Financial Instruments Business Operator Kanto Finance Bureau Director (Kinsho) No. 1960 / Membership Association Investment Trusts Association, Japan Membership No. 012-02323 Go to GogoJungle Home
icon List
柴亮太
09/22/2026 21:33
Release： 09/22/2026 21:33
Update： 09/22/2026 21:33
柴亮太
Investment style : Trading System
Investment career : More than 10 years
Investment amount : 10 million - 50 million JPY
Representative of RO LLC. My main line of work is AI agent research and development. I have about 25 years of investing experience, but it’s not a brag-worthy track record. I am not the type to sit still in front of a chart, and I would often cut losses when I looked away during critical moments. Through that repeated pattern, I have melted tens of millions of yen. What remains is knowledge and a record of my failures. “If I can’t watch it myself, I’ll build something that can watch for me.” With that in mind, I am also working on developing AI agent EAs. Not to predict the market, but to tell me when not to enter a position. To avoid silently expanding losses when prices move against me. Listing on GoGoJungle is a testing ground for that research and development. The tools I publish here are, as of now, completed products, but not the final finished version. They are things that run on actual accounts, record losses, and are cultivated while being improved each time. What didn’t go well is also shown as is. Therefore, for those who purchase, it isn’t a “buy and end” relationship, but a partnership to accompany the research. It isn’t that you are not winning because of your efforts. I have faced the same points and repeated the same losses many times. Those who stumbled in the same place, this time I hope we can turn to the winning side together and continue with verification. If you have questions or requests, please don’t hesitate to share. The feedback we receive will be reflected in the next version.
Article cover image

Unify GOLD13 to M5 original: disclose the deployment of time-frame isolation and the regression of icon display

Financial AI / AI Agent Research and Development · Indicator Foundation

Unify GOLD13 to the canonical M5: Deploying timeframe isolation and publishing icon display rollback

To prevent histories and judgments from mixing when switching timeframes, I made M5 the canonical execution timeframe. Meanwhile, since icon rendering on the VPS showed rollback, I report separately on deployment success and display success.

This article is a development log of the indicator GOLD13 I developed and the automated trading software (EA) I wrote, researching and validating them with AI agents and live accounts. The validation process is also published in a 24/7 weekday trading live stream.

Purpose of this change

In MT4, even with the same indicator, re-calculation and restoration run for each timeframe such as M5, M15, H1. If the execution timeframe is ambiguous, it can restore histories from other timeframes, cause duplicate judgments of the same event, or move the positions of past icons with subsequent ticks.

Therefore GOLD13 sets M5 as the canonical execution timeframe, i.e., the only execution canonical. Non-M5 instances become passive displays and do not proceed to event generation, ledger changes, or signal judgments.

Scope of deployment

  • Non-M5 is passive before ledger restoration
  • Stop signal calculation for non-M5 at the entry
  • Separate restoration and duplicate check by timeframe
  • Persist realtime icon anchor
  • Unify band calculations to a confirmed M5 bar
  • Input names, types, order, and default values unchanged

The EA, entry policy, AutoTrading, receiver, account settings, profile, template, and the actual chart timeframes have not been changed. This deployment targets only timeframe isolation.

Verification of VPS deployment

Deployed to a Singapore VPS on AWS Lightsail; MetaEditor on the VPS reported 0 errors and 0 warnings. I restarted BigBoss MT4 in the usual way and confirmed GOLD13 BAND_UPDATE and BANDTIP_CREATED, and EA positions = 0. At final verification, open position 0 and pending order 0.

I recorded candidate MQ4, pre-deployment backup, VPS rollback MQ4, and SHA256 of the generated EX4. Although the raw SHA changed due to BOM, CRLF, and leading newline differences in the VPS rollback MQ4, the normalized content matched the candidate. The EX4 build artifact is recorded as a result, not assuming deterministic generation.

What could be verified in natural events

After reboot, a natural M5 warning SELL event occurred. On the EA side, it was rejected as stale, and no orders were opened. This verified that the signal engine and the handoff path were functioning.

However, the production chart also connects FUNDAFX_EA5. Triggering artificial timeframe changes solely for validation would unnecessarily affect live operation, so LIVE round-trips switching from M5 to other timeframes and back have not been performed.

Display rollback found after deployment

In subsequent VPS REALTIME logs, event generation continued at SELL 16:15, BUY 16:48, 16:50, 16:55, 17:00. Meanwhile, each event reported BMPFILE setting errors, and only the signal icon on the chart failed to display.

Reason: the difference candidates for the timeframe isolation did not include six resource declarations, and OBJPROP_BMPFILE referenced an external image path rather than an embedded path. Image resolution conditions differed between local environment and VPS, so rendering failed only on the VPS.

What moves, what is broken

Signal engine, event generation, band updates, and EA connectivity did not stop. What is broken is the BMP icon rendering on the VPS. I will not describe this as “indicator fully stopped” nor “no problem” by conflating the two.

Display is an essential function for users to verify grounds of decisions. Even if order logic is operating, if evidence on screen is missing, operation quality cannot be deemed acceptable. I will treat function, display, and recording as separate success criteria.

Current status of the hotfix

The minimal hotfix restores only the missing embedded resource declarations and reference paths, and has been verified in an isolated environment with 0 errors and 0 warnings. However, deployment to production has not been performed. To fix the display without altering other logic, I need to save the current MQ4/EX4 and chart inputs and proceed afterward.

After deployment, I will verify MT4 reinitialization, disappearance of BMP errors, re-display of existing signal icons, continued event generation, and ledger/band/inputs maintenance independently. If anomalies arise, I will revert to the previous version and not change entry state arbitrarily.

Items to review during timeframe round-trips

  1. Record M5 signals, markers, ledger, and bands
  2. User naturally moves to another timeframe
  3. Ensure no events or ledger mutations occur non-M5
  4. Return to M5 and confirm past anchors and histories match
  5. Confirm no duplicate events and no icon movements

Until this round-trip is complete, LIVE_VALIDATION for timeframe isolation will not be considered finished. Offline regression PASS and actual screen verification are separate tasks.

Meaning as AI research and development

Even if AI financial judgment accuracy improves, if the indicators used as input have histories that change with timeframe switching, the foundation of learning and validation collapses. AI, EA, GOLD13, recording, and heartbeat must share the same timestamps and state to form a single validation line.

This time I disclose including the issues. By distinguishing the deployment facts, the confirmed operation, the detected rollback, the undeployed hotfix, and the ongoing LIVE verification, it becomes a change log rather than a mere completion report.

Acceptance criteria for next report

  • SHA and compile results of the production hotfix
  • Disappearance of BMPFILE errors
  • Display of existing and new icons
  • Continued event generation
  • Non-operation outside M5
  • History match after M5 recovery
  • Open / pending and EA running state

Even if some items succeed, it does not mean the entire process is complete. I will specifically verify display restoration and continuation of order logic as separate checks.

Monitoring to avoid missing display rollback

If the log shows many event successes but nothing is drawn on the chart the user views, the validation experience is broken. Therefore I will separately verify the number of events generated, OBJPROP settings results, actual number of objects, and how it looks on screen. I will not consider log success alone as the completion criterion; I will align screenshots with the live display.

Image resources may resolve in the local environment but fail on the VPS. I will check whether images were embedded at compile time, whether the runtime path points to an external folder, and whether file name case and separators match, and add VPS-specific conditions to testing.

Migration design to preserve history

Even if icons reappear after the hotfix, if the past signal timestamps or prices change, it’s a different issue. Instead of deleting and regenerating the existing ledger, I will compare object names, anchor time, price, and direction before and after deployment. Only new events will be corrected, and histories should not move.

For the band, I will verify not only the current value is drawn but also that the last closed M5 bar continues to be used. If you return to an forming bar, the boundaries may move later, and the description of live delivery and saved images will not match. Display stability directly affects the reproducibility of observation values given to AI.

Rollback conditions

If a compile error, initialization failure, event stoppage, EA connection abnormality, ledger missing, or inputs change occurs, I will revert to the MQ4/EX4 saved before deployment. I will not move execution paths in an uncertain state just to hurriedly fix display. After rollback, I will also retain the cause log and the failed version SHA.

GOLD13 Sales Page
Weekday 24-hour Trade Live Stream

This article is a research and development / live-account verification log and does not guarantee profits or future performance.