[R] E-Mail/Post Threading (was: Bonferroni p-value greater t

(Ted Harding) ted.harding at nessie.mcc.ac.uk
Thu Mar 29 22:39:20 CEST 2007


On 29-Mar-07 19:21:12, Marc Schwartz wrote:
> On Thu, 2007-03-29 at 19:38 +0100, ted.harding at nessie.mcc.ac.uk wrote:
>> On 29-Mar-07 17:15:27, Marc Schwartz wrote:
>> > [...]
>> > Just a quick heads up here, that deleting the body text of
>> > a message or changing the subject line, does not alter the
>> > 'linkage' between posts.
>> > 
>> > There are standards for how messages are 'threaded' and largely
>> > have to do with the e-mail headers, not the e-mail content.
>> > 
>> > A couple of quick references that might be helpful:
>> > 
>> > http://people.dsv.su.se/~jpalme/ietf/message-threading.html
>> > 
>> > http://www.jwz.org/doc/threading.html
>> 
>> This above, of course, is a good reason for not replying to an
>> existing message when you want to start a completely new thread.
>> 
>> However, I'm wondering what is the best way to start a new thread
>> which legitimately branches out from an existing one.
>> 
>> For example, someone posts a message which discusses at length a
>> method of isotonic binary regression, and in the middle of this
>> describes a curious approach to obtaining confidence bands for
>> the regression. I'm intrigued by the confidence band issue, get
>> some ideas about it, and want to start a new thread to develop
>> just this aspect.
>> 
>> However, to do so I want in the first place to include several
>> quotations from the original message. This, of course, is most
>> easily done by replying to that message -- so that it gets included
>> in the reply -- and editing this included message, and changing
>> the subject.
>> 
>> But that stays in the old thread, which I don't want. Now of course
>> one can copy over the old text into a brand new blank message,
>> and edit it up into a simulacrum of a "reply" -- all the usual
>> "On NN March 2007, XXX wrote:" ... as well as the "> " inclusion
>> markers, etc.. But that could be tedious. Nevertheless, perhaps
>> it is the right thing to do -- unless there's a work-round using
>> the "reply" mechanism?
>> 
>> Best wishes to all,
>> Ted.
> 
> Hi Ted,
> 
> The general approach, if "relatedly digressing" (also described
> in various 'netiquette' guides) is to do what I did here, which
> is reply to the post in question, but change the subject header
> by using "New Subject (was: Old Subject)".
> 
> This enables you to easily engage in the sort of editing that you
> describe, and still links your reply back to the original thread,
> under the presumption that your reply is in some way related to
> the subject matter of the original thread.
> 
> Since most e-mail systems (list managers, MUA's, etc.) thread
> based upon the headers and not the subject, as described in the
> above references, unless you generate a completely new e-mail,
> your reply will be linked to the e-mail and thread to which you
> are replying.
> 
> It's pretty much a dichotomous situation.  Use 'reply' and you get
> linked to the old thread. Use a 'new' e-mail and you start a new
> thread.
> 
> If you are truly moving in a new direction, I would be tempted to
> start a new thread and perhaps to make it easier for readers,
> include a reference/link to the post in question. That way, you
> keep your new e-mail in a separate thread, while 'virtually'
> linking it back to the original that raised your interest.
> 
> HTH,
> 
> Marc

Thanks for the clarification, Marc. Nicely put.

Now that I visit the R-help archives website, I can see that these
"subthreads" get hung off the originals in that display (just as
out present subthread is, as seen there), which is a good way for
it to happen (and ideally suited to the sort of situation which
I described). And the way the subject was changed keeps it nice
and clear as to what happened.

Best wishes,
Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <ted.harding at nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 29-Mar-07                                       Time: 21:39:17
------------------------------ XFMail ------------------------------



More information about the R-help mailing list