===== y2k.chat =====
~~hero-subtitle welcome to the other millenium.~~

y2k.chat is a retro-styled, vintage-computing-accessible chat service, written with the goal of having a reasonably secure, fun and usable multi-user and direct chat experience on older machines, like AIM, MSN, Skype or Yahoo was - but informed by modern development and user experience practices.

==== y2k.chat stands on the shoulders of giants. ====

y2k.chat is built on revolt.chat. (possibly using [[https://github.com/spacebarchat/spacebarchat|Spacebar]] instead? though I'd prefer to bring more alternative client support to Revolt.)

our software stack involves three main parts: a **service**, a **frontend**, and a **client**. Sometimes a **backend** may be involved, between the service and frontend/client.

=== service ===
the service is built atop the [[https://github.com/revoltchat/backend|revolt.chat backend]], and associated components. this gets a base data model, API and the necessary mechanisms in place for the rest of it.

=== frontend ===
revolt.chat has [[https://github.com/revoltchat/frontend|its own frontend]], which we will include at the start, but this is //far// too heavy to expect to run on anything less than a modern, high-end client machine.

we'll provide and customize this at first, but our plan for a web frontend is to write something lightweight and accessible to browsers like Netscape Navigator 4, Internet Explorer 5, and the like, in PHP.

realistically, all we would expect for a frontend, is a place to download the applications for your machine.

=== clients ===
y2k.chat will maintain forks of open-source clients for other, similar chat services, with the API retooled for revolt.chat:
  * iProgramInCPP's [[https://iprogramincpp.com/discordmessenger/|"DiscordMessenger"]] [[https://github.com/DiscordMessenger/dm|source code]], a Discord third-party client built for Windows 98-and-up machines (WIP for NT 3.5!).
  * not-nullptr's [[https://github.com/not-nullptr/Aerochat|AeroChat]] for Windows XP, Vista, 7
  * dosdude1's [[https://github.com/dosdude1/discord-lite|"Discord Lite"]], another third-party Discord client for Mac OS X 10.4.
  * ayn2op's [[https://github.com/ayn2op/discordo|"discordo"]] for the commandline, on machines that can run Go programs
and perhaps others from https://github.com/Discord-Client-Encyclopedia-Management/Discord3rdparties

y2k.chat may also have gateway services and customized clients on lighter protocols, like XMPP and IRC. We won't actively support non-y2k.chat clients, and we can and may make incompatible protocol changes for user experience or ease-of-use reasons, but we currently don't see a reason to block the use of other clients at this time.
  * adding Revolt support to [[https://github.com/42wim/matterbridge|matterbridge]] to support other platforms
  * [[https://web.archive.org/web/20250123171248/https://amendhub.com/jcs/wallops|Wallops]] (archived link, original is down as of time of writing) - an IRC client for classic Macintosh (system 6)

=== backends ===
these sit in between the revolt.chat API backend itself, and a different API. these would be bridge protocols, and not necessarily direct.
  * IRC: we'll be making a bridge of sorts to support IRC through revolt, in order to more easily support existing IRC chat clients on older platforms.
  * OSCAR: we may consider bridging to AIM's OSCAR protocol, as well.
  * Discord: While implementing the Discord API would be interesting to support other third-party Discord clients, it would require using a different platform from revolt.chat. we may look into using [[https://github.com/spacebarchat/spacebarchat|Spacebar]] as an alternative, though.

=== voice and video ===
it's not a non-goal for us, but implementing it - especially video - will be a challenge. we will likely be using mumble as a basis technology for this. don't expect 4K60 video if at all - think more like 2002 webcam quality.

==== faq section: ====
   * Q. why?
   * A. chasing the iteration and advancement of personal computing and smartphone technology has caused untold ecological damage, and has generally been a stupid pursuit for marketing and fads. this was true back in the early 2000s, too, but we managed to do it with 1/64 the amount of resources, often less. refocusing on what made the personal computer actually personal, and not marketing strategies to increase shareholder value before an IPO, is a good strategy towards making better software, better computers and better services.

===== see you in the last millenium =====