Pages: 1 2 3 4 5 6 7 8 9 10
Stage 5, day 14:
The day has just started and I feel optimist that it is going to be a good and productive day today. One of the thing that I did was to review the BASE instructions document and something did strike out.
In post #97, I did mention that I felt like I was falling for distractions. There is a script in BASE named:
- Disconnect From Distractions
Coincidentally, it is included in every stages except 4 and 6. Given that I became distracted just as I completed stage 4, this could explain what I did experiment (but I'm getting back this bad habit under control now...).
Stage 5, day 15:
I have finally broken the inspiration dry spell by writing down over 500 lines of code for a class that will populate an order table in the GUI app.
In the meantime, I did let my trading bot server component run unsupervised for about a week. It did pretty well despite not all my improvement ideas in. It does about 1% of profit per day. I'm still limiting its daily budget to $100 since I'm not done with its implementation so it isn't much money but my confidence in it is growing. With the given result, I'll probably have in my hands something that generates a non-negligible income before the end of BASE stage 6!
Stage 5, day 22:
I am developing new functionalities in my trading client. There is a doubt that I am currently working on the highest priority task. My trading server appear to be stuck for the last few days. The current justification is that until I develop the next trading algo version, despite the last profit strike, there is a risk and I need to improve the visibility of what is going on in the trading server side with those new client functionalities but there is a small voice that tells me that this might be an error.
Hopefully, all the tasks will be completed soon and this doubt will be a thing of the past.
Yesterday, I got my smart ringbell security driver so that I could open it to press on the setup button. It did turn out what is broken is the WiFi interface. The ringbell has essentially stopped to be smart. I bought the ringbell at Best Buy something like almost 2 years ago. I did retrieve the invoice and I didn't took the extended guarantee. Damn. I almost regret that decision as buying a new one is $300... This time, I took the extended guarantee.
Few days ago, I just noticed that my Marketing Biz did generate $1.2K in a single day. AFAIK, this is an all-time record!
Another decision that I kinda took impulsively yesterday. I did purchase a 5 stages reverse osmose water filter to clean out the fluoride put into municipal water. Fluoride is put to supposedly a better dental health but it has several health downside such as the pineal gland calcification.
I now have complete distrust over public health agencies and the decision that they take for our good. I'm very happy with the idea of having fluoride free water to drink for all my family very soon.. Very cheap purchase as well. From a USA company that build the filter in the US with great pride.
Stage 5, day 25:
Extraordinary productive day. GF is extremely supportive. She is usually more demanding in terms of time spent in family than now. I believe it is the whole CoVid crisis going on. The economic uncertainty that makes her take care of the house chores and the kids while I'm focusing on completing a possibly brand new income generating business. Honestly, this is the perfect setup and when I will succeed, she will trust even more my priorities and become more supportive to my mission and when things settle down, she will also realize how beneficial it is for everyone.
Beside that realization, I created a new base class. When I write many classes that end up using the same pattern over and over and identifying the pattern, by itselft it is a small victory.
Last feature that I did put in place was dependent of some server events that weren't happening. I thought that the server was stuck in some weird state. No, it wasn't. After having waited for a week, something did happen last night and the new code did behave ok. Not perfect because there are few glitches but pretty well for untested code that got executed for the first time.
I am almost done with Stage 5. It appears a little bit uneventful but I feel that some things did happen under the water...
idk if it could be it, but we did play to Trivial Pursuits in family yesterday. It seems like taking a break from the project did energize me.
Stage 5, day 32:
Stage 5 is almost over. It was a pretty uneventful stage. I have stumble on some features in the last 2 weeks and it is not moving fast. It is at the same time complex and a bit uninspiring. I maybe do some procrastination over it.
I missed a day during stage 5 so I'll extend stage 5 an extra day. I will initiate the final BASE stage on Tuesday.
I have totally cut off MSMs from my life. I was listening to radio in background while I work but recently, I'm reacting to what I hear as I feel that I'm fed up with lies and manipulations blatantly with an intensity that I have never felt before. Finding a good substitute that helps me focusing and isn't too much distracting and is also a endless stream source was hard to find. At first, I was listening to alt info medias but sometimes what they discuss is so interesting and novel that it distracts me away from what I'm doing.
I did stumble on something quite interesting and I'm pretty sure it could of interest of members of this forum. I have stumbled into a "militarization" of music during the WWII era. familiar names (Rockefeller and Bill Gates Sr. (Yes the father of the Bill Gates that we hear about almost daily these days...) on the british side were pushing alongside the germans (Goebbels name is mentionned) were pushing for changing the standard instrument tuning frequency from 432 Hz to 440Hz. Apparently the 432Hz frequency has several positive effects in terms of the mind and biology (DNA is mentionned). Those benefits may or may not have been verified but some very influential people did see it important enough for their interest to have the frequency standard changed.
https://globalfreedommovement.org/432-dn...-of-music/
Whether or not the theory is true, since it is so easy for anyone to give it a try, it is not important that the theory has been validated scientifically. You just give it a try and see the effect that 432Hz music has on you.
To my great surprise, YT and spotify contain an almost infinite number of songs using that 432Hz frequency. I did try it out myself and I find the experience very positive...
Yes, 432 hz feels so good
Stage 6, day 3:
Today was my birthday. I decided that it would be a special day and it appears that only because I decided it to be that way, it has even when not much out of the ordinary did happen.
I got my boost from starting a new stage. I love this effect. I have finally completed the feature that I have been working for roughly the last 2 weeks. I had very minor debugging to do and I have still some polishing to do but it feels good to have finally vanquished the mountain.
Since this is my last stage, I have started to think what will be my next program. I did grab WM2 during the last IML sale. I'm definitely more a sex guy than a man-woman friendship guy. WM wouldn't have been my first choice of program but I'm curious to see what it could do to me. I'm just going to give it a try with an open mind and ready to be surprised by something unexpected that I might like finding.
I was thinking doing UMS to keep going in the same direction that I have started with BASE but I feel that doing that could be unhealthy especially after the lockdown and a possible second wave. I am currently starving for social interactions and enjoying myself a bit during summer and fill my mind with good experiences will probably help me stay productive during the next fall/winter sprint.
UMS while waiting for the next DMSI release could also be another option because WM is another 6 months commitment...
99% sure that I'll end up with a AM6 stage 7 refresher then go through WM2.
Stage 6, day 8:
Time is flying so fast.
So the feature that I said was finished 5 days ago is now fully polished now. Starting tonight, I'm going to return on working on the server and restarting it.
Since I'm starting to feel the end of BASE approaching, I'm starting to summarize my overall experience.
Beside minor distractions, I have been very focused on the goal for a good 6 months. At times, it has been very challenging with the unexpected lifestyle change forced by the lockdown.
I think that I would have been focused without BASE but I feel like it did really help me to put the extra achievement that makes the difference between ok and truly amazing.
The most fascinating aspect of the benefits of BASE is that I have been all in on my new trading system project. You would expect that this could have negative effect on my other biz income. Quite the opposite. My income from my other biz has been stable and might even have increased a bit. My debt in the form of bank margin and credit cards have melted by a very impressive amount without any conscious effort
So wow. I'm inches away from having a new profitable biz that I built in less than six months and while I was 110% in that project, my debt did melt by several thousands of dollars. It is safe for me to say that BASE did work for me.
The most awesome aspect is that several BASE components are sold as standalone components that have all been upgraded lately. I believe that the new OF is one of them.
When/If BASE is updated with all the latest IML techs, this program is really going to be a beast...
I think that I may do small screen capture to showcase what I have done so far. If I do, I will post the video here probably...
Stage 6, day 10:
I am returning working on the server code. I'm having some difficulty to get back into it. I haven't looked at that code for roughly a month. A month ago, I knew everything from that code. I forgot a lot of the small details in the last month. It makes it harder for me to find my way into it. My ease to walk into it will come back but it makes working it a bit uncomfortable.
Currently, I am on a lead to uncover a subtle bug that I have spotted while I was reviewing the last debug logs. I'm uncovering also a recurring theme behind a lot of the bugs that did slip my attention are related to orthogonality.
A big word that means that a lot of concepts should be matched with its opposite concept. ie:
init/terminate
open/close
do some treatement for BUY order/apply it too for SELL orders
Stage 6, day 11:
I did learn from a mistake that I did. While reviewing the debug logs that are 1 month old, I'm investigating bugs that I did fix back then but due to my mistake of not documenting these fixes , I can't tell if the problem was fixed or not just by investigating the code.
I'll try to leave a bigger paper trace of the changes so that I do not repeat this problem again. On the bright side, this allows me to leave more comments in the code on aspect that weren't obvious when returning to it after a month of absence.
Also, by looking very hard at the already fixed code, this did allow me to spot other very small glitches. Nothing that would have break havok if left alone but still, this did allow me to slightly improve further that part of the code.
Stage 6, day 12:
I did review 32,000 lines of logs out of 60,000. I'm not done yet and I have fixed few more very small issues such tweaking code to prevent it logging false errors.
but I cannot wait for completion for restarting my server because doing so is leaving money on the table. Reviewing the logs made me realize that the server is getting pretty good at generating profits. I'm better letting a good but imperfect profitable software than waiting for perfection before restarting it...
The exchange added support for the australian dollar (AUD) this week and added few pairs around this currency. I need to add the support for it on my side (few minutes max) and I'm going to relaunch my server!
Stage 6, day 15:
Yesterday, the gym did reopen. Today, I'm rediscovering the satisfying feeling of sore muscles!
I pretty much completed the debug logs review. I have been able to make sense of all the observed problems. That was a very good feeling to isolate the problem issues out of a sea of numbers. I might have received help from BASE 'Be better with math' script!
I'm a bit concerned about 2 discoveries.
1. In few occasions, I did notice that the exchange was taking a bigger amount from the account than the advertised fee rate. This leads to "Insufficient funds" errors when my system try to use the balance that should be in the account based on the local fee calculation. (querying the exchange about the account balance between orders would unacceptably increase the system latency...)
2. They round more decimals than advertised in their doc (ie: use 8 decimals instead of 10) and they appear to pocket the difference (on top of the money they make out of the fees).
Currently, I have pending support ticket to get explanations about that.
Another bug occurrence that did bite me very nastily. A rare problem did throw an exception from the execution engine thread. The result of that was that the engine has been dead for days without me being aware of that. The effect of a dead engine is that there is no trade executed at all. I'm currently making sure that I detect this situation and shutdown the whole thing instead of creating a false impression that all is good. I'm also looking for ways make the code more robust instead of generating an exception. This is again an error returned by the exchange in response to a request that I did sent to them that did cause the exception on my side. I think that simply retrying the request 2-3 times before bailing out should be able to eliminate 99% of this rare exception...
As I was fixing all those things, another unexpected problem did pop out. I hate when service providers do that. Exchange did upgrade their API from version 1.0.1 to 1.1.0 without any prior notice and basically the new version is not preserving backward compatibility and did essentially broke my software.
At least, it doesn't appear to be a too big problem to modify the code but I still think that providing a short notice prior of the change would have been a nice gesture...
Stage 6, day 16:
The robustness issue that did lead to a dead execution engine did lead to some nice refactoring.
I did launch the server. After few hours, a trading opportunity did finally trigger my system. It was the first time that my Alpha Model did trade Yen. The position that did trigger the execution did disappear and because the Yen market is a low volume wide spread market, the juicy trade did become a loss.
But nonetheless, the server trading did generate new stimuli that did make my client new features crash. TBH, I only did very minimal testing on the new features but this result did convince me that I had to create a small test framework to reproduce at will the problem. This is a great opportunity to use the best software engineering practice that I did learn in my previous profession.
I have a pretty clear roadmap on how to do that.
I did notice a recurring pattern while I was reviewing the last month trades. I did extract the relevant info out of the huge log file so that it would be easy to validate this idea. It seems that my intuition is good. he he... BASE that keeps giving. I have discovered a specific situation where crafting orders to get the maker fee can make a big difference in profitability.
This idea did lead me to some good refactoring in my order class. I did add few lines of code to implement this new promising idea. Now, it is up and running. I'm waiting for the next opportunity detection to see if it improves the system profitability...
Pages: 1 2 3 4 5 6 7 8 9 10