Support Board
Date/Time: Sun, 18 Oct 2026 12:06:03 +0000
Post From: ACSIL Study for Automated Trading kicked off more than safeguard allowed
| [2026-09-04 03:42:45] |
| User158823 - Posts: 40 |
|
Thanks for quick response, cmet. Following up on the hotkey hypothesis, I'm attaching my complete automation footprint so everything can be reviewed in one place: 1. SierraChart_HotKeys_Master.ahk — my only hotkey script. Facts relevant to the held-key hypothesis: every hotkey is a single-fire mapping to one Sierra Chart customized keystroke shortcut (numpad 0–9, ., =). There are no loops, no timers, no Send/repeat constructs anywhere in the script. At the time of the incident I was not using the numpad layer — my manual trading is mouse-on-DOM. 2. Sierra Chart Desktop Switcher.ahk — manages virtual desktops and re-enables script 1 on focus changes. It contains no input injection into Sierra Chart (no Send/SendInput/ControlSend); its only timer is a 2.5-second one-shot tooltip cleanup, and its only PostMessage targets my own AHK script's suspend menu. 3. LogiKeepAliveActionsRing.ps1 — a watchdog restarting the Logi Options+ background agent (every ten min; set to five min at incident time) because the Actions Ring plugins hang. It only kills/relaunches named Logitech processes — no keystroke synthesis, no interaction with Sierra Chart. Along with the strategy .cpp/.dll already provided, this is everything running on my system. On cadence specifically: accounting for the OCO children (per the TAL, each retry is a parent + attached children tagged to my twconfig), the burst represents ~14 parent submissions per second sustained over ~9 seconds — faster than Windows key-repeat (~30/s ceiling on a held key, producing identical single events) and impossible for a mouse button, which does not auto-repeat. If review of these scripts shows anything that contradicts this, please, I want to know. If they're clean, my remaining lead is the Trade Window trigger-setup path — the resting orders showed as orange badges (trigger-setup color) rather than red/green placed-order colors, suggesting an incomplete trigger-arm transaction at the moment the burst began. I'm also happy to run a SIM session with all of these scripts exited if that would help isolate the cause. Thank you kindly. Warm regards, Gary Date Time Of Last Edit: 2026-09-04 03:58:25
|
Private File Private File Private File |