Define a solution to support a Multi-Server Analysis Solution.

When bulk recalculating, properly order all Analyses. Not just Analyses within one template but across a multi-template recalculation.

Add an option to define dependencies in AF templates for element deployment. For example, if an expected table entry doesn’t exist, notify the user somehow. This will help more novice users to deploy new elements not knowing all of the initial items that needed to exist for Analyses to execute properly.

For the Analysis Evaluate button, allow the user to specify the Evaluate Time (default *) rather than using current values/time.

When trending data in PI System Explorer, have the attribute/description vertical not horizontal. Maybe left or right of trend (configurable). Having 8 attribute names loaded horizontally across the top is not convenient.

PI System Explorer Management Tab should have better search criteria. For example, select specific elements in which to rerun an Analysis. Maybe use Element = something OR/AND Element = something else. Or search by element path.

In the trend, allow hide using the checkbox. Currently the pen is removed entirely. Just hide it and allow me to re-add it via the checkbox and not by right-clicking on the attribute again. Add a remove option also.

Trigger analyses "On Any" but also allow for a MaxTime and trigger after MaxTime has passed. Helpful to keep the timestamp current.

When right-clicking and opening a trend or time series data, don't run the previous query. It can take too long. i.e. I loaded one year of data from a daily total tag (365 values). The next attribute I trend is a one-second tag (31 million values load). Either default to one-hour or load nothing until the user enters the start-time/end-time and clicks refresh.

The bottom trend of the Analysis Preview Results window shows 8 pens. More pens would be better or at least allow me to choose which 8 inputs/outputs I’d like to view.

The bottom right legend of the Analysis Preview Results window needs to expand properly when resizing the trend so we can see the full name of the inputs/outputs. The horizontal scrolling of this section is not convenient.

When the last line of an expression is a comment (i.e. //My Comment Here), an additional variable is added to the expression.

When editing an Analysis Expression, only 4.5 lines appear. It would helpful if more lines appeared (greater than 20 lines).

When hovering over an attribute’s orange diamond, a one-line preview of the Analysis appears. It would be helpful if the name of the Analysis appears on the first line then the details on the second line. Clicking on the first line could bring you to the Analysis directly. We have 100 Analyses. This change would be helpful from a support perspective.

A use case for this request would be to provide a way to overcome the limitations and inconsistent behavior observed when rounding timestamps by PISDK client tools. Considering that AFSDK Clients would provide better subsecond timestamp handling.

The limitations mentioned above are described in the following article: https://customers.osisoft.com/s/knowledgearticle?knowledgeArticleUrl=KB01402

Would like to see the ability to create XML or other flat files from PI Notifications as was available in previous versions.
We have seen cases where the network that the PI system resides on is locked down to the point that only flat files can be transmitted one way, eliminating the ability to connect to an SMTP server and utilize PI Notifications to send information to other systems.

PI EFgen interface is built in such a way that it is not possible to configure events without Windows-Admin rights on the machine hosting the interface. This is not relevant in term of security and we would like to let key-users configure events with only PI-admin rights.

This request is similar to what is available in PI Batchgen interface, where you can configure batches with PI SMT and you don't need any OS-admin rights.

It is currently complex to mainbtain a link betwwen the ARF and Pi Data Archive UOM Databases

The arrays in Expression and Event Frame Generation analyses are useful, but use cases are emerging where more functions are needed to reduce/calculate the arrays in more streamlined fashion. For example, Concat function should support array of characters to construct a string.