[Mailman-Developers] Re: GET vs POST (was Re: subscription confirmations)

Les Niles les@2pi.org
Wed, 18 Jul 2001 11:54:36 -0700


On Wed, 18 Jul 2001 08:27:15 -0700 Chuq Von Rospach <chuqui@plaidworks.com> wrote:
>On 7/18/01 7:56 AM, "Les Niles" <les@2pi.org> wrote:
>
>>> There is much reason not to comply with the published standard: people
>>> are stupid.  Shame, isn't it?
>> 
>> I'm not sure which stupidity you're talking about.  Are you
>> concerned that people will launch the link from the email but then
>> not push the "confirm" (or "cancel") button?
>
>I won't use the word "stupid" -- because it's not really the right word. But
>unskilled, or naïve is.

Absolutely -- I would never refer to the set of users that have problems
with the mailing lists as "stupid."  Unskilled, or unfamiliar with this
electronic communications paradigm, yes, but the individuals are almost
never truly stupid (which is a gross and mostly irrelevant measure
anyway).  I happen to run mailing lists where most of the subscribers
have demonstrated some significant technical savvy, but nothing to do
with computers.  A small but non-zero percentage just couldn't deal with
the mail-back confirmation, even with a lot of handholding.  That's why I
hacked up the confirm-by-GETting patch in the first place.

One of the not-uncommon characteristics that I've noticed about novices
is to adopt a semi-random reading/browsing style.  That is, given a page,
whether a web page or an email message, containing chunks of text and
various links, they'll tend to read the text blocks and visit the links
in a somewhat random order.  This is not unreasonable given the design of
a lot of web pages, and is almost encouraged by the hypertext model in
the first place.  Since links in email are rendered as just a textual
URL, such a user will have about a 50-50 chance of clicking the
confirm-subscription link before reading about what it does.  I guess
that's the main reason I think that launching a page with a couple of
big, fat, clearly-labeled POST buttons to do the actual confirm/don't
confirm is a better UI.  Unix hard-cores like myself -- the sort who use
aliases because the 3-character command names are too much to type --
will chafe at the extra mouse click that's required, but we'll deal with
it.

  -les