How accurate are timers?

EvanReese

I was looking into using startTimer() and printTimer() in a feature, and realized I could call them in separate features, which could be a great way to profile my feature tree with more detail than the "Regeneration times" panel, since I could "start" and "check" my timers to evaluate sections of the tree. In some preliminary testing it seems like the values I'm getting from printTimer() are a lot different than the regen times though. Can someone explain why?

ilya_baran

Yeah, startTimer and printTimer don't work when they're in different features because they're measuring wall clock time, and we don't always regenerate from the beginning.  What happened was you edited some feature in the middle and we regenerated starting from that point some time later, and printTimer printed the time elapsed since the start of the previous regeneration.  What are you trying to do?

ilya_baran

Yeah, startTimer and printTimer don't work when they're in different features because they're measuring wall clock time, and we don't always regenerate from the beginning.  What happened was you edited some feature in the middle and we regenerated starting from that point some time later, and printTimer printed the time elapsed since the start of the previous regeneration.  What are you trying to do?

EvanReese

Ah okay this isn't likely to work as well then. I was seeing if it might be possible to make a Feature Timer feature so I could get more granular Regen data about my model. This would especially help to profile a part studio when it's not one egregiously large whale of feature bogging it down, but instead a bunch of little piranhas that don't rank high on the Regen list individually, but still add up to a lot, like a folder with 20 fillet features.

MBartlett21

@Evan_Reese
One way you might be able to work around it is to force a full regeneration (make an empty sketch or some other feature and drag it up to the top of the feature tree).