Swhack Weblog

AaronSw: A great OS is one that inspires you to do things Right.
AaronSw: It needs to have taste, and not cut corners.
sbp: A great OS is one that doesn't crash, comes fairly cheap, lets you play with its innards if need be, has protocol and location independence to a certain extent, is bugless, comes with Python pre-installed
AaronSw: It's not really possible to do great work in an OS that makes you sick all the time.
AaronSw: Try comparing some Windows apps with some OS X apps some time.

sbp: I'd suspected for a while that the Websafe palette is just a PITA. This article explores the World o' Websafe

AaronSw: This was how I first evisioned the Plex working (sorta).

AaronSw: Another oldie-but-a-goodie. Gelernter has moved away from the nitty-gritty and is now focusing on the problems of UI -- how do we get this technology into our lives. Man, so many interesting problems and so little time.
AaronSw: I had this epiphany (current UI sucks!) after reading Ted Nelson's "The Future of Information" in the days before school in the Summer of 2000. I put together a little hypertext after reading it, the only remains of which seem to be What's Wrong With Software? I really liked Ted's spiral calendar design -- it'd be very cool if Lifestreams ad
AaronSw: I love this line: "Any well-designed next-generation electronic gadget will come with a 'Disable Omniscience' button."
AaronSw: Lifestreams reminds me of Paper Trails: "These vapor-trails of crystallized experience will represent our first concrete answer to a hard question: what is a company, a university, any sort of ongoing organization or institution, if its staff and customers and owners can all change, its buildings be bulldozed, its site relocated ‹ what's left? What is it? The answer: a lifestream in cyberspace."
AaronSw: He also mentions the harness-human-power idea I came up with a while back: "Software can solve hard problems [...] by delivering the problem to exactly the right human problem-solver."

AaronSw: Describes the MIT Haystack project. A very interesting use of RDF to organize your personal files. Ties in well with Be's metadata filesystem and my thoughts about the Plex for desktops.

AaronSw: Very much like the Plex. I'm emailing them to see how we can work together.
AaronSw: See also their CETIS interview.

AaronSw: Bruce Sterling explains why our crypto situation sucks (ignoring P2P) and why the war on terrorism is a hoax.
AaronSw: It's definitely a fun read.

AaronSw: --
AaronSw: Our design sucks, but we'll build it anyway.
AaronSw: Our code sucks, but we'll ship it anyway.
AaronSw: Our system sucks, but we'll tell folks about it anyway.
AaronSw: And we'll try our best to make it suck less.
AaronSw: --
AaronSw: It's just like Dave always says. Your stuff is never going to be perfect, but you still need to get it out there. When in doubt, ship it out.
AaronSw: As Torvalds says, software isn't planned, it evolves.

AaronSw: A very interesting interview about how RDF, Dublin Core, librarians and Wikis can work together with Zope. I can only hope that Zope3 is flexible enough to let this stuff take off.
AaronSw: Man, Rael just has his hands in everything!
AaronSw: Although the story is rather linear, the building of these things is rather hypertextual. We're building the Plex when we don't have the Semantic Web, the Semantic Web when we don't have integrated metadata, integrated metadata when we don't have the two-way Web, etc.

AaronSw: (i.e. those not linked from /DesignIssues/.)
AaronSw: Anonymous, DistObjApps, More
AaronSw: Old DesignIssues: DosDonts, FunctionTraverse, Identified (especially interesting in like of the "semantics of Resources" discussions),
AaronSw: Intelligent_Navigation (essentially describes Google -- in Feb. 1995), LinkToLiving (shades of XPointer), MOO-WWW,
AaronSw: ManyIndexes, ProtocolProblems, Protocolcomms, RComments, TracingLinks
AaronSw: Note that many of these are linked from other places, and are not really "lost", but I include them all for completeness.
AaronSw: TimBL seems to have started a book in 1993-4.

sbp: """as Dyson points out, the first patent was granted in 1449, so maybe it's time for a bit of a shake-up"""

sbp: Using genetic algorithms to invent things - optimum design
sbp: An important precept is that you have a machine which can not only invent optimum things using genetic algorithms, you can also avoid the trees that are patented
sbp: cf. Genetic Algorithm
AaronSw: It needs to have taste, and not cut corners.