My company currently has a few floating licenses for the Intel C++ Professional compiler for Linux, so I'm familiar with how the licenses are set up for that product. However, we're considering obtaining additional licenses for the IPP libraries for a different set of developers. My questions pertain to how that product is licensed.
If we purchase a set of single-user licenses, it's my understanding that the developers for whom they are purchased can use them on any given machine (linux, of course) as long as they do so on only one machine at a time. I'm responsible for setting up our development environment, so I need to make sure we stay within the bounds of the licensing agreement. If flexlm monitors the license usage, then that means I don't have to take any additional action to do that monitoring. Is flexlm the monitoring mechanism for the single-uers licenses? If not, how is that to be accomplished?
According to the IPP license faq (http://software.intel.com/en-us/articles/intel-integrated-performance-primitives-faq/), the libraries can be deployed "on any number of machines on which the application is built and/or tested as long as there is only the number of licensed copies in use at any given time." I understand how building against the libraries would require a license, but if we're testing our developed software in an environment that simulates that of our customers, then I don't understand how that takes a license. We can redistribute certain libraries according to the licenses terms, and that should be sufficient for our test environments as well. Can someone please clarify exactly what actions require a license?
Lastly, I was able to find a reseller (it wasn't easy) that provides floating licenses for IPP. However, the reseller provides very little information about licensing, and looking through the IPP license faq mentioned above, it specifically states that floating liceses are not available except with the purchase of the compiler.
Can you confirm that the floating licenses for IPP work in a similar fashion to those for the C++ Compiler (ie, they use flexlm; they're machine specific, etc.)?
There is NO license management enforcement for any of our single-user licenses (including Intel IPP, MKL and TBB) at run time , other than for installation. But for Intel C++ Compiler, the flexIM monitors the license usage at runtime.
The EULA does not place any restrictions on the number of copies of the Redistributables that users may choose to distribute.
Hope this clarifies.
The information about the redistributables doesn't really answer the other questions, however. I was not concerned about how many copies may be distributed. What I was asking was how is it that TESTING the application requires a license? I understand how building an app requires one, but it seems to me that testing such apps should not, so I'd like clarification on that statement in the EULA.
I was also asking how the floating license for IPP works. It's clear how it works if you're using the Intel Compiler, since running the compiler will cause it to request a license. But with the IPP libraries, we'd be linking (with gcc) against the IPP libraries; as far as I know, that won't require running any application that would know how to request a license. So how is that accomplished?
I'm probably pushing my luck, but I do have yet another IPP floating license question. As far as I can tell, they're only sold in "2 packs" ... 2 floating licenses per purchase. Because we do work for the government, we have a number of isolated networks: each network has no connection to any other (nor to the "outside" ... such as the internet). Thus, there's no way for a single license server to work on both networks. Some of these networks are very small, and only one or two developers work on them. If we had a "2 pack" of floating IPP licenses, could I split one up and use one license on each of two independent networks?