<a href="http://www.briefing.com/Investor/Public/MarketSnapshot/PageOne.htm">Briefing</a> gave a good heads up on the likely timing of Jobs's announcement. Sometimes the Page 1 has some good insights. So. I installed 11, thinking that I'd best look it over before spending a considerable amount of time on Ninja just yet. Maybe my current backtesting headaches will be improved by a 64 bit program. I can deal with the chart formatting issues, but now I'm getting "undefined function call" on scripts that worked easily in 10.6. Something as simple as this: NoDoji = wma(Math.round(Math.sqrt(Length)),efsInternal("RetailDayTraderisaBonehead",Length,sym(Symbol)); Is it due to calling efsInternal as a series parameter of a wma? Something wrong with the function called by efsInternal itself? It's driving me crazy. I used to be able to decode errors in my scripts based on descriptions in the formula output window. Now there's no way to tell what exactly is undefined: function A, function B, or the call itself? It hasn't even made it to my playsound calls. I can't imagine what else is going to be wrong. I will likely spend as much time figuring simple flaws like this out as I would learning Ninja's code structure. This first attempt does not bode well.
Thank you for the reminder about briefing.com. Haven't been there in a coon's age. I think they pissed me off for some reason. Terrible being old. Can't dismember why. The wife asks "What did you do this morning?" "Damnfino. Obviously I didn't die." The main reason your simple code mayn't be working in 11 is the symbol call, of course. Calling the internal as a parm, I don't know. Just as a quick check you could hack out your own MA code and try it in that. But there is the larger issue of trying to use al-gore-rithms you find on ET. I have no doubt that NoDoji is round, but does she squirt, much less what her squirt length is? "Retaildaytraderisabonehead" I think probably got assigned the value "null" higher up in the code. Ninja's code structure doesn't look all that bad. I assiduosly (or acidulously) avoided learning C for some reason I can't now remember. So I guess I'll just do it. The fact that the formula output window doesn't wrap may explain why some efs doesn't work. THEY couldn't figure out what was wrong, either.
I'm a big Esig fan. Everybody's got the issues that pertain to whatever things are important to them in the software (EFS coding, etc.). For me, I've simple needs. What pisses me off is that they could not manage to simply at the least to carry over ability to format quote screens. The ability to color various cells, flash certain cells, etc., is very limited, something that just by itself will keep me in V10. What I specifically refer to: If you go to quote screen properties then to colors in V10 you see a large array of choices for cell coloring and flashing. They removed nearly all of it in V11. For the life of me I can't understand why if they didn't want to add functionality to this, why they didn't just leave it alone instead of removing the user choice. I know this may not bother many of you, but this small detail is a real hassle for me. I didn't think I was asking much, asking for nothing additional, just not to drop user choice from properties. I had mentioned this multiple times during beta, and following full release.
They probably wrote V11 from scratch instead of bootstrappingt V10 code. Hence missing functionality and lots of bugs.
Thanks for sharing what is important to you in screen setup flexibility. It is indeed the "little" things in 10.6 that make it so useful to a REAL trader, that is, someone who needs to get the information shoved into his head without trying hard to watch for it. The whole point is to make your experience of the market personalized and pleasing to the eye and brain. In my screen I do not even have to focus my eyes to "get" what I want to know. It leaps out at me in the specific colors which appear simultaneously in various locations in "helper" studies. For example, I don't want to look at actual numbers, unless price is close to a stop. I want to know from one location that by color coding "the cumulative volume is good" and from another that "but the volume right now sucks." Since I am usually badly hung over, I want to get a color cue a minute before a system trips that "all conditions are felicitous to enter a trade in the next minute", elsewhere that "the upcoming trade is a short" and finally a visual cue that says "look alert, dipstick, you're on in 60 seconds or less."
I'm sure that is the case, however you can be certain they reviewed every single feature to see what to program into V11, and decided that this was not worth carrying into 11.
I think they said that somewhere on their bulletin board. But there is SOME good news in that. For example, I love the clock they added. I time my entries to the second to keep track of my slippage and to minimize it. To do that I have a simple one-second chart running squoze down so I see only the time. And I also adjust my windows so that the IB clock shows. Assuming their clock is right, which I haven't yet checked, that clock will simplify my screen.
There was not an ounce of that that didn't have me on the floor. Thank you for somehow getting me to laugh while I try to shake off this dread of change that will force me to make a hard decision. Good call on the sym() -- in transcribing I accidentally left out a variable that evaluates whether it's OHLC4 or close of the symbol. Apparently you can't do that anymore (i.e., assign eval(close) to a variable and have that variable determine the OHLC/open/close of smthg). I'm finding more errors related to how functions receive and evaluate parameters -- they just don't seem as robust in their handling. This is the last kind of change I would have expected -- I thought all of this was native javascript and the main issues were appearance (just as relevant, but something I can adapt to). Plus it's ridiculous the damn formula output doesn't wrap. Even notepad wraps! I think my plan of attack will be to take a month and untangle all of these errors and see if my primary script will run, while studying Ninja at the same time. At that point, I'll have a sense of how much of a paradigm shift toward "pretty toy" eSignal has made. They seem to want to price themselves at the institutional level and yet drop functionality to also appeal to the very amateur retail trader. As a former prop trader eSignal-user who's now been retail for a long time, this seems like a great recipe to dump their market share on anyone standing by the road. Lifetime license for Ninja: $1k. Less than 8 months of eSignal, and that's not even considering data costs.
You are quite welcome. But while you were laughing I was crying. My vaunted system told me not to short this morning. Your observation about Notepad wrapping reminded me that in 11 they call Windows Explorer when you want to retrieve or save a script. I like the open/save in 10.6 much better, where it gives you a vertical list to scroll down. Your analysis of price vice functionality is spot on. Charging too much for too little. I am sorry you told me you are former prop. My talking to you reminds me of an anecdote about George Gershwin and Maurice Ravel. Gershwin wanted to become a more "serious" composer, so he went to France to ask Ravel to take him on as a student. Ravel was confused because he much admired Gershwin's jazzy compositions, and had experimented with that style himself. Ravel asked Gershwin, "How much money are you making?" When Gershwin told him he reportedly sputtered "You should be teaching me!" So if you want any more advice from me, well, you can afford to pay for it.