Disadvantages of using Solver Foundation for constraint programming

What are the disadvantages of using Microsoft Solver Foundation fo开发者_运维技巧r CLP? Solver does have some support in the Express/Standard versions, but would imagine that one would need to buy the expensive Gurobi / Knitro add-ons to accomplish anything more than the most basic constraint programming.

Purely considering it's CLP capabilities, how does Solver compare to ECLiPSe?

The main disadvantage is that Microsoft Solver Foundation is discontinued as a standalone product as mentioned here:

As users have pointed out, Microsoft has not been active on the Solver Foundation forums since Nate left. We have been quiet while we have gone through restructuring and planning. Some would say we have been too quiet. We know we have very loyal and enthusiastic users who want to know the future of Solver Foundation. So, here is a long overdue statement about our plans for Solver Foundation.

The current 3.1 release of MSF will be the last release as a standalone install. We are working hard on integrating Microsoft Solver Foundation into a larger analytics framework that will help users build both prescriptive and predictive analytics. We look forward to releasing this new product for your use as soon as we are able to do so. This new product will provide a migration path for current Solver Foundation users and partners.

We would like to continue to keep the current forum open to the community to discuss MSF until the release of the new product. However, Microsoft will be providing limited support of MSF in terms of monitoring the forums and providing bug fixes during this transition time.

We have been responding to email and will continue to do so. If you have feedback on issues/bugs/improvements, we welcome your feedback via msfsupport@microsoft.com. Please check back on the forum for future announcements with regards to the new analytics product.

ECLiPSe project on the other hand seems to be a much more established project in the area of constraint programming with substantial number of publications. It is also open source, so there is no possibility for a vendor lock-in.

I'd also recommend considering:

• “Logic” and Constraint Programming Extensions for AMPL
• Google CP Solver

Re: model sizes, For Solver Foundation Express (basic 'free' version):

• Linear or quadratic programming 50,000 non-zeros
• Mixed integer programming - 1,000 variables, 1,000 constraints, and 5,000 non-zeros
• Constraint programming - 5,000 total terms
• Non-linear programming - Unlimited

The Standard version has roughly double the capacity, and the Enterprise/Academic versions are 'unlimited' with no real restriction on core/CPU's used.

Re: advantages A huge (IMHO) advantage of Solver Foundation is the object model approach to defining your models. Makes for easy-to-setup models programmatically in C# or to any language to which you have bindings, or you can use AMPL and MPS inputs if you want. There is a also a custom model definition language (OML) which I haven't used, but looks pretty good.

Re: disadvantage You're tied into a proprietary setup, if that matters to you. For a GNU LP, check out GLPK. I've thrown some >20K contraint models at it no problems, and it has a pretty active user group.

From: Installing Solver Foundation