Skip to content

Comment on Talkita Lets You Chat with Users Browsing the Same Sites as Youparent

Comments

ThirdVoice had the problem JonnieCache identified, you generally either had a wasteland (long tail) or so many elements on the page it was as useless as a YouTube comment set. And those were persistent comments; seems like this will have the same sensitivity-to-conditions problems where a page(/domain/whatever atomic unit) either has 1000 users or 0, only since there's no persistence it all has to happen at once so you'll trend even more towards 0 and have an even-smaller sweet spot. (20 notes on a page? Pleasantly active. 20 people talking...?)

Since you mention it, I was the one who first worked out how to block ThirdVoice; their little image tags were all assigned the same class and so one CSS rule made them all display: none. They "fixed" that, so I made a JS script that ran after theirs that could look at img source and remove them if they were Third Voice's. At that point they changed their license to forbid those who downloaded the program from "interfering with another user's experience of the product", and while I would happily have refrained from downloading it and clean-room implemented something, it was never an issue; their day in the sun had passed.

Technical detail: ThirdVoice was implemented as an Internet Explorer plugin that worked in the context of the page itself, so all the DOM nodes were accessible within the page once ThirdVoice "fired", which, in a 1999 browser, was a very noticable event. If in fact it was still around today, it would be notorious for breaking otherwise-functional websites by changing the DOM in ways the website didn't expect. They also had the usual bad HTML filtering, so their product allowed you to insert cross-site scripting attacks into arbitrary sites. I was able to convince <p onmouseover=""> to work, another group discovered <bgsound src=""> worked as soon as you opened the note with no further interaction. Fun times.

(And I count myself vindicated by time; this approach never took off, and the one I advocated did, which was sites that aggregate links and collect comments and communities on their own. Sites such as, oh, say, Hacker News. Or less crowd-panderingly, Slashdot, MetaFilter, Digg, Reddit, etc. Please forgive me if I take a moment to toss that in here, ten years later.)

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.