It's been just over a month since I was last paid.  It's starting to hurt now.  I've actually started to look for a real job.  Not sure I want a real job though.

I'm going to run a special this month as Orange Room Software where I will work on any project for $150 per day.

So, anyway, not wanting to get a real job has me thinking.  I came up with a great idea while searching for work on the internet.  I can probably do this in less than 90 days.  So I just need to get $15,000 to start it.  That's really all I need is three months of my bills paid.

I'm pretty sure that I can do this.  I'm pretty sure that it will make some money too.  At least keep me paid for awhile.  So the question is funding.

I have been looking at Kickstarter.com and thinking about posting up my idea there to get funding for it.  But here is where I start to sound like all the people I used to work for; what if someone takes my idea...  Worse, what if I don't get the funding and then someone takes my idea.

I know that just sounds stupid, but now that it's my idea it doesn't sound so stupid to me anymore.  I used to cringe when I heard that from partners on the many projects I worked on.  I would say, "nobody wants your idea."  But this time, I think, someone out there would want my idea.  Of course my idea is probably not unique and there maybe a site out there already doing it.  I've been searching but I can't find one, but still...

I think what I'm going to do is start working on my idea right away and get a prototype working.  Then I'll think about funding.  I think I could do a prototype in about a week. (A week in Phill time is more like three weeks.)

Here goes nothing...

I've been working on some cool stuff for the last six months.  One thing about being the boss is that I get to call the shots on the technology that we use.  I chose Rails 3, MongoDB, and PusherApp for our service.

Rails 3 is awesome.  A really big step forward I think.  I'm also running on Ruby 1.9.2. This combo has really made development fast and easy.  I find myself writing hashes the Javascript way, {symbol:value}. I like it.

MongoDB using Mongoid has been really nice to work with.  I originally wanted to use CouchDB but the Rails support just isn't that great.  I gave it a second try in another project but still it's not as complete as Mongoid.  Also not having to do migrations is awesome.  It just works.  It took some time to figure out a few things, and Mongoid kept changing along the way, but the latest 2.0 version is pretty solid.

PusherApp saved my life.  If I had to write the socket stuff for this auction, it would have taken another three months.  Pusher is a direct drop in and go solution.  Even when Heroku went down, Pusher was still up and running.  I'm using Pusher to push updates to the clients during the auction.  It is also used for our chat system.  I love it.

I started with everything running on Heroku, but when they went down I moved the whole thing to Linode.  It took me like three hours to do and I was back up and running.  I was able to get a lot more speed out of the system on Linode.  Local MongoDB running and now Resque is really fast.  So fast that I'm sticking with Linode for now.  I tried to use the new CloudFoundry but my app would not run on their platform.  Not sure why.  I think in the future I'm going to install CloudFoundry.org on my own Linodes and have my own cloud servers.

So now that the Auction app is done I've been working on another car app for a guy in the Bay Area.  I'm using Rails3, MongoDB, Resque, and Linode.  To me it seems to be a winning stack. I have about 80 hours development into it so far and I'm just over halfway done.  I'm going to try to finish it this weekend ;-)

So, what's next?  I don't know.  I don't have anything scheduled.  But next week I'll be looking again.  If anyone needs a Rails 3.1 app... I'd be happy to go for it!

You might have noticed that I had a whole lot of duplicate posts from Twitter. A plugin that I was using, FeedWordPress, has a bug since a few days ago. Not sure if it was caused by a WordPress update or the changes Twitter is doing. But it broke. So I turned it off for now.

Jane and I have been thinking about moving back to the bay area.

While we were in Santa Cruz last week I went to an interview at a company in Burlingame. Spent all day interviewing and I think it went really well. I'm waiting to hear back.

I have to say that living way out here is starting to wear on me. Not having good Internet and nowhere to grab a bite to eat is getting old fast these days.

It was really nice to walk to the store or a restaurant in the city. Not to mention the weather was much better.

It will suck for wheeling. I'll have to drive 4-5 hours to do that from SF. Probably will have to put my buggy in storage.

Yep. I think it's time for a change.

We finished it.  We finished the auction system.  We have had two relatively successful auctions.  I say relatively successful because the auction system worked great, but we didn't have many bidders online.  

In the last auction we had 129 registered members.  We generated about 50 leads.  The system had no glitches or crashes for the whole event.  I think it went great!

The problem we have now, after two auctions, is that we are out of money.  We have one more auction coming up in two weeks and that's all we have scheduled.  You see, we lost a partner last month, Jeff, who was our sales guy.  He was the pivotal partner to making this whole thing work after the software was finished.  Since then we have lost money and had to let go of our sales team.

What we really need now is funding.  We need a few hundred thousand to get a top notch sales team together and start moving this business forward.  In three months we could be generating enough to keep the business running.  Three more months is all we need.

Well, for now I'm going to go back to consulting.  I have one gig lined up, but that's only going to get me a quarter of what I need to keep my bills paid.  If only I could sell my house I could pay off all my debt and start new.  But for now at least I still have my house and that's good enough for me.

Image credit: RichardTheRough (http://richardtherough.deviantart.com/)

I'm constantly getting jobs that need to be done very quickly.  They also don't want to spend very much money building the applications.  It's very hard to explain that good software does not come cheap.  They don't really want good, they just want done. So, I take the jobs and deliver a cheap fast project.

This is where things go bad with the fast and cheap method.  They never come fast.

After shipping the project, there is always something wrong with it that requires going back in and fixing it.  The customer usually balks at the extra charges to fix the problems.  They tell me how it's my fault that there is bugs or design issues and that I have to fix them for free.  Usually I just fix a few things to get the customer out of my hair and out of my life.

But the majority of the work out there is Fast Cheap.  Everyone wants it NOW and they don't want to pay anything for it.  But, then get upset that it's poor quality.

I'm taking a stand right now.  Fast Cheap is off my menu.  It's going to be Good plus fast or Good plus Cheap.

If you want a app built for cheap and you want it in 3 months, it's going to take a year.  Deal with it.

I've been looking for something better than TextMate for a long time.  MacVim is OK, but it's not Mac enough yet.  Kod is looking good, but it's still not ready yet.

Sublime Text 2 Alpha steps in today...

I just checked out the new Sublime Text 2 Alpha text editor.  It has some pretty nice features started.  Here's my good/bad list:

Good:

• Loads really fast


• Long lines don't lockup the system


• Auto-remove trailing whitespace


• Line numbers don't scroll off the screen when scrolling sideways

• Preferences don't work (maybe I'm doing it wrong)


• UI Color scheme is really bad

I decided that I need to put myself on a tight schedule to fit in everything that I want to do.  I don't read enough because I'm always working.  I'm working too hard because I don't have the knowledge to work faster.  I have to spend a lot of time either with trial and error or researching the problem to find a fix.  This just sucks away time from my day.  I also think that I would have better health if I took time to relax more.  And of course I would make more money if I can get my work done faster.

7AM: Wakey wakey, hand off snaky.  This give me time to get ready for work.  I could even read the morning news while eating breakfast.

8AM: Start my first work block.  I have found that four hour work blocks are ideal for what I do.  That is enough time to really get into my work and resolve problems.

12PM: My first break.  Lunch, news, twitter, exercise, whatever.

2PM: Back to work for another four hour block.

6PM: Quitting time.  I can sneak in some TV or whatever.

8PM: Reading time.  I really need to stick to this one.  I just don't read enough.  I have a lot of books that I need to get through.  Technical manuals and also some fiction to round it off.

10PM: Back to sleep.

Had a blast this weekend at Reno Collective doing the March Madness Hack-a-thon.  The Think Kindness project was pretty successful.  We didn't quite finish all the features, but we have a real good start and made it all the way to beta.

Things that I learned:

• It takes lots and lots of time to setup development environments


• Select the platform before starting the clock, and stick with it


• Don't sweat the details until beta time

There is a all weekend hack-a-thon going on this weekend at Reno Collective.  We are doing a new project for Think Kindness for tracking Kindness cards.

So they have these plastic cards with a serial number and web site address  on them.  When you do something for someone, like buy their dinner, you give them the card.  The card receiver will then go to the web site and punch in the number.  They can optionally add a message and location.

So after millions of cards are giving out we can track to see how they spread and stuff.  It's all anonymous, we are just looking at the locations.

I'm thinking of something that is really simple.  Just a serial number and message entry with a location button to tag the location using geo tracking.