Mouse Gestures in Javascript

christodd writes "I have become big fan of mouse gestures, a feature included in Opera, Mozilla, and MyIE2. There's even a plugin for IE. Other programs like StrokeIt and Cocoa Gestures are also based around the concept. I can't believe nobody else has thought of this before, but what about mouse gestures in javascript? Turns out that it is incredibly simple to implement, and really handy for those 'feature incomplete' web browsers. Unfortunately, for the total user experience, we'd have to upgrade the whole internet..."

Comments

Browser Level == Better

[dolo666]

I'm all for the idea of faster, better, stronger ways of browsing. I happen to think that mouse gestures and browser level code should be based in the browser, and controlled by the browser. Like look at all the gestures you get with Mozilla.

(mo: Don't invent the wheel: we have it already)

The problem I forsee with the jscript use, is a misuse of the mouse gesture jscripts by unethical sites. Because it's the planet Earth, and The Internet, half of the sites will impliment this correctly, the other half will use it as a joke, or for annoying adverts (browser interstitials) and thus cause the whole thing to be crap.

If it's at a browser level, websites can't fuck with it. So ideally, browsers will want to add the ability to block javascript mouse control, and promptly add this cool feature at a browswer level. I'm all for the idea of mousegestures, but I'm against the ability to tell a website to fuck off using them. (mo: KISS).

Re:Browser Level == Better

[Zeinfeld]

The problem I forsee with the jscript use, is a misuse of the mouse gesture jscripts by unethical sites.

I think that is the problem with Jscript full stop. why netscape thought it a good idea to allow any site an almost arbitrary level of control over my browser is beyond me.

The idea of doing mouse guestures or any other browser extension in JScript, except as a demo is idiotic. The whole value of these systems comes from consistency. Apple do know some things about UIs, the value in the Apple UI is that every program work the same way and you don't have to spend lots of time relearning.

If I go to one site that has mouse guestures and then another that does not or worse implements them a different way ... yuk!

But back to the original issue, Jscript sucks. The command set should be partitioned according to the security considerations. Popping up a window has a significant security impact, it can be used to launch a trojan. The toolbars on the browser window are my toolbars, no web site should be able to disable them.

I use the feature of IE that allows Jscript to be turned off by default and enabled selectively site by site. But this is not as effective as it could be because you often come across idiotic sites using jscript for everything - including navigation. The idea being to force the site designers idea of a user interface down the user's throat.

Insert code with proxy...

[KrispyKringle]

Or you could have your filtering proxy (like Proxomitron or Privoxy) insert the JavaScript code on every page. Though personally, I'd just use a browser that suppots it.

Re:Oh yay!

[DeadSea]

I tried some of the mouse gestures on that page and they conflict with other things for which I use the mouse. For example, try selecting a paragraph of text on the page. It activates one of the mouse gestures causing the page to scoll.

Really annoying.

Pie menus

[Russ Nelson]

Actually ... mouse gestures are better implemented as Pie Menus.
-russ

Why this is stupid.

[cybermace5]

Honestly, it's a pretty stupid idea. Let me illustrate why: scroller mice. Once you get used to a scroll mouse, then you have to use a computer that just has a normal mouse, it's a major pain.

If you depend on every web page to implement mouse gestures, then you'll get this effect from page to page while you're browsing! It would be annoying to no end. And it's not an easily visible thing you can check for, unless each web page also uses some kind of cheesy "Gestures Enabled" logo. And each site might implement it differently, so that strokes mean different things from page to page. I repeat: stupid idea.

A user interface tool should be just that: part of the user interface. Just like a keyboard or mouse, gestures take time to become accustomed to. A user interface feature needs to act the same way no matter what you're doing.

Re:RSI

[rotomonkey]

Actually, my personal experience with mouse gestures has been the opposite. I have been using them fairly extensively in Maya for about six years now without problems. Because the gestures need only be approximate, I can execute them with a much more ergonomic arm movement than I can when I need to get the cursor to a specific pixmap to open a menu. In addition, the mouse actually travels considerably less distance when gesturing.

Granted, Maya can be an incredibly complex interface, with common tools appearing on a sub-menu of a sub-menu of a menu. You could make a strong case that a browser is a much simpler interface that doesn't need it, but I've found that I do start to develop wrist problems when I'm repeatedly using that nested menu item I'm too lazy to make a gesture for. When I'm gesturing, everything is fine.

For a pure coder, who spends a majority of their time pounding keys, gestures might seem frivolous, but if you mainly mouse, as is the case in graphics work, gestures can actually help reduce RSI.

Re:"next page" tag in HTML?

[Idarubicin]

Opera has already implemented this functionality in its newer versions. There's now a Fast Forward button on the toolbar. It's pretty good at figuring out what the next page should be. In Slashdot's site, it pulls up the comments for the next story; on Google it returns the next page of results. I don't know anything about the algorithm behind it, however.