According to KB00740 - PI ACE support for Visual Studio 2012, 2013 and 2015 the ACE Wizard will no longer work and be supported in Visual Studio 2015
That's quite a statement.
That would mean I would not be able to create an new ACE module after mainstream support for Visual Studio 2013 ends, in 2019.
Or when we move all software development to Visual Studio 2015, hopefully before 2019.
Will Osisoft provide another tool to replace the ACE Wizard or does this mean the PI ACE will be End of Life.
根据KB00740-《 PI ACE对于Visual Studio 2012,2013 和2015的支持 》,ACE 引导将不再被Visual Studio 2015支持，以及不能正常工作。那是一个声明，意味着我们将在2019年主流2013支持结束后，不能再创建新的ACE模块。或者说，我们还有希望在2019年前，迁移所有的开发到VS2015平台。 OSIsoft将提供另外一款工具替代ACE或者这就意味着PI ACE生命的结束。
Yes, you are right. PI ACE won’t be supported in Visual Studio 2015. We have another product to replace PI ACE called PI Asset Based Analytics. Product management found out that the majority of the use cases of PI ACE can be migrated to PI Asset Based Analytics even considering the fact that this product currently does not provide a real programmatic experience.
Remember that PI AF SDK is now our preferred data access product. It has better performance and more options than PI SDK. PI ACE uses PI SDK under the hood while PI Asset Based Analytics utilizes PI AF SDK internally.
If you feel that PI Asset Based Analytics won’t solve your problem, creating a Windows Service which uses PI AF SDK in a good way to go. This is what we do with our own internal projects.
Hope it helps!
Let me jump in here for a moment. As Marcos already mentioned, and in case you're not aware, we released a new calculation engine in 2014 that is tightly integrated with AF. We call the new feature "asset analytics". It is not a product by itself, but rather an added feature to AF. This of course allows us to leverage the features and capabilities of AF for context, UOM conversion, consistency (via templates), etc. The data that we have is that many of our customers use ACE primarily for the scheduler and it's not absolutely necessary to do the math in ACE. The feedback we have received is that many of our users are migrating from ACE over to asset analytics as they much prefer a configuration experience over a programming experience and the performance of asset analytics is acceptable for their needs. This of course assumes the provided functionalities in asset analytics satisfy their needs.
Meanwhile, our current recommendation is that if you are using ACE and the new asset analytics does not satisfy your needs, please contact product management (that would be me ) and we would evaluate your needs. If your needs are applicable to a wider customer base, we would consider adding it to our standard product. This accomplishes many things - allows us to better understand, in details, what our customers are trying to do and to make it available to a wider customer base as a supported feature if it makes sense, help our customers to move to a configuration experience rather than a programming experience for ease of use, and (selfishly) decrease our support load. It's very difficult to provide support when given a bunch of ACE codes that don't work to the customer's satisfaction and we have to figure out why.
Lastly, we fully understand there will be situations where a programming experience is required. We do have a backlog item to provide extensibility to asset analytics. However, I would like to use customer requirements to drive our backlog priority and if the need is there, we would move up the priority.
So while we're on this subject, what is preventing you from moving your ACE code over to asset analytics? Would you like to share you use case? You can do it privately if you don't want to post it in a public forum - just send me an email.
We are currently on PI 2012, that's what stopping us at this moment.
According to Life cycle management, we should upgrade next year to a newer version.
AF is installed, but we are not really using it at this moment.
We use ACE, and the MDB, quite extensively for calculations and the configuration of ACE contexts.
PI ACE itself is used mostly for scheduling and one of the advantages is that some of our end users can create, maintain, start and stop contexts using the Pi ACE Manager and PI SMT or Module database builder without the need of going to the "IT department".
We don't really use the ACE wizard a lot, only for creating new ACE modules. We have other tooling for testing and debugging and I think I already know what to do to create a new ACE module using PI SMT. Technically we are probably covered
Some of the simple calculations can be easily converted to Asset Analytics, but there are also calculations with 20+ inputs and as many outputs.
Again, technically, it would probably be possible to convert these into the Expressions of an Analysis, but I'm not sure of the readability and maintainability of such calculations. We would have to look into that.
As far a I know there is no end of support date for PI ACE it self.
But if the tools for creating and maintaining PI ACE modules are phased out that should be a trigger for me, as a developer \ application manager, to warn my (internal) customer of this change and discuss how we should deal with it and give some options.
I would rather plan these changes somewhere in the next 4 years, than rush these changes in the last few months because Visual Studio 2013 has to be removed for development machines for some reason.
That was the main reason for my question, to make sure it is placed on the Application Life cyle Management calendar for the correct reasons.
该文档由下列讨论生成：ACE wizard and visual studio 2015