Tighten etlas prose.

1 files changed, 15 insertions, 18 deletions

diff --git a/_log/etlas.md b/_log/etlas.md
index 9e2d191..7c50c72 100644
--- a/_log/etlas.md
+++ b/_log/etlas.md
@@ -20,20 +20,20 @@ DHT22 sensor.
   </tr>
 </table>
 
-Stocks: Two weeks EOD prices from Polygon.io (max SRAM holds). Flask app on VPS
-relays feed, lets me swap symbols via config. FastCGI + httpd + htpasswd.
+Stocks: Two weeks EOD prices from Polygon.io (max possible). Flask app on VPS
+relays feed; manages watchlist. Backend: httpd + htpasswd + FastCGI + Flask.
 
-gui_plot_stocks() triggers WDT. vTaskDelay() prevents that. Stepped graph was
-easier to code for the bit-packed framebuffer (not enough memory for much
-else), but it's messy--too messy.
-
-Note to self: Refactor gui_plot_stocks(). Bresenham’s line algorithm perhaps?
+gui_plot_stocks() triggers watchdog; vTaskDelay() required. Stepped graph was
+easier to implement (under memory constraints), but the logic is hideous. Note
+to self: Refactor with Bresenham’s?
 
 News: Channel NewsAsia RSS. MCU parses XML directly. Character glyphs stored as
-bitmaps in sprites directory.
+bitmaps in header files.
 
-Weather: DHT22 via PWM. 26µs/50µs/70µs pulses too fast for standard ESP32 APIs.
-Compare relative pulse widths:
+Weather: DHT22 single-wire protocol. 26µs/50µs/70µs pulses are too fast for
+standard ESP32 APIs. Bit-banged relative pulse widths: (Ported from <a
+href="https://github.com/Fonger/ESP8266-RTOS-DHT" class="external"
+target="_blank" rel="noopener noreferrer">ESP8266</a>):
 
 ```
 static inline int dht_await_pin_state(int state, int timeout)
@@ -91,15 +91,12 @@ static inline int dht_get_raw_data(unsigned char buf[BUFLEN])
 }
 ```
 
-Ported from <a href="https://github.com/Fonger/ESP8266-RTOS-DHT"
-class="external" target="_blank" rel="noopener noreferrer">ESP8266</a>.
-
-After flashing new versions, display would stop responding. Didn't observe this
-with e-reader. Issue occurs spontaneously. Couldn't find root cause. Making a
-change like adding a comment or tweaking delay in refresh() (then changing
-back) fixed it. Noisy electrical connections?
+Ghost in the machine: epd_init() stalls intermittently on first refresh() after
+flash. Toggling delay values in refresh() resolves it. If first refresh
+succeeds, it remains stable. Root cause unknown--suspected power supply issues
+from powering display via MCU.
 
-Once running, it runs: August 2024 - December 2025.
+Uptime: August 2024 - January 2026
 
 Commit:
 [a92c86a](https://git.asciimx.com/etlas/commit/?id=a92c86ac1592c2137d3d1fec1668eacc2d0ca581)