Hi Chris, this is a good idea, indeed. But, what we should also consider is that the cost function could vary between protocols and drivers. So this is something we have to keep generic.
Julian Am 18.06.19, 11:09 schrieb "Christofer Dutz" <[email protected]>: Hi all, as I mentioned, I am trying to get a colleague of mine, Veronika, on board with our team. She has a Data Science background and I think she would be a great addition to our little team. Currently only the S7 driver is able to request multiple items in one request. But the optimization I implemented there is very strictly tied to the S7 driver and can’t be used outside. So with all the code-generation and hereby theoretically all protocols being implemented the same way, we would like to try to come up with a more generic approach. This will probably require modifications of the generated code. For example would it be good for us to not only provide “getLengthInBytes”, but also “getEstimatedResponseInBytes” for some objects. In general we discussed, that it would be great to model this optimizer as layer between the PLC4X layer and the protocol layer, so it would translate the PLC4X objects into optimized protocol objects. I’ll be creating a container project in the sandbox/code-generation for her to start working on. Chris
