GamParse 1.0.4

Discussion in 'The Veterans' Lounge' started by Beimeith, Apr 16, 2014.

Thread Status:
Not open for further replies.
  1. Spankage Elder

    Could there be a separator added for "downtime" inbetween fights?

    Example: [IMG]

    This should make it much easier to find the start/end for combining adds between wipes on events without damaging yourself so it puts "itself" on the list

    (I know it lists timestamps in fight list tab, but it's somewhat inconvenient to use since overview tab is clicked most the time)
    Smokezz likes this.
  2. Spankage Elder

    Also if grinding the same raid for a few hours with tons of add waves can also make staring at hundreds of timestamps a bit mind numbing.
  3. Azrael Elder

    Can you open source the code? Who currently has the rights to it? I have some features in mind I'd like to mess around with myself. I wouldn't mind dedicating some development time either. Would love it if the code could be somehow made available so others can contribute. Even a heavily restricted license would be fine, I just want to mess around with it for my own personal use.

    A few features/bugs:

    1) When you have log rotate setup through gina, and the log file is rotated while gamparse is running, gamparse will crash. Obviously, this is because the file gamparse was reading from was moved, but the program should exit gracefully. Seems like an uncaught exception, shouldn't be too hard to fix.

    2) Adding the timestamp to the fight navigator, and making the entire left sidebar width-adjustable. This way it can be shrunk to appear identical to the way it is currently. Optionally make the columns that appear in the fight navigator user selectable (ie, right click the top, check the columns you want, resize to your convenience). Obviously this is very similar to the fight list tab, but from a UI perspective the navigator is way more useful. Bonus points if I can even custom format the date string, and sort the columns by time, hp, dps, etc as I see fit.

    3) Add an option on the log search to return X lines before and Y lines after the matched line for context. Similar to grep -C, -A, and -B options. Depending on how the log search is written, this could be kind of a pain, especially lines before.

    4) Add a send summary button that copies a summary of your outgoing healing to the clipboard. Currently can only quickly copy incoming heals.

    5) On the spells & discs tab, when selecting your character, the comparison window currently shows you how many times each spell was cast. Would like to see columns Total Damage, Average Damage, Min Hit, Max Hit, # resists added to see my per spell breakdowns. I would also probably switch the placement of the Spells over time window that takes up the bulk of the screen with the comparison window. There's no reason the spells over time window needs to take that much space, the same information can just as easily be browsed in the slimmer left area. Additionally, after adding the damage columns, make sure the column headers are sortable. Unsure if spell procs/augs are already included in this list as I don't use one right now, but if not they should be added.

    6) On the Player DPS tab or Spells tab, add summary statistics for total resists and total reflects. Unfortunately I don't think the logs tell you specifically which spell gets reflected, but at least knowing the total number in a fight would be useful information.

    7) Somewhere, I would like to see breakdowns of incoming spell damage. Spells that were cast on me, how much damage they did (min / max / average / total), and how many times I resisted especially.

    8) Graphs obviously need a lot of work - a ton of information in gamparse, and probably a lot of cool graphs that can be made, but I find this section sorely lacking atm.

    9) Regarding scaled dps debate - I fail to see why you can't just make this an option in the settings. There's even a section already called "Send to EQ Options". All you have to do is add a checkbox to make scaled the default or not. Simple, and satisfies everyone. Alternatively, it's just as easy to add a second button for copying scaled on the overview tab.

    I've read the entire thread, and personally, I disagree with you completely in that "scaled dps is all that matters". That is a valid opinion of course, and you may have your own reasoning for supporting it, but none of your arguments here are convincing enough for me.

    IMO, the only number that matters is total damage, not dps or scaled dps. Neither dps stat is really that relevant, particularly when evaluating a single named, raid, or even raid night. What counts is your actual contribution to killing the mobs - ie, the amount of damage you did. If you are looking at dps and not total damage when looking at a single event or raid night, you are already doing it wrong.

    Furthermore, the value of dps is that it allows you to compare performance across multiple events, scenarios, raids, whatever. If I see similar dps in multiple events, I know that this player is able to contribute effectively, even when the circumstances vary. If I see major discrepancies in my own dps, I can figure out what went wrong (or right) and accordingly adjust what I'm doing.

    It's also very valuable to the guild as a whole - for example, if evaluating a trial member, if I see this player able to provide consistent dps across the entire trial period, even on varied events, then I know he/she is a capable player, able to adjust, and be an asset to the guild even when the circumstances might not be optimal for his "maximum burn". Additionally, it allows me to compare his relative performance vs other members of the same class, even when these others are not on the same raid or even guild or server. This information is useful to see on every parse - I can see that X class generally outputs Y dps in certain events, and this will be true even when the raid is heavy or light, independent (mostly) of the fight time.

    Finally, scaled dps - the value of this statistic is that it shows the relative contributions of each member of the raid for a particular parse. This information has some value of course, but is completely unnecessary on most parses for several reasons:

    1) We already have this information, in the form of total damage. It actually already exists in gamparse, all you have to do is check the "Show % of damage" option, and....like magic you have exactly the same information that scaled dps would provide.

    This is simple math - using your own example:

    /ooc A Greedy Miner in 43s, 2620k @60938sdps --- Beimeith 1956k @45485sdps --- XXXX 424k @9858sdps --- XXXXX 241k @5596sdps

    Okay...so you did 45k out of the 60k total dps. What's the value in that? Well, that's obvious - the value is that it tells me you did 75% of the damage. Guess how else we can see that?

    /ooc A Greedy Miner in 43s, 2620k @60938dps --- Beimeith 1956k @72439dps (74.6%) --- XXXX 424k @18430dps (16.2%) --- XXXXX 241k @6332dps (9.2%)

    Look at that, scaled dps and I didn't even need the new version! All I did was enable the % damage option. To be honest, I didn't even need to do that, I could have just looked at your total damage 1956k, and compared it to the total 2620k, and come up with a very reasonable back of the envelope approximation of your contribution at a glance.

    2) It completely removes the ability to compare your performance across different raid events, nights, etc. That scaled dps value is only relevant for that single parse, and has no use when comparing to previous or future parses. What if you have a light raid force? What if you are raiding with friends, or helping another guild, or raiding with an open/public force occasionally? That 45k dps number you did is only relevant when compared to the other 53 toons that were there that night. It tells me nothing about how you do on a night by night basis, or how you do compared to the same class in another guild. If I'm an officer and I've got some chat logs, I can right now quickly glance through the pasted parses and see which members are consistently contributing. I know that you typically do 75k dps, regardless if we are in dh2 or bixie2, regardless of whether we have 30 people, 54 people, whatever.

    With scaled dps, I can't even compare the same event twice. Okay so you did 45k this time, 65k that time....how did you do? Well maybe the dps just sucked the second time, so your contribution was equivalent since you did a larger relative share of the damage, or maybe the dps was just as good, but you actually did improve. I really have no way to tell.

    3) It removes a useful piece of information and replaces it with something that can be trivially calculated. If I really wanted to, I could go through my chat logs, and compute the scaled dps myself, just by using (or computing) the % damage you did times the raid dps on the parse. But I have no way of actually finding out the players actual dps, unless I have the original parse. It's not always the same people pasting it, and now in my logs I lose one of the most useful pieces of information in comparing performance over time.

    4) Lol at guilds asking for parses as part of an app...hey look, here's my scaled dps at these events...too bad that's completely meaningless in comparing me to other apps / current members, because you have no idea what the raid was like.

    Don't change it or if you are really convinced that scaled dps is still what you want, then give me the option to choose.
    Smallpox likes this.
  4. junglenights Augur

    The link in the OP isn't working.

    Here's the one for 1.0.3, for those just visiting:
    http://gambosoft.com/Pages/Downloads.htm

    What're these arguments about scaled dps? Isn't scaled dps just saying you did 3% of the damage for a fight as opposed to 34k? Or is it saying something else? If it means solely the percentage of the damage you committed then I can see hte potential problem in that as two different raids can be very different. It doesn't tell you how long the fight lasted or how many hp's the raid boss had. 15% might only mean 45k damage (very weak raid mob). And it's valuable to see how well they burn or sustain.
  5. Beimeith Lord of the Game

    I do not think opening the source code is a good idea at this time.

    Feature requests should be made here: http://eqresource.com/board/index.php?board=312.0

    1) Already fixed for 1.0.5

    2) I don't really see a benefit to duplicating the columns from the fight tab in the fight navigator. Maybe on being able to customize the datetime format. You can already sort things by the column though currently there are limited sorting options, but the sorting has been much improved in 1.0.5.

    3) Maybe.

    4) Maybe. I -think- there is an issue with the amount of healing you report doing being higher than the -actual- healing you did or something? I'm not 100%.

    5) You mean you want to see a Total/min/max/average/resist for each spell you cast? I'm not so sure that's reliably doable. The actions over time is as large as it is because it displays more than just yourself. Select multiple people from the overview box and you will see what I mean. Procs are not listed because there is no reliable way to do them all.

    6) Its on my list. It does tell you which spell is reflected, it doesn't tell you which spell you fizzled, but that should be possible to find for yourself at least, I'm not sure about other players.

    7) Good idea. I'll see about it.

    8) It's on the list.

    9) What has been done is individual times have been added to raw DPS, and it has been made optional. Example:

    All options on:

    /ooc An archaic trilobite in 17s, 623k @36662sdps - Rogue 276k@16256sdps (17272dps in 16s) [44.34%] - Magician 151k@8897sdps (15125dps in 10s) [24.27%] - Monk 126k@7385sdps (13950dps in 9s) [20.14%] - Warrior 40k@2354sdps (4002dps in 10s) [6.42%] - Cleric 30k@1770sdps (3009dps in 10s) [4.83%]

    Without labels:

    /ooc An archaic trilobite in 17s, 623k @36662 - Rogue 276k@16256 (17272 in 16s) [44.34%] - Magician 151k@8897 (15125 in 10s) [24.27%] - Monk 126k@7385 (13950 in 9s) [20.14%] - Warrior 40k@2354 (4002 in 10s) [6.42%] - Cleric 30k@1770 (3009 in 10s) [4.83%]

    Without SDPS:

    /ooc An archaic trilobite in 17s, 623k @36662 - Rogue 276k@(17272dps in 16s) [44.34%] - Magician 151k@(15125dps in 10s) [24.27%] - Monk 126k@(13950dps in 9s) [20.14%] - Warrior 40k@(4002dps in 10s) [6.42%] - Cleric 30k@(3009dps in 10s) [4.83%]

    Displaying the time with the raw dps is not optional, nor will it be made optional.

    Total Damage is useful, perhaps -the most- useful bit of information, but it isn't the end-all-be-all. You can do a short burn and then stop completely and still beat someone in Total Damage. What it won't do is show that's what you did. Player A did 12 million Damage, and Player B did 10 million damage. Did both players reach their full potential, or did one of them slack ? There's no way to tell just by Total Damage.

    You have it completely backwards. You cannot compare raw dps between fights unless they were killed in the same amount of time, and that rarely happens. Quick, answer in 5 seconds without a calculator: Which is better to do: 234,567 dps for 27 seconds or 245,678 dps for 25 seconds?


    If you guessed the second one, you're wrong. Those 2 seconds may not look like much, but they make all the difference. To accurately compare the events you need a common frame of reference. That means either multiplying the damage by the dps (aka total damage), or equalizing the time and recalculating the dps (aka sdps). Total damage is already shown, and even then, it is limited for comparing fights that have approximately the same damage. Player did 5 million damage to Mob A and 10 million damage to Mob B. Which fight did he do better on? You don't know, because you need more information. Maybe Mob A only has 6 million hp, and Mob B has 100 million, in which case he did way better comparatively on Mob A.

    Sdps can be compared between all fights because it is an average. Player averaged 40k dps on Mob A, 33k on Mob B, and 20k on Mob C. Obviously he did better on Mob A, and possibly slacked on Mob C. You don't need to know anything else to tell that. (Though other information still has uses for complete fight analysis).

    As already stated, you have this completely backwards.

    And to get any use whatsoever from raw dps you have to base it off the time of each individual person, so its just as useless to you unless you break out the calculator as well. (Which is why I've added the time to 1.0.5).

    Again, wrong, but congratulations on being a shining example of why I made the change in the first place: you don't know what you are talking about. Other people at least had legitimate reasons they took issue with the change, (namely, kiting/offtanking skewing it, which is valid), you are just talking nonsense. You use raw dps for comparing between events and/or with different raid makeups. Are you joking?

    What is this I don't even? If your average contribution to an event was 45k one time, and 65k a second time, you did better the second time. You can tell that regardless of any other information.

    Maybe the event was killed faster the second time so you didn't have as much time to settle down from the high of your burn at the start. Maybe you just did better the second time. You don't know which it is from looking at sdps, but it doesn't matter in determining which time you did better. But hey, guess what? You don't know that by looking at raw dps either. Lets try that again:

    You did 45k raw dps one time, and 65k raw dps a second time. Did you do better the first or second time? Answer: You have no clue. Maybe the event was killed faster the second time so you didn't have as much time to settle down from the high of your burn at the start. (Wait that seems familiar) Maybe you just did better the second time. (Hrm that seems familiar too). Or, maybe you engaged for less time so it skewed your number higher. You don't know which it is from raw dps, and guess what? It -does- prevent you from determining which time you did better, because there is no way to compare them just looking at dps.

    Parses people post are completely useless 99% of the time because they lack all context. Context is important, raw dps numbers are not outside of very, very narrow circumstances. If you are using raw dps to compare parses people post on the forums, you have bigger issues than I can help you solve, and I might suggest not using any parser at all until you understand the fundamentals of data analysis.

    lol at any guilds that ask for parses at all, especially if all they are looking at is raw dps which tells you next to nothing itself. If I'm going to ask an applicant for parse information, I want to look at cast logs.
  6. Beimeith Lord of the Game


    It's fixed. I was doing some summer cleaning of old files and it somehow must have gotten mixed up in it.
  7. junglenights Augur

    Thx.

    I read some of hte posts above but I still don't grasp what scaled dps is. Don't try to explain it to me, lol. I probqably won't evne install a parser. I did a long time ago or a short while. Raiders seem to love parsers a lot. I could posit a theory it has to do with the sheer amounts of text flying by in the window and htem trying to figure out how effective everyone is. For exampe, say I'm in a raid and have autoattack turned off. It's probably less likely they'll know than in a small group setting.
  8. Branntick Augur

    @Beimeith

    Would you have any interest in developing a plugin for arguably the most sophisticated parsing program ever made? If we could get a functional plugin, it would literally be the most amazing thing that has ever happened to the EQ parsing scene.

    http://advancedcombattracker.com/

    That said, Scaled DPS is the most important number. What IS important is having a means to break parses at specific times to facilitate encounter phase changes. ACT is built to parse a line associated with the command "/act end" which will end the current encounter and begin a new one. It is easily the most functional parser ever made, and I'm pretty sure this option could be implemented using a custom channel to post the "act end" line in to and have it parsed.

    I'm not sure of the extent of your programming ability, but I would sincerely love this. Even so, I greatly appreciate that you've axed the highly inaccurate "raw dps" from being the default export option.

    Thanks.
  9. Branntick Augur

    Scaled DPS is the total amount of damage you dealt to an encounter divided by the total amount of seconds since it was engaged (and registered by the parser).

    Raw DPS is the total amount of damage you dealt to an encounter divded by the amount of seconds since the moment you personally engaged the mob.

    Scaled DPS is the most useful value here for comparative purposes, which is honestly all parsing should be about.

    Raw DPS is (in most, but not all) cases an artificially inflated representation of your contribution. It does however have its uses on mobs that have long transitional phases between actual burn phases.

    Even taking the merits of Raw DPS into consideration, it still utterly all over the idea of comparative parsing.

    Edit: I should say, if you take a small snippet of a parse, for instance a burn phase (as you would be able to do with the program I mentioned in my previous post, ACT), you will notice that the Raw DPS and Scaled DPS numbers are nearly the same.

    This method is inaccurate in 90% of cases though.

    If I stand around and then walk up and smack a mob for a million damage in the last second of its life, it will display as me having done a million DPS for the entire fight, despite the fact that my actual contribution to the mob's HP was less than 1%. I usually displayed the % of damage dealt in conjunction with raw DPS to display this disparity, and I argue that scaled DPS alleviates the necessity of this while still allowing accurate comparative parsing. At the end of the day, Gamparse is simply a simple, not-feature-rich application that isn't a good medium for analysis anyway.

    Edit 2:

    I might add that I was a very competitive high-end raider in EQ2, and that I wouldn't recommend using raid experience as an argument for the relevance of raw DPS. Please take that elsewhere.
  10. Honno New Member

    I keep seeing "if soandso runs up and hits the mob for 1 million dmg then stops attacking it'll look like they did 1 million dps" who the hell does this? lol. i don't wanna be compared to Xwarrior who's been offtanking a mob for 3-4 minutes while I'm off dps'ing other mobs. I want people to see my raw dps for when I burn each one
  11. Serriah_Test Augur

    It's a shame SOE couldn't you know... like... wrap the log tokens in some xml or something.
    It'd be so easy to write up a really awesome parser... though log files would be about 5x larger.


    I mean... it's not like they don't already have tokens in the chat boxes... like player names you can click on, loot links and quest text.... ;)
  12. Harabakc Augur

    Then you click on the player dps tab and see how many swings they made. It's stupid kinds of simple to figure out if you spend 30 seconds looking at the parser.
  13. Kunon Augur

    When you read a parse
    Total Damage > DPS = End of Story
    Axxius and Smallpox like this.
  14. Beimeith Lord of the Game


    That's called an example. The point is unless your engage time is exactly the same engage time as everyone else / the full length of the fight, it will be skewed, and therefore very little actual use.
  15. Warpiggs Augur

    This.
  16. Branntick Augur

    Exactly.

    Please stop interpreting exaggerated examples as something I'm saying would actually happen. Two posts above, someone said that total DPS is the only thing that matters. You do realize that is what the purpose of scaled DPS is? It's simply a more practical (easier to read/comprehend?) manner of interpreting someone's actual contribution to a fight.

    I could cut any long parse to show a snippet of when I was burning my hardest dps discs, but it doesn't make it an accurate representation of my contribution. The purpose of parsing is comparison, not how artificially high you can make your numbers appear in some ideal circumstances.
  17. Adaire Lorekeeper

    I'm taking a bit of a break but I was perusing the forums a bit and came across this post.

    1st off thanks B for taking this on and the multitude of bs moronic posts here is a shining example of why you need to be thanked!

    I didn't read through all 7 pages but I did see where you were asking for suggestions and apologies if this is a duplicate request but could you add an easy way to paste into EQ a sorted by class cast count? Currently if I want to see how each cleric did as far as who casted what and when I have to go through the listing and manually type them in. Perhaps make a sortable column by each class and then allow an export to EQ to just give a raw # of spells cast highest to lowest?
  18. Hazimil Lorekeeper

    Thanks for taking this over, one feature I would like to see (and apologies if it already exists and I'm just not seeing it), would be to also capture the damage by mercs. Be interesting to actually see how good/bad they really are.

    Thanks
  19. Corwyhn Lionheart Guild Leader, Lions of the Heart

    Tried installing Game Parse but got the following error:

    + Downloading file:///F:/Game Parse/GamParse-1.0.4/Application Files/GamParse_1_0_4_1/GamParse.exe.manifest did not succeed.
    + Could not find a part of the path 'F:\Game Parse\GamParse-1.0.4\Application Files\GamParse_1_0_4_1\GamParse.exe.manifest'.
    + Could not find a part of the path 'F:\Game Parse\GamParse-1.0.4\Application Files\GamParse_1_0_4_1\GamParse.exe.manifest'.
    + Could not find a part of the path 'F:\Game Parse\GamParse-1.0.4\Application Files\GamParse_1_0_4_1\GamParse.exe.manifest'
  20. SaderakhBertox Augur


    https://forums.station.sony.com/eq/index.php?threads/gamparse-1-0-5-test.211941/
Thread Status:
Not open for further replies.