Michael Hanke
Michael Hanke
Thanks @jsheunis! Based on all we have heard in here (so far), I propose the following action: We close this issue, because we do not think that `wtf` represents a...
This issue is categorized, and the resulting test disabling wrong. The failure can also be triggered when running the test on a crippled FS (`tools/eval_under_testloopfs`). Here is the result when...
If this is done, it should come with speed considerations. The idea was to terminate as early as possible when any modification was detected.
This relates to and is in part caused by #6875 (not just for subdatasets). We could distinguish `modified` and `dirty` with relative ease, if we would consistently **not** report a...
https://github.com/datalad/datalad-revolution uses pathlib to avoid this code entirely
The release of 0.12 with this fix is imminent (#3993) closing.
Well, the issue with the test itself remains, actually. It is still featuring `@skip_if_on_windows`. Although the usage of helpers like this is much reduced, we simply ignore whether they work...
The issue title poses a grand question. The failure seems to happen on a smaller scale: `run-procedure --discover` finds the `crawler` extension to be installed and assumes it to be...
Thanks for pointing this out. It is good to keep it in mind, but hard to act on it. What would we do, if someone ran `save` in a dataset...
Looking in more detail at the severity of this issue for datalad-use specifically: ```sh % datalad create super % cd super # create works % datalad create -d . --...