Introducing the Unofficial Orbus Armory


#21

We have a mechanic whereby to affiliate correctly against a fellowship, you should have 3 out of 5 members from the fellowship present in the run.

Secondly, we only started updating a few days ago… So if you happened to have a run before a few days ago that is not in your top 10… You might not see all the names against it… That is just how it is as the API has the last 10 run records available.

Thanks,


#22

We have 5/5 in the same fellowship in the run.


#23

Again… As I have explained, we don’t have the data to show this, so unfortunately it will not reflect. For all future runs it will.

Thanks,


#24

I have record of the runs using our bot.

Edit, or maybe not…anyways this leaderboard isn’t correct.


#25

The data tracking started on 12th of April, at around 9:32:19 AM (GMT+3). If those characters included in the run were looked up after that, and they had the run in their last 10 runs as provided by the official API, they would show up.

Since they didn’t, they won’t show up on that run, and currently there is no way to bring them in to the run either. If at some point new endpoints are added to the official API to dig up the entire history of a character, or fellowship, the situation will be rectified.

Again, none of this info is new, its all explained on the F.A.Q page.


#26

It does make the leaderboard inaccurate, no matter what you write in your FAQ.


#27

Hey @aRkker thanks for doing this, it’s really nice of you to put all these efforts into doing this. And the new fellowship leader-board is really nice too ! It got me excited to go back and play hard to keep EK as #1.

I have a couple question.

  1. Is there a way you know of that could help you with data retention ? If you need sever space I might help you with that.

  2. Else what is there a way to make a bot to search each character name in a loop ? ( stupid way to make it work )

  3. Finally, with the data available at this moment, would it be possible to show who in a fellowship as been inactive for more than X time. That would be helpful in the long run, and seing EK with 21 members when 7-8 are doing daily made me sad for them. I want them to have more credits :stuck_out_tongue:


#28

Very cool! Thanks for doing this. The update is amazing and I can’t wait for my guys to see it!


#29
  1. Currently data retention isn’t a problem for now. The data we’ve got is still small, and I don’t see it growing super large in the near future. But I’ll keep that in mind :wink:
  2. I’m not sure I quite understood the question here, can you try re-phrasing a little?
  3. Technically with the data available the only way I can see activity being calculated is based on whether a character has done shard runs in the recent times. But that’s a bit unfair for characters that don’t participate in such activities. I suppose we could split the members number into people who actively contribute to the PVE progression and people who don’t. In the case of EK it would show up something like: Members: 8 (21), meaning 8 “active” and 21 in total. I’ll look into that with the patch 1.2 coming in the next week

#30

You could diff the timePlayed property.


#31

That’s actually a really good idea. Some kind of a threshold that you need to have played X amount of minutes/hours per week or you’ll be flagged as inactive


#32
  1. Since the PVE leaderboard only show data from players whom looked at their profil (or at least looking at your profile update the leaderboard) would a webscrapper who search for each players name in the API solve that issue.

(sorry if it’s not clearer, english is not my main language.)


#33

Ah, yes it would solve the issue, but I feel that @Riley_D would rain hellfire and brimstone on my house if I were to do that because of the massive amount of traffic it would generate :laughing:


#34

What if it scrapped once a day on a non busy time (middle of the night)


#35

I’d have to ask if that’s okay, because there might be server-side data transfer/CPU cycle tied costs which would go up if I was to do that. I’ll keep that in mind and see if we can come to some kind of a solution for it


#36

Well, if you stored the data in a database locally each time you scrap and access this database when people search for their name, that could theoretically reduce the number of time you make a request to the API. (I did my name like10 times this morning haha) and you could have more data retention for each player since you can store more than 10 runs.

But yeah, that could be a much bigger project than intended. And I already think you are doing a way better job than expected for a community project. :slight_smile:


#37

There is a caching system in place for repeated requests. If the data is less than 10 minutes old, it will drop it from the cache on my server. If the cache is older than 10 minutes, it’ll refresh it from the official API. The system sort of retains more than 10 runs per character, because it generated run-entries to the database and links players to it, so technically we do have a bigger than 10 runs saved per character


#38

So yeah, you actually did a yeah better job than I could have ! :smiley: Thanks again for taking time out of your day and doing this ! :slight_smile:


#39

If you are interested in tracking fellowship progress, then why on earth wouldn’t this be 5/5?


#41

That is actually a really good point. That makes it so you can’t have 2 ringers coming in to help a guild. It should be 5/5 in the same guild to be called a guild kill.