As I noted before in Tags and Divergence and attempted to address in Emulating Errors for Tag Convergence, tagging as practiced today could use more convergence features. While thinking about this and wiki-related problems, I came up with stickers.
I came up with the idea of stickers when I remembered the opening chapter of Snow Crash in which a girl tagged a bad driver's car with a sticker. Stickers on wiki entries? Stickers on Flickr pictures? Stickers on links?
Stickers are like graphical tags that users can attach to text or images. While textual stickers can be applied likes tags are, similar to the way adjectives work, I think graphical stickers offer better user experiences. An open system can allow users to create custom stickers and variations of stickers (like icons with modifier pieces) to help users create a graphical language. Time and effort needed to create new graphical stickers is not a liability but a convergence feature.
I think the best way to use stickers is to combine it with limits in availability and time. So a user gets N number of stickers of various types to start with and will get M more per week or month and each sticker type has specific time limits (meaning they come off after a while).
I am still not done thinking about stickers but I thought it was interesting enough as is.
Bank of America just announced the SiteKey service, a very effective protection against phishing and pharming. The service will launch in Tennessee next month and roll out across the country by the end of 2005.
SiteKey service is based on the patented 2-way 2-factor authentication solution from PassMark Systems. For the most part of last year, I've been helping them build their core authentication technology. This year, my focus at PassMark has been the forensic technology.
Oh yeah. Hurrah!
News.com has a screenshot of a SiteKey screen for those more visually minded. That was actually a screenshot of this page at BoA.
uitags is an interesting open source project that tries to package solutions to common webapp UI problems (i.e. blog post/comment input and session timeout) into a taglib. I think what they are doing is valuable but they need to allow developers to try alternate, possibly browser-specific, solutions beside their own.
For example, telling the user that session is about to expire is not very helpful because the user has only three choices, a) do nothing, b) submit incomplete form, or c) manually save incomplete content and then try again.
A more friendly solution monitors user activity to decide automatically whether to transparently rekindle session or save the entered text (without formal submission) using AJAX techniques. If saved, saved input box can be replaced with a friendly message with a link to renew session and content. If saving content scraps at the server is not an option then there are other techniques such as client-side Flash Shared Object storage.
This weekend is the start of our camping season this year. This year, the first camping trip will be at the Pfiffer Big Sur State Park. We've been that five times or so but I have no problem with repeating a good thing. My wife, on the other hand, wants to try new camps. Sheesh.
My only worry is that my cigarette craving might get real bad during the trip. Last time I quit smoking for longer than a week, I stayed away for 6 months but I started smoking again while camping at Clear Lake. Oh boy, it was bad. I chewed through a full box of nicotine gum before throwing the towel in. I hope I fare better this time.
Benjamine Booth updated his simple recipe-style RCP tutorial for Eclipse 3.1M7. FYI, RCP is a platform for building Eclipse-based rich clients and Eclipse 3.1M7 is the first version of Eclipse 3.1 releases to make it relatively painfree to create RCPs. Before it was like laying dodo eggs which is a problem if you are not a dodo.
BTW, eRCP might interest you if you are into mobiles. eRCP is embedded version of RCP so that you can run java apps using SWT UI on mobile devices. Looks like Motorola is one of the sponsors behind this project.
At 43, I now have to work harder to concentrate. 11 steps to a better brain sounds good to me.
Modafinil can keep a person awake and alert for 90 hours straight, with none of the jitteriness and bad concentration that amphetamines or even coffee seem to produce.
90 hours?!? Longest I went without sleep was three days and, at the end of it, I threw up before plunging into ZZ land. Not sure if I actally fell asleep or simply fainted my way into it. I am happy to say I can't duplicate that feat now but wouldn't mind giving Modafinil a try. Mixing Modafinil with Levitra would be interesting too. LOL.
My tax dollars are being well spent by the Department of Homeland Security to protect America against copyright infringement violators. Huh?