What you're reporting is correct.
We have only exposed the permission levels in between the original 4 levels.

That is correct. No message is printed to the terminal, but the help text does mention deprecation.
See #6417

I am not sure if you would call this a feature or bug report. We have a kubernetes cluster that does not allow systemd for processes. Is it possible to start iRODS without systemd?

One of our users has created a data object and a collection with the same name, both of which belong to the same parent collection. From what I can tell, the user created a collection named Week5 Lab in the collection Bio352. Next they uploaded a data object named Week5Lab into Bio352. Finally, they renamed the collection Week5 Lab to Week5Lab. I think they were using Jargon 4.2. I cannot reproduce the issue using iCommands.

Duplicate of #6267

We have not released the kerberos plugin for 4.3.0 as we intend for kerberos requirements to be met by the new pam_interactive plugin.
You can install iRODS 4.2.11 if you need kerberos functionality today.

rsZoneReport fails when generating the zone report because it attempts to invoke the following python function which was removed during the transition to Python 3.

For now, this can probably be worked around by avoiding including irods/rodsClient.h in baton's header files. Any structs needed for function declarations can be forward declared.

I was able to build baton against the irods-{dev,runtime} built from the main branch. That fix will be included in 4.3.1.

I hope I am not missing anything here. As an rodsadmin user I can list all available rules in delay queues by iqstat -a. As normal I can also list details of a specific rule that belongs to me by iqstat ruleId. Furthermore, I can list the details of a ruleId that belongs to a user that I know by iqstat -u user1 ruleId. However I couldnot see the details of available rules that I could list without knowing the user accounts that made those rules.

iqstat -la lists all.

What did you try to do?
run iget -f to retrieve a file to a non -existent dir
Yes, that is weird, but not a bug.
That has to do with the OS and reference counting. See https://stackoverflow.com/questions/12359598/delete-current-directory-on-linux

The implementation for delay server migration only considers the name returned by gethostname.

Yes, this is due to ichmod honoring the data_owner_name. It should not. This will be fixed alongside #6579.

open() currently will honor the kw.DEST_RESC_NAME_KW keyword.
kw.RESC_NAME_KW and kw.RESC_HIER_STR_KW are also honored by open() - so it may be that there is no reason to also support kw.DEST_RESC_NAME_KW as it is designed for use with replication (with a pairing of source and destination).

Maybe we make a session.acls method that is "nearly synonymous" except it behaves more consistently than session.permissions (ie the default for the report_raw_acls parameter would be True). and at some future point, we'd deprecate session.permissions.

Querying acl on (getting permission of ) an object doesn't bring g prefix specifying group based permission although there is already an granted group permission exists. AccessManager behaves it like a normal user. On the other hand ils -A shows it is a group user.

#411 adds a user_type attribute to the iRODSAccess.
This will have the value "rodsgroup" in the analogous situation where ils -A would print a "g:" prefix.

Created an issue in the Logical Quotas plugin repo , including how to reproduce this same error -1205000, with the only plugin installed or active being logical quotas:
irods/irods_rule_engine_plugin_logical_quotas#84
In that issue, istream is used rather than PRC to evoke the error, but the same effect can be achieved in PRC with this put script and a single argument specifying a very small file to upload:

Good clients must report the server version they are maximally compatible with else they run the risk of using old behavior (e.g. bad packstruct/xml encoding - apos vs graves).

When adding a new user to a research-group, one can see all kinds of user email addresses available in Yoda including personal email addresses. These addresses have not been included by the user themselves. The image below is from my test account in Yoda DGK at UU. The same is also true for Yoda@WUR which shows a much longer list of users when you for example search on gmail to add a user. Users have indicated that this is undesirable to be able to see private email addresses of others. I'm not sure how this relates to the GDPR either, I can imagine the system does not require showing email addresses to the user to function correctly (and hence data limitation is not applied?).

At WUR we have a different data classification terms in our security policy
It would be nice to be able to have customised set of data classification terms for use during group creation and metadata editing.

We will lose this modification if the default schema is upgraded, so it might be a good idea to not put the enumNames in metadata.json itself (the same goes for the default value of affiliation).

See DOI 10.48338/vu01-2lt5v4
The metadata contains a Date field with "Date Type" "Updated". This was set to 2022-12-08Y11:57.46.000000.
When you try to update the dataset in DataCite Fabrica you can't update the DOI and it shows an error message stating:
"To save this DOI, first resolve the errors with these properties: date."
This field is not in the Yoda metadata so is probably automatically set in the Yoda to DataCite conversion.
Removing the decimals 2022-12-08Y11:57.46 resolves the issue.

When we publish a dataset the value of the Publisher field in the DataCite metadata is set to "Utrecht University". In our case this should be "Vrije Universiteit Amsterdam"
A quick search suggests the Publisher value is set in https://github.com/UtrechtUniversity/yoda-ruleset/blob/release-1.8/json_datacite41.py lines 153-155
This can't be changed at all by us and is difficult to change for SURF.

When trying to changing the subcategory of an existing datamanger group, I get the error
"Datamanager grup names may only contain lowercase letters (a-z) and hypens (-)