lano1106 OF 5.75G journal

[lano1106]

Cycle #9, day 4:

I had a lot of dreams last night. There was a good chunk of it (at least the part that I remember the most) was about sexual attraction....

When I did wake up, I did found out that a trade did occur during my sleep.

Disappointingly, the found opportunity has nothing to do with what I have added yesterday to the system.

The opportunity was detected as a 3.16% opportunity but due to the changing conditions, it resulted into a 0.36% result after all. This is disappointing but at least, it did stay in the positive territory.

Inspecting this trade logs made me add a very minor improvement to the code. Nothing that is a game changing but it is an improvement nonetheless.

[Shannon]

7,000 steps to the top of the mountain.

[lano1106]

Cycle #9, day 5:

Something pretty fantastic did just happen.

My dynamic trade plan execution did work.

The execution was at the plan step #3 out of 3 steps and conditions did change which would have resulted into a small loss.

The engine did detect a better path and it did modify on the fly the plan to replace the final step with an addition 3 steps to become a 6 steps plan.

Guess what, I did return to square 1 with even a small profit...

The nice thing is that by adding steps, this is generating volume...

I need more as my 30 days volume did continue to decline... it is currently at 15.5K...

[lano1106]

Cycle #9, day 6:

I added a new feature to my trading system. From the UI, you can get a trade done for you by the execution engine.

You specify, the currency of origin and the amount, the destination currency, and upon receiving the request, the execution engine will analyze the hundreds of possible paths from A to B to pick the optimal one and it is going to execute the plan by using the same strategies developed for the automatic trading.

Everything is pretty much done since yesterday but I'm not ready to deploy it yet. Be able to remotely initiate trades requires some security measures...

Security isn't currently bad but there are few things that I can do to make it better and I'm a bit paranoid about that...

On a different topic, I had a dental surgery yesterday... That wasn't fun and I could have waited much more before having it done as it wasn't by far an emergency... Curiously, I didn't have any hesitation to take the appointment. I felt a mild anxiety few minutes before the surgery but that was it... It is done... and it was much easier than I expected...

I attribute merits to OF for the way that I have handled the surgery situation... Possibly, it was the best possible way for my health in the long turn.

Finally, I'm starting the cycle off days tonight. I'll start cycle 10 on Monday evening and by having completed cycle 9 means that I have done a third of the program. I'm planning 27 cycles for reaching 8 months of usage.

From here, I'm not sure what I can expect from the program.

Will I become even more fearless?
Do I still have major fears to eliminate that I'm even not aware of?
Or will the rest of the program will consist of solidifying the acquired freedom from fears...

[lano1106]

Cycle #9, day off #2:

I bought MLS for my GF who has returned to the university and she is experiencing anxiety and mental blocks that stops her from learning her courses content.

Since, she started the program few days ago, she did report a lot of funky dreams... To me, this is the #1 sign that the subliminal is doing its work...

It feels the program is helping her. She appears to had a good day of study today...

Concerning my trading project, I have put a lot of energy to deploy fresh ideas to make stuff happens...

I haven't been very successful and my 30 days trading volume kept melting.

I had this hypothesis for some time that my algo is very sensible to market volume. I have became very serious about validating that idea.

I have inserted the weekly and daily exchange volume since 2020-02-24 in a spreadsheet.

My first automated trade did occur on 2020-03-27. To correlate the volume spikes with my good trading strikes, I did reread my BASE journal that covers the period from January of this year up to end of June.

As a side note, reading back that journal was very enjoyable. I did make me remember some of the very memorable moments during that period....

To make a long story short, the months of March, April and May were good for my system and during that time the weekly exchange volume was always above 1.5G.

400M+ volume days appears to be very good.

After that, June and July have been horrible. This is to be expected I guess. Summer time is always quieter due to vacations.

The surprising observation is that August have been busier than September. I would have expected the opposite. The last 2 weeks have been only marginally bigger than the average week in June-July.

It seems like binance is calling me which is the world busiest crypto exchange.... Too bad their API isn't as good as the one I'm currently using... I guess, I'll need to get used to it...

[lano1106]

Cycle 9, day Off #3:

Last night and today have been very productive days.

Many small improvements have been added to the system.

A glitch in a trade this morning made me discover a way to still improve it.

Tonight, Bitcoin price did drop by about $250 in minutes. Needless to say that my system was happy...

As I am writing this entry, I was thinking that it was maybe an indication of the return of a high volume trading day at the exchange... Nope... Still a slim day with a shabby $204M volume...

Another glitch did happen during this second trade...

It appears to be something very fundamental that I did overlook and this behavior probably is there since day one back in March... That could be a game changer to fix that one... but I feel too tired mentally to look into it tonight...

But I have it all recorded in the logs. All the hints that I need are there... They don't give much details but simply a clear evidence that something is wrong, it is enough for me to know that there is something to find to put all my will into finding what it is...

So bottom line, glitches don't bother me at all... In fact, I believe that they make me happy. They give me feedback on how I can improve my work. They are opportunities to learn. Learning experiences and indication that I am progressing in my journey are really satisfying...

Learning is fun.

I'm going to bed soon... I have some sleep to catch up. Yesterday, I went to bed very late and because of that... I kinda dragged some tireness with me all along the day... I want tomorrow to be different. Plus: I'm eager to start cycle 10 with returning to listening to the OF loops!

[lano1106]

Cycle 10, day #1:

I have been hunting my bug for the whole day. I haven't found any plausible explanation for what I am seeing in the log.

It is a total mystery.

In a nutshell,

1. Event y happens
2. change data inside a structure
3. Store aside the first element of that structure in var x
4. Go 4-5 functions deeper
5. Dump the whole data structure where x is from and x is now nowhere to be found.

This is a very serious issue. This could be a major reason why I'm not detecting as much viable trades that I would like to have. Obviously, one or more of my assumptions are wrong. After a full day of static analysis, I haven't been able to identify what could happen.

I'll need to add more logs to have better picture and hope that the glitch happens again and that I'll catch new hints about the problem. The tricky part is finding out where and how to place those new traces. This is the part of the code that process millions of updates every 15-20 minutes. Adding traces at the wrong place will flood the logs and finding out what is happening will just be impossible if it is a once in a week occurence type of problem...

[lano1106]

cycle 10, day 2:

I ended up adding 2 clever traces in the code. I need to move on or else this bug is going to drive me insane.

If the problem happens again, I'll catch it and I will extra info on the issue which should hopefully give me new insights into the issue.

Yesterday and today, the markets are very quiet. Yesterday, the exchange volume was only 184M...

I think that 2 trades have been detected but my system attempts to grab the first position of theses did fail. It does happen. Despite, having put a lot of energy to make my system as fast as possible, 100% success rate is impossible.

I expect that between now and next week-end, things will get more exciting. The reason being that a lot of investment portfolios get balanced at the end/start of each month.

I haven't rebooted the server because the last trade is still pending... From this state, I have some code that can be tested with the right event.

I made a really strange dream... My family was threatened by a murderous 10 year old boy that invaded our home. The bastard did injure the dog with a kitchen knife. I was trying call 911 to get some help but that did not work... I have been forced to use the rolling chair that has belonged to my grandfather that I did receive in inheritance to push the mad kid against the wall and I did use his own weapon on the poor devil as self-defense...

Super weird dream... I don't like dreaming about violence.... There must be some stuff preoccupying me that stir anger in me to dream about such things...

[lano1106]

cycle 10, day 4:

Yesterday, I met a malicious person. The harassing municipal inspector is back after an absence of almost a year.

I felt something similar to fear but I'm not 100% sure... It was perhaps adrenaline without fear... Bottom line, I'm 99% of the time at home and his visit was exactly while I was at the dentist to have my stitches removed. I have learned about his visit by receiving a notification from my smart bell that I have installed because of him.

I drove as fast as I could because that means it was my gf that was there and she didn't know how to behave with him.

I confronted him and I felt that he was afraid I was moving forward him and he was stepping back... Maybe it was because of covid or me but this observation was satisfying. I hope that he got bitten back by DRS because this person 100% negative...

As expected, there is no point to argue with him. He is hostile and isn't interested at all in discussing and finding a solution. They want to initiate a legal action against me... I'm going to require from the city a security for cost because I have all the reasons to believe that they are acting maliciously with unjustified hostility.

I bought a new laptop for my GF. I pretty good small machine. When I saw the machine, I felt that it was the one I needed. The processor just got release in March of this year, therefore the machine is even newer than that. Probably a new model from this summer.

So, gf has bleeding edge technology which should be good but it is not always that good for a Linux user. I have installed Linux on that machine only to find out that the sound chip is bleeding edge too and not recognized and supported by the OS yet. So everything works fine except the sound... There is not much I can do about it except report the bug to the Linux developers.

I got a similar issue 2-3 years ago when I purchased my first laptop with USB-C port. Beside Apple computers, it was quite the first PC featuring those ports back then. Guess, what? they didn't work well at first. I did figured out a way to work around that issue and few months later they started to work without me noticing when because of some software update...

This is the same situation except not having sound is more annoying...

2 days ago, I made my longest bicycle ride of the whole year... Idk, maybe 30 miles... And the not so unexpected result of that is that I have dreamed about it last night.

The only difference was that in reality I was alone and it was in the city while in my dream I was with my family and we were in a forest.

[lano1106]

cycle 10, day 5:

The day is almost over yet I feel importat to record an entry in my journal to keep track of where I am in the cycle.

Today went extremely fast...

Idk, if I am drifting away from my main goal which is that have a working and profitable automated trading system but because few days ago, I have completed a major milestone for that project. I gave myself the permission to put it aside for few days to take cares of all the other tasks that I did put aside to dedicate myself to my project.

Installing Linux on my gf laptop was one of them (And a kernel dev did help me to resolve my sound issue too!).

Other side tasks that I have is:

Get my cryptocoins from crypto forked from BTC in 2017. That is BTC Cash and Gold. Gold isn't worht much and possibly isn't worth the trouble to get those back but BCH is worth 300-$400 dollar each. I have few hundreds dollars sleeping on that chain. The concept is if you owned BTC at the moment of the fork, you automatically get the same amount in the new chains.

So I downloaded the BCH client source code... and I'm realizing that it is not only the blockchain that did fork but the source code as well. This is worrisome because it appears to be conflicting with the BTC client. I don't want to install a new client that will corrupt my local BTC data... I'll need to invest some time to read the code and make some changes to few parameters (at least the exec names and its various working directories..)..

The other project that I have is to install some miner on the computers that I have at home. So many of them are turn on all the time just doing nothign and wasting electricity and dissipate heat...

Monero could be a good candidate for that as it is designed to be worthwhile to mine with CPUs... I just need to find a clever way to do that. The basic Monero client requires you to be a full node in order to mine... It just doesn't make sense to have as much Monero full nodes than I have computers in my house... So, I started to study the Monero daemon source code and I have stumbled into something very cool... The JSON library that it is using is EXACTLY 14 times faster than the one that I am currently using!

This is something that is of interest to me considering that my system is processing millions of JSON messages per hour... Having this performance boost in place might have a positive influence on its result or at least will make it more scalable when I add other exchanges to it...

Actually, I started to read this library code and it made me see something in my own code to eliminate possibly unnecessary copies at the minimum and might even be the source of problems (such as the latest elusive issue that I did talked about few days ago...)

So bottom line, I trust the plan and my intuition... It seems like I'm stumbling on useful stuff no matter where I look... As if success was inevitable...

[lano1106]

Cycle 10, day6:

Integrating the new JSON library starts to work.

The author didn't touch it for about 5 years and it is still perfect.

My first issue with it was that it didn't support my specific situations:

1. I receive JSON payload in non terminated string. I could copy them into a new terminated buffer but the whole point is to minimize memory copies. The old currently used lib support that by offering a function allowing you to specify the buffer length.

2. When sending out JSON message, I need to know exactly how much space the JSON encoding will need to request a msg frame with a payload of the correct size from the network protocol layer. The encoder must then output into that payload space without terminating the string with a NULL char.

The library design is so neat that I have been able with code to adapt it to my exact need. I intend to share those custom classes to the author once I have successfully tested them all... Who knows maybe he is going to like what I have done with his work and decide to integrate my classes into his official lib... I think that he may have moved on to new challenges and perhaps not interested anymore in maintaining the lib.. but I'm going to let him decide...

It took some time to create a whole new pattern how my code handle JSON (it took the whole day). I had my first prototype ready to test at the end of the afternoon in a non critical spot and it did work like a charm. I need to polish a little bit more the pattern but once it is well refined, it is going to be a breeze to deploy it everwhere in my code and totally pull the plug on the old lib.

I am very thrilled to find out what it will do in my system hotspot where millions of JSON messages are treated. The new solution does not allocate a single byte from the CRT lib whereas the current libs allocate/free dozens of buffers for every single processed msg...

The last few days exchange trading volume still confirm the downtrend... The volume now is literally a quarter of what it was in spring... Either, investors are uncertain of what happens next in the market and remain on the side or the exchange volume shrinking is the reflect of the real world economy...

Either way, since my system doesn't perform well in low volume environment, it is becoming obvious that I need to integrate binance which is the biggest crypto exchange... I'm postponing this move because I know that this will require some colossal effort to pull it off...

In the meantime, Speeding up my system might allow the system to catch more detected good trades. It detected few one in the last day but they didn't last long enough for my system to catch them (ie less than 100 ms)...

So now, I'm going to start the off days and I'll start cycle 11 on Wednesday evening. What I can say about the cycle 10 6 on days is that they have been very rich in dreams. I believe that I made very big dreams in every single on day of cycle 10 including last night...

Subconcious is currently very busy shuffling things with the OF subliminal suggestions it seems...

[lano1106]

Cycle 11, day 1:

This morning, a trade did occur and the elusive mysterious problem that I have described in the last cycle day 1 and 2 entries has shown its ugly face this morning.

With the added traces, I have finally understood what was going on and it is a very fundamental flaw. It isn't generating problems but it may be a root explanation for the lack of detected trading opportunities...

The fix is baked in... I'm not sure if it is wise to install it right away because I'm expecting that the fix may generate a ton of activities which will be distracting to what I wanted to accomplish today...

I'm going to finish what I want to complete and installing the fix + what I'm working on, is going to be the self reward for the good work.

Yesterday was the last day my gym was open since starting today the gov did shut them down again for an undetermined period. It is funny, that I did receive a good amount of female attention by 2 cuties that I did see for the first time... To top it off, I did some dream again and it was a little bit sexuality themed... but TBH, I did watch a little bit porn before going to bed to relax myself and make sure that I did do insomnia because of my work preoccupation... that might have seeded my dreams a little bit since I did just tease myself a bit without going all the way with it...

[lano1106]

Cycle 11, day 2:

I have installed the fix... It must have an effect but not the desired trading activity explosion...

[lano1106]

Cycle 11, day 3:

I had a nice surprise this morning upon waking up. I went see what my trading system did while I was asleep to find out that it did 2 trades by using the full potential of what it is capable of:

that is concurrent trade execution and dynamic trade sequence modification and everything went flawlessly.

Replacing the JSON library is a much bigger effort than I initially expected. I thought that it could take 1-2 days. I have spent almost the whole week working on that...

but I'm almost done.. I started to change low risk/low frequency used code first to make sure that the new lib is working as expected. I must have now migrated roughly 75% of all the JSON code... I kept the high impact, most used code for the dessert... that is what I need to finish... Maybe tonight... Maybe tomorrow...

Also, I just realized that I am approaching the first anniversary of my trading project. I did start coding it last year in November... It is quite amazing that I have been able to create so far... and the best is still to come...