> >Wait, *everyone* cannot write survey papers. There aren't enough
> >sub-disciplines.
> I disagree. Most areas in which I've performed research suffer from a
> lack of good surveys, rather than a surfeit.
Now that you mention it, I realize this is true. Dang.
> >I think he's nuts. Most survey papers I've read have been awful.
> Must be a CalTech phenomenon :-), most of the ones I've read have been
> very good.
I poured through some of the _ACM Computing Surveys_ back issues
today, and I gotta agree with Jim. Almost every paper in that journal
is solid and well-written.
> >This idea is really batty. Why should we trust graduate students to
> >write authoritative surveys on anything???
> Um, because nobody else has enough free time?
Another good point. Plus, grad students are hungrier than anyone to
produce good stuff.
> >I have enough trouble getting comments out of people under the
> >current circumstances. If people are continually having to read survey
> >papers too I will never get a free moment of their time.
> Agreed. Just make sure their advisor has read it, then put them out as-is.
> You'll quickly discover which are gold, and which should be composted.
Right. Plus a good editorial review board can work filtering magic as well.
> On the other hand, nobody reads ACM Computing Surveys as it is (although
> every time I do read a paper from there I learn a *lot*)
Same here.
> so I doubt others will read these surveys much either. But, on the
> off chance someone has too much time on their hands, these surveys
> should be more accessible.
I just want to say that writing a survey paper is a major pain. Rohit
and I are dying of exhaustion from our trust paper
http://www.cs.caltech.edu/~adam/LOCAL/trust.html
but it within epsilon of being finished (Rohit: I'm now done with 1-4;
6 and 7 are in the homestretch; and 8-10 are done now, too) and I'm
becoming quite proud of it...
----
adam@cs.caltech.edu
Ah yes, ketchup. God's other food. Oddly, it's not wonderful on
cranberry sauce.
-- John Dobbin