4 ms·
I think AIM is a little different... It was born out of the original AOL dial-up walled garden service where they built a bridge to the Internet where people co
by a-dub 6y ago
I think AIM is a little different... It was born out of the original AOL dial-up walled garden service where they built a bridge to the Internet where people could download standalone clients and communicate with themselves and old school AOL subscribers. I don't know the internals, but I do know that much of the original AOL backend existed at least through the early 2000s which provided the instant messenger and authentication services [0]. It wouldn't surprise me if they were shutting down all of that infrastructure which I'm sure had a non-trivial cost to operate.
I think the thing to keep in mind is that running centralized backend infrastructure (ie; store and forward, user state, directory) for an instant messaging service is not cheap and if the platform is open and allows for user supplied clients, there's almost zero direct monetization opportunity without stepping into unethical privacy breach territory. That means they're relegated to just pushing a brand, and it seems like a very expensive way to do that. Moreover, AOL transitioned in its later years into a digital journalism company, and as far as I know AOL itself as a consumer brand disappeared.
[0] Worked on a whitebox set top box browser platform (ala Android) in the early 00s specifically with AOL as a client. Amusingly, the big feature they were trying to push was onscreen chatrooms for people who were watching the same thing live, which I guess if you think about it may have led to ways to monetize chat.
- icedchai 6y agoToo bad they didn't open source it. I'm sure most of the costs were in the employees required to operate and maintain the system. I am skeptical the infrastructure itself was that expensive, unless they were forced to keep it running on obsolete hardware. 90's tech is generally pretty light weight...
- deaddodo 6y ago> unless they were forced to keep it running on obsolete hardware. If you read blog posts and the like from people involved in those projects (MSN Messenger and AIM/ICQ, specifically); the companies would take advantage of protocol updates or other scheduled downtime to failover to newer hardware. So I doubt they were ever stuck on any “obsolete hardware”.
- a-dub 6y agoAOL launched around 1991 if I recall. I remember watching with great interest (I think I may have been in some kind of beta program) as it was obvious (at least to me at the time) that digital communications with a graphical UI was the future. (At the time in the US, there were a few commercial information services that were textual, General Electric Network Information Exchange and CompuServe were the big ones... followed by Prodigy which basically looked like teletext or a Bloomberg terminal for the home- it was clear that something better could be done, which ended up being AOL). The original AOL client shipped for DOS with Digital Research's GEM widget kit, it wouldn't surprise me if they used additional DR hardware and software on the backend (VAX anyone?). When I worked on the AOL project, I would see e-mails from AOL ops (from AOL screen names, no less!) about how they would take scary and unique sounding things like "access rotors" down, so I suspect that given how early the launch was, the backend was extremely bespoke simply because they were making things up as they went along in order to launch a new and unique thing. It wouldn't surprise me one bit to learn that it ran on VAX, HP3000, AS/400 or perhaps something even older and larger. You didn't hear about horizontal scaling nearly as much back then, most scaling consisted of "buy a bigger computer."
- icedchai 6y agoDigital Research (DRI), the creator of DR-DOS / GEM, was not at all related to Digital Equipment Corp (DEC), the producers of VAXen and the like, despite similar names.
- ballenf 6y ago> stepping into unethical privacy breach territory Given oath's laughable privacy policy, I'm not sure any stepping would be required. Granted the shutdown came before the Verizon/Oath takeover, but somehow I doubt the decision came down to "we could make a lot of money off this, but it will require relaxation of privacy practices, so we'll just shut it down instead."