The Lair

Ok, so the work I needed to do this month is mostly complete. Therefore, time for some entertainment. I pushed out a couple of improvements for Grapevine.

Added The Academic to the list of sites you can scan/read off the Grapevine. A couple more in testing, should be ready for deployment soonish. I’ll probably drop support for some of the clear duds (Newsvine Vine and LankaPapers don’t seem to be updated anymore, pointless keeping them on).

Also allowed changing display preferences without fiddling around with the URL. (Thanks to everyone who prodded me about it). Just click on the link and the Grapevine page will reload respecting chosen preferences. Remember though, that as with Ach; preferences set on the .com site will be ignored on the .org version of the site.

So, inbetween doing real work; I’m still tinkering around with Grapevine. There is a reasonable amount of “oh, ok. cool. I did this.” satisfaction (dare I say smugness) at having gotten it to a point where I can just leave it alone and it will (mostly) just work… Most fun, however, has been the weird and wonderful stuff I’ve had to do to make it actually work the way I want…

Some sites don’t have RSS feeds

Horrific, I know… but true. There have been various cost related arguments made elsewhere about the expense of an RSS feed. I saw a case being made here (can’t remember where I saw that link, but it was some local blogger). Extremely popular sites like Slashdot implement their own policies on how often an RSS feed can be updated by a client (Slashdot’s policy, original story is also worth a read). Even more interesting is the prediction made here that RSS would become mainstream in two years. The article was written in 2004. I think the prediction has come true, more or less… perhaps even sooner than the authors dared imagine…

Read the rest of this entry »

grapevine. The Ach empire is expanding… *snigger*

I think I can understand why my mother was so fond of the phrase “idle hands are the devil’s tools” when I was younger… I used to hear it a lot. Instead of doing real work (which, this month involves the wonders of the JMF among other things); I’ve been building what I call the Grapevine. It was suprisingly easy to cobble together too.

Nothing fancy… Grabs a few news feeds from mainstream SL/South Asian sites and throws them into a single page for a quick read/scan. I opted to use a slightly different layout from the traditional Ach “river of news” style but I think this works better for the type of feeds included. Like most of the other things I do, this is nothing more than another personal itch that I wanted scratched. Of course, the guts of this will probably evolve into Ach2 but that’s all way into the future … then again, I might start next weekend if I feel really bored.

Comments, suggestions, any more news feeds anyone wants added (so long as they’re vaguely SL-an in content)… lemme know. There is at least one other feed that I need to include in that list (it’s already there, but not visible to the general populace yet) and I’m working out a small kink that prevents including Google News and Yahoo News feeds from being added at the moment. Updates every half an hour, more or less.

In a slightly related vein, I’ve long been ambivalent on the subject of mixing code with presentational elements. Some of the most popular webby languages (PHP, ASP, Coldfusion) employ this approach, even though there are templating systems that help avoid the inherent messiness of mixing code with HTML… There is a similar problem for JavaScript heavy pages. It’s tempting to put in reams of code in onclick handlers and the like and I’ve done it frequently in the past. The main reason I employ this approach is readability, strangely enough. The intent of an anchor or button is immediately apparent on reading the source code - even if there is a lot of JavaScript attached. The alternate approach is to load everything via DOM or XPath operations. This leaves the HTML of a page unchanged, all the additional JavaScripty bits are added automatically after the page is loaded into a browser. I can see why this is neat, but it’s so hard to figure out which element has JavaScript actions associated with it that I tended to avoid the approach. This is why I’ve been wary of using (otherwise nifty) libraries like Behaviour.

Last week (or perhaps the week before), I suggested the use of Thickbox at work. It solved a UI problem that we had but introduced this strange new library (strange and new to me, at least) called JQuery. After a fair amount of frustration - it is rather different from the conventional heavyweight JS libraries - I’m beginning to finally get it and it’s pretty damn slick.

It’s also about … oh … one fifth the size of the current effects library that I use so that’s a hefty saving in bandwidth and page load times too.