Pat Gunn (dachte) wrote,
Pat Gunn

  • Music:

Kaiser Dukkha

This is my last day in Miami for a conference.

Hotel Hilton: sucks. For a computer networking conference, having the entire conference network sharing a single T1 with very few APs (all unreliable) results in many frustrated geeks and broken demos. Random sounds of construction during talks and the hotel being in the middle of a slum (only restaurants in walking distance are a Checkers and a Burger King) do not help.

Last night some of us went to the beach, which was pleasant. It's kind of nice having people to do things with, even if it's only temporary.. and swimming in the ocean was pretty nice too. Afterwards we went to a cuban restaurant with very limited vegetarian selections, but that happily funneled me towards the black bean soup, which was cheap, a "local specialty", and excellent. My mild embarassment at ordering a cosmopolitan was lifted when I spotted someone else in our group having one too.

There were also some cuban girls around where we were eating, one of whom I was attracted to in just a physical sense - that was a bit odd - I almost never have significant attraction to people unless I get both physical and mental signs and know them for awhile, and in the light of my current level of depression and persistent crushes on a few people I know, it was weird to feel that kind of longing. I think maybe given these circumstances, distant from my everyday life, with good weather, beaches, and company-without-baggage-of-people-actually-knowing-me, I'm feeling a bit better. If only I could hold onto this, I'd bear the not-terribly-interestingness of much of the conference and the lousy internet - it's not good enough to fix myself much, but it's a nice pause in what feels like a decline. Having a nice long hot bath made a good end to last night.

I've had a number of daydreams here of orthoganal directions, maybe a bit like Stephen King's 「The Langoliers」, or the idea of stepping into a slanted mirror and having my dinner there at the restaurant.. maybe it's more of a coming in from those directions than going out - a reminder that random points in a more-dimensional space are lonlier than those in a less-dimensional one, but n-dimensinoal points don't see the full space..

It seems there's a lot of interest in embedding OpenVPN into various parts of the network testbeds, which I'm glad to hear about. There's also finally some dissent from what was the overall feeling in the community that Layer 2 tunneling is the most interesting thing in the world.

Other discussions: forcing Cisco to adopt OpenFlow, which is difficult (against their economic interests to reduce cost-of-switching) but worthwhile (both in terms of standards - doing what snmp should've done, and added capabilities). Also, some concern about cloud computing a la Amazon and Google, incompatible APIs for that kind of thing. In general, a lot of people here "get it" that interoperability is something that's a major concern that we will always fight vendors of various services about. It is not enough if your app is opensource if it builds on an API where there's only one vendor providing the other side - upgrades happen, and strategic "no"s happen. In a way, Stallman was right to point out these things as a concern (and trying to address them with the GPL3), but licenses only go so far. If you build your app on Google or Amazon, and they either upgrade their systems and break your API (or cancel the project) or decide to compete with you and pull your plug (in the hard or soft sense), you're not going to have anything useful in your source if you have to try to replicate their architecture (patents?)..

This conference's gift-gimmick is a combined mini-flashlight/laser pointer. Nobody here can stop playing with theirs, which is a mark of a successful gimmick. It's useful too - good to point at things on other people's projected slides when asking questions.

There are people here who look rather similar to .. my uncle David, Gene Wilder, and a few others. ¡Uncanny!

April Fool's day is kind of irritating for people who follow world events and news of various sorts. It made the day feel like a moderately-out-of-the-way article on Wikipedia, to the extent that I could eke any browsage out of the broken networks here...

I am amused that here, like everywhere else, there are intellectuals of a certain kind who stand around shaking their head about how sad it is that government plans are often hampered by investment-type-people looking to profit off of every programme, manipulating every solution for their personal profit because if they don't everyone else will... but these same intellectuals immediately cry fascism (omit regular rant about how idiotic that use of the term is) when legal or structural changes are suggested that would prevent this. It's like they love the problem.. or perhaps they like the complaint.

Daydreams about fiction:Big difference between systems of magic in fantasy books and modern technology is that no large-scale systematic exploitation of these abilities is possible. Wizards are "superheroes" with inherent powers, rather than engineers of generic facilities in the universe. If we were to want to design a system where that property is preserved and make it well-thought-out (a magical "physics" resistant to science-commodification), what would it look like?

Preserve: training in learning wizardry, because of long narrative tradition (sorceror's apprentice, so on).Danger: Universities of magic - we want to avoid that Pratchett-esque kind of thing because if people had or happened upon a progressive-commodify-it science mindset, they would eventually push it forward into recreating science, probably.We do sacrifice the coolness of a University of Wizardry, but that doesn't really fit the genre anyhow. Daydreams of that can be detached from this effort/continuityIdea: Combination of personal particulars (personal-nonchangable interface with 「hidden APIs」 of the universe) and mentoring makes one a wizard, and a partial match with one's mentor is "good enough" for learning the basics. Buildable mechanics are based on those personal particulars, and not generic or significantly abstractableEval: This gets us part of the way in that it encourages diversity and discourages commodification, but there is still an exploitable hole - persistant school that would exclude all people who are not within its scope of sameness for those personal particulars. We don't stop the emergence of a single dominant school with a large fringe of self-taught amateurs outsidePatch: Strong subtlety in personal particulars at high levels so no more than the basics can usually be transferred in mentor-mentee relationship. Perhaps slowly shifting APIs of the nature of things that partly "key" to one's establishment/training when it is done that unpredictably alter the nature of the craft.

I will try to resist analysing this in terms of irritation at software engineering types and that Capability Maturity Model rubbish.

Fennec foxen are amazingly cute.

I like this site, which attempts to rate public statements for truth. Their detailed analyses of each statement are sometimes interesting..

Being here reminds me a bit of the way things used to be for me.. Tonight's return to harsh reality looms... It's so tempting to just pack up what I have here and walk out of the life I've known.


  • Testing functions in Perl

    (Nothing particularly profound or my-idea-centric here, and I was tempted to post it to my personal blog instead, but it's worth trying to learn…

  • Abstract strategies for abstraction

    There are a few purposes of abstraction in programming; one of them is to construct a uniform API that is independent of the backend that can work…

  • Statistical Software Components

    A few months ago I mentioned my big library of useful generic C/Perl functions (libpgunn). There are plenty of other general-purpose libraries out…

  • Post a new comment


    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded