world changing

i'm curious, in the confused and exhausted state that i'm in, what few points i would consider to be necessary to change the world. here's a brief list that i'll be evaluating under more lucid conditions:

* legalize drugs
* legalize sex-work
* remove merit from education (i.e. don't delineate based on test performance)
* make all education free - at all levels
* require all education curricula to include farming
* require a foreign language (and that should exclude english in relevant locations)
* make all education wiki-based - so students can write their own textbooks
* put all political and public information online (wikileaks, sunlight foundation)
* make all religions pay tax (i.e. no tax benefits for contributions to religious charities)

that's my list of 9. i wish there was a 10th.

hope

the ilugc [indian linux user's group, chennai] have been having perl programming sessions at the workplace. they've been progressively declining in terms of attendence, and progressively evolving in terms of the quality of presentation and focus of the attendees.

today i spent a lot of time describing my visions and aspirations - basically the ashok jhunjhunwala vision - to the attendees. trying to instill the vision and leadership that they need to contribute to society here. i was hoping to not have wasted my breath.

today one of the attendees at the end of the class said he was very glad to hear my insights because he had never thought about his programming career as having opportunities for the local population. he had only thought about programming for programming's sake. this person works for jhunjhunwala's company.

i'm thrilled that he gave me this feedback.
and sad that jhunjhunwala has not instilled the vision into his own employees.

but i have hope (outside of the usual gymkhana club stupidity, which i feel is going to crop up regularly until i leave india).

dreams, fantasies and visions [continued from last post]

if you have not read the last post, pls read it.

yeah, it has nothing to do with this one, but still, might give you some perspective :)

of india's 1 000 000 000+ people almost 50% are under 30 (i can't remember where i read this, but let's assume it's true for now). this is an astounding figure. over 500 MILLION youngsters. it's either damn scary or damn useful, depending on how you look at it.

there are some downsides to it right off:
* there's a massive imbalance between men and women
* income disparities are huge
* education disparities are excessive

of the three the only real problem, at least as i see it, is the first issue - the gender imbalance is going to either result in war, or some other kind of political unrest unless men have a useful outlet. and that has to be driven by leadership and vision.

this brings me to the the other two points. provided the proper vision (and the corresponding leadership) i believe they are will take care of themselves. people are willing to work for very little - as can be seen by the web 2.0 revolution. they also can accomplish a great deal with their own self-effort given minimal education. see all the dot-com heroes. life in india is cheap enough to allow people to pursue their visions with minimal renumeration, provided they see the long-term potential.

if vision and leadership can be provided to the unwashed multitudes - the youth who have the energy, the drive and the ambition - i believe this nation has the ability to change the world. without vision and the associated leadership we're doomed to repeat the mistakes of previous empires, regimes and institutions.

today i have high hopes that i can contribute to putting the seeds of change into the mindsets of a few of the youth here.

but i fear that my optimism may be short-lived since the simultaneity of opportunity and fascist adherence to the idiocies of a bygone era are too apparent. specifically my previous rant about institutional ignorance such as that found at the "gymkhana club" (and their dress code) vs the recognition by youth here that there are possibilities for them to create magic that meets the needs of the current populance. i don't know if the youthful exuberance and desire to create magic can fight the bureaucracy-adled stupidity of incoherent and irrelevant rules which are going to be instilled on the youth - to make them kow-tow to those who decide how things need to work.

i realize this all sounds vague and distracted. i need to digest my ideas into something more concrete.

70 hrs and counting

my average work week since jan 1st has been 70 hrs per week. the last time i clocked that kind of hourage was back in the late 90's when i was young and stupid. my body does not appear to be able to deal with this kind of abuse any more.

i've long held that productivity decreases past the 35hr per week mark. i believe i have managed to conclusively verify this in my case, and in the case of people around me. at the very least efficiency drops dramatically after the 35hr mark. i think people can still get things done, but it's at a diminishing rate of return.

i have a chance right now to reflect on what i've accomplished (and not) over the last 2-3 months and here's the short list:

a) accomplishments:
* installed and running openfiler - need to get it ha'ed and clustered for full reliability
* running trac, and have customized it to some degree. need to hack it some more.
* running openHRM, which is just barely acceptable, and needs lots of hacking to make it a usable hr system.
* running subversion, and have the in-house developers using it regularly.
* running rsync to do regular mirrored backups, but this is still a far cry from ha!

b) failures:
* versioning-integrated filesystem. i had high hopes that i could have this implemented, but the lack of a high-capacity data storage system has hindered this extensively. i'm hopefully going to have collabnet help me with this.
* automated ftp synchronization. this is at most a 2-day programming task. that's if i had 2 days of non-distracted development time. which i've not.
* building a better rdf/html comparison tool
* automating the processing of image files to different sizes and resolutions. this is really a *very* simple job given the power of imagemagick

one of the things that i should have (but have not) added to the "accomplishments" side is instilling vision. i realized today that this specific function can never be underestimated. all workers, whether they be programmers, drones, whatever, need to have a vision of something greater that they are striving for. they cannot continue to do the mindless repetitive tasks that we ask of them day-to-day without the golden light at the end of the dismal dank tunnel. in many cases the tunnel ends up being the end itself - and people get dejected and decide that they just can't take it any more and quit.

this is basically the reason i've quit the last 3 jobs i've had. i did not get the sense that the person above me had the vision of doing something great that i could strive for.

i have been working on instilling a vision of what's capable in the current field i'm working in (e-publishing to be sufficiently vague) into some of the employees. i have 2 successes in spite of having spewed my spit at over a dozen. i think this is a high response rate :)

i know that this is of value because one of the guys, saravanan, came to me today and talked to me about the vision he has for automating certain processes which we spend inordinate amounts of time doing right now. and he kept grinning about what he referred to as my "dreams" of what we could accomplish. i'm quite confident that these dreams are well within achievable reality within the next 9 months. then i gave him an inkling into the real "dream", which i currently still consider fantasy, that i was aiming for. that got him very very excited, and i think it lit a spark in him to go create something amazing.

this leads me to my next post ...

catalyst mvc reducks on osx

Yeah, I'm back at hacking MVC after having blogged abut it over a year back. I tried the Rails and I lorved, but didn't have much of a chance to muck with it. Turns out that Perl, though icky in many respects syntactically, is just more prevalent and the user-base is just that much more active than Ruby. Seemingly.

Anyway, now I need to build a quick and dirty and maintainable (ha, oxymoron in most cases) database app that's web-connected for my current gig. It'll have to be maintained by mostly non-techies. So I'm thinking MVC is a necessity. The few techies who *can* work on maintenance duties are self-taught Perl "programmers". Note the quotes, and keep Chris Farley in mind.

Building Catalyst for OSX (my OS of choice at present, when my Ubuntu host is not available) is turning out to be much more of a nightmare than building a rails app ever was. The standard issue:

perl -MCPAN -e 'install Catalyst;'

bombed with many errors. Then I found out that macports had a lot of this already port'ed, so to speak. Doing some

sudo port install p5-catalyst-*

should have made for a lot of happy-happy-joy-joy (which, incidentally, dates me), but it too fails with silliness like:

Error: Target org.macports.activate returned: Image error: /opt/local/bin/corelist already exists and does not belong to a registered

So, I'm currently debating whether to continue the struggle to make this work, or just give up. Obviously I'm not going to give up, but this really is unfortunate. Why the HELL does it have to be so hard to install something like this? I'm trying to figure out if Apple is to blame, Catalyst is wonky, or my machine is just so mucked up with all the crap I've been installing that things are just totally out of whack.

Now it seems that the fact that I installed some of the modules using CPAN and the rest using macports is actually causing some confusion to one or the other (I guess their databases aren't too intelligent). So I'm having to encounter errors, then continue installation in spite of the snafu's and I'm hoping that this ultimately leads to an actually working install!

In any case, this is what I'm going to be having upleasant dreams about tonight!

catalyst mvc reducks on osx

Yeah, I'm back at hacking MVC after having blogged abut it over a year back. I tried the Rails and I lorved, but didn't have much of a chance to muck with it. Turns out that Perl, though icky in many respects syntactically, is just more prevalent and the user-base is just that much more active than Ruby. Seemingly.

Anyway, now I need to build a quick and dirty and maintainable (ha, oxymoron in most cases) database app that's web-connected for my current gig. It'll have to be maintained by mostly non-techies. So I'm thinking MVC is a necessity. The few techies who *can* work on maintenance duties are self-taught Perl "programmers". Note the quotes, and keep Chris Farley in mind.

Building Catalyst for OSX (my OS of choice at present, when my Ubuntu host is not available) is turning out to be much more of a nightmare than building a rails app ever was. The standard issue:

perl -MCPAN -e 'install Catalyst;'

bombed with many errors. Then I found out that macports had a lot of this already port'ed, so to speak. Doing some

sudo port install p5-catalyst-*

should have made for a lot of happy-happy-joy-joy (which, incidentally, dates me), but it too fails with silliness like:

Error: Target org.macports.activate returned: Image error: /opt/local/bin/corelist already exists and does not belong to a registered

So, I'm currently debating whether to continue the struggle to make this work, or just give up. Obviously I'm not going to give up, but this really is unfortunate. Why the HELL does it have to be so hard to install something like this? I'm trying to figure out if Apple is to blame, Catalyst is wonky, or my machine is just so mucked up with all the crap I've been installing that things are just totally out of whack.

In any case, this is what I'm going to be having upleasant dreams about tonight!

duct tape - not so universal after all

ran into a cultural idioma problem 2 days back. there's no such thing as duct-tape in india. that's mainly because there's very little duct-work in india. why do you need duct-tape if you don't have duct-work, right?

obviously those familiar with the universal applicability of aforementioned tape will disagree with that statement. i mean we need duct-tape to build everything - and duct-work is just the least of it!

so anyhoots, there's no duct-tape in india, so what the heck do u use as an allegory when you standard issue "it's like duct-tape" just falls on blank stares? i'm stumped. i thought maybe fevicol or cello-tape, but it's just not the same, now is it?!

btw, just for the record, i'm no longer a fan of the duct-tape. i have joined the legions of hackers who have sworn allegiance to the gorilla vs the duck, and i'm not looking back (even at $8/roll, which is just crippling).

moment of inter-zen

i've had a vision of what can be done with truly searchable video for a few years now, ever since i heard about the fraunhoffer institute's idea to create mp7 (i believe it was around 2000 or 1999, but can't recall exactly). one example scenario:
imagine being able to create a storyline, in text, and have it dynamically generate different video streams. take, for example, the outline: "children playing; sunset in calcutta; beach hawkers; calcutta street". imagine being able to stipulate this sequence, and have it directly translate to video segments which can be spliced together. take that one step further and imagine that this outline can create hundreds of different video streams - all from different cinematographers.
this is the first in a long list of uber-cool ideas surrounding a project called pad.ma. their about page states that they:

.. see PAD.MA as a way of opening up a set of images, intentions and
effects present in video footage, resources that conventions of video-
making, editing and spectatorship have tended to suppress, or leave
behind. This expanded treatment then points to other, political
potentials for such material, and leads us into lesser-known
territory for video itself... beyond the finite documentary film or
the online video clip.

Dense annotation is the key to making this happen - and this has been an elusive idea in video. it's even more elusive in audio. at least our contextual perception of video and text is similar. audio and text are entirely separate beasts that really need to have more work done to bridge the gap.

pad.ma is developed by 2 german hackers working with the alternative law forum, and a friend of ilya's and lawrence's (sanjay). the moment of zen here for me is that pad.ma is something i've been thinking about for a while, and ilya told me about it and i found out that it was an alf project, and told lawrence how cool it is, and lawrence forwarded by email to sanjay, who is a pal of ilya's and so just as sanjay was asking lawrence how to get in touch with me ilya's sending sanjay my contact information!

should be interesting to see how pad.ma is coded, since it'll all be released under CC/GPL.