Veterans advice request!

[Peskanov]

Hi there; I have seen there is a good quantity of veterans of the computing world here, and I wonder if some of you can give some bussiness advice.

A pair of years ago I started a company to create videogames for small devices like Nintendo's GBA, Symbian mobile phones, etc...
For starting, I programmed a special emulator, which (combined with a GNU debugger) greaty helps development. The emulator contains all the information of whatever happens in the emulated machine in the last minute or so.
This way, when we are testing software (using the emulator) and have a crash, we can easily get to the origin of the bug. It's interesting because it helps with very nasty bugs, the kind that left you scratching you head... :)
Also, beta testers can use the emulator to save every bug they catch. The whole state of the machine with the story is saved in a crunched file.
Another plus: subtle bugs like gfx animation glitches or sound 'clicks' are also easy to review; just rewind the history a bit and look again.

Now the question:
When we started using the tool, we realized that it had the potential to be a commercial product. But as you now the difference between a internal tool and a commercial product is very big, so we would have to invest heavily on it. Now I am pondering the following posibilities:

- Working it a bit more and releasing it as an open source project (which benefits do you thing I get this way, apart from good karma?).
- Create a commercial product, specially aimed at the embedded world.
- Keep it internal, use it only as a competitive advantage over other portable videogame companies.

Opinions?

[epepke]

Originally posted by Peskanov
Hi there; I have seen there is a good quantity of veterans of the computing world here, and I wonder if some of you can give some bussiness advice.

I probably have more veteran scars than many, and I say, none of the above.

Write a paper about it with the goal of getting it published in an ACM or IEEE or some European journal. It's free advertising, and publishing research lends a company an aura of respect. You will probably have to release the source as part of it, but as it is for research purposes, you won't have to clean it up.

[Iconoclast]

Originally posted by Peskanov
A pair of years ago...
Damn that's eloquent, I think I'll steal that phrase and make it my own.

Originally posted by Peskanov
Now I am pondering the following posibilities:

- Working it a bit more and releasing it as an open source project (which benefits do you thing I get this way, apart from good karma?).
I don't think you'll get any other benefits, well somebody will probably write a cool and snappy front end for it which will save you guys from having to do it. However, I think you'll still need to spend a fair amount of time cleaning up the source before you release it anyway, do you really want people from outside your company seeing all those nasty hacks you put in to save on development time since it was always only going to be an in-house tool? I've never subscribed to the new age hippy open source movement, I guess I'm just a capatalist b'stard.

Originally posted by Peskanov
- Create a commercial product, specially aimed at the embedded world.
Don't forget you'll also have to deal with packaging (unless you sell it via download only), and you'll need to spend money on advertising so developers know this cool new tool is out there. Then you have to decide if you'll support the product, that's going to be more expense. You should do a search of the internet and see if there's any products already out there that you might have to compete with.

Originally posted by Peskanov
- Keep it internal, use it only as a competitive advantage over other portable videogame companies.
This sounds good to me. You're obviously producing some good, stable, high quality products because you have access to this piece of software.


Here's another idea you may not have thought of: You could approach Symbian and ask if they'd be interesting in licencing this product from you for distribution with their SDK. If you can make a deal with them then you can get a slice of pie every time they send out an SDK to a developer.

I'm thinking Symbian are starting to feel a little heat, Pocket PCs now support a cut down version of Microsoft's .NET Framework, and that means that any idiot who knows how to write VB code can potentially be writing apps for devices that support this compact framework. The CF development environment is extremely good, there's a 100% compliant emulator, so a target device is not even required except for final testing. Developers can also run the debugger on their desktop machine, stepping through the code on a target device installed in the cradle or via an IrDA link. I'm thinking Symbian will be trying to get any competitive advantage they can at the moment.


[edited to add:}

Though I have a Symbian SDK and emulator sitting on my work machine for use in an upcoming project, I've not had a chance to take it out for a spin to see what it's capabilities are yet. For all I know it may do everything the .NET stuff can do, and more, but my point in the paragraph above is that more developers have .NET installed on their machines than Symbian, by one or two orders of magnitude, and it seems natural to me that many of them will start to dabble in portable computing applications. So, when their boss asks them if they can build apps for portable devices they'll use .NET CF. I did.

[Peskanov]

epepke,


Write a paper about it with the goal of getting it published in an ACM or IEEE or some European journal. It's free advertising, and publishing research lends a company an aura of respect. You will probably have to release the source as part of it, but as it is for research purposes, you won't have to clean it up.


Thanks for the idea; I didn't think about that, the academic world is strange to me.
I don't have academic credentials; however one of my bussines partners has, he could sign the paper. However I am not sure about the benefits for the company, the videogames world is a bit closed and separated from the rest of the industry. I think the people we contact in our bussines is also unaware of academic credentials.

[iain]

Originally posted by Peskanov
[B]- Working it a bit more and releasing it as an open source project (which benefits do you thing I get this way, apart from good karma?)./B]Releasing it as Open Source will not bring immediate financial rewards; but may be the better approach. I notice that your software would not be doing the good job it is if you hadn't had access to software someone else wrote and released as Open Source; so if you do go commercial, remember that you are building on something that someone else has given away for free (no problem there, but credit where it's due).

I don't think you need to clean up code to release it as Open Source (programmers know what source code is like). You can use the sourceforge site for code management and so on.

You get the benefit that your software will be improved without you having to make the improvements. Programmers will take a look, suggest and make improvements, but you retain overall control of the product. You really should end up with a better product this way.

You don't have to worry about packaging, marketing or getting the program super-polished before releasing it. You get more beta-testers built in.

Of course, we all have to make a living, so how can you still make some money out of this. Well, you may not, but then again the risks are lower. If you go down the propietary route and things don't work out, you could lose a lot of money.

You can make money in several ways though.
- you can offer support on your product, or be paid by companies to make improvements (e.g. add certain features that they want).
- you could use the model that mySQL uses and have different licenses for commercial and non-commercial use, making money from commercial companies.
- you might find that if the software is good, your reputation as a programmer is enhanced and this leads to other opporunities.

Overall, the open source route is well worth considering : you get better software with lower risk and still have the opportunity of making some money or, at the very least, making good contacts that lead to future opportunities.

[Peskanov]

Iconoclast,


You should do a search of the internet and see if there's any products already out there that you might have to compete with.


We found none interesting so far, only a few studies done in the MIT, and rumours from some old strange hardware in Silicon Graphics.


You could approach Symbian and ask if they'd be interesting in licencing this product from you for distribution with their SDK. If you can make a deal with them then you can get a slice of pie every time they send out an SDK to a developer.


That's a good idea, we also thinked about that. However, Symbian does not charge anything for their SDK's.
Anyway their development kit is horrible; maybe they would pay for a decent emulator, their current setup is nearly unusable.


For all I know it may do everything the .NET stuff can do


We needed weeks just to make sound streaming on a symbian device, and we still have some ugly bugs to iron there. Developing in symbian can be a little nightmare!
However there is a big resistance to microsoft product on the mobile market, nobody wants that wolf inside!.
Microsoft only has an important share on PDA's market, not phones, and is probable that PDA's will disappear being integrated with mobile phones.
Interesting times for these markets...