Tuesday, January 22, 2008
Also an awesome Google sneak was in tonight's episode, I entered kambamthankyoumaam in to Google and I was rewarded with an awesome page that was straight out of the episode.
Obviously as a result of this posting I'm hoping to show up in that search results, which as of right now (11:55pm PST) is showing exactly 1 result, that being the one I linked to. Van city players, indeed.
Monday, January 21, 2008
Secondly is thine non-credit card status. There are laws that protect credit card holders from fraud and limit their liability to $50. While most banks are fully willing to refund and replace money promptly when fraud amounts are low, they generally take a bit longer when the amounts are significantly high. Most credit card companies can take upwards of 90 days to resolve fraud charges in the $3-5000 range.
The root of the problem is banks are not liable for fraud on credit cards. The responsibility lies completely with the merchant. In addition to the 3% transaction fees they pay, card companies regularly "charge back" to the merchant. Charge backs can be as high as 5% of the transaction cost. So now your cost of doing business is 8% - and there is no easy way to increase security.
The solution? Turn debit cards in to real debit cards that only take online PIN-authenticated transactions. This will reduce costs for everyone since the 3% credit card charge is built in to all prices even if you pay cash.
My friend Sacha points out that PIN transactions are not immune from professional fraudsters. This reminds me that the problem with "secure" systems that really aren't is convincing the powers that be of the legitimacy of your fraudulent transaction. A real life example is the "anti theft" car systems that aren't undefeatable - despite manufacturer claims. At this point you are in an uphill fight to regain your week's meal money.
Thursday, January 17, 2008
But instead of rail against that, I'll cherish it with the introduction of Ryan's Five Minute Thoughts. Designed to be readable (and writable) during a Youtube video, or (in my case) a build, they are the ultimate in cognitive adaptation.
Today's 5 Minute Thought (5MT): Comcast's Tiered Service.
TimeWarner is introducing tiered internet plans - presumably in an attempt to either restrain bandwidth usage or raise funds to, and this is highly speculative, build a better network. TW claims that 5% of the users consume 50% of the bandwidth usage. Bandwidth usage is surely only going to rise, with the rise of downloadable video from iTMS and Netflix. Right now video downloaders are a leading-edge case of internet users. Bittorrent and iTunes HD downloads is the future of media consumption.
Cable operators are in an interesting position of offering "unlimited" bandwidth that really isn't. Looks like TW has decided this is untenable and decided to take this leading-edge case as exactly that. Right now it's 5% that consumes half, but soon it will be 25% that consumes 99%. At that point you can't realistically cut off a quarter of your userbase. Given the dim outlook for cable companies and subscriber growth it might make sense to plan for a future where you can accommodate huge bandwidth users without cutting them off.
Core message: Watch out for the leading-edge case users. They might become a large chunk of your business, and if you can't scale to handle them, someone else will for you.
Monday, January 07, 2008
But I am not alone in this, a fascinating paper entitled "Where are the Software Engineers of Tomorrow?" talks about these worrying trends. The conclusion is a weakening math requirement combined with the "assembly" style of Java programming is ruining an entire generation of developers. Choice quote: "We are training easily replaceable professionals."
Demand better CS curricula! Even the ACM is not the answer here - their CS curriculum guide is part of the problem too.
Tuesday, January 01, 2008
I got a tripod for christmas. This is the results.
So now that it is January first, I guess a Project 365 update is in order. Basically put, I didn't complete the project. However, according to my metadata I took something like 7-8x as many photos in 2007 as I did in 2006. So something was a success there.