Monday, April 15, 2002

From Peter Drayton: However, when Dave says "...three useful ways to view the Web: 1. Search. 2. Heirarchy. 3. Time." I think he is over-emphasizing time as an organizational tool (maybe because Radio focuses on time-oriented content?). Information architecture people talk about there being 2 primary modes of navigation: 1. Searching. 2. Browsing. IMO hierarchical and time-oriented views merely faciliate different modes of browsing - other modes based on geography, rankings or proximity also exist. There's even a flow between the modes: Google displays the ranked results of a search, and one then browses the results. I think the lack of a time component is often a problem on the web. Nobody bothers to update their old pages to let us know what has changed, where new or better pages are, or even if the whole page is obsolete. If we ever get this holy grail of Google on the Desktop, I think time will play a greater role. The linking algorithm won't work, so Google must use a more traditional best match algorithm. But how recently a file was changed or a message was written is a really big deal to me when I'm searching through my own stuff. I wrote an inverted index for a large number of text files a while ago, with words being scored based on frequency and importance. But when I boosted the scores of recent documents, users found the results considerably more useful. In my mail archiving, I've given up browsing completely in favor of a magic bag approach (I wish I could remember where I read that phrase). You just toss everything into a magic bag and when you put your hand in, the thing you want is on top. I love that because when I save something, I don't want to spend any time anticipating how I expect to retrieve it. For web searching, Google is the magic bag. And now for my mail, I don't put things in folders anymore. Everything I want to keep is forwarded to a POP address I use only for archiving. I have a program that grabs everything from that address, saves it, converts each message to an HTML file, pulls out the attachments and links to them, and stores everything where Index Server will find it. So now I only search. Where I used to have a folder called "SVG" that I could browse, now I must search for "SVG". So what? 8:34:58 PM

Joel Spolsky talks about his .NET strategy [via Peter Drayton's Radio Weblog] First problem: we don't know enough about .NET to write good code. As usual in any development environment there are many ways to do any given thing, and we haven't quite learned the first way, let alone the second way. So the quality of .NET code that we can write is not good enough to ship. ... So our first priority is education, which we will accomplish by doing all future in-house and web-based development in .NET -- basically, all the software that nobody is paying money for. I like this strategy. I've been using it myself, every chance I get. A test data generator, an HTTP test driver, a mail archiver, various test programs. And of course every new COM component must play well with .NET (see Adam Nathan's book for a good chapter about that). Even for those of us in companies that are slow to embrace change, there's a lot we can do to be ready. But why (oh why) am I still trying to convince people that a C++ class is not a good unit of reuse? 8:18:05 PM

© Copyright 2002 John Sands.