For the past few months, I've been reading articles from the Notices of the AMS, something I never did earlier. This is a direct consequence of reading In Theory and Ars Mathematica, both of which will regularly highlight articles from the latest issue of the Notices.
I'm amazed at how many of the articles I find interesting enough to peruse, or even print out to read offline. Most of these are technical articles, on specific topics in mathematics. Quite often, there will be discussions on topics tangentially related to theoryCS or even things that I'm interested in. For example, the May issue of the Notices has an article titled, "How to generate random matrices from the classical compact groups". Anyone who messes around in geometry and graphics will immediately see the connection to the standard problem of generating a random rotation, and in fact there's a very interesting explanation of the group theory underlying the standard procedure for generating a random orthogonal matrix.
A second article with the whimsical title, "If Euclid had been Japanese" discusses the constructibility of points using folding operations rather than "Euclidean" constructions. Apart from the obvious origami connections, this is particularly interesting to me because I've been experimenting with a short presentation that tries to explain algorithms via origami (folds are like steps in a program, etc..).
I could go on like this. Every issue has at least one article that is both acecssible and interesting to me, and I'm not even a mathematician ! Why can't we have such a delightful magazine for computer science, or even for theoryCS ?
SIGACT News does a valiant job. And it's hard work to manage a newsletter, along with all one's other activities. But it comes out far too rarely, and one doesn't often find the kind of short vignettes that entertain and illuminate at the same time. I enjoy the algorithms column and the geometry column. I will make valiant attempts to read the complexity column, but it's often too "structural complexity" for me. But in a way I am unable to articulate, I find SIGACT News somehow less stimulating than the Notices. I wonder if others share this view as well ?
Update (5/3/07): I think I just had an epiphany. The above post is the equivalent of my standing in front of the Model T, wondering why my horse-buggy can't go any faster. In other words, we already have excellent publications that write expository surveys and cute vignettes that connect the community together. They're called blogs !!
Maybe you can find something interesting in the Bulletin of the European Association for Theoretical Computer Science
ReplyDeletehttp://www.eatcs.org/publications/bulletin.html
deid
Why can't we have such a delightful magazine for computer science, or even for theoryCS?
ReplyDeleteEasy, people quietly optimize for do what is rewarded and avoid what it isn't.
Ask yourself: when was the last time the hiring/promotion committee at your university attached any value to a well written article to SIGACT News? and henceforth when was the last time *you* wrote a survey or nice article for SIGACT news?
Well, to answer the second point first, (a) I've never been invited, and (b) I write numerous posts on this blog that are mini surveys in their own right.
ReplyDeleteAs far as optimization goes, at least the NSF does pay more than lip service to outreach and educational efforts. That is one place where the rewards are concrete. Secondly, what is tenure for if not for the ability to do things that don't have immediate reward ? I give Lance a lot of credit for kick-starting theoryCS blogging with very high quality posts, many of which could have been expanded into survey articles.
Maybe rather than complaining about the lack of publications, I should be happy about the proliferation of blogs !
I've never been invited
ReplyDeleteMost SIGACT column editors frequently print an open invitation for submissions.
I write numerous posts on this blog that are mini surveys in their own right.
We are talking about SIGACT News-style (or even AMS notices-style) surveys, not informal short minisurveys.
at least the NSF does pay more than lip service to outreach and educational efforts.
A good SIGACT survey takes about as long to write as a SODA/STOC/FOCS paper. Does NSF consider those two equal? I rest my case.
what is tenure for if not for the ability to do things that don't have immediate reward ?
Tenure allows you to do things that have later rewards (e.g. proving Fermat's Last Theorem). It still gives you no room to do things for which there is no reward.
I give Lance a lot of credit for kick-starting theoryCS blogging with very high quality posts, many of which could have been expanded into survey articles.
I'm thankful for the blogs and they surely have provided a much needed forum for the community to exchange views. I'm doubtful as to their impact. Furthermore, as good as Lance's postings were (I surely miss them, though I'm sure Bill will do fine) they can't be compared to AMS notices.
I think the eco system of math is
ReplyDeletemuch larger compared to CS and
they are all "theoreticians". And
more value is attached to
explaining difficult things and
coming up with new proofs for
known theorems. This is what happens
to a mature and large discipline.
CS is not there yet.
It's a little unfair comparing the Notices with SIGACT News. One is the premier publication of a large society; it has a staff of professional editors and a substantial budget. The other is the newsletter of a small special-interest group, created almost entirely by volunteers.
ReplyDeleteThe more appropriate comparison would be between the Notices and the Communications of the ACM -- another professionally edited publication with a staff and a budget. And when I look at what CACM has become in recent years, I echo your question: "Why can't we have such a delightful magazine for computer science?"
Hi Brian,
ReplyDeleteactually when I started writing this article, I had CACM in my mind as a target. In fact I was planning to start a series of Brad Delong-ish "Why oh Why can't CACM be a real magazine" articles :). But I've given up on CACM, and to a lesser degree on the ACM. ACM's identity is as a trade organization, and CACM is its main magainze, so I can't hope that CACM will ever go back to its glory days.
But you're right about SIGACT News. Maybe I am being unfair to compare the two. I actually didn't know that the Notices was a professionally run magazine.
Suresh,
ReplyDeleteI love browsing through/reading the Notices of the AMS too. It is a great magazine, packed with high-quality contributions. Did you ever read Wigderson's apology for Sudan's work that appeared in the Notices? If not, see
http://www.ams.org/notices/200301/comm-wigderson.pdf
I really think that the Bulletin of the EATCS is the closest thing to the Notices the TCS community can offer. Moreover, like the Notices, and unlike SIGACT News, it is now freely accessible to everybody. (This is a very good decision that the EATCS council made at ICALP 2006 in Venice.)
The BEATCS covers pretty much the whole spectrum of TCS, and I believe that, under the creative editorship of Vladimiro Sassone, it is turning into a high-quality magazine.
I encourage the readers of your blog and you to contribute pieces to the BEATCS. I am sure that I speak on behalf of all of the column editors for the BEATCS, when I say that you are most welcome to contact us with suggestions for surveys and articles.
Keep up the good blog!
Luca
P.S. Selected columns from the BEATCS are often turned into book chapters appearing in thematic volumes edited by World Scientific Press. This could be a reward for writing a column for the BEATCS.