Fix chat module bugs #1561
Fix chat module bugs #1561
Conversation
|
Hi @SriramS-77, Thanks very much for this! Two small administrative things, could you add a "Chat: " prefix to the commit title? To be clear I mean the commit title not the PR title. Also if possible could you rebase on master to remove the "Merge branch 'master' into master" commit? I'll give this a quick test shortly but it looks good, thanks! |
| @@ -42,29 +42,47 @@ def __init__(self, mpstate): | |||
|
|
|||
| # create chat window (should be called from a new thread) | |||
| def create_chat_window(self): | |||
| print("Trying to create chat window") | |||
There was a problem hiding this comment.
we should remove this debug output
| self.chat_window.close() | ||
|
|
||
| # unload function override for module interface | ||
| def unload(self): |
There was a problem hiding this comment.
I've tested "module unload chat" and it doesn't seem to work I'm afraid. Please tell me if I'm doing something wrong. I get the message shown below. Perhaps it's because I'm using WSL2?
STABILIZE> Unloaded module chat
Trying to create chat window
Reloaded module chat
(mavproxy.py:7420): Gtk-CRITICAL **: 21:57:40.028: gtk_box_gadget_distribute: assertion 'size >= 0' failed in GtkScrollbar
[xcb] Unknown request in queue while dequeuing
[xcb] Most likely this is a multi-threaded client and XInitThreads has not been called
[xcb] Aborting, sorry about that.
python3: ../../src/xcb_io.c:175: dequeue_pending_request: Assertion `!xcb_xlib_unknown_req_in_deq' failed.
SIM_VEHICLE: MAVProxy exited
SIM_VEHICLE: Killing tasks
There was a problem hiding this comment.
I have only tested on Windows so far. I will look into it.
There was a problem hiding this comment.
I looked into why it was not working in Linux (I tried in Ubuntu to recreate the error). It seems wxPython is built on top of GTK in Ubuntu, which is much stricter than the Windows version. It is not allowing reinitializing app or rerunning mainLoop in the same process. I tried different approaches to getting it working in the same process, but keep getting seg fault.
I think it would be better to launch a different process for the app using multiprocessing. What are your thoughts?
There was a problem hiding this comment.
Another approach would be to:
- Hide the chat window and clear the data (unloading).
- Show the window along with new assistant thread (reloading).
I think this might be better, since there is no need to rework the implementation for handling mavlink acknowledgement.
There was a problem hiding this comment.
Hi @SriramS-77,
Either way sounds good to me. By the way, do you have an OpenAPI key? I just checked the list of ArduPilot sponsored API keys and I don't see you there. If you'd like one, please just PM me on ArduPilot's discord server, I'm "rmackay9" over there too.
Fixes:
Added "chat hide" feature, to interact with existing "chat show" feature.
Demo Video:
Git_PR_Demo.MP4