It was (predatory) pricing policies like this that led Knuth and the entire editorial board of the "Journal of Algorithms" to resign en masse, and start a new journal with the ACM.
This happened again, more recently, in the machine learning community. Here is the resignation letter from the editorial board of the _Machine Learning_ journal, as they left to start an open journal:
http://www.sigir.org/forum/F2001/sigirFall01Letters.html
I consider the ACM just as bad. Their membership pricing is restrictive for casual users and their conferences are worse.
I also consider their restrictions on who is able to become a member insulting (though they don't apply to me). Wait, so a physicist who's taken up programming in the last year isn't supposed to be able to read papers on computation?
RE: restrictions on who is able to become a member:
"Membership Qualifications
You must satisfy one of the qualifications below:
1. Bachelor's Degree (in any subject area); or
2. Equivalent Level of Education; or
3. Two years full-time employment in the IT field." --- https://campus.acm.org/public/ProfQJ/qjprof_control.cfm?form...
I know a number of OSS contributors that would fall foul of that. Think for instance of somebody that dropped out college and is working in something non-IT, but still likes to sling code. It's just a really useless requirement, especially given that you have to pay $200 on top of it. If there were real, you know, proficiency requirements or something that would be one thing, but anybody who's had any job in IT for two years?
For those interested, the statement from the board can be found at http://www.cs.colorado.edu/~hal/s.pdf, and Knuth's (long, fascinating) letter to his colleagues can be found at http://www-cs-faculty.stanford.edu/~knuth/joalet.pdf