Error with sentry-sdk when saving MDA

[stantashev]

When saving an MDA acquisition—either from the GUI or programmatically—if “Send error reports to pymmcore-plus developers” is checked in the Error Log, the acquisition fails to run and no data are saved.

Instead, the acquisition stops immediately and produces the following error:

TypeError: Thread.start() takes 1 positional argument but 3 were given

File "C:\Users\Presse Lab\uvtest\Lib\site-packages\sentry_sdk\integrations\threading.py", line 115, in sentry_start
return old_start(self, *a, **kw)

The pymmcore version I am using is:
pymmcore-gui: 0.0.1rc0
python: 3.13.5 | packaged by Anaconda, Inc. | (main, Jun 12 2025, 16:37:03) [MSC v.1929 64 bit (AMD64)]
platform: Windows-10-10.0.19045-SP0
pymmcore-plus: 0.16.0
pymmcore: 11.10.0.74.0
pymmcore-nano:
core-version-info: MMCore version 11.10.0
api-version-info: Device API version 74, Module API version 10
adapter-path: ~\AppData\Local\pymmcore-plus\pymmcore-plus\mm\Micro-Manager_2.0.3_20260126
useq-schema: 0.8.3.dev4+gebc44baf5
pymmcore-widgets: 0.10.2.dev24+g4f17aa0cc
qt: PyQt6 6.9.0

[gselzer]

@stantashev thanks for the error report! However based on your report alone I'm not sure quite what the error is. But:

if “Send error reports to pymmcore-plus developers” is checked in the Error Log,

Perhaps this is the issue you're facing?

If so, it looks like this can happen when you set an interval of 0s in an acquisition with a time component like in the image below:

I can reproduce the error sent to sentry with that setup. Was this along the lines of your acquisition parameters?

[tlambert03]

interval of 0s

yes! (I was chatting directly about this with @stantashev and can confirm there was indeed zero interval)... which has been problematic (due to a useq bug, which, incidentally, is fixed in v2)

but I bet you're absolutely correct @gselzer... the really weird thing though is why that would manifest as

TypeError: Thread.start() takes 1 positional argument but 3 were given

File "C:\Users\Presse Lab\uvtest\Lib\site-packages\sentry_sdk\integrations\threading.py", line 115, in sentry_start
return old_start(self, *a, **kw)

I will say though: I have seen that exact TypeError before, in a similarly seemingly-unrelated context. So, I have a feeling it's not exactly a sentry bug. alas: I can't find the issue where I found it before.

[gselzer]

(due to a useq bug, which, incidentally, is fixed in v2)

I am very excited to see v2 start to make its way into broader usage (including from our widgets!) I've been tempted to try myself at pushing on that, but it's low on my priority list 😅

[gselzer]

We may want to transfer this issue to pymmcore-widgets, though (or even upstream to useq? unsure)

[tlambert03]

i think the crux of this issue is not the underlying bug as much as the fact that it presents itself to the user as a sentry problem... with the real bug hidden. That is an issue here in this library

[stantashev]

@stantashev thanks for the error report! However based on your report alone I'm not sure quite what the error is. But:

if “Send error reports to pymmcore-plus developers” is checked in the Error Log,

Perhaps this is the issue you're facing?

If so, it looks like this can happen when you set an interval of 0s in an acquisition with a time component like in the image below:

I can reproduce the error sent to sentry with that setup. Was this along the lines of your acquisition parameters?

Hello Gabriel ,
Sorry for the late response. Yes, the initial acquisition was with 0s interval, but the bug is present even with 1s interval between frames (see attachment).

The issue gets "resolved" when uninstalling sentry-sdk.

[fdrgsp]

Hi @stantashev, just out of curiosity, did you try with python 3.12?