Closed SafeteeWoW closed 4 years ago
As for the transmog info - if people can't be bothered clicking a button, do you really think they'd appreciate the implications of adding this? Not to mention it doesn't matter if people don't care
The goal of this PR is not doing auto button clicking for raider who care or don't care about transmog. Non-council cant see the voting frame, thus cant really see the changes by this PR. This PR is for the ML who never wants an item to be wasted.
Furthermore there's no reason to use resources on scarcely used stuff, hence the opt-in nature of EU (which unlike the other addons can't be load on demand).
- This is more likely that the version of ML's (RC and EU) is incompatible with the raiders if they are two pieces of the addon instead of one.
- Some scarcely used stuff should be completely removed instead of putting it in EU. I don't think EU should contain useless stuffs either.
- "legendaries" column is useless (everyone has 2, not considering Aman'thul).
- "ilvlUpgrade" is useless. This is not in Legion.
- "Titanforged" column is partially useless and not working. It is incorrect to use
[3337] = "Titanforged"
(bonus id scan) to determine titanforged. Instead, tooltip scan should be done (Implemented in RCLootCouncil-EPGP. See Module\customGP.lua for reference)sockets is a bit useful, if the guild wants to balance the luckiness of raiders. This is a small bug that this is calculating #gems instead of #sockets and it is not properly calculating if there are gears with more than 2 sockets, for example, crafted legendary. I personally don't think this is useful.
- RCScore. I don't think anyone will use this. The formula makes no sense. Such score (DPS/HPS performance, eliminating the affects of ilvl) should be viewed by ilvl% of WarcraftLogs. There is no way to come up with a good formula to calculate those.
- traits. Dont think this matter at the moment, but I do agree this matters in the early days of Legion expansion.
- upgrade and Pawn. I know Pawn is a good way to determine whether a gear is an upgrade. But pawn is very misleading if you use its default stat weights instead of doing simulation on your equipped gear and enter the simulated weights. Your stat weight is highly depend on your current gear. For example, although Fire Mage prefer crits. But if your gear is full of crits, the stat weight of crit is no longer highest because everything will diminishing return when you stack it. The current EU just applied same Pawn stat weights rule to everyone with the same spec in the raid, not considering their current gear, it does not tell whether it's an actual upgrade for them or not. This is just misleading and useless. Pawn column is only useful if raider send their own custom stat weights of Pawn to ML.
- Overall, I personally think only spec icon(already in core) and bonus rolls in EU are actually useful. So only adding bonus rolls into core is left and you agree it is essential. Then there is nothing useful left in EU, so I don't think it's worth to make a separate EU.
- RCv3 should have a separate file for scrollcols. There are too many duplicate code to handle scrollcols currently. lib-st is super old. I think we should add some useful custom functions into it. And I want to add the feature to resize/reorder the columns by mouse, if that is possible in WoW. Any PC application with tabs can drag the tabs to reorder it and it's just dumb that user need to enter give number index to each tab in order to reorder the tabs.
It would add yet another comm layer, which the current system probably can't really handle many more of - not to mention the root problem is solved by people clicking a button.
- The comm adds 5n+6 bytes, not too much
- I have considered to make this feature as auto rolling transmog when clicking pass if appearance not collected. But I think ML should differ ppl who really wants the mog and ppl without the mog who don't need the mog, so I give up the idea. But the problem in my group is that most ppl don't care about mog, so most of them click pass instead of mog, so I want to differ among those ppl who don't care about transmog, who has the apperance, and who don't have appearance, so the gear is not wasted. So currently, when everyone passes, I simply throw it to some random guy, and later he says I already have the appearance and I am absolutely don't need it, is there anyone else who may need it or its appearance? I don't want this to happen and this happen fairly frequently that I don't want to spend time asking who may need the appearance.
But most ppl, at least 50% I think, use and are used to the default response color. Although I agree I shouldn't worry too much about it. P.S.: Is the default response of relic buttons still grey?
Except it's very possible for non-council members to see the voting frame, and either way the council members still will. And no, the column shouldn't only be shown to the ML. This PR is for people interested in transmogs - imo an item is not wasted if no one wants it.
Yup. It's definitely possible to do drag and drop - it's just very time consuming to do, hence the current implementation. Same goes for click resizing, albeit easier.
My point here was just that the original problem is already solved, and again, imo an item is not wasted if no one wants it/can use it. The only way I see this being an issue is when a ML really cares about it, but can't get people to install another addon or don't think the issue is big enough to ask them. That's not ideal, but I care more about not bloating the core addon however.
Yes.
As for the EU comments:
While Legendaries
surely were more useful in the begging, Amanthuls does make it really useful again.
ilvlUpgrade
just a remnant.
Titanforged
Why? It has always registered correctly every time I've checked. Also note this intentionally included warforged items as well.
Sockets
I intended it to only include socketed gems. One could argue the name or intent is wrong, but the idea was to show the amount of sockets benefiting the candidate. True it misses the 3rd potential socket.
Traits
my guild requires a specific number of traits on mythic. Not to mention it makes it easy to see who's slacking.
Pawn
I expect people using it would set it up correctly, and be aware of the limitations of stat weights. Otherwise the default behavior is for people to send their custom stat weights to the ML (or at least the calculated score). When first installing EU, it does not register ones current default stat weights, but that's on the todo (and also the primary reason why EU is still tagged as beta). Personally I don't use stat weights to determine gear overall, but still find it very useful for comparing individual pieces (for which I use pawn all the time). And while you're right stats are worth less the more you have (which is included in the calculation, if people have set it up), it's still very usable for comparisons as you can quickly filter responses based on it.
RCScore
. This has yet to be released, as I'm not 100% happy with the current formula. The intent with this is to bring a rating (not unlike warcraftlogs ilvl%, see below) into the game. This could be very useful for comparing player performances, further reducing the guesswork of where a gear would have most effect. Of course this will never be 100% precise due to the use of in game damage meters (although Details! is not far off) and the users setting of those, but the calculated scores are relatively imprecise, thus usable.
The warcraftlogs ilvl% is percentile based, and while it could be duplicated almost 100% in-game, that would require an up to date extract of the data the percentiles are based of (i.e. every single log from the past 2 weeks). It's doable using their API, and it might even be possible to estimate the results to avoid having to update frequently, but all in all not something I'm looking for.
Instead, a much easier approach is to calculate how far a candidate is from the proven max dps, based on ilvl. In general, this would be dps / maxDpsAt(ilvl)
% which is very easy to create a formula for.
The current formula is an approximation of the average dps/hps made by the top 99 percentile on warcraftlogs. It does take ilvl into account, and it's actually a pretty good approximation (only about ~0.5% off). The problem with it is specs are usually +-10% off each other which makes it very imprecise when comparing specs far from each other on the dmg meters. Furthermore there's a lot of discrepancy between fights, which makes an average estimate vary wildly from fight to fight. To this end I've made a small program that pulls the top n-percentile (most likely 99 due to data) for each fight for every ilvl of every spec. This is linearly approximated (on average this has about 0.15% error rate) and extracted as an invokeable lua function.
But most importantly: shared by all of the above: they're completely optional. Don't like, don't use it. The only reason they're in EU is because I believe not everyone will use them.
See also issue #149
I don't think this feature should be put in EU, because it happens very frequently that I need to know the transmog info so an non-upgrade piece for the raid will not be wasted.
I would like to move BonusRoll column from EU into core, because this is very very important for the loot decision, and this is true for any guild, casual or hardcore, unlike Pawn score, or #sockets equipped.
Now it's time to talk about how this PR works.
Current problems