Mike Medved

MT writes to disk all the time, because we use memory-mapped files to keep the market data. That is how we keep the memory footprint of MT reasonably low. Some (most?) anti-virus software just stupidly monitors EVERY disk write for who knows what. That slows things down quite a bit. Unfortunately we cannot just disable that on our end. That is why we ask you to disable it on your anti-virus by excluding MT.

Make a rule-based scan for price above SMA50. Then convert it to advanced. That will generate the C# code. Then you modify it to your taste. I am not sure what you mean by "8% away from tagging it". 8% is huge in stock-price terms

Ok, first of all, it is good to know the ORA window is being used. I have seen so few comments on it, I wondered. I will think about how to make that option in the window. Making it a toggle in each row would make it too busy...

Ok. So I added the "Round to Lot Size" option - that would show 990 as 9. Will add the bolding of columns too.

It depends on your computer, really. 2500 - that's more than the max # of simultaneous RT symbols allowed by most sources, so that would be a natural limit. The scan itself (once it initializes) only works on incoming data, so it really should not be THAT CPU-intensive.

Ok I can see not caring about the exact number of shares (but if we do round to 100s, we should round down because if you want to buy 1000 and you see 10, you assume 1000 is available but if it was 990 originally, you're out of luck). Rounding the prices - I don't know. So you may see a bunch of 10.1 prices where before you would see 10.19, 10.18, 10.17, 10.16? How does that help?

Neither I nor Jerry have removed any posts here. I see your post just above from Thursday at 11:22am...

1. Ok but what do we do with 2,285? A bit of a problem with L2 in general in MT is that different sources give different granularity data. Some are true book sources. Some aggregate. When they aggregate it is a lot more likely that the aggregate # of shares for a price level will not be divisible by 100. 1.5 The # of decimals - that's automatic - but how would you change it in the snapshot above, for example? 2. Bold - will see about that.

When you say put it on chart - do you mean in the "Quote" box?

Something in your MT is trying to use IB. Look at open portfolios' data sources, indexes sources (if you have index panel shown), any other windows' data source.

Any glitches at all?

1. There are "historical charts" which would give you daily, weekly candles as far back as the stock goes. And there are "intraday" charts which give you data in 1 min, 5 min, 1 hour etc. candles. 2. The "intraday" charts can handle up to 120 days of data *but* the "backfill" for most sources doesn't give you that much data. Though the data is stored locally, so if you let's say backfill 20 days today, then 20 days 10 days from now, you will have 30 days full of data. etc. 3. The order entry - you can trade from order ticket, but also from charts or DOM by visually dropping the orders on the chart and moving them to modify them. 4. We also are about to release the hotkey and saved orders system that will allow you to place orders by using a fairly flexible pseudo-scripting system. So you could have a hotkey that says "place a bracket order at these offsets from current price for half the current position". We will be releasing it soon but if you email support@medvedtrader.com we can enable it for you to try.

Ok first of all you should not have the finishing ; in the if line above. If you have that ;, it would just stop the if there and always execute the bracketed code. Second: gratz You found a bug in MT's C# preprocessing of the Any function. It was converting the if statement above to: if ((MACDHIST_Histogram[0] >= 0.0) && CCI_Line.Any(0, 4, (i, x) => { return (CCI_Line[ i ].CrossesUp(-100,0)); })) and it should have been if ((MACDHIST_Histogram[0] >= 0.0) && CCI_Line.Any(0, 4, (i, x) => { return (CCI_Line.CrossesUp(-100,0)); })) and third - there is more wrong with that line. It really should be: if ((MACDHIST_Histogram[0] >= 0.0) && CCI_Line.Any(0, 4, (i, x) => { return (CCI_Line.CrossesUp(-100,i)); })) but this will take some more fixin'. Doing that now.

You could make a paintbar that calculates this. Simple code like this var MaxLoss = 250; var RiskPercent = 5; var Amt = Math.Floor(MaxLoss / (RiskPercent*0.01) / Close); SetColor(Color.Yellow); SetYValue(Amt); Then show the paintbar on the bottom chart.

Give me the code that gives you the error Not just the errors.

This will be fixed in the next beta.

This is usually a datafeed problem. They send those quotes, MT records them. You can remove them by using spike filtering in charts - it's in the right-click menu.

https://www.medvedtrader.com/www/frontend/production

There is a thread where you could give your opinion of MT. Please do so only if you've been an elitetrader user for a while, don't just create an account to post, that's considered spam. And be honest we don't just want empty praise. https://www.elitetrader.com/et/threads/best-charting-software.356495/

Drag does work on option chain window. You have to drag the little empty gray cell on the left of the option display row. You can even drag many option symbols at once - highlight many options (by clicking and dragging NOT in that cell) then drag that cell.

You know you can drag it...

How are trading increments decided for crypto pairs? Is there a doc somewhere?

You can, with some restrictions. The paintbar allows you, instead of painting a color on the chart, to draw a line instead. See the SetYValue function. So you can code your own logic for your indicator, possibly using other indicators as well, and generate the line.

Ok sell half of current position is easy - Pos / 2 The other - the DefaultQty should work

No, seriously. C# doesn't include in its libraries the "AlmostEquals" kind of function for floats. So I had to write it myself. It is a non trivial problem. The most precise method would be of course to decode the IEEE float into its exponent/fraction components and then compare them. I settled for a different method that unfortunately involves calling log functions etc. It is a little slower but I think more clear. But I of course screwed it up. Fixed now. Will be in next ver.