Call for Application Build Time Metrics

[CaseyM]

In another thread, I've been trying to find causes of ballooning (among other things) build times and there have been some helpful suggestions, but it strikes me that we, as a community, don't necessarily have a good feel for how long we should expect it to take for large AF projects to build.

Would other be willing to share some rough metrics about their projects and their build times? I'll start with the most recent stats for my current project:

(All number are approximate)

• Total LV Classes: 1350
• AF Actors: 150
• AF Messages: 925
• VIs: 8250
• Type Defs: 903
• Type def clusters used in class private data: Very much
• PPLs: 0
• LabVIEW Version: 2019 (64-bit)
• Platform: Windows 10 (64-bit)
• Other notes: Pretty new hardware with lots of RAM and a fast SSD.
• Application build time: 2 hours

Obviously something like build time can be greatly affected by computer hardware, but if we get enough data, maybe we can find a trend helps us identify root causes of ballooning build times and other IDE performance degradation.

-Casey

[Mike_Le]

• Total LV Classes: 700
• AF Actors: 150 (~130 concrete, ~20 abstract or interface classes)
• AF Messages: 464
• VIs: ~8000
• Type Defs: ~400
• Type def clusters used in class private data: Yes
• PPLs: 0
• LabVIEW Version: 2016 (32-bit)
• Platform: Windows 10 (64-bit)
• Hardware: SSD, i7-7700HQ 2.80 GHz CPU, 16GB RAM
• Application build time: 4 hours
  

Build time is increasingly out of control and unclear to me why. I don't think I have a lot of circular dependencies, but definitely a large amount of static linking that I'm working on untangling with more abstract classes, dependency injection, etc.