Tales from the Smart Client

John Gossman's observations on Avalon development

Enabled comments

Curious how long before I get comment spam.
Published Saturday, January 22, 2005 6:54 PM by JohnGossman

Comments

 

Nick Parker said:

Why not add a CAPTCHA control, this seems to do the trick. I've recently done this on my site to control comment spam.
January 22, 2005 8:24 PM
 

Matt said:

I could offer you a Rolex if you want
January 22, 2005 11:24 PM
 

Bearded Geek said:

Solution to >99% of BlogSpam: Captchas + Greylisting + Time-limitations

CAPTCHAS

These are the weird images you see - a mixture of alphanumeric characters on LSD. Computers are extremely bad at parsing them, humans are OK. For just avoiding 'nuisance' you could quite possibly limit it to just a couple of letters, especially when combined with

TIME-LIMITATIONS

Set minimum time limits both for posting after first viewing a page, and on subsequent postings, based on subnet addresses.

GREYLISTING

Get one of the free (preferably Free) Greylisting programmes: SpamAssassin is an excellent first choice. (Yes, it can even be used with LookOut or LookOut Express).

Feed it a copious quantity of ham posts and spam posts.

Set your blog to 'moderate everything'. Let anything in the top 30% be automatically sent through. Everything in the next 50% gets scanned once a day. The rest gets binned. (Obviously customizable!)

Amalgamate your ham and spam databases with other like-minded folk to get the Bayesian filter working even better. Remember to feed your filter regularly (and largely, automatically!) to keep it up to date. Try experimenting with greylisting for non-spam purposes, it's very useful!

CONCLUSION

Blog spam is largely a solvable problem. The rel=nofollow thing is cute but it doesn't actually prevent blogspam, it merely makes it less profitable.

RANT

blogs.msdn.com is very visible, and yet MS, a software development company, didn't even set something this public well first time around. It doesn't really give much confidence in the management, and speaks volumes of the arrogance that comes with monopoly-based market share. For God's sakes, allow nested comments, sort out the blogspam, have a well separated TrackBack section if necessary. It's not hard!!
January 23, 2005 2:40 AM
 

Andrei said:

See? You got some spam already! :)
January 23, 2005 6:58 AM
 

John Gossman said:

This is actually a personal experiment: I literally am counting the hours until I get SPAM.
January 23, 2005 10:46 AM
 

zzz said:

Bearded Geek: "blogs.msdn.com is very visible, and yet MS, a software development company, didn't even set something this public well first time around..."

Well only thing I really find confusing about the whole msdn blogs is why even go bother with the rss xml technology, when even the basics have been forgotten.. Where is the metadata/keywords about the technical subjects talked about? Certainly there can be progress in search engine tech that can reduce the need for proper keywords but if you ask me, from the blogs.msdn start, there should have been some "policy" of putting proper keywording to technical topics talked about - just to show that even if rest of the world is too lazy abouut it, we here at MS will do it.

I am referring to the J.Paoli videos - or perhaps I've understood them wrong, as in "no need for metadata"?

http://channel9.msdn.com/ShowPost.aspx?PostID=33709
http://channel9.msdn.com/ShowPost.aspx?PostID=33806
January 23, 2005 1:13 PM
 

Troy said:

8 days is what I typically see. I turn off comments after 7 days, on the 8th there shall be comment spam in the form of Poker Tables.
February 10, 2005 12:31 PM
New Comments to this post are disabled

© 2009 Microsoft Corporation. All rights reserved. Terms of Use  |  Trademarks  |  Privacy Statement
Microsoft
Page view tracker