Recount or Details style in game DPS parser

Discussion in 'The Veterans' Lounge' started by Turinbar, Mar 7, 2019.

  1. phaeril Augur

    While we're talking client performance, has there ever been any consideration to do things like trimming inactive (or way under the old thresholds) guilds? Guilds might not be the best example, even the servers with the most guilds only come out to like 1.1MB for the guild list, and maybe that is only built once on first zone in entry.

    Side note, hats off to whomever did the original PFS implementation. I thought I'd be all clever and see if I could improve load times by repacking all of the art and models and such with lower compression, at best I saw only a 5% increase in load speed, and that was only in zones with very large files. So I went the other way and tried to ramp up the compression, changed my mind about 90 minutes in and I had barely gotten a quarter way through the files, lol.

    Now I'm wondering if it can support variable compression levels to reduce the number of blocks for big files but not waste allocation and decompression time on the little ones.
  2. I_Love_My_Bandwidth Mercslayer


    For some odd reason, in Win10 v1803, my parsing performance while raiding cratered. It actually brought my overclocked i7-7700k to its knees. EverQuest was lagging badly. By the end of the raid my screen was 3 minutes behind the raid. I looked at everything I could think of. Nothing helped. I was beginning to think my trusty Samsung 850pro SSD was to blame.

    I opened ProcMon and noted that when the slowdown hit, the Antimalware Service Executable (ASE) CPU usage jumping to the top of the usage list then back to near the bottom. It was just a blip. But noticeable. I expanded the ASE tree and saw HUNDREDS of Windows Defender Antivirus Services running. Every time the log file added a line, a new service was opened to scan it. Which lagged the entire system so badly the game client was essentially put on very low priority CPU access.

    I created a live scanning exception rule for the EQ log folder and tested. No more lag!!

    I have since moved the game folder to its own M.2 NVMe drive. It's fast!
    Scornfire likes this.
  3. ~Mills~ Augur

    I did not do all this work a few others did for the most current versions of dots in a few other threads in which you participated. Its not rocket science taking something you already do from 6-9 seconds of cast time and dropping it to 2 seconds concentrates what you do more in line with the nutso that everyone else has become. To reel in sustained some to offset the front gains you reduce the overall dot stack power down from what it currently is. This was not done for anyone else. He took dots that were not used typically or a dot where only one was used often or took a class that might typically used 2 dots worth of power and then took each dot individually and multiplied it by 6 or 10 or 20. And then we wonder why it was and is still broken for some classes. Who maintained all their melee, pets or nukes in addition to those giant increases.

    What is pretty OP is the current state of every former burst dps class as well as just about every hybrid, 2 healers and 2 tanks. You can't hold necromancers to some ancient ideal when the rules for everyone else went out the window years ago.
  4. kizant Augur

    I agree that there's a problem and it should be fixed. I only disagree that it's easy to implement one of these overly simplistic solutions that have been posted or that players could have helped determine whether changes actually worked properly without being able to get accurate parses. No, I don't trust necros to do that on their own.
    Xianzu_Monk_Tunare likes this.
  5. ~Mills~ Augur

    There is no way you can honestly believe that they test things out via parses to any real degree having witnessed the hundreds of things that have entered game over the years from every single dev at some point. Things pass a general sniff test, which obviously varies greatly by each individual dev, and is then sent into the game to be actually tested. Where it can sit for weeks, months, even years if ever at all being corrected.

    Having played the class for decades and seeing what greater minds than mine have suggested I'll lean on those suggestions over someone randomly winging it or over zealous forum questors who may or may not be good with numbers. I don't need a parser to see mobs melting or the duration it happens for when related to other dps classes or hybrids or even healers and tanks at this point. When some classes in a good group setting can burst for over a million dps and maintain over 300k dps in a group setting indefinitely the lets err on the side of caution with necros suggestions can take a hike.
  6. kizant Augur

    I don't think you're reading what I'm saying. It's the Players who needed the ability to parse things accurately so we can provide feedback when changes don't have the intended effect. I'm sure the devs have their spreadsheets and do their best but it's a complex game with lots of bugs people take advantage of. Seeing DoTs reported at the mob instead of the player and being able to break each spell down nicely is a big step for the rest of us to help provide feedback.

    In my first reply I was trying to point out that if you want changes to be made then you need to be very specific and ideally provide some evidence to support the effects you think those changes will have. Here you seem to have done the opposite of what I was trying to suggest. Your justifications have gotten more generalized and you've added unrelated problems to the discussion. I don't think that's going to help. And you should at least go vote on the issue tracker bug related to the DoT Revamp. It has zero votes.
    Xianzu_Monk_Tunare likes this.
  7. ~Mills~ Augur

    Players in general are tainted and skewed. People frame and pick things out that suits their agenda more often then not. So just because there is a new or more calibrated tool doesn't change anything in the slightest. Its just a different tool. The devs determine whats what. Dots have been able to be parsed for the longest time. And its been close to a year with sphere being parseable. So changing where dots are reported didn't change anything as far as accuracy.

    No other class has had to present detailed complex breakdowns of what they wanted, how it would work out, as well as branch out what possible effects might occur from said changes to move forward. They asked or in some cases didn't ask and got revamped dots. Or ask for AA, spells, changes, items, clicks etc and thats pretty much that outside of a general explanation in many cases. So once again someone, in this case you, is trying to hold necros to one standard when it wasn't the standard used for anyone else. In this case necros took it upon themselves to come forward with a well thought out and reasonable suggestion that tries to prevent past mistakes and concerns and very intelligently tries to tackle the problems faced while maintaining the status quo so to speak. Giving necros a bit more burst and similar sustained relative to what every other dps class and even non dps classes have become. You are taking what they already are currently doing just speeding it up some and then offsetting it being sped up by lowering overall dot stack power some. Without any evidence or seemingly understanding you claim this is OP with a straight face.

    I am to lazy to look but was there even a issue bug tracker or was it the flavor of the month that is now when this revamp process started 3 years ago? Regardless a non intuitive external tool that was not used for any other dot revamp should not be the justification for why this process has seemingly halted to a stop.
  8. Brohg Augur

    No other class started out as already #1 damage on nearly all raids…
    Scornfire and Xianzu_Monk_Tunare like this.
  9. kizant Augur

    It changes everything.

    If it's as easy as you suggest then post the specific changes you would recommend. You're just being obtuse. Some changes are easy. It didn't take a lot of evidence to get the type 18/19 procs changed. All it takes is an obvious problem (I cast mez and a damage proc breaks it) and an easy solution (add a damage restriction like all other procs have). But in your case you're asking for an entire revamp of a large number of spells. Changes that will have to be made going back many expansions and your requirements are vague at best. And we just had a revamp of a much easier set of spells and it was a complete mess. And yes plenty of people send the devs detailed spreadsheets and breakdowns and suggested changes when there is a problem they think needs to be addressed. Just because you don't see them posted on the forums where the clueless people can argue about them doesn't mean it doesn't happen.

    Citing what was done with druids and shamans in the past is not helping your case. You can't point at something that went poorly to justify doing the exact same thing again. You should be asking yourself what could be done differently this time so those mistakes aren't repeated.

    And I really don't understand why it's hard to suggest something. Just take 1 spell line that you're currently using multiple levels of and figure out what it would look like if they no longer stacked. Save yourself just 1 or 2 buff slots and change 1 spell line. See how it goes. Use the fact that we can parse things better now as a justification why it's OK to take a chance on the change. Get votes for it and get people to collect before and after parses to show it worked. At least try be reasonable about it. Ranting about it every few weeks without adding to the discussion is obviously not getting you guys anywhere.
  10. ~Mills~ Augur

    So prior to the changes if you stood by a necro you got 1k. After the changes if you stand near the mob you get 1k. Yep game changed! You are trying to make it sound more complicated or make it sound as if you are speaking from a place of wisdom. When instead youre posting your opinions with less knowledge basis than I and then trying to spin it that I am coming from a position of weakness and bad info on the subject. All the data that was needed has been around for almost a year, nothing in that regard has changed.

    I just can't fathom how you are being this obtuse when in general you seem very knowledgeable. I am doing the exact opposite of what you are claiming I am doing. And am in fact doing exactly what your bolded quote above is saying and yet you keep repeatedly posting. I am reproposing an idea that many have already posted as a solution to prevent the same mistakes and poorly handled revamps of the past from being repeated. It is completely different and yet here you are trolling. This isn't my idea I am just rehashing what Menown, Forcallen and a few others have already done. I don't need to repeat what they have done just because you keep harping for me to do so.

    You just have to be trolling at this point.
    Druids or shaman had dot A. They did not use dot A or only used dot A or only used the 2 most recent versions of dot A. When revamped they got dot A damage x 5 or 10 or 15 or 30. So they got massive bonus damage added to that dot for zero reason. This mistake was then repeated for dot B, C and D. So they went from using 0-3 dots sometimes that maybe did 6,000 in dot stack power to wanting to always use 4-5 dots that do 55,000 in dot stack power. The same happened for enc, rangers, beastlords.

    Necros have dot A. They use the 3 most recent versions of dot A. They are suggesting you Make dot A = the damage of those 3 most recent versions. Necros also already use dots B, C, D, E, F, G, H, I, J..... Repeat the same thing for dot B. Then to offset the fact that casting dot A and B now equals what used to be casting 3 dot A's and 2 dot B's you make new dot C only be 70% of the combine dots it is made up of. For dot D you make it 60% or simply make dot D remain what it is with the newly added block restriction from stacking older versions. So their casting dynamic goes from casting 18 dots to 8-10 and to offset this change you take them from the 100,000 in base dot power they had been using and reduce it by X% . This nets them reduced dots, shifts their burst dps potential up some from less casting but is offset by a smaller overall base dot stack.

    This works through any era or period is is no more time consuming than just winging and saying lets take dot A and make it do 5 times its base power but also remove their epic, change how dots work with focus effects, alter how dot damage is displayed.

    AKA completely different from how ever other dot revamp was done, reasonable and addresses the issues faced while trying to prevent the stupid that every other dot revamp had or has in some cases.
  11. Tatanka Joe Schmo

    Re: Shammy/Druid DoT changes.... did it really go that poorly? (just speaking from druid perspective here, no clue on shammy changes)

    Everyone seemed to be thrilled with the initial change. More damage, less spells to mem, etc.

    Then, they tuned down the damage, and didn't touch the mana usage. Yes, people complained (me, included), but further analysis showed they were still our best damage/mana spells, and still out top damage spells, and still better than what we started out with.

    Again, did it really go that poorly?
    Xianzu_Monk_Tunare likes this.
  12. kizant Augur

    It took 3 tries and it's still not what it should be. Plus the only reason devs toned down the damage is because players complained. For months druids were able to double the DPS of other DPS classes on a bunch of events. It was very very broken. Most people still don't realize how broken it was. And the necro case is more complex.

    You try to follow Menown around to get a decent parse. He's not dumb enough to just stand there. The only reason data on druids and shaman were easier to get is because they had to stay in group healing range. But then you'd still miss damage on their nukes. So, you would never see the full picture either.
  13. Waring_McMarrin Augur


    Shaman dots are still a lot less mana efficient after all of the changes that have been made
  14. Tatanka Joe Schmo

    Exactly my point, it was a great success.

    In the solo/molo game, druids were very pleased with the changes. In the group game, mobs died quicker, so everyone was happy. In the raid game, even though mobs died quicker, some players were sad their was broken, sob, sob.

    Yes, a great success ;)
    Xianzu_Monk_Tunare likes this.
  15. Fanra https://everquest.fanra.info

    I have on my wiki: "By default this [/loginterval] is set to one second.".

    Am I wrong? What is the default setting, please? Because your post implies the default setting is not 1 second.

    Thank you.
  16. Zenshai New Member

    The default interval is 5 seconds--which is why folks recommend using /loginterval 1 so that it can be useful for Gina/GamParse.
  17. Windance Augur

    From a performance point of view an in game DPS window would be much faster/less resource intensive than having to write out all the combat messages to a file, then having the 3rd party utility load them into memory to be processed.

    On the other hand, having the external logs is nice so you can reach back and look over days/weeks/years of progress.
    Tucoh likes this.
  18. Turinbar Augur

    Thanks for all the replies, did not realize it was the add on framework and other backend stuff that doesn't really make it possible. You really get spoiled w Recount / Details and its user friendliness.
  19. Beimeith Lord of the Game

    The default log interval is 0, which is write immediately, which is how it existed for 15 years before they added the interval command.

    Changing to interval 1 delays writes to once per second, massively reducing the number of times the file is opened, written to, and then saved from thousands of times a second (literally) to 1.


    Also, for what it's worth for those asking, when Gamparse is in "monitor" mode (aka it isn't loading old data, only live/new data) it works by every second it opens the file, reads up to 10,000 lines, stores the last location it read into a variable, parses those lines, then grabs more the next second.

    If it is not in "monitor" mode (aka it is loading old data) the loading interval is reduced lower than 1 second until the whole file is loaded and parsed, then it changes to "monitor" mode.
  20. Tatanka Joe Schmo

    Which is interesting, because I have Gina triggers set up for buff wear-offs, using the default (have never changed it).

    Very frequently, wolf form triggers, even though the buff has a few seconds left. Which is annoying, because you cast the group form, which won't overwrite, but does blow the cooldown. I assume this is because the game is sending the log message too early, as opposed to an issue with /loginterval, which would manifest as messages getting there late, not early.