Mike Medved
-
Posts
1,543 -
Joined
-
Last visited
-
Days Won
124
Content Type
Profiles
Forums
Blogs
Gallery
Downloads
Events
Posts posted by Mike Medved
-
-
Somenos - if you could point me to a good concise description of the Darvas Boxes algorithm, I would appreciate it. Everything I see so far is kinda vague and contradictory.
-
Regarding CPU, I'm surprised my older PC does as well as it has. Nevertheless, I see plenty of PCs for sale with the newer 4th gen i7 Intel processors (4 cores/8 threads). However, how much memory in a newer machine like this would be usefull when using MT? I see many PCs available out there with 8, 12, 16, even 32GB of memory. How much memory would MT actually use with about 50 symbols in portfolios and perhaps 7 historical charts and say 15 intraday OHLC charts that pull a total of 5 days of intraday OHLC backfill data, and 2 days OHLC TICK backfill data?
Thanks again,
Tim
Everything (and I mean it) that we do in MT is with the eye towards speed/efficiency of execution and reduction of memory usage and memory fragmentation. Everything that can be cached is cached. All chart updates try to draw as little on the chart as possible, and only new stuff. The indicator calculation is done using very efficient state machines. Data storage is done using memory-mapped files, which are memory-usage-efficient. And, of course, everything (like, every window for example) is in its own thread, so that the smoothness of user experience is not compromised. No hiccups (or as few as possible).
The actual memory usage is very hard to predict, since it depends on a LOT of factors. Every indicator you add to the chart adds memory. If your frequency is 15 sec, you use more memory than if you used 1 min. Tick data on hugely liquid stocks can take a LOT of space (when you have 100 ticks per second for 60 days, it adds up). If historical charts have their last candle updated from L1 feed, that adds to CPU usage. To add to the complexity, this is C# - so everything is done using the arcane magics of the internal memory manager, so when it runs a garbage collection you suddenly see memory usage drop etc.
What we do is add features, test, and see what happens. If we see bottlenecks, we solve them. It's an ongoing process.
-
I just added Average True Range.
-
Basically, if your Y axis is set to Dynamic, or Dynamic Padded %, that setting will be retained. If it is Absolute, or Sliding Window, it will be reset to Dynamic when switching symbols.
-
That's a bug. Will fix.
-
ELD = Tradestation Easy Language Document = custom indicators
Cool. We will make it possible to do custom indicators. In fact, every indicator in MT will be "custom" - there is no distinction in how system indicators are done vs. user-written ones.
I think I made the framework where you do the custom indicator as easy or easier than other frameworks I have seen. I do have to document it though before we release it
-
It is actually very important for us to provide the tools that traders like you employ in their trading - but it has to be done without overwhelming more casual users.
We value your input, really. If you can think of a feature implementation that would help you, please put it here. Preferably one per post
-
The horizontal padding is easy enough to do, but I am not sure how to do the UI for it...
-
1
-
-
Jerry is a softie.
-
The reason why I wanted to do the help as Wiki is that this allows easy collaboration of several people in editing the help.
Making a help file is a major undertaking. QT's help file tool several years and a LOT of effort.
-
Lew, the chart types have to be the same. That is, if one chart is time-based, the other has to be time-based. If one is tick candles, the other has to be tick candles.
-
If you're referring to my: "BTW - you are aware that you can make the annotations work across charts - that is, you draw on one chart and it appears on all charts of that symbol (and of that type)?" - it is a setting in General Chart Settings in MT - towards the bottom - "Trendlines Unique Per Chart". Uncheck it.
-
Babyface.
-
And Jerry's profile pic. is from back then.
-
Lew, the problem is that it is a "trace group"... Not a "draw group". But I guess we could put another option in settings to have it work that way. Sometimes, though, you can have WAY too many options.
BTW - you are aware that you can make the annotations work across charts - that is, you draw on one chart and it appears on all charts of that symbol (and of that type)?
-
A show of hands please...
How many of you guys are familiar with Wiki editing process? We're thinking of dispensing with the traditional help files for MT and going with a community-driven Wiki.
But that begs the question - what percentage of our users will be ( a ) familiar/knowledgeable enough with Wiki markups etc. and ( b ) will participate/contribute to the Wiki?
So I would like to gauge this. Please respond. Thanks.
-
vireya, unfortunately that doesn't tell us much. Tell us how it crashed. Maybe an error window came up and you can tell us what was in it. Or at least give a description. Something like "I clicked on news, and smoke poured out of my computer, it said in a tinny voice "Oh mommy!" and died".
Something like that would help us pinpoint what happened.
-
We will (eventually) have all the indicators QT had and more. Just making sure all the infrastructure for them is in place before going full speed - since we don't want to have a change that would necessitate changing 100 source files - one for each indicator.
-
1
-
-
That has been done. press Ctrl-Shift, then move the cursor over the chart.
-
Once we have MT out in production we will look into it for mobile. I have had some experience writing mobile apps. But MT has to be out first.
-
Saty, what I am trying to say is that with MACD, the main thing you look at are signal crossovers - that is, the histogram value...
I will try to show all 3. I will try to change it so that the values shown are moved around so they don't overlap.
-
Isn't the histogram's value what the MACD is all about?
-
Hm I will have to think about how to implement that. The X axis is tricky...
Also - I looked at the definition, I understand how the thickness of the candle is related to volume, but it seems a bit awkward to specify how it is related (that is, the "look-back" period) - do most people just use a "standard" lookback? Like - 40 candles or something? Also - why are the candles H-L ones and not OHLC?
Darvas Boxes
in Feature Requests
Posted
Ok - I was sufficiently curious to research this and code the indicator, since it was a good exercise to test whether the indicator infrastructure is sufficiently versatile to deal with something like this. It is.
It will be in the next release Jerry puts out. Please check on it - there is some vagueness in various algorithms, especially on the point of how to start the box - whether the start has to be some kind of local maximum or it can be any point as long as it fits the rest of the algorithm.