Hey folks,
One of our users here posted his question to the Computational Science StackExchange web site (“SciComp” for short). There he received some useful answers from people who are not part of the CVX community. I expanded upon my answer there as well.
I want to invite everyone who visits this site to consider joining that site as well. While this site is focused specifically on questions related to CVX and TFOCS, the SciComp site welcomes more general questions of computational science interest. And because it has been around longer than the CVX Forum, it has more active users. For example, Johan Löfberg, the developer of the well-respected YALMIP package—another MATLAB-based optimization modeling framework—is a frequent visitor to SciComp. I intend to visit frequently as well.
Please note that I am not trying to discourage anyone from posting their CVX or TFOCS questions here. This remains the first place to go if you’re having difficulty with CVX and TFOCS and it is not a bug report (which should be going to the CVX Support site instead, of course). But if your problem is more generally with convex optimization or modeling, then perhaps SciComp is a good option.
The SciComp moderators have created a special “CVX” tag that you can use to tag posts. So for instance, to see all of the CVX-tagged questions on the site, you can follow this link http://scicomp.stackexchange.com/questions/tagged/cvx Of course, there is only one such question at the moment.
A point of etiquette: it’s probably not good manners to post in both places simultaneously. On the other hand, if you post in one place and are not satisfied with the answers received, you can certainly post to the other—but make sure to link to the first post in the new one, so people can see what help you have already received.