December 2003 - Posts
browsing through some stuff at science magazine online,
Science, 19 December 2003: Breakthrough of the Year . dark energy causing (apparent?) acceleration of the universe expansion, and left-handed materials that can lead to an inverse-doppler phenomena. i dunno, just making connections here, there still could be sources of observational anomalies that haven't been fully considered.
length of day is all uphill from here. ramblings...
- it's time to get back to my reading activities (mags, books, newsletters, all fields). i'm done enough with my web site for now; basic content "channels" are in place, and the basic tools and structure are there to do any changes that i might want rather quickly. still a few outstanding larger projects, though. but the reading is what gives me ideas, which turns into content here. not much of that so far.
- time to do some holiday visiting again. lazy over here; since thanksgiving, been mostly sitting in front of this dang machine, playing around with code and stuff. thought about heading to maryland, but will just go up to michigan, where everyone is turning up eventually. not a long stay this time. sounds like everyone is going their own way for new year's, so i will probably head back here by then.
- seems a former employer is still seeing rocky times. perhaps the money would have been nice, but it's sure good to have escaped those other headaches. of course, there are ways to do both. just shake my head, and move on. this has happened with several former employers, getting caught up and tossed out in the middle of transition processes. within a year, the companies belong to someone else. hmm.
- one thing i occasionally ponder while riding chairlifts is what would happen if a cable broke. well, something close to that happened at a local ski area this weekend. a large tree uprooted, fell onto a lift line, throwing some people around. a couple people were seriously hurt. i wonder what the fallout of this will be (there's a bad joke in here, i know). i had also been wondering if that area was open yet, thinking i might head out for some skiing if it was. sometimes laziness pays off ;-)
this one crossed my inbox today:
Could it be that Doctorates don't pay?
CNET News.com: What's up, Doc? Not the number of science Ph.D.s
The number of doctoral degrees awarded in U.S. science and engineering programs continues to drop, but women are earning a growing share of them, according to survey results published Thursday by the National Science Foundation.
About 24,550 science and engineering doctorates were earned by students attending U.S. universities in 2002, down from slightly more than 25,500 in 2001, according to the NSF, a federal agency that supports science and engineering research.
and of course, i can never resist opening my big mouth:
apparently, there's no demand for them, at least in the academic world, where it's normally the mandatory rite of entry. so there's the premium there, for "suffering" through years of graduate and post-doc work. the other two markets, commercial and government, have a different problem: they want phd's, but aren't willing to offer enough of a premium in compensation in the great majority of cases to make it worth anyone's while. it doesn't help that advanced degrees have been getting a lot of bad press lately. for the corporate world, it only makes sense from a compensation standpoint to get a master's degree or two - one technical, and one managerial.
which kind of brings us to the underlying problem here: a great deal of people looking for phd's seem to have the pay and position aspects as their primary focus. the frequency with which i heard comments like this in my recent brief encounter with that world was alarming. they think about careers and salary, and pick their field of study accordingly, sort of like playing the stock market with their lives as the single share they can buy - what will it be ? not really different than the majority of us, eh ? so maybe it's not a bad thing that these types are being demotivated somewhat. that way, maybe there will be less of an academic grind to actually get the good teachers and researchers we need in those positions, instead of a nasty war of attrition that breeds for the wrong qualities. but it is possible that the system is irretrievably poisoned already.
not that this is really new or anything. perhaps the problem has been that schools for years may have over-marketed the need for science and math graduates. which normally means bachelor's or master's holders, at most. the problem is that it is natural for the kids who have bought that message to buy the whole package in those fields - there, recognition and status come with the phd and university affiliation. perhaps the correct scholastic message to send should have been simply increasing math and science requirements for all programs of study, an adjustment in the base level of exposure to these fields for everyone, rather than an intensity of focus in electives for some. but that doesn't always work, because such technical aptitudes are not widespread - just declaring them so doesn't make it so. and then we end up with frustrated students, having exposed them all to something they cannot easily do. and that seems to be a taboo today as well.
it always bothered me as a kid when we did our talking about what we would be when we grew up, and someone would always say they wanted to be a doctor or a lawyer or a dentist, and we'd ask why, and they'd say "for the money". that will make cynics out of any halfway observant intelligent person. likewise in the sciences (i didn't find a lot of them, so there's no childhood memories of those discussions). so it seems that what advances we've made over the years - centuries ? - have been in spite of ourselves and our baser instincts, not because of them. now _there's_ an interesting alternate history twist - what if the profit motive had not reigned supreme ?
which brings us to a discussion on forms of government and economic systems, which this ain't about ;-) or is it ?
anyway, look at all the engineering folks that graduated in the 80's, and computer science in the 90's. promises were made and grand futures were painted there as well, and now we're sending their jobs overseas. and there's a corresponding issue with the proportion of foreign student enrollment in the us as well. _here's_ a question: what is the _real_ academic message we should be sending today, if we were really honest about things, as they are and as they are likely to be ? not "things as we want them to be", but "things as they are". ie, the best practical advice for survival in the world as it stands, not someone's idealistic view of it.
after that, one can talk about where they might rather want to be, and possible means of getting there. and that's not going to be anywhere near as simple as "teach more math and science".
there's lots more where that came from, and there's lots of loose ends in those comments, but well, there it is.
(where did amazon.com get their name, anyhow ?)
- first off, the wishlist queries aren't showing data any newer than the beginning of this month. this is a known issue on the discussion boards.
- for some reason their xmlhttp interface seems to have quit working when i run it from wsh. still works from the browser, though. i discovered they had some typo-ed header tags, nasty stuff, that might have been hanging the connection open and preventing further talk until it timed out.
- there were other comments on the boards about making sure to use valid developer tokens, as they were supposedly going to finally enforce them, starting a few days ago. however, i notice that the xmlhttp links entered in the browser don't need correct tokens. this kind of throws a red herring in the mix, causing more confusion with the other problems.
- apparently people are complaining about all kinds of stuff, like interface changes, reliability and performance issues, you name it. ahh, just in time for the holidays, let's break all our associates applications ! which, some astute person observed, will show up on amazon's bottom line, when all their associates can no longer funnel business to them. or wait - maybe they want to keep all the profits themselves ? hmm...
- so i decided to go the soap route, and since i'm using wsh, will use the ms soap toolkit 3.0. turns out that dang thing _still_ has problems with typed arguments when using its “high level” api. i first ran into that almost three years ago (with an earlier version, of course). so i had to use the “low level” api, and some namespace and type guesswork. so that works fairly well.
- however, i noticed that the xml docs returned from soap and xmlhttp are slightly different. now that's plain rude. and they still both are missing things like total page counts, which means you have to check error messages for each page, looking for the one message that indicates the end of the result set rather than some other error. wait - did i say “one” ? turns out the message is slightly different in soap and xmlhttp too. that's even nastier. so that required still more code changes.
- things might have gone just a bit quicker had there also not been a slight error in the soap response - starts with some digits, so it's not well formed xml and therefor causes the trace tools to cough it up.
- anyway, got everything working finally. or have i ? the original reason i was using xmldom.load on the server side was to work around a winhttp proxycfg issue, which seemed to stem from some dll hell problem. but i think the soap toolkit uses winhttp again. uh oh. well, i'll find out if it works tonight after the scheduled run. what a mess.
- doing all of this for a remote server doesn't make it any easier either. nor does tiny bandwidth. more like band-narrowth ;-)
a bit of a marathon session here. learned two things:
1. this would probably have been a hell of a lot easier in dotnet, if you can believe the examples of dotnet/amazon webs service use on the web. the soap toolkit has some major drawbacks. yeah, i know it's deprecated. but there's also no dotnet script engine for wsh yet. so something originally simple now becomes a full-fledged app. and does dotnet use winhttp ? i would guess it does. but that's not a soap issue.
2. perhaps using the non-ms soap tools that are out there would have worked better as well. however, i avoided them since there was no single download with all the \pieces needed. crappy packaging.
speaking of packaging, the ms soap toolkit is perhaps _too_ inclusive: more than half its size is due to the the presence of msxml4. like, hey, i already got that. and you also are forced to install it - even if it's already present.
so not much has really changed in terms of soap interop over the last few years. it's supposed to be a consistent glue for forming heterogenous systems, but it turns out the glue comes in all kinds of different flavors.
whoa, i forgot about this guy: james burke. well, not forgotten, really, just his name hasn't been present in my conscious mind in a while. but man, was his “the day the universe changed” influential on my younger mind. i had already started doing my broad based reading, and making connections between what seemed disparate information, but i really started to become aware of what i was doing when i saw his show on pbs. loved it, the way things always seem to link around in big circles, the multiply related parts in unexpected ways, etc. powerful stuff, a wonderful conception of the universe. anyway, he was mentioned in that gartner summary i mentioned in my last post, so i was like, cool !
apparently he's started up this knowledge web project, and was looking for volunteers, according to some guy at the conference. man, tempting. their vision is a retrospective historical take on knowledge that is similar in interesting ways to my initial ideas for my “history project”. although similar in many ways to a lot of other projects out there, their vision, and their, uh, “conceptual sponsor” seem to be uniquely advantaged. what i would like to see is a synergy of all these different groups pursuing different angles of the same project. it's a shame to see things so fractured, yet exciting to see the work being done, and finally depressing to realize that i'm not doing it myself. get off your ass, me ! ;-)
i can see my amazon wishlist building already ;-)
a james burke fan site
(note for the targetted link people: this is all done in .text's design mode, without dropping into the html view. so i have to switch over to start inserting “target=_blank“ attributes on all the links. there, are you happy now ? ;-p )
(to the tune of “in heaven there is no beer“)
wow, just got an email from someone about a recent gartner group application integration conference, and he mentions that in longhorn, there will be no com. now _that_ will be interesting. can't wait to see the replacement architecture. and what will the legacy strategy be ? i can think of several. at least in the projects that i have worked on, which admittedly were all pretty mickey mouse, the capabilities of com+ were pitifully underutilized. such a wonderfully rich set of functions. it seems that they will now be encapsulated in higher abstraction building blocks, (implemented differently, of course), which is probably what was needed all along (as anyone who has listened to me talk about the future of software engineering and architecture has heard before).
lots more on that topic, but what i really wanted to post about is...
ode to the average joe ;-)
http://www.stud.ntnu.no/home/alexann/
“in space, no one can hear you scream”
(source: anna)
similar to the "french military victories" google spoof, try this:
so _that's_ how they do that. hmm, sounds like fun. ideas ?
(search for “google bomb“ or “google easter egg“)
more fun to try:
“weapons of mas destruction”
re-read this scientific american article: Parallel Universes. a lot to argue about there, which makes it nicely thought provoking in a number of different areas. it's good to see so many different things mentioned in one place.
anyway, the thing that was bothering me, was with his “level I” multiverse concept: if the number of repetitions of a particular volume in the universe is inversely proportional to that volume, then don't we end up with a whole lot more than two thumbs per head ? so now i look at my thumbs kind of suspiciously - are you mine ? what did you do with all my other thumbs ?
they aren't talking. yet.
“goom bee” ?
416 isn't one of the standard ftp replies; see http://www.faqs.org/rfcs/rfc959.html (rfc640 is the orignal, and perhaps easier to follow)
perhaps this particular ftp server publisher invented their own code: http://www.wordspy.com/words/GOOMBY.asp, perhaps meaning “get off of my ...” whatever. but then the code doesn't really fit the rfc:
“4yz Transient Negative Completion reply
The command was not accepted and the requested action did
not take place, but the error condition is temporary and
the action may be requested again. ...”
“x1z Information - These are replies to requests for
information, such as status or help.”
and maybe i should just get a life ;-)
actually, i think it's my friendly neighorhood host playing with my head, and 416 is a “kick-off”code with a customizable message. paybacks are a bitch ;-)
now here's something for you ee types to puzzle out:
http://www.livejournal.com/users/quicksilvre/87715.html
direct explanations to her site, or explain here and link there, or there and here, or, um ....
(ps - is it a coincidence about the double-e's ? i don't think so.)
frosty the snowman gone evil. airgate skirt, hockey net shawl, and a beer belly. the head arrangement ? must have been some rogue teletubby gene. whaddya mean, snowmen ain't got genes ? oh right - he's wearing a skirt ;-)
how embarassing.
my niece comes into my brother's kitchen last week, wants to send pies to her cousin at saint john's college on march 14th. my triple-ms engineer brother, and me, the physics/math major and on/off grad student and software engineer, can't figure out why. his wife is laughing at us - that's pi-day. uh-oh, endangered geek credentials here. then my niece says they have to be square pies. again, we were like “why?”. as my niece patiently points out, everyone knows that “pie are squared”. except, apparently, for her red-faced dad and uncle.
our geek cards are revoked for the day. later we try and redeem ourselves by comparing exponents of pi and the natural logarithm in full public view. it doesn't work.
my heat was gone when i got back here yesterday - again. dang. this time the pilot wouldn't stay lit. and the last time i lit it, apparently there was some leak in the control valve assembly that caught fire and burned the valve and some nearby wires. total burned area was a few square inches. but i'm thinkin', damn, if that valve had gone, i would have had a raging supersized blow torch in the basement. wtf, who the hell makes gas valves out of flammable material anyway ? great way of making a small problem worse. so everything's shut down now, waiting for the gas man. again. at least they should hire a cute gas lady to send around. never heard of one of those. then i could get, uh, serviced ;-)
turns out that little space heater i bought a couple years back was a great investment. pop that into the "things that shouldn't be" barrel.
odd thought: i wonder if a hair dryer could be used as a space heater substitute ? aren't they basically the same things with different form factors ?