Windows: BLE workflow serial notifications are delivered twice
维护者通常 1 天内回复
还没有人认领这个 Issue。
评估
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 新手友好度
- 35/100
- Issue 类型
- 缺陷
- 描述清晰度
- 需要澄清
- 活跃度
- 冷清
- 技术栈
- javascript, python
- 领域
- embedded-iot
调研方向
Start by tracing connectToSerial() and its notification subscription path in the web editor. Compare whether notifications are duplicated on air or only after reaching the editor, using a BLE trace if available; check first connections and reconnects on Windows. Done means the source of the duplicate delivery is identified and the terminal no longer shows each notification twice.
由索引模型根据 Issue 内容生成。
描述
Claude wrote this:
Over the BLE workflow serial connection from Chrome on Windows, every notification arrives twice. Consistently, not intermittently. The same board and the same firmware do not show it from Chrome on macOS.
Not investigated beyond the observation, so the cause could be anywhere: the peripheral sending twice, the Windows BLE stack delivering twice, or the editor subscribing twice.
What is known
- Seen on Chrome on Windows; not reproduced on macOS with the same board and build.
- The board was a Feather nRF52840 Express running CircuitPython 10.3.0-alpha.4.
- It duplicates the data the user sees in the terminal, so it is user-visible rather than merely wasteful.
Worth checking first
- Whether the duplication is on air or above it. A sniffer or
btmon-equivalent trace would settle immediately whether the peripheral sends one notification or two. - Whether the TX characteristic's CCCD gets subscribed twice —
connectToSerial()in the web editor removes and re-adds itscharacteristicvaluechangedlistener before callingstartNotifications(), so a double subscription is a plausible shape for this. - Whether it survives a page reload, and whether it appears on a first connection or only after a reconnect.
Notes
Observed during the three-platform testing behind adafruit/circuitpython#11178, and originally recorded in a comment there that has since been removed because most of its content was superseded. This observation was not, and it is unrelated to that PR's changes.
Also unexplained on Windows, and possibly related: Chrome spins while reading boot_out.txt for the board-information dialog. That one is noted as pre-existing in #11178.
- 主要语言
- C
- 星标
- 4.6k
- 派生
- 1.4k
- 平均合并
- 1 天 7 小时
- 30 天内合并 PR
- 166
环境准备
这个项目没有提供开发容器、Dockerfile 或贡献指南，环境需要你自己搭建：先看它的 README，通用步骤见我们的新手贡献指南。
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
adafruit/circuitpython 的其他 Issue
-
board breaks api
难度 2/5 1-3 小时 新手友好度 68/100
adafruit/circuitpython#11099 · 3 条评论 ·
维护者通常 1 天内回复
-
storage usb zephyr
难度 4/5 3-5 天 新手友好度 48/100
adafruit/circuitpython#11531 ·
维护者通常 1 天内回复
-
silabs
难度 5/5 一周以上 新手友好度 20/100
adafruit/circuitpython#11507 · 1 条评论 ·
维护者通常 1 天内回复
-
Feature/API request: portable camera capture across ESP-IDF, Zephyr, and parallel interfaces可能已有人在做 @tannewt 于 1 天前认领。 未关闭circuitpython api displayio enhancement
难度 5/5 一周以上 新手友好度 25/100
adafruit/circuitpython#11505 · 5 条评论 · 已指派 1 人 ·
维护者通常 1 天内回复
-
难度 4/5 3-5 天 新手友好度 38/100
adafruit/circuitpython#11472 · 1 条评论 ·
维护者通常 1 天内回复
查看 adafruit/circuitpython 的全部 Issue
相似的 Issue
-
IO.get_env on Node truncates names at embedded NUL可能已有人在做 @Yi-111-a 今天认领。 未关闭
难度 2/5 1-3 小时 新手友好度 82/100
HigherOrderCO/Bend#1449 · 1 条评论 ·
-
bug C/C++ code
难度 2/5 1-3 小时 新手友好度 70/100
webarkit/WebARKitLib#84 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 82/100
OpenPrinting/cups#1751 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 72/100
es-ude/OnDeviceTraining#488 ·
维护者通常 1 天内回复
-
enhancement good first issue
难度 2/5 1-3 小时 新手友好度 66/100
维护者通常 1 天内回复