Forum
Player Stat Tracking, Class Based Leaderboards etc
Created 23rd December 2014 @ 13:24
Add A Reply Pages: « Previous 1 2
Quoted from skyro
[…]
oh look it’s another all shittalk no constructivity retard poster
<3
Quoted from Popcorp
Gentleman Jon made something similar before, but then he stopped.
Never doubt Gentleman Jon 8)
Quoted from Admirable
[…]
Never doubt Gentleman Jon 8)
He’s never stopped working on this :)
Quoted from xiiirog
[…]
bro what do you have in mind , anything to help with ?
my plan was to parse the logs.tf logs that you can access through their api, mainly to explore some stuff I wanted to experiment with like distributed systems with dynamo or mongodb
i dont know if you can help…
Quoted from Popcorp
etf2l coders are already busy with new site design and Swiss System. Gentleman Jon made something similar before, but then he stopped.
There were a few different problems, the main one being that large scale collation of league logs requires an automated method which means checking is extremely time consuming. Also player habits such as uploading manually combined logs from multi round maps such as Gravel Pit or Turbine in addition to automatically uploaded logs means that double counting is very hard to avoid. There are others of a similar nature.
That’s not to say a robust solution isn’t possible, but it’s superficial simplicity is an illusion.
Quoted from Gentleman Jon
[…]
That’s not to say a robust solution isn’t possible, but it’s superficial simplicity is an illusion.
Wouldn’t it be easier to start with server logs, not logs.tf? It seems to solve the issue with double-upload, etc on serveme servers.
Quoted from olBaa
Wouldn’t it be easier to start with server logs, not logs.tf? It seems to solve the issue with double-upload, etc on serveme servers.
But logs.tf has already done all the hard work by getting a very wide installation base, years of debugging to make the logs parsing itself very robust, an extensible logging syntax agreed, integration into tftrue, mass online storage available to all, etc.
Overcoming the problems identifying the appropriate logs in logs.tf is preferable to trying to replicate the important parts of that infrastructure IMO.
Quoted from lexs
[…]
my plan was to parse the logs.tf logs that you can access through their api, mainly to explore some stuff I wanted to experiment with like distributed systems with dynamo or mongodb
i dont know if you can help…
well i dont think i can then.
Add A Reply Pages: « Previous 1 2