Developing Upwards

Today was a very troubling and frustrating day for both myself and one of my best clients. This is my declaration of ownership for the my own failure to make today not happen. The short story is right after declaring the "make the site more stable" milestone complete and shipping out an invoice, the site spent its most unstable day ever being frantically put on stilts and duct taped to the wall by myself. For the long version, read on. I had already spent roughly a week and a half working on an impromptu milestone in the project to increase the reliability and stability of the site, as well as beinggreenlit to apply hours to better build, test, and deployment processes. This is a good thing and it still stands as such. Now, the site wasn't fragile before, but a couple incidences understandably gave concern about long term quality. We had a few instances of corrupt MySQL logs, ran out of space on ourEBS volume, and embarrassingly I've had occasion to deploy code and f...

I am running Windows 7 via VirtualBox, and I skipped Vista completely, so some of my comments might also apply to Vista and thus be outdated. Too bad. You can probably expect a few other short pieces as I find something I like and something I don't. So, we see Internet Explorer here trying to help you out and tell you the download seems safe. Of course, it also lets you report that the download is, in fact, unsafe! This will no doubt be fed back into their SmartScreen Filter service, and when enough users report something, future users will be warned on downloading whatever bit of malware it might be. What a great way to protect your users. Now, the only obvious place to report the download is right here, in the download dialog box, which disappears as soon as the download completes and you can open or run it and actually discover anything threatening about it to warn others about. My Windows 7 review will eventually be the composite of many small pieces. I'll build up a score ...

An update to my previous post : I so called it .

This is all about the strange and confusing state of win we see repeatedly today, where a brand or concept gains such control and mindshare that no one even recognizes them as a thing anymore. Few people think of Q-tips as a brand versus just being the name of a thing. Pepsi Co executives probably grind their teeth thinking about movie goers ordering "just a coke" when Pepsi products are prominently and solely for sale. Most Internet Explorer users go beyond not understanding what IE is, they don't even understand what a browser is! Today, I want to talk about something newer and more specific, and less sure. The direction is visible that Mercurial is being given steps (it is important to phrase it this way, as I'll explain) to not matter, and that is precisely why they will win. CVS still matters, which is precisely why it has lost so utterly in the imaginary battle for geek mindshare. If you are using CVS, it is important to remember that along the way, because it a...

This is just a quicky tutorial for anyone else hitting the problem I had: installing Google Gears on 64-bit Linux. Google has not released support for this officially, but are apparently working on it and while there have been no official beta releases to try it out, someone posted a 64-bit build on a forum post . This will give you an XPI, the packaging format used by Firefox extensions, but it won't work right away. Firefox needs some special instructions to make a link actually install something and the forum post doesn't include it. There are no obvious ways to tell Firefox, "Install this XPI at this URI," so what are we to do? It turns out that it will initiate an installation if you select the XPI from the "Open File..." option in the File menu. So, download it locally and then open it in this manner, and you'll restart Firefox with a working Google Gears extension. Enjoy. I've got to get back to work now. So little time to post these days!

The infamous Zune 30GB failures were traced to a leapyear issue, and apparently they use some code we can see in the Freescale codebase. Take a look at the following sample of code, which determines the year from the day number (counting from January 1, 1980). I don't know about you, but the infinite loop is immediately obvious.  On a leapyear, the main loop continues when days = 366  and the incrementing is never reached, because days > 366  fails. Am I naive to think that even a casual code review would have caught this in a moment? year = ORIGINYEAR; /* = 1980 */ while (days > 365) { if (IsLeapYear(year)) { if (days > 366) { days -= 366; year += 1; } } else { days -= 365; year += 1; } } UPDATE: Fixed formatting issues. It looked fine when I posted it, honest!

This is a two part posting, hopefully. The first part is a request from me to my Rails-using readers. (Do I have Rails-using readers?) Can you leave lots of comments telling a non-Ruby guy like me what importance the Rails/Merb merger I've been hearing about means for the wider web developer community?