Glad that helped! Here is some direct insight to what I have seen recently with TS10. I went with TS because they have the raw data, autotrading software and brokerage wrapped up in one bundle. Speed to getting a trading system up and get my capital working was my main driver. The few test trades I performed showed the brokerage/manual order entry fast and slick.
I'm almost 4 mo in on learning TS/EasyLanguage from scratch with the intent to mechanize my trading and take majority of my (huge) human error out of my trading. I've been trading and programming in the real time enviro 25+ years. My TS account exec is decent and the tech support guys on the forum are well above average in knowledge and response time. Many times I get same day responses which is great this day and age. Account service for the way I use it is 9/10.
There have been painful times using EL/TS due to wonky charts and relatively arbitrary limitations costing me a bunch of wasted time. IE ... I have noticed a few times that the most recent candle is not anchored to the right side of a chart and will just march off the right side of the screen. I had a trendline alert occur today and when I looked at the chart, I had to jiggle the scroll bar (which was already entirely to the right) to get the current bar to become visible again. Also, if you use the 'squeeze' button to compress the bars, the chart is anchored to the left rather than right side. Click enough and the chart gets squz up on the left side with most of the screen blank. Jiggling the slider (that is already fully right) will get the screen filled again. That behavior should be an easy fix. That it made it into production tells me that their current programmers don't actually use their own software to trade because these basic GUI programming misbehaviors are annoying as hell. I am keeping my previous charting package on subscription, because it functions very cleanly without a dozen extra jiggles per chart I want to view. I go through a ton of charts all day long and this misbehavior eats me alive.
I tried putting together a dual instrument strategy to trade two correlated instruments. I found you can apply a trade strategy to the first (data1) instrument, but not the second one. I also found out that you are prohibited from performing intrabar order generation AT ALL on "dual" data charts, but you can for single instrument charts. I'm just starting to figure how I am going to work around that, but having a strategy with 30 min bars that must wait for the end of that bar to launch an order clearly isn't workable. if you have a trigger condition a couple min into a 30 min bar, the market isn't going to wait 28 min for your software to get an order out at the end of the bar. There may be something I am missing in terms of an alternate way, but if you try to start a strategy with IOGeneration enabled in your EL code on a chart with more than one instrument, TS will turn your strategy right off and pop up the error log saying you've been summarily disabled. Thanks for playing.
TIP: If you attempt any strategy or special indicator work ... stop ... and search out & become an expert in "maxbarsback" for TS charts. This little gremlin running around in the background will eat you alive if you don't get out in front of it. This is a process that has to be cared for manually in TS. I've had several charting packages and all of them automatically took care of not drawing indicators or triggering until enough bars are available for a computation. For not being easier and more bulletproof, right now, I have to rate TS/EL10xx 4/10 for my needs. Hope to see some improvement at some point.
I had hoped to be up and auto trading by now, but these workarounds are costing me 2-3 months of trading with their software. Not to mention trying to continue discretionary trading w my old broker while trying to get a workable strategy up and running.
IMHO, there is a problem here from an architecture/GUI perspective or new revision performance situation. I would be happy to pay a couple hundred a year towards the software platform so they could get a trader/programmer on staff that personally used the software. From the outside looking in and what I have heard others say, it appears that mgmt views TS 'free' software as a loss leader to get trades into the brokerage and not a revenue center or at least neutral, breakeven center ... or as I have found with other brokers ... management sees discretionary traders as their 80% and algo folks like me as the 20%. Hope I am wrong because I think mechanical eventually will be the only sensible way to deal with the trading day.
All that said, this route is still probably going to be faster than me duplicating a lot of what they already have working with my own code. We will see in a few more months. That is always the fallback.
T
