And I dare saying you have no clue about what good software in trading does or can do. Like the dude at McDonalds saying this is the most elaborate food in the world, and chefs are totally overrated.
Having come from NT I have spent - with some partners - years working on our own software stack simply because at least at htis time (and this includes NT 8 coming up) there was no decent offer around below a price that you can buy a large house for.
Simple things like:
* A trading calendar that can handle exceptions and changes over time*
* Multi computer backtests that can be analyzed and compared after months (because the data gets actually STORED, like in a database that is accessible by standard tools).
* A runtime that is capable of running 100+ strategies at the same time, including doing more complicated risk control than "runs or not".
* The ability to queue backtests so that a long weekend is not wasted for all the computer capacity (not like for example NT where you have to manually schedule it once the last one is over) and where backtests do not run hugh memory footprints.
* A good tick by tick simulator hat is in fact not totally awkward slot.
* The ability to run trading strategies on a trading SERVER. That is a machine where no one is logged in, but you do essential things like start/stop strategies from a web browser or a command line on another machine. Like, you know, with server software in general - Apache, MS Exchange, Oracle Database Server somehow do not run as interactive tools in a logged i n user space.
* A decent API that has a concept of a logical TRADE, as opposed to orders. Trades with a beginning, an end, an assigned risk profile, a maximum profit and loss, that can be tracked later in analysis. And possibly form a hierarchy. Getting time overlapping signals decomposed into separate logical trades can be seriously complicated.
* An environment that does not force me to abandon the basics of software development to write so useless things like strategies or indicators. Stuff that for example can make me do a UNIT TEST. Can you imagine that I am now working against indicators I know are correct because I have unit tests for all of them? And can you imagine that during the development we found out that standard core indicators in both NT at least are calculated wrong? No joke. Some more complex ones have "optimized" the underlying base indicators and - replaced them with some more convenient variants, but fail to document this. This blew in our face in the unit tests.
There simply is nothing sensible. And especially not NT or MT - which have awkward development environments. NT does not properly allow using Visual Studio (for C# development - though they falsely claim that is their own langauge, it actually is not), while MT does not even bother with anything such idiotic as source control integration. Or checking that the numbers actually do add up
So, no - doing strategy development for years now and gave up looking for anything remotely decent around 4 years ago.