Okay,

I made 3 pictures based on MT4 to explain the problem I had with your software. The catch is that the ideal product for me should adjust:
(i) thickness of bars
(ii) spacing between bars
when a trader changes resolution.

When the algorithim in a charting platform for (i) and (ii) is incorrect this results in trendlines being untrustworthy.

Look at the screenshots (a picture is worth a thousand words) to get the idea of what I am talking about. However, whatever inconveniences I may have experienced while testing your demo platform, I still think you have the best business concept in the world: 1 single platform that supports many data feeds and many brokers at the same time. No other platform does it the way yours does

And it was my purpose by starting this thread to show that there are software developers who are capable to think the way a trader does, which I think is a compliment for you guys.

I just checked 5 minute resolution of the same instrument on MT4. Unlike 15 M it plots trendline below the point indicated with the arrow. One does not know which trendline one should trust.

Now you guys can do the same excercise on your platform and introduce changes. If your platform can achieve a much better accuracy than Metatrader (which shoudn't be too difficult given Metatrader inaccuracy), then you will be one step closer to a success.

[DugDug]

I looked at the trial - and admittedly fairly briefly over a day or two. For me it was simple I just did not like the feel of it - so it is more a personal vibe than anything else... a bit like apple v PC. It looked like it had a lot of potential - functions, bells and whistles and for those who like beta testing I would say go for it. For me - I have other projects to do in my spare time at present.

[MidKnight]

Originally Posted by Tresor »

Then change the resolution from 1 minute to 3 minutes to 1 hour to 4 hours, etc. What you will notice is that point C will vary GREATLY from one resolution to another resolution. E.g. when on 1 minute resolution the ray crossed the price bar at 122, on 5 minute resolution the ray may very well cross the price bar at 132, and on 30 minute resolution the ray may cross the price at 112. You do not know which trendline to trust.

This is a result of incorrect bar thickness and / or bar spacing on respective resolutions

Hi there,

Unless I totally misunderstand what you are saying here, the problem you describe has nothing to do with error in the charting package but is a sad fact of trendlines. By changing the interval you are adding / removing bars and that is going to change how your trendline interacts with the bars. It's no error with the charting package, it is just a fact with trendlines and IMHO, really makes trendline concepts flakey.

With kind regards,
MK

[BlowFish]

I have not looked at this platform (yet). Anything I would look at now would need similar data handling capabilities of NeoTicker (particularly there tick precise technology...very important).

Nice bonuses would be ...Easy Language support. Drawing and visualisation of Ensign (it just look so damn pretty). Ninjas chart trader.

Originally Posted by MidKnight »

By changing the interval you are adding / removing bars and that is going to change how your trendline interacts with the bars. It's no error with the charting package, it is just a fact with trendlines and IMHO, really makes trendline concepts flakey

Adding / removing bars on one hand should be offset / compensated with adequate decreasing / increasing bar spacing and bar thickness.

I tested a number of charting packages,some of them can do it with acceptable quality, yet some are disaster.

[pfsoft_support]

Tresor,

After investigating your interesting (and ass-kicking!) comments, our conclusion is that the logic of how our algorithm works means that what you are describing is not a programming bug, but rather correct functioning of the system (i.e. the way it was designed) so as the be logical and useful.

Please see the attached screenshot with 1-minute candles. In this example the line does NOT touch any candle, but if this chart is converted to 15-min intervals then the line will touch the new single candle which will occupy the area of the red box (note in this case, of course, the 15 candles in the red box, will be converted to just one candle when changing the chart period/interval).

Note also that this is true for whichever of our 5 line types are used (e.g. freehand):

However, we will try to look how MC works, but it would be good if you could explain the logic you think it uses. We would greatly appreciate your further input if you have ideas how it would be possible to usefully change the logic here…

[Kiwi]

As Midknight said, this is really a function of how trendlines function. I don't agree that it makes them flakey - it illustrates one of the reasons that price often respects the zone around a tl, rather than the exact intersection.

If you try to automatically change for time (by aggregating bars to create the higher timeframe connection when lengthening the timeframe) you need to consider how it will work in the shortening direction as well ... you may need to record the point of contact on short timeframes internally so that when someone lengthens and goes back it looks the same.