tysond's Blog in Malaysia

With the latest security updates (including WMF out-of-band release) a lot of discussion has been going around on the web about "how long does it take to code a fix" and "how long does it take to test it".

If you come from a small scale software shop, or only have a couple of customers, it might seem like all fixes take about 2 hours and can be released the afternoon after you compile it.  After all if it doesn't work just apologize and try something else the next day.  Ahhh the good old days.

When I used to work on a compiler used by only a few dozen people, it took several hours just to run the automated test suite on just one platform, and gearing up for release took weeks of testing.  If there was a bug, if we were lucky we could turn it around in 24 hours.  Realistically we took weeks to fix most bugs.

One time when I wrote an online commerce site (many years ago) I had a bug go live that could reveal personal information (VERY VERY bad bug).  I stayed awake for about 2 days testing the fix to the bug (I fix the symptom in about a minute but I wanted to be very sure that my real fix worked).  It was terrible, as I knew there was no way to sleep until it was fixed.

But things are of a bit of a different scale if your site or software works in different languages, runs on different systems, support millions of users, etc.

Here's an interesting article about how much work can go into a "five-lines of code" feature.

http://blogs.msdn.com/ericlippert/archive/2003/10/28/53298.aspx

Interestingly, none of this thinking is really new - Brookes wrote the Mythical Man Month some time ago and really not a lot has changed in that time.  Old hands in software development know about this stuff.  But maybe entire generations have grown up without learning the complexities of large projects.  I personally think it would be instructive

Another good article here on fixing bugs:

http://headblender.com/joe/blog/old/001280.html

And a video on "bug triage" - how do you decide what to do with bug reports when they come in thick and fast?

http://channel9.msdn.com/ShowPost.aspx?PostID=26641

Ahh... all that complexity... it's almost enough to stop me from missing late night code writing sessions...  :-)

So it's about that time when many of you may be considering new years resolutions.

In the past, I've considered learning DVORAK keyboard, for the (supposed) advantages in speed.  Every year or so I consider it, but have never put it on the list of New Years resolutions.  Always on the "someday/maybe" list.

Today I stumbled across an article from Reason magazine talking about the controversy of QWERTY vs DVORAK, and sure enough, there's enough conflicting data out there on the web that now I must say I no longer believe there is a compelling advantage in one over another. A few percent increase in speed isn't going to help me get much more done in a day.

It's very interesting to read about how myths about QWERTY persist - surely if QWERTY is so bad our friends who have made the switch will be able to provide a nice study from a reputable source that shows a really clear study of advantages of switching?  Instead we have a real live internet mud-slinging going on with DVORAK sites rallying to the cause and QWERTY continuing to thrive regardless.  People are duking it out over reports that are 75 years old.  Surely there's a statue of limitations on this kind of thing?

So this year I think I'll look for New Years resolutions that are a bit more concrete in their payoff.  Learning languages, losing weight, saving money, more charity work, that kind of thing

I've often wondered if there is a business model betting against people's new years resolutions - you can I could agree on resolutions, I place a bet against your resolutions, we put the money in escrow.  Come end of year, we weigh you, look at your bank balance and credit card statements, that kind of thing.  If you keep your resolutions you win the bet!  Think of that incentive!  Of course if you don't... I keep your money. 

Somehow I think it would be a real hit.  Anybody out there doing it?  If there's a guy out there selling pixels for a dollar each surely someone is cashing in on broken promises?

It appears my grandmother found my video on BlitBlog.  If you haven't checked it out yet, BlitBlog is a series of 5 minute videos about Microsoft Malaysia.

Of course, being a grandparent means she can leave comments anywhere she likes and just write "Grandma" as her name.  My family is quite devious, they have a filter setup to see whether my name pops up anywhere online and will be immediately notified about it.  SMSes will arrive, the house alarm will go off, the cat will get an electric shock - whatever is required to alert my ever vigilant family that someone somewhere is saying something about Tyson.

I'm not sure if they have found this site yet but it's only a matter of time.  Once they find it, like teenage kids with a can of spraypaint and not enough to do, they leave their messages and names scrawled all over the place.

The good part about this eternal vigilence is that I don't have to call them to tell me what's happening in my life.  They just know.  And come the next family gathering, everybody else knows too.

So this evening I managed to spend the evening at a CRADLE networking event held at Modesto's in Sri Hatamas.

It was a good opportunity to chat to some of the entreprenurial companies in Malaysia about the great ideas and systems they are building.  Some of the cool things I talked about included a system for remotely locating and disabling your car (now it just needs to find it for you in the car park) as well as a company that offers English lessons via SMS (many Malay speakers are interested to improve their English and it's very convenient to do this via SMS).

I was pretty impressed by the way MAVCAP/CRADLE people introduced people with common business interests and encouraged them to get to know each other.  Some people did some very blatant marketing (good idea to bring your brochures along but perphaps giving them to everybody is a bit of a saturation effort!) while others just schmoozed.

You can read about the Microsoft's partnership with this program - it's a nice adjunct to our existing programs for startup software companies.

Next time I will remember to take photos

Very strange I haven't been blogging here because this site wasn't visible when I first posted.

Now it seems to be ALIVE - and has been for months.  I really should have paid more attention!

So perhaps it's time for a bit more posting.

I'm starting this blog as I've just moved to Malaysia and want to have a way for friends, family, colleagues and people I work with to keep up with what I'm doing and what I am doing.

It's been a few years since my family actually understood what I do for a living, so I'm not expecting everyone to understand everything.  But I guess they can just look at the pictures.