
You know you're on to something when changing your home page interface sends the entire Internet into paroxysms of interface nitpicking.
"But I liked the tabs!"
"This new, leaner design just shows once again that Google is head and shoulders above the rest of the search engine crowd. In fact, I'm going to name my firstborn child Google."
For my part, I didn't really notice the change until it was pointed out by every weblog RSS feed in my aggregator... and even then I didn't find it all that earth-shattering. I mean, you go to Google.com, type in your query, and hit enter. If you need to search Groups or Images, you click those links first. It still works the same as it has for as long as I can remember, and they've added Froogle.
The links are in the same order and close enough to the same position that the tabs used to be in... does having to click a text link instead of a tab really make Google suddenly unusable? According to some of the people on Slashdot, Metafilter, et al who clearly missed their calling as Google interface engineers, it does.
I also don't understand complaints about the new advertisement formatting; I rarely give Google's ads so much as a glance, even with the new format - the interface is completely consistent. Ads live in a skinny column on the right, so unless you're looking for a sponsored link you know you can safely ignore that side of the page no matter how much its links may look like search results.
Really, it's not The End of the World.
My parents got their first relatively modern computer in early 1996, a Compaq Presario with something like a Pentium 75 and a 1.1 gig hard drive, running Windows 95. My Dad to this day has not been bitten by the Internet bug, but Mom most definitely has. During the six years I spent on the other side of the country, I'd say that more communication between myself and home took place via instant messenger and email than did by phone. She cut her teeth on that Windows 95 system, and in 1999 it was replaced with a nice Dell Pentium III running Windows 98.
One new hard drive, OS reinstall, and a failing DVD-ROM drive later, that computer is still running Windows 98, with only one log-in. It's true that Win98 did implement a half-assed sort of user management arrangement, but it was always my impression that it caused more problems than it solved.
So, one user account on the computer whether it's Mom (which, 99% of the time, it is) or myself or one of my siblings checking email or doing a bit of surfing while visiting the homestead. That means that my poor Mom inherits the after effects of every single OS/software configuration tweak or application experimentally installed by one of us... document types that used to open in one application will suddenly open in a new, unfamiliar one, or not at all. Things behave just *slightly* different, and since nobody really documents what they changed when, it's hard to put everything back the way it was.
Anyone who's owned a computer for an extended period of time knows what it's like to have everything just so, only to have something like the RealPlayer installer come along and take over all of your file type associations, and install 18 different apps that all run in your system tray. It's damned annoying!
Since returning to the east coast I often find myself fielding the tech support IMs resulting from these issues, and I've adopted a very hands-off approach to Mom's computer. I touch no control panels, no preference panes, and install no software unless it's warranted by the troubleshooting I might be doing.
I am hesitant to recommend that my parents upgrade to Windows XP (especially with the $200+ price tag for a new license), but I know it would make a lot of these issues go away if my brother, my sister, and I each had our own accounts and preferences that were completely compartmentalized from my parents'. I use my PowerBook conspicuously and talk up Mac OS X at every opportunity when my parents are around and the subject of computers comes up, and if I get wind that my parents are thinking of buying a new computer I'll go into car salesman mode if it means I might get them to switch... so many little obstacles and annoyances that come from day to day Windows usage by non-power users like my folks would just go away in OS X.
I've managed to use *nix-based operating systems for three and a half years without taking a written stand in the never-ending Vi versus Emacs flamewar, but having just been forced to use Vi to edit a crontab, I would like to go on record as saying that in the 21+ years that I have been using computers, I have never had such an instantaneous, overwhelmingly negative reaction to any other piece of software. I simply can't stand it, and it angers me when I'm forced to use it.
It was true the first time I tried Vi when getting into Linux in late 2000, and it's still true almost four years later. Using Vi makes me want to hit something. Why can't I just move the cursor, enter some text, or delete a few characters just like I can with pretty much every other editor written in recent memory?
Well, I know there are well-established historical reasons for Vi's infuriating separation of command mode and insert mode, but that doesn't make them any more tolerable for me. I'll also be the first person to admit that Emacs' key-bindings are arguably just as obtuse as Vi's editing model, but at least in Emacs you can open up a document, move the cursor and insert/delete text pretty much as you'd expect.
I don't buy the classic rebuttal that "People who hate Vi just can't be bothered to spend a few hours learning it," either. I don't care if Vi has a small memory footprint, or how fast it is, or how powerful it is once you stop being a wuss and pretend to enjoy incessantly jumping between command and insert modes. I think this quote from 6 years ago sums up my feelings best:
Below is the pattern used in this card, showing the radial grid used to lay the piece out.
The fisheye lens effect would be further enhanced if the circles of the grid became progressively farther apart as they spread from the center, but I can't think of an easy way to accomplish this in Illustrator.
Building on yesterday's experiment with oblique origamic architecture, I tried using a grid whose vertical lines radiate from the same origin as the horizontal lines: (The Illustrator 10 "Polar Grid" tool is very handy for this)
The radial lines do compensate for the distortion caused by the angled horizontals, which makes the end result much more subtle... at first glance you might even assume the above design is a standard 90 degree card photographed with a very wide lens.
(Again, pardon the floppy paper, mangled folds and rough cuts - this is a proof of concept model)
I recently saw some of Hannah Biemold's oblique origamic architecture by way of the Evermore Origamic Architecture mailing list, and decided to have a go at the oblique technique myself. (Please pardon the flimsy paper and rough cuts - this is just a proof of concept)
The key, as Hannah mentions on her page, is making sure all of your fold lines converge at a single point, which I think needs to be on the main fold line of the card (although that point need not occur within the bounds of the card itself.)
Once you have the grid laid out, you can design just as you would a straight 90 degree card.
I like to lay card patterns out in Adobe Illustrator, but to get a better feel for designing oblique cards I drew this one by hand - here's how the back of the card looks without having erased any pencil lines:
It's especially interesting to see how the vertical lines of the folded card become distorted as the distance between the horizontal lines increases. With some experimentation it should be possible to figure out how to angle the vertical lines to compensate for some of the distortion.,
I'm currently trying to familiarize myself with Zope as a tool for building and managing web applications.
Zope is not, as is commonly misstated, "a programming language like ColdFusion", nor is it "like PHPNuke written in Python". It's rather more complicated that that - the best way I can think to describe it is "A self-contained, web-based application for building other applications which includes the ability to do embedded scripting like ColdFusion or PHP, which in turn allows you to build things like, but not limited to, PHPNuke."
I first heard of Zope a good two years ago, and after reading all kinds of gushing praise for it all over the place, I installed it somewhere and took a look at it, found it pretty much inscrutable, and dropped it.
I took a closer look at Zope last fall, and once again installed it on my Linux box. That time I even worked my way through the official tutorial, which uses DTML to construct a simple Elvis sightings database. Then I dropped it again.
Within the last month I've renewed my interest in Zope, having come to the conclusion that my own evolving attempts at a modular, extensible framework in PHP are always going to be falling short of where I need and/or want them to be. I can continue my one-man development efforts and maybe in a few years I'll have something I can call stable and more or less consistently usable, or I can invest some of that time in learning something like Zope, which has been around for a good long time and has an extensive developer community, and seems to have already handled most of the obstacles and use cases I keep encountering in my own work.
So once again, I worked my way through the tutorial. It's the sort of tutorial that kind of teaches you some core concepts, but mostly just teaches you how to copy and paste your way to the end of the tutorial. When you're done you have a form that submits to a database, and a page that pulls from the database, and you understand how it all works, but you see no compelling reason not to have done the same thing with PHP or PERL.
Obviously, the tutorial is meant to introduce you to mechanics, rather than the process of building a more complex application. I can sense the compelling reason for using Zope to build more complex web-based applications lurking right around the corner, but I'm having a hell of a time reaching the lightbulb-flashing-on-over-my-head point. This is the problem I've encountered with all of the so-called "frameworks" I've looked at since finding myself in the business of building web applications; the common goal of all of them is to make the process of building and managing solutions easier, but because most web frameworks tend to be scripted applications themselves there's a steep initial learning curve to tackle while you figure out how to build things through the layer of abstraction that the framework imposes on development.
Universally, the problem has been lousy documentation. I can usually get a technical sense of how a framework works, but that's usually where the documentation peters out. I see how Zope Page Templates work, and I could spend a few hours creating forms that submit to other ZPT files which then "do stuff" with the submitted info, but I would love to see a more advanced tutorial that works through database interaction and user authentication, covering those subjects in depth as they're introduced. Few things are more irksome than being told "Don't worry about what all this means right now" after a code listing and then getting referred to an appendix at the end of the exercise.
Short of paying somebody to teach me this stuff, I suppose there's nothing for it but to keep banging my head against it until that lightbulb goes on. The nice thing about Zope is the complete web based GUI it provides for building things within the framework. Most frameworks leave most of the repetitive dirty work of including libraries and correctly coding method calls up to you, and in the end you're still left with a pile of files in a directory somewhere. Everything is self-contained and consistent in Zope, and I expect that will make the crucial difference.
Note to self: Don't use lump charcoal in mid-summer when it hasn't rained for a few weeks.
When I began seriously researching banjos online a couple of years ago, I discovered that there are a lot of people out there who have graciously made wonderful MP3 recordings of themselves freely available for download.
In large part, it was this body of free music that fueled my original decision to learn the clawhammer style; I never would have known the difference between frailing and Scruggs style had I not heard it for myself.
I hope I've made it clear before I continue that I am grateful to all of these kind people for the time they've spent making the recordings and putting them online.
That being said, why doesn't anybody set the ID3 tags in the music files they put online? Any halfway functional MP3 software will let you do it, and it takes all of 10 seconds to set the names of the artist, title of the song, and album/track information if applicable. Genre I don't so much care about, since since everyone always rubber-stamps bluegrass/old-timey music as "Country" anyway.
Proper tagging can mean the difference between me keeping the track in my iTunes library, where I can easily find it by genre/artist/title, or throwing it in the trash because I won't remember what 'sndrp_fl.mp3' is 5 days from now.
If I keep the song in my library I'm likely to keep listening to it and coming back to your web site for more, which I'll be able to do because I'll have the "Artist" information right there in the file, and maybe even a well-placed URL in the "Comment" tag. Heck, I might even order a CD from you.
Sure, I could tag the info by myself, but I'm just the consumer, and I'm lazy... and if a nerd like me is too lazy to tag your MP3's, Joe Q. Banjo-enthusiast probably will be too.