The online home for Greg Leedberg, since 1995.

Monday, September 15, 2003

Big day yesterday!

Well, yesterday was a big day.

I released Billy 4.0 (finally), which I have been designing for over 2 years, and programming, in some form, for about a year. It was actually somewhat emotional for me, as this project has been a big part of my life for so long now -- no other version of Billy was in development for nearly as long.

But it was a happy occasion, as well. It feels good to design something that is beyond you at the time, and then keep at it and follow through and actually see the project get finished and released. And, to wit, it had a somewhat successful opening day -- there were 331 recorded downloads yestersday, and Billy was only posted at around 1pm. Not too bad.

But it was indeed crazy. I had been testing Billy 4 for about a week prior, by giving it out to just three of my friends. Between the three, they had Windows 98, 2000, and XP Home covered (and I myself have XP Professional). They managed to find 2 significant bugs that I had not seen until that point -- one of which was a crash bug, ew. Good thing I tested.

Then, at about 10pm on Saturday (the eve of the release) I was playing around with "Beta 2" of Billy 4.0 at my girlfriend's, and discovered that:

1) The section numbering in the manual did not come out right when viewed under Opera 7. Instead of numbering correctly (1, 2, 3, 4, ...), each section was numbered within its own sub-list, so the numbering to the user was 1, 1, 1, 1, ... and so on.

2) If you restarted Billy (File | Restart), after he prompts for your name, there was no USER> prompt.

3) There was one sentence in his mind file that kept causing *really* strange responses to be generated.

and 4) I had spelled "chatting" as "chating" in the welcome dialog that displays the first time you run Billy.

Now, I got back home around midnight or so, and had a big decision. Do I push back the release, try to fix the bugs, or leave them be? I did NOT want to push it back, I had all the history files and web pages already written with 9/14 entered (and at any rate, it's already been over 2 years since the last version of Billy... I can't wait any more!). For a short amount of time, I had decided to just release it with those known bugs, and follow up with a 4.01 release soon thereafter.

But then I remembered how much it stinks to get a newly released program on opening day, only to find it riddled with bugs. I just couldn't live with myself if I went down the Microsoft path, and released it with bugs I knew about but could possibly be fixed with a little work.

So, I stayed up till 1:30am, redoing all three manuals (Billy, Advanced Billy, and Link shell), fixing the little restart bug, correcting the spelling mistake, and removing that one sentence (which entailed re-generating the default mind file). Then it was time for bed.

I got up the next morning, tested the new Billy a little, had my girlfriend install this "Beta 3" to make sure it worked on non-development systems, and then released it at around 1pm, just 12 hours after I had been cramming to fix bugs. And I celebrated with a Mountain Dew.

Ah, it feels good to achieve a goal. Especially one which requires writing a 10,000 line AI program... ;-)

Now, go try it!