You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, unpackPyBundle() extracts pybundle.so on every startup of the main acitivity. If there is a previously running service waiting on select.poll(), it will crash during the extraction.
This was not a problem in the earlier versions of p4a, where unpackAsset and unpackPyBundle were together as unpackData.
Any ideas on how we can implement versioning for unpackPyBundle() ? This comment alludes to taking a hash of the file, but in some cases, the bundle is many megabytes in size, and would take some time to compute. in which case, can we use private_version string to implememnt disk versioning similar to what is done in unpackAsset?
Here is my take on resolving the issue. If this is acceptable, I can submit a PR.
03-17 11:38:42.590 881 881 F DEBUG : backtrace:
03-17 11:38:42.590 881 881 F DEBUG : NOTE: Function names and BuildId information is missing for some frames due
03-17 11:38:42.590 881 881 F DEBUG : NOTE: to unreadable libraries. For unwinds of apps, only shared libraries
03-17 11:38:42.590 881 881 F DEBUG : NOTE: found under the lib/ directory are readable.
03-17 11:38:42.590 881 881 F DEBUG : #00 pc 0000000000001f30 <unknown>
03-17 11:38:42.590 881 881 F DEBUG : #01 pc 00000000000033e0 /data/data/org.test.myapp/files/app/_python_bundle/modules/select.cpython-38.so
The text was updated successfully, but these errors were encountered:
Versions
Description
Currently, unpackPyBundle() extracts pybundle.so on every startup of the main acitivity. If there is a previously running service waiting on
select.poll()
, it will crash during the extraction.This was not a problem in the earlier versions of p4a, where unpackAsset and unpackPyBundle were together as
unpackData
.Any ideas on how we can implement versioning for
unpackPyBundle()
? This comment alludes to taking a hash of the file, but in some cases, the bundle is many megabytes in size, and would take some time to compute. in which case, can we useprivate_version
string to implememnt disk versioning similar to what is done inunpackAsset
?Here is my take on resolving the issue. If this is acceptable, I can submit a PR.
Example service.py
Example main.py
buildozer.spec
Command:
Spec file:
Logs
The text was updated successfully, but these errors were encountered: