Kymermosst's Journal: Slashdot AJAXy stuff.

So, over the last few months Slashdot has introduced more AJAXy features. It's buggy crap, if you ask me, and lately it has only gotten worse.

Often, Firefox shoots to maximum CPU utilization and a few seconds later reports that a script on the page is still running, and asks if I want to kill it, which I do and then can no longer post comments and use other functionality.

Maybe I'm just becoming a Luddite when it comes to the web, but I happen to have *liked* having the web being just thing to read rather than what is becoming today: an interactive, buggy, often eye-torturing mess of "applications."

Thank God Javascript doesn't run in my Usenet news reader.

"Improvement"

[johndiii]

Few developers seem to be able to resist the urge to "improve" something that just works. I don't exclude myself from that group, though I try to be very conscious of exactly what I am breaking. For many, it's the urge to try out the latest technology. In many cases, it's an honest desire to make a product better.

The problem is that it is almost impossible to forecast usability without doing real testing with real users. That isn't just asking them what they want - that seems to end up making things wor

Re:

[Shadow Wrought]

You likely know more about this than I, but I'm wondering about the whole D2 thing. The server load makes more sense than any other reason I've heard for its existence, other than an "Oo, shiny!" facotr that it is "new." It seems that if they they really wanted to improve the "user experience" they would tackle things like making the firehose less buggy and making the search function more usable. So the question is, how much more difficult is fixing the search and firehose than is adding D2? In other wo

Re:

[Kymermosst]

making the search function more usable.

I think that the Slashdot search function works exactly as implemented:

SELECT * FROM stories ORDER by RAND()

Re:

[johndiii]

You've been looking at the code! :-)

Re:

[Degrees]

I think it's a little of both. From what I hear they did need to re-work the original code to use CSS. I can understand that.

Of course, once you are comfortable with the new way of presenting you data, the urge to tweak is going to be strong. ;-)

The idea of sending only the diff to expand a comment would be really appealing. I don't know which came first, but I can see where the idea of saving bandwidth would be a driver to take the interface to the next level.

I too run into the problem that large pages