence it has to be escaped to be used
in a URL, which is what prefixing the Message-ID stripped of its angle brackets
is supposed to be.

BTW this shows that $ is supposed to be safe but I think that I have seen
that OE usually escapes it too. E.g. $ represented in news: links by %24

Quote:

>
> I hope you noticed my post yesterday claiming that news: links invoke the
> default news server, not the one currently active.

Yes. That's the same as OE.

Quote:

>
> Whilst we're at it, why does your post have no Content-Type or charset
> header when mine do?
> --
> Noel

Probably a difference in our Send options.
I suspect it is just that for News, Plain Text I have UUEncode selected, not MIME.
Also ICIM I do not use Reply to messages in the format in which they were sent
and in International Send Settings I have: Western European (ISO) for my
characterset and I have checked: When replying to messages always use English
headers.


Robert
---


My System SpecsSystem Spec