Phoenix 2007 (new thread) - page 98
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
write the phoenix into a eld easy language document and this can be optimized on tradestation. I hope there is someone in this forum that can do that. I would run a perfect optimization on this in a very short time and we could use the phoenix on a live account for a long time
Sharing is the only way we grow together. It is important to share the code. My larger concern is having literally 50 versions of the same code, and no settings files for anyone. Does anyone want to handle code merging and testing? We can have huge sections of code commented out or turned off/on with quick settings but we honestly need to keep the code tied together.
I think we should post versions of phoenix in this forum - but ONLY ONE VERSION of the code. If we all paste code snippets, then the code merger can make weekly releases.
I have been going over the previous suggestions, and looking at the code. Hendrick said he would do certain things, research others, and think about more. We either need to A) get the fix for mode 3, B) fix mode 3 ourselves or C) disable it. Too many people test with mode 3 thinking default mode = best mode.
Settings files should be shared here as much as possible, I know I'm late submitting mine but the code I do for food has been calling me. Some standards on settings files are needed, but so are some modifications now that the settings files are more obviously tied to a trading mode, version, currency and timeframe. These will all take code changes.
I have some development idas of my own for phoenix, most are just slight modifications to the ideas submitted here in the forum. Others are things I have talked about, etc. I don't want to make radical changes to Phoenix, we might kill it from feature creep and everyone would remember how "Hendrick's official releases were the only ones that worked." I'd rather make minor modifications to ensure Phoenix stays alive, and then slowly integrate larger changes. Confucius said, "You cannot call yourself a good son unless you can walk in your father's shoes 10 years without making changes." That said, Confucius loved tradition more than change, and we live in an era of exponentially increasing change. My personal preference is to treat this as an inheritance that we protect and keep for a few weeks, fixing bugs and extending code in small ways.
Since there are so many ideas, a software roadmap might be useful, but since the changes are so easy to write, it might be better to just paste code snippets.
Does anyone want to handle the merging? One Official Community version of Phoenix released every Friday or Saturday?I dont think changing the code is a good Idea. Hendrick said so himself. If we are going to continue we have to stick to two principles.
1) find the bugs and squash them.
2) No more features.
The code needs to be repositoried somewhere, one person needs to be responsible for bug fixes. Anything else needs to be simply optimizations of the parameters, or adding additional currencies.
If we really want to do good by this system we would focus our collective programming knowledge on creating an outside optimization engine. Of course that would probably take a mathematical genius(which I am not). It would be good if we could come up with some client/server system so that those of us with multiple systems(didnt some one say he had a 32pc farm?) that would allow for faster crunching of numbers. But that is really out of the scope of this probably.
My main point is that We have a working system(with a few possible bugs) its time to lock it down, fix the bug that people are saying is in the level 3, then only optimize the settings and add additional currency pairs.
Oh yea and to keep all the unncessary chatter down(can you email me this ea, how do I attach this to a chart etc etc) I think we would need to start a new thread and keep ALL the information in the first post. I know on other non-forex forums(ok ok I am a LOSTie) the do this quite amazingly. As relevant information is posted inside the thread it is also copied to the first post of the thread so all the Lazy people wont clutter the thread with uselessness. I dont know if NewDigital is able to give moderatorship status to a single thread or not, but that would allow someone to delete extraneous files so that only the real version is available.
I also realize that english is not everyones native language so sometimes its hard to really be clear with explanations. I would be willing on a trial basis to be the editor of the information. Who ever takes charge of the project could email me the details that need to be cleaned up and explained and then I could do so and then the PM could post them.
Does anybody fix Phoenix 5 Mode 3 ?
What the heck is the LARGE font for ... do you not think that people will notice your normal writing.
And in regards to Mode 3 then what the heck do you want to fix. It was an experiment just like Mode 2 that turned out to be not such a good idea. Personally I would just simply ignore it and let it die a natural death.
After all, you can't expect all of the ideas to be positive. That's what this forum is about. To thrash out ideas ... to trial them ... and dispose of the ones that aren't producing the desired results.
Does anybody fix Phoenix 5 Mode 3 ?
Flying Dutchman and his crew HerbertH went for a... vacational trip...
As father and assistant of this EA, only them can do something to fix, repair and improve Phoenix....
In the meantime, like everybody here : WAIT
I dont think changing the code is a good Idea. Hendrick said so himself. If we are going to continue we have to stick to two principles.
1) find the bugs and squash them.
2) No more features.
I don't think Hendrick said exactly that. Adding features for the sake of having more features is poor design. At the same time, Hendrick said he was going to add several features and then did not release the code. I am trying to get the original fixes still.
I agree that all of these things need to be done. It appears you want a 'stable' build only. Open source projects like the Linux Kernel have 3 versions: stable, unstable, nightly build. They release milestone nightly builds as unstable testing versions, and separately maintain a stable version. Sometimes bugfixes that go into the unstable version appear in the stable version as well. Right now, I consider 5.6.3 to be "unstable" since mode 3 can't make a profit in backtest reliably. There is more than simple bug fixing necessary to change that.
I have a friend with a doctorate in physics testing Phoenix right now. We were discussing genetic algorithms for parameters, neural network for confidence, and bayesian filters for logic processing yesterday. He is also a math genius with access to large computer clusters working for the Department of Defense doing Quantum PHysics. He just doesn't have the time build the entire framework himself, although he has written some genetic algo lately for other projects. This goal isn't unattainable, it just isn't practical this week. Give it a few months and I should have AI in DLLs coming up with parameters. It is possible for us to pool our CPUs and compute data by breaking up the work into sections. I might make a series of 10 market conditions that every indicator would need to predict, and we use GA to pick the parameters to find which settings most accurately predict the market conditions for different time frames in the current market. We could have one person sign up to do testing on 4 indicators, someone else does testing on 10, someone else 2. It is possible, not sure it is necessary.
That sounds like the appropriate path for developing a "Stable" branch of code. The code posted above would be part of an "unstable" or "nightly build" branch that people need to test and refine before putting those features in "stable". "Stable" should be something people can run on real money accounts in all modes. "Unstable" for live testing to find out which features work and which ones don't. "Nightly build" or "weekly build" would be collections of all features that were submitted that week. I think we can all have the kind of development we want (bug fix, improvements and new ideas) with this strategy.
I also agree. Who wants to do this? We should probably have two threads, new users and development so the two types of users don't get confused. Maybe the titles "Phoenix - FAQ, Stable, User support -Read Post #1" and "Phoenix - Development+Suggestions - MQ4 in Post 1"? I'm willing to run the second thread but not the first one actively. We don't need moderation as much as "first or second post edited by one person regularly".
Thanks for the answer, depictureboy.
As far as reading goes, this thread is pretty long. In past couple of hours, which is how long I've known about this system, I only got through the first 20 pages. I am willing to read and think about it, I was just curious about my question and didn't want to wait until I find the answer while going through the entire thread, which I will eventually be doing anyway.
Also, for past couple of hours I've been stretching my CPU backtesting the system. I'm willing to read and learn, only I am a beginner with no experience of live trading and designing complicated EAs like this. I don't want to make a quick profit off of this, I just want to be around, watch, and learn. This business is fascinating for me. I hope I won't be one of those who spoil great systems and wonderful communities around them like this one.
As far as building the team, as I said I can't help you with designing and debugging the system, as far as this I will be grateful just for an opportunity to watch you think aloud and learn. But I could help you with backtesting anytime you need. I am also in web programming, if you need to have someone create a website e.g. for effective sharing of files or with other features, I would be happy to be the one.
I hope I explained my point and let's get go get the pips
If you read through the thread you would see that there are plenty examples of what you are asking for. Its a shame really. I guess even with Hendrick leaving and giving his reasons people just dont get it.
There have been many a great thread/system die simply because people are unwilling to read and want to be spoonfed. Or the system goes commercial simply to filter out all the "spoon feed me" crowd.
this is the first EA I have really seen work. Usually its a system like Mouteki or Vegas 4h that dont have an EA that gets drowned out by all the "BUILD AN EA" crowd...
Malcik this wasn't totally directed at you. I am just tired of all the great systems dying because no one wants to do their own work.
Forex isnt a game, you gotta do most of the work yourself. The threads full of people who cant grasp simple concepts or are too lazy to read a thread and want the author to cater to them is the one more reason 90% fail.Does anybody fix Phoenix 5 Mode 3 ?
Work, Learn, trial & error, that should help you to fix Mode 3. Nothing impossible in this world.
My main point is that We have a working system(with a few possible bugs) its time to lock it down, fix the bug that people are saying is in the level 3, then only optimize the settings and add additional currency pairs.
I'll try demo next week, fix GJ different signal buy & sell instead, time filter for both signal. Will soon upload backtest report.
Maybe this thread need to be close, this thread going bad.
My advice:
1. Learn how to optimize and how to program an EA, not just ask never tried.
2. Tried to understand the system
3. Fired your broker (it's killing with that spread not include the commision, indonesia always have commision $20-$50)
4. No advice, just try learn try learn and try, then ask to the forum.
I created a forum thread for development, but nobody had created one for Q&A from new users.
https://www.mql5.com/en/forum/175560
Please reserve the 3-5 slots for people who are willing to edit their posts regularly with current information. If we have 3 volunteers, then we can start posting suggestions and information.
"Phoenix - FAQ, Stable, User support -Read Post #1" still needs to be created. I will volunteer in position 3 or 4 on that thread, but I would prefer someone else creates the first post.
I wouldn't mind creating it, but maybe I'm don't have enough knowledge about this system to answer all of the Q&As. Except for the "Why Phoenix doesn't make any profit on 1M chart" kind of thing, or how do I set the MM and PrefSetting to run the backtest, or how to get char history into my MT4.
Let me know if I could be of any help.
I created a forum thread for development, but nobody had created one for Q&A from new users.
https://www.mql5.com/en/forum/175560
Please reserve the 3-5 slots for people who are willing to edit their posts regularly with current information. If we have 3 volunteers, then we can start posting suggestions and information.
"Phoenix - FAQ, Stable, User support -Read Post #1" still needs to be created. I will volunteer in position 3 or 4 on that thread, but I would prefer someone else creates the first post.Now we'll see more victims than profit to share source code. That's not educated people, them use to cheat others to make money. If they use mql4 for study or use it in real acoount for himself that's ok, but we cann't close our eyes about this victims. Like Einstein smart man, die with regret. We don't want to do this same to Hendrick, do we? Sorry Hendrick, you are Einstein in programming and forex, and i now you recently very upset with those people in your pm. They used you to make script greed to phoenix, don't they?
fikko & daraknor & other people that contribute the scripts to hendrick are the best.
I totally agree with you. Something is going wrong in this thread. I dont think its a good idea to further develop this EA in this forum. There are too many greedy people who trying to make profit or selling this EA without doing anything for the developers or the community. Open your Eyes. Dont get used by them! This thread contains everything to make conclusions by yourself. There is only one last thing that you could do for Hendrick and this nice EA - further develop it with people who deserves it and do it outside of this forum (via EMail, Messengers etc.). Dont get used by them.
regards,
David