Some features requests :)

Nov 9, 2013 at 2:09 PM
First of all, I'd say very nice job on this parser - I really like it.

And onto the request..
Top of the list would be a target view. That is to see how that damage or healing done was divided onto targets.

Another thing I thought would be nice, if one could use regular expressions for the combat log filter.

And finally, not very important and seems more like an omission. Why doesn't the threat generated from taunts appear in the combat log?
Coordinator
Nov 10, 2013 at 2:16 PM
The target healing view would be helpful I will add it.

The missing taunt amounts are a bug with the combat log. They are still used in the calculations. I will fix the display in the combat log in the next patch.

Thanks!
Nov 13, 2013 at 4:52 PM
Thanks for the reply :) Would hope to get the view on damage too though.

Meanwhile, I seem to have come across a bug in terms of effective healing. I assume it is calculated from threat, and I don't know if the guard on me was a stick in the wheel - but it looks odd:

18:22:34.804 You heal Dashira with Kolto Probe for 1503 (488 threat).
18:22:35.027 You activate Recuperative Nanotech.
18:22:35.027 You spend 26 energy.
18:22:35.028 Faerlys gains Recuperative Nanotech.
18:22:35.028 You heal Faerlys with Recuperative Nanotech for 783 (254 threat).
18:22:35.029 Týranous gains Recuperative Nanotech.
18:22:35.030 You heal Týranous with Recuperative Nanotech for 783 (254 threat).
18:22:36.107 You heal Faerlys with Recuperative Nanotech for 783 (254 threat).
18:22:36.107 You heal Týranous with Recuperative Nanotech for 1559 (506 threat).
18:22:36.504 You activate Kolto Probe.
18:22:36.504 You spend 9 energy.
18:22:36.505 Faerlys gains Kolto Probe.
18:22:37.123 You heal Faerlys with Recuperative Nanotech for 783 (254 threat).
18:22:37.123 You heal Týranous with Recuperative Nanotech for 783 (254 threat).
18:22:37.195 You lose Faerlys's Guard effect.
18:22:37.801 You heal Dashira with Kolto Probe for 1503 (1085 effective) (488 threat).

If you look at the top entry and the bottom one, I heal the same amount and generate the same threat - yet am not credited full effectiveness. I was so puzzled about the reported effective healing during the fight that I was trying very hard not to overheal. There are plenty more examples in mentioned log like the one listed above.
Coordinator
Nov 18, 2013 at 9:32 PM
Yes, Parsec uses threat to calculate effective heals. Guard reduces your threat by 25% so it is tracked by Parsec and used to calculate effective healing.

What you are seeing in this circumstance is one of the unfortunate aspects of SWTOR's combat log. Things in the log sometimes occur slightly out of order. It seems to me that things that come from your client vs things that must be verified server side have a chance to be logged at different times than they actually occurred.

Parsec thinks you are no longer guarded when the heal at 18:22:37.801 takes place. However, on the server side this heal must have taken place prior to the guard effect being removed. Parsec calculates your healing coefficient unguarded at .45 instead of .325.

If you were to find a healing entry after 18:22:37.801 you may find that your threat is reported higher by SWTOR since you are unguarded.

Unfortunately Effective healing calculations are subject to this kind of situation because of SWTOR's logging process.
Nov 19, 2013 at 7:23 AM
Thanks for the reply. But looking a bit closer it turns out that there was no heals in that log dealing 1503 pts., that ever created more than 488 threat.
Looking at a different combat log, I found a mix of 488/676 threat for that amount of heal. I noticed that guard places 2 effects on you (even if they can't be told apart from the combat log). So it would seem like it's easy to get out of guard's range, and thus lose the 1 effect, but that doesn't affect threat. Only when losing both effects, does it seem to impact threat. In such cases, however, it does appear like the threat can get adjusted already when losing the first effect.

So while 100% accuracy is likely utopia, I do think it should be possible to get closer. But I would think just reversing it (that is only adjust threat effect if both guard effects are lost), would come a lot closer. I understand if it's no priority and all, just wanted to pass the observations and hope that maybe something can be done :)

From same log as shown in previous post:
[17:56:11.204] [Guard {1776136480620544}] [ApplyEffect {836045448945477}: Guard {1776136480620544}] ()
[17:56:17.251] [Guard {1776136480620544}] [ApplyEffect {836045448945477}: Guard {1776136480620544}] ()
..
Guard is lost and reapplied many times, but never is it lost twice in a row and never does it appear to affect threat.
..

From a different log:
[17:00:33.025] [Recuperative Nanotech {815232037421056}] [ApplyEffect {836045448945477}: Heal {836045448945500}] (783) <254>
[17:00:33.026] [Recuperative Nanotech {815232037421056}] [ApplyEffect {836045448945477}: Heal {836045448945500}] (783) <254>
[17:00:33.416] [Guard {1776136480620544}] [RemoveEffect {836045448945478}: Guard {1776136480620544}] ()
[17:00:34.105] [Recuperative Nanotech {815232037421056}] [ApplyEffect {836045448945477}: Heal {836045448945500}] (783) <352>
[17:00:34.105] [Recuperative Nanotech {815232037421056}] [ApplyEffect {836045448945477}: Heal {836045448945500}] (783) <352>
[17:00:34.543] [Kolto Probe {1014415440740352}] [ApplyEffect {836045448945477}: Heal {836045448945500}] (717) <322>
[17:00:34.820] [Guard {1776136480620544}] [RemoveEffect {836045448945478}: Guard {1776136480620544}] ()

Here both guard effects are lost, but threat adjustment kicks in after the first. Still, I think granting too much effective heal on those few heals would be much closer to accurate. I didn't have that many logs to look at, so can't say with certainty that it always works like this..
Coordinator
Nov 19, 2013 at 2:18 PM
You are probably right. How nice of them to make each guard effect have the same ID. That just makes this so much easier!

/sarcasm

I will do some testing, confirm and track a guard effect count instead.

The logs will still be subject to the out of order effects notifications. Specifically out of combat notifications almost always occur before the last few attacks are written. This type of thing makes this SWTOR log file parsing more art then science :)
Coordinator
Nov 19, 2013 at 4:07 PM
Turns out that Guard places one effect on you. Then when you leave the 15m PVP range it places another affect on you. Then when you re-enter range it removes the second effect.

The threat is only affected if you lose all guard effects.

I have changed the guard tracking logic to increment and decrement guard effects and to report true for IsGuarded only if the guardcount is > 0.

This seems to have done the trick. This will be included in the next release.

Thanks for the assist :D