...more recent posts
Here's a new Samsung flat panel display I'm keeping my eye on for MB.
Here's a massive XML-RPC howto.
Not counting vacations (and even counting some of those) this past week seems like I've been away from the web more than ever before. I guess I was ready for a little break, and then my monitor situation gave me the excuse. I read Bruce Schneier's new book "Secrets and Lies" which isn't really meant for me, but was fairly interesting nonetheless. Mostly though I've been thinking about what I'm always trying to think about: the next version of the mythic "perfect web publishing system." Of course this beast doesn't really exist because "publishing" means so many different things. Anyway, it's been a productive week. At least in my head. This slashdot article about this project aiming to make mySQL into file system really got me started. I'd always wondered about this possibility, but since I'd never heard it mentioned I figured there was some basic problem that I wasn't seeing. Maybe not. Now I'm thinking about my project under the name FFS (for Fluid File System - I stole this from someone posting in the slashdot thread,) or alternately as VSFS (for Very Slow File System) just so I remember the main problem I am facing. One quick (although slightly misleading) way to describe what I'm thinking about is to ask this question: What if every Get request to your web site (no matter what the URL requested) 404'd, except instead of serving a static page informing the surfer that the page couldn't be found, the 404 page is a script that parses the path info (of the requested URL) and uses the directory structure of the request to form an SQL query on the database and return a page with posts that seem to fit the request? Hmmm.
Still broken.
Sick of IT yet? Today the speculation turns toward sterling engines. And oh yeah, now there's a whole site dedicated to monitoring the event.
Nope, still broken.
I don't want to jinx my good luck by speaking too soon, but I've rebooted my computer 3 times today and the monitor has come back on every time. I'll try shutting down when I go for dinner and see if that works too. Not sure how it could have fixed itself (or even that it really has,) but here's my theory. There are two batteries in the monitor, one big one that powers the storage of all the preference info for the on screen display settings, and a smaller one that serves to just protect the most important low level stuff that allows the monitor to work at all. The problem I was having was "solved" by unplugging the monitor for > 2 hours, which I think was just draining the main battery and forcing some sort of internal reset which would allow the monitor to work again. But the problem was in the secondary battery which has a much longer life. This one had never run down during the unplugged times until this weekend when it was off for 4 days. Maybe this finally cleared out the problem? I can hope right? In any case, if it is actually fixed, this will be good evidence that procrastination can sometimes pay off, as I would have spent at least $200 to have someone else solve this (potentially) simple problem. Keeping my fingers crossed...
Interesting article by Eric Davis on wireless technology and the changing notion of space.
"Though we "know" that electromagnetic modulations of the spectrum are no less material than waves of electrons cruising along a wire, wireless nonetheless amplifies the experiential sense that we live and move in a world of invisible intelligences, a magic world verging on telepathy. Simply put, the more the physical apparatus disappears, the more we are simply listening and responding to voices in our heads."If you like this sort of thing, definitely check out his book Techgnosis. Worth a look.
In case you want another link to a good Jorn Barger explanation of why XML, and all container-based hierarchies don't really work, here it is. This is exactly what I was trying to think about this weekend. I might be getting somewhere in the sense that the question is coming into focus. Getting the question right seems like at least 85% of the battle (according to my top secret rigorous scientific testing....) Anyway, what I want is something like atomic storage of all "digital media bits" (as if they can all be classified together like that) and then a dynamic containerization (hierarchicalization? folderizing? treeifying?) of these bits depending on the context of the question I'm asking. Here's an example: I have a picture of MB I took this weekend. It is one of, say, 35,000 pictures I have on my computer. I store it in a directory structure that looks like this "/documents/travel/2001/january/montauk/pictures". This looks good. My hierarchy at least has several levels, so it's clearly a step up from having all 35,000 files in one folder called "/pictures". But now imagine that next year for MBs birthday I want to go back and pull up all the photos of her to make a little collage. In that case I'd would have wanted the picture filed in "/documents/photos/friends/MB". D'oh. And worse, for practically every situation I can think of, I would want the file hierarchy to have been different. So that is the problem, and the solution (at this point in time) seems to me to involve storing all the files (what I called "digital media bits" above) in a giant blob, or cloud, or .net, or whatever you want to call this non-hierarchical storage space. And then I want tools that dip into this storage space (I guess "cloud" is the most popular term right now) and construct container based (i.e., folders inside folders like /documents/travel/2001/january/) file hierarchies based on the context of the question I'm asking ("Select all pictures from january 2001" gets you one type of file hierarchy - namely, /pictures/2001/january - while "select all pictures of MB from the last year" gets you a different one - "/pictures/mb/2000/".) Of course, this amounts to something like teaching the computer how to think, so I'm not saying I have any insight into how to generate these on the fly hierarchies. But you have to have something to think about.
More on IT, the mystery product from Dean Kamen's DEKA research. Here's a patent (granted in 1999) that some people are saying is IT, although others are saying that this patent is for a wheelchair device which has already been released to the public. Huh? Looks like a pretty strange wheelchair if you ask me. In response to all the recent hype (which you can only suppose DEKA was somehow the instigator of,) they released this press release intended to deflate some of the more outrageous speculation. It reads, in part:
"DEKA is currently working on several exciting projects. The book proposal referred to one. However, the leaked proposal quoted several prominent technology leaders out of context, without their doubts, risks and maybes included. This, together with spirited speculation about the unknown, has lead to expectations that are beyond whimsical. We have a promising project, but nothing of the earth shattering nature that people are conjuring up."Maybe it is that weird scooter thing after all. $2000? It better be fast. Or something.