ahh, glad you didnt say the jump will be to metastock. the next step in reality doesnt exits, but it is created by vision, innovation.and innovation does not comes from consensus. how many polls/user feedback steve jobs take before launching iphone?
no platform/software is perfect. rather i say, every users need is unique and sometimes (if not always) no one software can meet all the requrements the users. but the uniqueness of ninja lies in ninjascript. personally, whatever nt lacked, i could fill the void by simply making an indicator. i am yet to come across any other software which gave me such flexibility.
Last edited by bukkan; October 9th, 2011 at 01:02 AM.
I think a lot of problems could be addressed if Ninjatrader spent more time making point releases (7.1, 7.2, etc) instead of working on multi-year, major releases which are giant, risk-laden infrastructural overhauls (NT8) that leave customers waiting for enhancements for dozens of months and long, bug-ridden beta cycles.
Another thing unique about their model is "pay once, get upgrades for life." Most software companies don't do this and instead, charge for upgrades on major releases but give free point releases/updates. The benefit of this is that the company is incentivized to make new major releases in a timely manner that can generate revenue, and the customer likes getting free point releases that improve the product in the meantime. Also, instead of paying a large lump sum of $1K up front for a license, you could charge, say, $299 for the software, then charge for major upgrades. This lowers the entry price but could lower the overall lifetime value of the customer if they buy just 1 version and never upgrade because they give up on trading or switch to another platform. However, you may be able to make money on the lower entry price with higher volume, but your support costs probably go up as well.
Another problem is that you have guys that don't want to upgrade and you have to support 4,5, 6+ past versions of software. That could be mitigated by end-of-life-ing older versions so you don't have to support those anymore. Also, it requires more complex source code management (multiple trunks for each major release, branches off of branches for support of point releases, etc) but it can be done, and is done all of the time.
I was surprised that for a 40-person company, 30 of them are support people. It seems like they should beef up the development staff. This could be a byproduct of their revenue model. For example, I bought a lifetime license, but odds are they will never make another dime from me even though I keep getting newer and newer versions of their software. Meanwhile, they have to provide the support that I expect from them during the lifetime of my patronage with them. I think they need a better way to extract revenue from customers to help continued funding of their support staff while at the same time properly funding R&D and innovation for newer versions.
The following 5 users say Thank You to shodson for this post:
good points, I think most people would subscribe to idea to get improvements more often and have to pay less for initial version. Also NT could follow model to give discretionary trading features for one entry price affordable to everyone, and give additional features (like development, backtesting) for upgraded, more expensive version. Idea behind is that not every new trader is programmer or will ever want to dig into code.
that's pretty obvious to anybody that they are understaffed with interface programmers. Adding one checkbox half of the forum raves about to NT interface is not a big deal for any programmer and changing object property to make for example toolbar floating is easy too. Instead they only manage to keep up with hotfixes. Also it looks completely strange for NT to ignore real commercial opportunities by refusing to add highly popular data feed connectors... So either problem is rotten brains at highest levels of management, failing to put priorities or they are short of programmers and lack of versioning protocols/workflow management.
Another thing - it looks completely strange to see all features and pretty complex NT functionality combined to one big fat program. This way for sure, if one single smallest part fails (because of say new addition community asks), whole program crashes. In contrast if NT worked more like client-server way, where different tasks are split to separate programs and they interact between each other i.e. data retrieval-storage is one program/service, backtesting - another separate program, Control Center and charting, dev. editor - again separate programs, this way each part could have it's own source trunk, probably independent upgrade cycle and adding one additional interface change or feature would not threaten whole program. Charting/interface=face of NT could be improved and beautified faster. So here we simply talk about lack of intelligent architecture from the very beginning and problems arising from that now.
im not sure if this was covered in the webinar... but did Ray answer why Market-If-Touched orders have never been a part of ninjatrader? That is one of my biggest pet-peeves for me. Every other platform supports it yet for as long as i can remember, Ninjatrader says "It is on the list of future considerations." This is from early 6 to current versions
For example, a buy stop market order is valid above the curret market price but will be rejected if submitted below the market. A buy market if touched order is valid below the market. So, if you placed a buy limit below the market, you are filled once your up in the queue, an MIT at the same price would trigger at market once the price was touched.