Wayne, once you're ready to go full-time with the project you really should consider moving further discussion to Nabble. You've already covered a lot of ground in the past few days, most of it really deserves a separate thread so a more focused discussion for each of the topics, issues, etc. can be discussed in a more focused manner. In the meantime, best of luck sorting through your current challenges in quick order. Can't wait to see what you've done so far and to get involved with this project.
Jprad, Your advice below sounds well based. "surrender" is clearly the wrong word both legally and from a P.R. stand point. Those agreements on committing are "two-way" not uni-lateral. But I haven't found an actual committer type agreement to read. Your advice confirms my research that it's absolutely necessary to get all this legal stuff straight before releasing any code. And my conclusion is the same as yours, specifically, I'm totally not qualified, even after reading all this intelligence to figure out how to do this. I need legal advice. But that will cost money. Money I can't afford to spare right now. I was planning a few days ago to "do-it-yourself" but the more I read, the more dangerous that appears to be. For example, GPLv3 website says to put their notice at the top of all the code. However, my impression is that you have to modify the GPLv3 with some clauses to allow dual license in the future and deal with committers regarding joint-ownership, that isn't included in the standard GPLv3. However, if you read their FSF website they approve of the Ruby license saying it's GPL compatible because it has dual license clauses in it. So I conclude you're correct. What we need is GPLv3 "compatible" license which allows for commercial later and deals with contributor rights in that case. Your point about surrender makes total sense too. I was using the wrong term. Joint ownership is the way to go. But again how to get a contract like that short of funds? I'm not some major corporation and it may be years before recouping any investment in this so shoestring budget has to be the word. What do you think of just copying a license from somewhere like db40 or MySQL to save money? Contracts can't be copyrighted so it's legal to do that. I know attorney's would advise against for multiple reason including their own self interest. Would it be crazy to find out the price of legal help and ask for donations to cover it via paypal? Thanks so MUCH for this perspective and the advice. I'd like to keep this ball rolling towards releasing the code. In the meantime, I'll keep working on my own uses for it. Sincerely, Wayne
Jprad, I agree. Here's the link to the Nabble forum. http://n2.nabble.com/TickZOOM-Discussion-f1641878.html Let's move any further discussion there. Except I'll post any significant news back on this thread on EliteTrader. Again, I want to keep this alive. But have to juggle priorities. Wayne
Hmm. I added the Nabble link in my signature for EliteTrader and it didn't appear in my last post. This is what I put in my signature: Does it reject it because of the link? Or what? There's no further help on that page. Wayne
The commonly used term is Contributor License Agreement (CLA). The MySQL example is one and there's a pretty good treatment of what they are at: http://www.oss-watch.ac.uk/resources/cla.xml You mentioned setting up an LLC for the project. You may want to consider running it as a non-profit because that will give you access to pro bono legal advice and services from the Software Freedom Law Center (SFLC) at http://www.softwarefreedom.org/ Regardless, there's a lot of good information contained in their publications that may help round out your understanding of the open source way of doing things. You can't modify the GPLv3 to suit your purposes, it's copyrighted by the FSF. As far as dual licensing goes, you really need to read the following: http://www.oss-watch.ac.uk/resources/duallicence2.xml I guess it ultimately comes down to what you want. If you view TickZOOM as always being "your" baby and the future is a one where there's a commercial offering that only benefits you monetarily, that's entirely fine, go for it. But, if that's the goal I don't think open source is where you want to go. You may be better off with an entirely different tact along the lines of contracted developer/trader types who would be compensated for their efforts via free access to your commercial offerings down the road.
Cool. Thanks. I made a thread there introducing this topic separately. I'd like to copy your your 2 and mine from here to over there. You have offered the most well informed and relevant information. Sincerely, Wayne
Jprad, Thanks for joining the TickZOOM forum. I answered your post with quotes in a thread over there. Please, avoid reading that for an hour or so. I'm curious if you receive an email notification of it because I'm still learning Nabble. If not, I need to see how members who wish to do so can opt in to get notified of any new threads. Sincerely, Wayne
Hi all: I am a experienced C++ and C# programmer . I'm sorry but I don't see the need for an Open Source Black box Trade Platform. I am using a existing commercially available C# based Trade Platform that I can do any of the things discussed in this thread...am I missing something? I would like to be involved if it makes sense...but I don't want to waste my time during over capabilities that have completed and proven.
Buttermilk, have you learned of the necessity of back testing using real tick data? If you trade a black box, it will be fed real time tick data. Plus, all stops and exits happen on a tick by tick basis. So you haven't truly back tested (at least in smaller time frames) unless you use tick data. If so, how much tick data can your current platform handle? Does it have a limit on the amount it can fit in memory? TickZOOM can handle 10 million ticks with only 200Meg ram. On a 4 gig box it can take 400 million ticks at once. Very soon (in a few weeks) that limitation will be gone also, so that it's irrelevant how much memory you have. It will stream the ticks from file during back testing which will eliminate even the few seconds load time before starting each test. How long does your tick back testing take? TickZOOM can currently load and do a single back test with 10 million ticks which is 5 years of GAIN capital tick data in less than a 45 seconds including loading and running it. That will soon be reduced to under 30 seconds. And then we'll cut it by 4 X again because have a design to implement to capitalize on quad core technology on a single time series back test to parrellelize and finish a complete back test of 10 million ticks in 10 seconds or less. A will probably be filed before releasing this technology. BTW, it already uses quad core during optimizing. Plus TickZOOM has other features never heard of in other platform data engines. For example, it already has unlimited simultaneously bar intervals available to strategies and indicators. At has features to allow you write very "EasyLanguage" C# classes while adding on the power of C# and TickZOOM. It has an execution server and a real time quote collecting server also so you can run your strategies and indicators as a black box system directly without code change. It will soon have the ability to run unlimited instruments simultaneously as well. NOTE: It has only limited indicators and other add-ons yet. It focuses primarily on the zoom speed data engine. Personally I find it wonderful compared to any of the C# platforms I tried commercially simply due to the awesome speed. I'm able to try out more ideas in far less time. For more info about what's great in tick zoom search elitetrader for "tickzoom road map" and read that post in this thread. Unfortunately, many of the other cool features are spread through the pages of this forum. I hope to condense this information into pages on the tickzoom.org site soon. (which isn't available yet). Sincerely, Wayne