I keep getting spam about "WATER! THE NEW MULTI-LEVEL MARKETING!" It cracks me up.
* * *
My new thing for today is mini-updates. Often I don't have enough material built up for a real entry, so I'm just going to update this page throughout the day.
* * *
Yesterday was a rough day. I spent nearly fifteen hours in front of the computer (in total) and only cranked out about four pages of computer code for my interviewing system. The problem isn't an emotional block or anything like that, it's just too incredibly complex to grasp in my feeble mind.
For example, (1) each respondent has a number of databases, which in turn hold a number of responses to questions. Each survey has a number of respondents, however, we can only interview one at a time so the interviewing system will have to be separate from the main code.
(2) Sometimes we need to be able to pre-load information about respondents off a sample sheet, so do we pre-initialize these respondents, or do we create them on the fly when they pop their little heads up? It makes more sense to create them on the fly, as they're not actually respondents until we contact them, but then when I consider that we need a mass e-mailer -- where do we store the respondent information about the mailing like whether it bounced, if the respondent clicked through but didn't start the survey, and the other stuff? We want to be able to access this information all in one place, so maybe we should pre-initialize these respondents. It's an uncomfortable compromise.
(3) I was also hoping that I could use the system to write quick little scripts for "cleaning" the collected data like:
Which works just fine, but I worry that I might make a mistake or typo and damage $6,000 worth of interviewing, so we need a way to back up respondents, or preferably to maintain a pristine copy somewhere of each original interview, and I'd only be able to modify the working copy. A log of all the changes that's been made would be nice too, but might be technically difficult.
And these are just a few of the decisions I'm having to make, and all-together they influence the design. But what's driving me nuts is that I can't grasp it all at once, or work on a design incrementally, because I think I need working experience with the system to actually be able to evaluate whether it's a good design or a bad one. It's like designing an all-in-one gym without ever having worked out before except with free weights.
I feel like the Dyson vacumn guy, "5 years and 5,127 prototypes later, I was able to say that the world's first cyclonic bagless vacuum cleaner has arrived."
It doesn't feel right to build prototype after prototype before being able to say, "I'm finished."
Software design doesn't work that way.
* * *
I just had rainer cherries for the first time. They're yummy, I wonder why they don't export these down to California.
Also, a beer bottle is the perfect receptacle for pits.
* * *
A dentist sent me a little refridgerator magnet and a "welcome to Bellevue!" note a few months ago. I kept the magnet figuring that any dentist was better than none, and set an appointment which would have been for next Monday.
The initial visit would cover a cleaning, a set of x-rays, and a periodontal exam to check for cancer, gingivitis, and those other lovely ailments of the mouth. I was figuring it would cost about $150, which I remember paying back in Los Angeles. But no, I called them a few minutes ago to ask about prices and found out that all this together would cost $365! What a bunch of assholes! Lankakitten was supposed to go with me too, so we would have ended up paying $730 and not had anything physically done to our mouths other than the cleaning.
That's fucking evil to prey on people who just want to keep their teeth. I'm going to write their head dentist a nasty letter.
* * *
I just found out that my dad's hospital bill for his double-bypass heart operation and a week in intensive care is already over $300,000.
* * *
My new thing for today is mini-updates. Often I don't have enough material built up for a real entry, so I'm just going to update this page throughout the day.
* * *
Yesterday was a rough day. I spent nearly fifteen hours in front of the computer (in total) and only cranked out about four pages of computer code for my interviewing system. The problem isn't an emotional block or anything like that, it's just too incredibly complex to grasp in my feeble mind.
For example, (1) each respondent has a number of databases, which in turn hold a number of responses to questions. Each survey has a number of respondents, however, we can only interview one at a time so the interviewing system will have to be separate from the main code.
(2) Sometimes we need to be able to pre-load information about respondents off a sample sheet, so do we pre-initialize these respondents, or do we create them on the fly when they pop their little heads up? It makes more sense to create them on the fly, as they're not actually respondents until we contact them, but then when I consider that we need a mass e-mailer -- where do we store the respondent information about the mailing like whether it bounced, if the respondent clicked through but didn't start the survey, and the other stuff? We want to be able to access this information all in one place, so maybe we should pre-initialize these respondents. It's an uncomfortable compromise.
(3) I was also hoping that I could use the system to write quick little scripts for "cleaning" the collected data like:
foreach $Respondent ($Survey->Respondents) {
if ($Respondent->Database("Closed Ends")->Variable("QUOTA") == 3) {
$Respondent->Database("Closed Ends")->Variable("QUOTA",4);
$Respondent->Save;
}
}
Which works just fine, but I worry that I might make a mistake or typo and damage $6,000 worth of interviewing, so we need a way to back up respondents, or preferably to maintain a pristine copy somewhere of each original interview, and I'd only be able to modify the working copy. A log of all the changes that's been made would be nice too, but might be technically difficult.
And these are just a few of the decisions I'm having to make, and all-together they influence the design. But what's driving me nuts is that I can't grasp it all at once, or work on a design incrementally, because I think I need working experience with the system to actually be able to evaluate whether it's a good design or a bad one. It's like designing an all-in-one gym without ever having worked out before except with free weights.
I feel like the Dyson vacumn guy, "5 years and 5,127 prototypes later, I was able to say that the world's first cyclonic bagless vacuum cleaner has arrived."
It doesn't feel right to build prototype after prototype before being able to say, "I'm finished."
Software design doesn't work that way.
* * *
I just had rainer cherries for the first time. They're yummy, I wonder why they don't export these down to California.
Also, a beer bottle is the perfect receptacle for pits.
* * *
A dentist sent me a little refridgerator magnet and a "welcome to Bellevue!" note a few months ago. I kept the magnet figuring that any dentist was better than none, and set an appointment which would have been for next Monday.
The initial visit would cover a cleaning, a set of x-rays, and a periodontal exam to check for cancer, gingivitis, and those other lovely ailments of the mouth. I was figuring it would cost about $150, which I remember paying back in Los Angeles. But no, I called them a few minutes ago to ask about prices and found out that all this together would cost $365! What a bunch of assholes! Lankakitten was supposed to go with me too, so we would have ended up paying $730 and not had anything physically done to our mouths other than the cleaning.
That's fucking evil to prey on people who just want to keep their teeth. I'm going to write their head dentist a nasty letter.
* * *
I just found out that my dad's hospital bill for his double-bypass heart operation and a week in intensive care is already over $300,000.
Do you suppose maybe they were talking about multiple levels of water? I just blew your mind, didn't I?