Please take this as constructive criticism: I think Collapse/Expand icons, defaulting to collapsed, would be far superior to pop-up balloons (I thought these were generically called "tool tips" by now, regardless of context.) Clicking on a Collapse/Expand icon or link is much more intuitive than hovering your mouse. Also, the minute movements to keep a potentially jittery mouse (as we've all experienced at one point or another) hovered over a small target are much more RSI-inducing than the single action of clicking on a small target.
As always, the rule is to ask yourself: WWTD? (What would Tognazzini do?)
Have you tried this? Your description doesn't conform to how we implemented this. We use the entire headline area (not just the link) for the pop-up balloon.
We preserved the old mode as the default, for people who don't want this method. The triangle expand/collapse we tried extensively and found it too fussy, and thus not something T might D.
As Glenn noted, we tried the collapsible triangle approach, since it would be familiar to Mac users from the Finder. However, it doesn't translate well to the Web, in my opinion, and it particularly fell down when used to reveal article summaries.
The problem is twofold. First, we assume that people will want to view more than one or two summaries, unlike Finder folders, where you generally know what you want to see. So it has to be easy to use for multiple headlines. However, and this is the second problem, collapsible triangles require constant clicking - you'd have to run down the left edge of the page, click click clicking. And worse, the target zone for an expansion triangle is very small. And, to go one step further, it becomes easy to click on the wrong thing, making it even more troublesome.
Hovering over a headline provides a larger target and requires only that you move the mouse pointer, not that you click each time you want to see an article summary.
Adam says just above this article: "Twitter is thousands of answers looking for a question that no one cares about." Then he goes on in the next article to claim: "users can of course scroll to see more headlines, research shows that many people don't."
I think this new innovation falls into the same category -- an answer looking for a problem that nobody has any problem with.
So your readers aren't intelligent enough to know how to scroll? Seems like you have a pretty low opinion of your demographic, I think. Or maybe you'd be better off without these folks in the long haul?
Okay, I realize this is opt-in (at least here at TIDBITS). But I vote we stop this wholesale "improving the internet" and stick with basic experience as envisioned by Berners-Lee, which is easily accessible (and I emphasis the "easily") content and information. Enough already with letting "designers" and wannabe engineers and proprietary commercial interests tinker with the structure.
The vast majority of our loyal readers get TidBITS via email, RSS, or Twitter, so when we improve the Web site, we're aiming at people who aren't yet regular readers, or who are coming to the site to find an old article. So it's actually a different demographic.
As much as many would agree with you on the (to quote a snarky comment in Eudora by Steve Dorner) "trendy 3D junk," the unpleasant fact is that it's the world in which we have to compete.
Hopefully we're adding just features that make the site easier to use or more flexible, rather than just changing it to take advantage of the latest snazzy technology.
Thanks for the comment. We usually work on design problems on TidBITS that are things we're frustrated with (since we use the site constantly) or that readers have specifically raised with us as issues that make it less likely for them to read articles (or comment).
In this case, because it's opt-in, we provided a solution for readers (and ourselves) who wanted to omit seeing blurbs, and just read headlines. We don't know the size of that audience (although I can check cookies in the log to find out!), but it seemed like a reasonable improvement for those that wanted to engage in it.
For all other TidBITS readers, the site continues to function just as it has for a couple years.
As Adam noted, we now have readers who see (or hear) TidBITS in many different forms, so we try to tailor any improvements for the specific medium, like the Web.