1
I definitely appreciate what is probably #2 on the list - the availability of static game data via API. This clearly can be used to create some pretty advanced mech workshop style applications.
An under-used feature PGI already has is the map heat-maps. They contain, if i recall correctly 3 facts (kills here, deaths here, movement here?), as a color coded image or color highlighting on a map top-down view image.
For that, what would be even better, though I wont call it #3 on my own wish-list, would be for those 3 facts (or others) to be made available with coordinates as a dataset. Pair that with a same scale map image, might allow for more interactive heat-map experimentation.
But then there is match results. I know I started looking in to MWO internals long ago to try to come up with a way of getting facts (as opposed to stats) from a just-completed match. I tried the geek-first-route and fired up the network sniffer only to come immediately on the encrypted packet content (duh). Considered briefly doing a tunnel proxy solution, but did not pursue since no-one in their right mind would a) want to run that on their machine or
allow their game to be manipulated willingly in that way. Rejecting anything that required manual grabbing of the info, and also an automated screen grab and scrape as just being way to problematic, I gave up on it.Ive seen some ideas how match results could be achieved via API, but not much detail. If I were to reqest something, I would imagine it to be something like: every match has a unique identifier, and that for every match I could via an API obtain the match setup including type, conditions, teams, and player keys (not callsigns, for privacy reasons), and the match 'events' (A killed B, etc...) to whatever detail was available, and of course the match results (A killed B, C and D, and so on). I am guessing on the need to obfuscate the player callsign via a player key, but the player key regardless is needed even if the callsign were also shown on match info (callsign changes later on, that sort of reason). Add to that a list of the last 100 (or whatever number) of matches played (by key) being available via a player profile API. If something like that were available, team or ladder reporting becomes almost too easy. (being the not-trusting sort and very protective of data integrity, I still would have a timing issue with - challenge issued, match played, player submits results - but what if the two decided to run a second match just for fun, or ran a non-challenge match earlier that week - whats to stop player 1 from submitting the better results - that sort of rabbit hole)
Just some thoughts. As always, I dont have time to read the boards in depth too often, so apologies if these ideas are only beating the dead horse.
Edited by Escobar, 02 October 2014 - 05:53 AM.
