Load chart by chart , with as complex template as shown below. With only 4 charts it is not so bad. Was able to run replay all morning, came back and did one more. Finally everything is working, but not going to trade live with 7b9. Will make trades on Laptop, running 6.5.
Interesting observation, Lap top has only 6.5 now, no registry problems not much fragmentation. Desk Top after 7b9 was a mess. Probably due to multiple installs with out super efficient disk & Registry cleanup.
I have added the output of MemTurbo with Replay running at 500x from 2/8 - 2/12 - 5 continous days, also attached. I am also running Norton Security Suite and Memeo Backup as I run NT7b9.
Most of what I was told by NT support was not the problem: No problems, even though there is an error message saying NT needed to shut down sometimes when I run Replay - I Don't Understand. It still runs.
Will find out what happens when I connect tonight (CME is open this Sunday night 'till Monday Morning)
Bottom Line: 7b9 works very well with some effort.
Last edited by ZTR; February 14th, 2010 at 01:07 PM.
Reason: Found Thread on NT Site & Added MemTurbo Output
The following user says Thank You to ZTR for this post:
I do feel for them because clearly they don't have the man power to pull off the scope of this project. I imagine there is also an issue that with something like locking chart refresh rates, that it effects their bottom line more to not have issues with this than it does to kill of the traders using their platform that this would effect.
What doesn't make sense with them though is new features like this piece of code posted in the bid ask thread there..
Add("AAPL", PeriodType.Minute, 1, MarketDataType.Ask);
With the 1 second time stamps and no historic onmarketdata, its hard to understand why they wasted time creating something like that...
I wouldn't doubt the bigger problem is there are some fundamental design flaws in the software that make what would seem like easy fixes hard to do.
The following 2 users say Thank You to darthtrader3.6 for this post:
Actually the more serious design flaw is in their software development process. They are obviously not using any of the standard development practices which any software company should be using in this day and age, such as unit testing, continuous integration, etc..
These tools are specifically designed to stop you from doing things like breaking one feature when you create another, and reduce the friction of creating releases.
This is a more serious problem than you might expect because no professional developer worth their salt would work for a company with such outdated development practices.
It is a well known fact in the software industry that one or two very talented developers can produce more code and of higher quality than a team of 20-30 less talented developers can in the same time frame. In fact having more people on the team usually ends up slowing the process if any of them do not know what they are doing, as the good developers then have to waste their time fixing their mistakes (which usually takes longer than if a good developer had designed it right the first time).
Having been a consultant for a number of software companies I have seen this time and time again, and I can almost guarantee you that this is the situation at ninjatrader.
The following 2 users say Thank You to sefstrat for this post:
So let's pretend that we have NinjaTrader's attention. I know Ray is aware of this thread, as is Dierk.
I would have to say that both of them are resistant to outside ideas in general based on what we've seen over the years. Perhaps one of them is really wanting change, but unable to gain the friction needed to make it happen.
We've got a lot of terrific and smart people here on futures.io (formerly BMT), with a lot of background experience, and we all want a better product. So write them (in this forum) some "open letters" on how to improve. Try to provide some specific action points (hard, I know, without knowing the detailed inner-workings of the company, but we can make educated guesses).
If nothing else you'll at least feel better having tried to help them change their ways. And my gut tells me there are some people at NT that are truly wanting change, so maybe with enough external support the reasons that change has not been forthcoming can be eliminated.
Due to time constraints, please do not PM me if your question can be resolved or answered on the forum.
Need help? 1) Stop changing things. No new indicators, charts, or methods. Be consistent with what is in front of you first. 2) Start a journal and post to it daily with the trades you made to show your strengths and weaknesses. 3) Set goals for yourself to reach daily. Make them about how you trade, not how much money you make. 4) Accept responsibility for your actions. Stop looking elsewhere to explain away poor performance. 5) Where to start as a trader? Watch this webinar and read this thread for hundreds of questions and answers. 6) Help using the forum? Watch this video to learn general tips on using the site.
If you want to support our community, become an Elite Member.
The following 3 users say Thank You to Big Mike for this post:
In my opinion the two things I listed above: namely using Unit tests, and continuous integration, would almost certainly eliminate the great majority of their problems.
If they are really wanting to improve their product and reduce the friction in their development process then my suggestion would be to hire an outside consultant who is knowledgeable in such matters to come in and help them, because I know how first hand how difficult it can be to change longstanding development practices in an organization.
Typically this is called a software process audit and/or software quality audit, where an objective outside party is hired to come in and observe the development practices. They then make suggestions and provide guidance on how to improve the development process and/or the software architecture itself, that may include them providing training sessions for developers or even assisting in implementation of architectural changes..
Unfortunately, Ninjatrader will probably never do this just like most other software shops in a similar position because it costs money, requires hard work, and does not offer immediate gratification. That is why these services are primarily only hired by companies where there is some regulatory oversight which requires them to have a third party verify their software, or by companies who find themselves in legal trouble after flaws in their software cause some kind of damage.
The following user says Thank You to sefstrat for this post:
I guess causing damage to your customers' trading accounts doesn't qualify!
Seriously though, it is obvious that they are not up on proper software development practices, and I suspect that their code would confirm that, if one could actually get a look at it. Given their somewhat indifferent attitude on a few other topics, I highly doubt that they would be the type of firm to bother going through the necessary steps to change their development process to improve code quality... I think we'll have to just keep hacking around their flaws for the indefinite future.
Last edited by FBJS; February 15th, 2010 at 10:25 AM.
So the question still remains... what are the alternatives to NT for multi-asset, multi-market, automated/discretionary execution? I'm looking at X_Trader/XTAPI again ($1,600/month) and TTFix($2,000/month where I am if I want some Guardian control). PATS has nothing that impresses me.
Its like there is a "best of the worst" in the sub $1k/month space, competition in the $2-5K/month space (TT, CQG, RTS, etc), and "real deal" packages and support in the $10K plus/month space.
My head has Ninja as the best of the worst now... in its price range.
Seriously what other packages are you guys looking at that offer connectivity beyond IB?
Favorite Futures: CL & 6e, looking at ES, ZB and AU again.
Posts: 2,100 since Nov 2009
Thanks: 1,104 given,
NinjaTrader Is way better than Pats or TS, but...
I have experience with PATS & TradeStation and I don't want to go back.
Last night I successfully ran NT7b9 all night and did some sim trades. Got out when I though the ES would close a @ 6:AM Pacific time, I was mistaken or I would still be holding a sim Short based on a Ladder chart showing an Al Brooks' Failed Break Up @ 1079.00
I am running a several custom indicators, modified by Zondor, to record just like Gomi's Recorder, as well as a GomVolume Ladder. Pretty much pushing NT7b9 as hard as I can. The best part: it did not crash once.
When it comes to NT philosophy - I agree with sefstrat
"It is a well known fact in the software industry that one or two very talented developers can produce more code and of higher quality than a team of 20-30 less talented developers can in the same time frame. In fact having more people on the team usually ends up slowing the process if any of them do not know what they are doing, as the good developers then have to waste their time fixing their mistakes (which usually takes longer than if a good developer had designed it right the first time)."
I used to work with programmers doing Magnetic Resonance Image Processing (3d Fourier Transforms and massive matrix calculations). We always had at least one Code Guru who drove the entire project. It was better to do in-house coding rather than rely on General Electric or any of the other vendors to get us what was needed. In fact GE bought the package developed at UCSF and it is now used world wide.
I believe that this model could work for NT. They should hire some of the super coders on this site to help them develop the best product for Traders.
Few months ago I've found Ninja trader. I was happy and I felt lucky - it is what I need - C# coding, tick data, powerful strategy analyzer. My first disappoinment was in GUI intuitiveness. Total chaos, it seems that somebody from the Mars or maybe the Moon developed it. Crazy.. I have never seen such a lame GUI. Crazy again.. But it wouldn't matter. You can used to such things. Sometimes powerful tools are not easy to work with. So I started to code a strategy, which I've already developed on Metatrader4. I needed Ninja for data and analysis reliability. I do not talk about help and order handling.. They ARE from another planet. So it took me less then a day to code that (I'm C# programer more then 5 years, so I've expected to write that code less then in 20 minutes). And again surprize - it appears that you cannot use tick data in strategy analyzer!!! Well you can, but there's a problem - if you use tick data, then you cannot work with any other time frame candles and if you use 1 hour candles you cannot access tick data. So if my indicator calculates values based on 1 hour candles, but I also need every tick data (for stoploss, takeprofit, enter position) Ninja is useless. Anyway I think I will fight with this junk. I'll write my own indicator, which would work with tick data but calculate values based on 1 hour. Anyway DO NOT USE NINJA TRADER. IT'S JUNK and CREATORS of this stuff ARE FROM ANOTHER PLANET. How can you trust your money to somebody from another planet?