FSP 1.1

blog coment spam

Comment spam is raging here on FSP these days.
There are from 1-5 legitimate comments a day, but now
over 500 total comments.

Up to now, I just reviewed the 'last five comments' on
Moveable Type's administrator interface, and any comment
with a suspicion-raising author name I look at and
depending on the comment's actual content, I likely delete it.

But moderating 600 comments a day is out of the question.
Time to look for a new strategy before turning comments off.

Here's a typical view of the administrator view today:

See also Invasion of the Spambots, by Sam Williams [Salon].

Needed: weedkiller

Wide open paved spaces, with no ditches, trees, walls, lamp posts
or other such hazards are great for developing and testing
car handling and driving skill.

Former and downsized airforce bases with sturdy concrete apron
pavement are particularly excellent venues. However, they still
need maintenance, crack filling, and weed control.

From the Southern Indiana SCCA:

Looks like lawnmower racing to me.

Auto Correction for Urls: strip the < >

Web browsers should be able to strip out surrounding < brackets > from URLs.
For example, take < http://omor.com/b/ > and convert it to
http://omor.com/b/ .

Many browsers have an URL auto-correction and auto-completion feature:
if the user-specified URL fails [*], the browser generates a derivative
list of similar URLs, and tries them.

For example, if you enter http://omor or even just omor in your
browser's URL box, the browser cycles through (not necessarily in this order):

http://omor.net
http://omor.org
http://omor.com

etc, and when http://omor.com succeeds, the browser loads that page
and stops.

An excellent complement to these auto-correction and completion features
would be an auto-deletion feature which strips out spurious < > , automatically
fixing
< http://omor.com/b/ > to
http://omor.com/b/

Why this matters: most URLS I receive via a-mail are padded with surrounding
< brackets > .

[*] If the domain register hijacks unregistered domains, the trial-by-error
nature of testing a sequence omor: http://omor.net; http://omor.org
http://omor.com ... will fail, because the incorrect URLS
will generate a valid web page, rather than the 404 error they should.

See 404 Controversy: Verisign address dns redirect finder.