Future Steam technology wrap-up
Seeing as it has been an extremely quiet period lately, with the team’s efforts seemingly focused on server- and client-side fixes and no relevant news of any importance, a wrap-up of what is currently known of Valve’s future plans for Steam seems worthwhile.
While I have it on good authority that Valve’s Steam strategy stretches at least two years into the future, its form is still a mystery. For now, we must make do with what little has found its way out of Valve’s offices. So, in order of certainty:
Judging from the dialogue layout file that betrayed the existence of Steam’s Peer-to-Peer download network, it is intended for the distribution of mods in a manner closer resembling the distribution of full games. This probably includes mods being downloaded to GCFs, and perhaps even the more advanced filesystem features such as content streaming, easy localisation management and backup. What isn’t so certain is the system’s flexibility: an open house is clearly out of the question, but both whether other custom content will be supported, especially during server map downloads, and the extent to which mod teams will be able to make use of the system, remain up in the air.
Either way, its release is bound to make a big impression and will bring modding that little bit closer to the accessibility of the games themselves.
Friend Programme Buddy Program was revealed in last March’s Game Informer interview. It will allow users to offer up to five of their friends simultaneous access to one multi-player game which they have registered. Their access is tied to the original account, and the originator is also notified through Friends whenever one joins a server. Nothing else is known at this point, and we haven’t heard a peep since Game Informer’s interview, but it is surely far too elegant an idea for Valve to allow to drift away. Keep your eyes peeled.
While not a feature per se, the long-overdue Atari back catalogue is a radical focus shift that presents a series of interface challenges. How Valve will handle large influxes of new content without swamping their independent clients remains to be seen.
Multi-player statistics are already collected, so is it really such a large jump to start analysing them before a user joins a server? Perhaps not technically, but there are social impacts of matchmaking which must be taken into careful consideration. Cheating attempts would unquestionably skyrocket should any form of matchmaking be implemented, the black market in Steam accounts would gain a powerful new sales driver, and what little good nature remains in the various communities would soon buckle under a sustained assault of scoreboard-obsessives. Implementations like those of Xbox Live and Halo 2 might offer some insight, but on the PC, and with games that began life without any such luxuries, matchmaking’s introduction could easily be disastrous.
Nevertheless, Steam is an ideal account-based platform for matchmaking, and with the detriments of cheating of so much concern to Valve, a solution of some form to unfair matchups is the next logical problem to tackle.
One of the more telling changes made for the Steam 3.0 interface was its sudden specification in several locations that it was the “Windows” client. Furthermore, Steam’s core authentication and account code has been running on Linux for years already in the form of dedicated servers and even a stripped-down account management client. The technology is there, Steam unquestionably has the influence to turn Linux’s gaming image on its head, and with a core Linux-ready package of Red Orchestra, Darwinia, X2: The Threat, Jagged Alliance 2 and the upcoming Defcon, along with the general tech-savviness of Linux users driving down the support burden, a port could even be economic from the get go.
The only potential issue visible from our admittedly limited perspective is the manner in which the open source ethos would limit the uptake of content server distribution; but with a Peer-to-Peer network in place, Steam on Linux becomes a potent proposal.
One of Valve’s goals with Steam 3.0 is to “look for ways to make Steam…solve problems for people”. What better way to achieve that than with plugins? A Friends SDK was announced along with Friends itself, and with Valve’s new on-staff SDK manager should be released fairly soon. A Steam plugin SDK for developers and modders is not an enormous jump from there. No word yet, however.
Free spectating was first confirmed to be a part of Valve’s plans in a rare post on the Steam forums by Gabe Newell – his only surviving contribution, in fact. Although he was speaking in 2003, there have subsequently been two low-key attempts at spectator events in the form of ‘Valve vs. Modders’ HL2CTF games, and SourceTV has been created and maintained, even despite the fact that almost nobody uses it; however, Free Weeks and Weekends have stolen some of the idea’s principle. Spectator events seem to be dormant for now, but their case has yet to be closed.