Blogs

So ummm... v3 is out :-)

10 months in the making, v3 is a complete rewrite of v2. For all intensive purposes they are completely different applications.

To put it succinctly, v3 is automated trading software done right. We took all our experience from making v1 and v2, added that to most of the customer suggestions over 3 or 4 years, and also got rid of all the weird / nasty / annoying bits.

This is just an alpha release though, so don't expect a finished product. Having said that, it should be completely usable. And if you get in on the ground floor, you can have a say in the direction the software goes from here.

Of course, there is bound to be that 1 bug which will take about 30 seconds to fix but which prevents everyone from using it... Just the nature of making software. In that case we'll have a fix out almost immediately.

New releases are always available on the downloads page.

Note:

• There is no upgrade path from v2 to v3. They are too different to have made that worthwhile. However, you can export your price history to text files using v2, and import those text files using v3. Users, accounts, trading systems, etc, all need to be re-entered manually.
• Use the forum for help. If you send us an email for technical support, our answer helps only you. Ask it in the forum and our answer helps everyone. Therefore, starting right now, all technical support questions via email will be answered with a canned response of "please use the forum".

Added in v3.0.0

• Not going to list it all because seriously everything changed.

We hope you all had a good 2009 and have an even better 2010.

Back in April or so we decided to tear down our software and start again. The rebuild has been a long journey. However at the end we are left with better code, better features, and a better product. Moreover, adding new features, new indicators, and new brokers to trade with from here on in will be much simpler.

But none of that really means much because you are yet to have anything to play with. And that's pretty much because when we near completion, I say "hey, you know what we should do is..." The most recent end to that sentence was "have it so dates show up in any timezone the customer wants, rather than just GMT". And one-sentence changes like that sometimes take a really long time. Like that one-sentence change did :-)

So I'm going to stop trying to make v3.0.0 "perfect", because it is just going to be an alpha release after all, and instead pump it out, use it as a base, and build on it incrementally.

Also in 2009 we got a new website, with forums. But of course everyone's scared to be the first to post in a shiny new forum :-)

Best wishes for the new year.
 

All existing customers have tested v2.0.9 and report that it works fine. And so it is now available for everyone, and we will begin the transition to the new website.

Remember that once we change websites, any previous version will cease to work.

To make sure nobody was effected by that, v2.0.9 no longer has a version check, nor a subscription check (and also we got existing customers to test it before making the change).

Another reason you might want to use v2.0.9 is because Backtesting and Alerts are now free for all.

This is of course all gearing up for the release of v3.

Here is the full list of changes:

Added in v2.0.9

• Removed version check.
• Removed subscription check.
• Backtesting and Alerts now free.

About 2 years ago I wanted to move house. So I looked at an apartment, and it was nice enough and had fantastic views of the beach. Next door was an empty yard. Just grass. I signed a 6-month lease.

On the very day that I moved in to that apartment, they started building an apartment block in that empty yard :-) I was stuck there for 6 months, and nobody would take over my lease because of the construction next door.

It seemed like it took a really long time just to dig the hole for the foundation and the carpark. But once the foundation was complete, the actual construction of the apartment block went relatively quickly (though I still cursed them every morning).

And that's the segue into the update on v3. I tore everything down and started again. And man, it seemed like it took a really long time to create the foundation. But now that I'm getting into the "fun stuff" (of getting TSATM to actually do backtests and autotrading, etc), it's all just clicking into place.

v3 completed its first backtest yesterday, 24th August. A milestone which probably doesn't mean much to you, but was certainly a big deal for us. Because so many little operations have to work, and work together, in order to spit out a backtest result.

Get excited! :-)

A quick summary before I get into the explanation: TS v3.0.0 will be a complete re-write. We threw just about everything out and started again. If you have any suggestions for new features; if there's anything in the current version of TS that's annoys you, that hampers your efforts, that prevented you from installing it or understanding how it works; please tell us. Either through the Contact Us page, or via the comments below.

If you, personally, wanted automated trading software designed just for you, what features would be in it?

We've completely re-designed TS. Nothing has been spared. If you have any suggestions, anything at all, now is the time to come forward.

To the explanation...

About 4 to 4.5 years ago, in my room near Shinjuku in Tokyo, Japan, I started making some software to help me out with my currency trading. It was just for backtesting at first, but things progressed, knowledge expanded, plans became grander.

Now I have a company with a few staff based around that software, and our goal here at Thinking Stuff Pty Ltd, printed and stuck up on the wall, is to bring automated trading to the masses. Regardless of how good you are with computers, we think you should have all the advantages that automated trading provides at your fingertips.

Just about every manual on how to create a successful business says to find a niche. Ours' is the niche of people who want to get into automated trading but don't have the computer skills. And even if you have computer skills, it doesn't necessarily mean you want to learn computer programming. And even if you know computer programming, I'm guessing you mostly prefer to just click buttons than write code. I know I certainly enjoy downloading something and using it a few second later. So really, our niche are people who want to get into automated trading quickly and easily. (TradeStation, MQL, etc, will always have their niches, and we don't pretend we are direct competitors of theirs' other than a small subset of potential customers).

Over the years though, we've run into speed bumps. People would ask "can it do this?", and we'd have to say "no, in trying to make TS as simple as possible to use, we had to make some design decisions that cut off certain types of trading".

And there was the legacy code. "Legacy code" means old code, that's been around forever, that nobody wants to touch because it looks really scary and would probably break everything.

Actually my buddy came around the other week and asked why some things were like there were, and I had to reply a few times "historical reasons".

From comments and emails we get, and me saying to my buddy "historical reasons" more than a few times, it was clear that things had to change. People both want to be able to do more complicated trading systems, and they want the software to be easier to use.

The legacy code was holding us back. PostgreSQL was holding us back. The version check and subscription system were holding us back. The lack of supported brokers was holding us back. The lack of custom indicators was holding us back. This website and the lack of forums were holding us back.

So I thought to myself "Sharky, if you were going to design this from scratch, how would you do it?" And because I now have 4.5 years experience in writing this kind of software, along with 4.5 years' worth of comments and emails from customers, I made a lot of different decisions to 4.5 years ago. A lot of problems solved. A lot of niggly little things solved. A lot of big annoying things solved. And some prettier code too :-)

The drawback to all this of course, is that the monthly release had to be abandoned for a couple of months. And the next release, being a complete re-write, will need very thorough testing from us *and you* before it's ready for use. But let me guarantee to you, that this is entirely worth the trouble, and the few steps back will be followed by about 100 steps forward.

So as I said in the opening, if you have anything you want changed; anything you think needs to be changed; any reason at all that prevented you from signing up (and by the way - there probably won't be a subscription fee from v3 onwards), now is the time to come forward. Don't assume anything. The more people request a feature, the more likely it is to get done. Thanks.