Webadmin 'Read-Write' perms

[will-moore]

This allows Admins to create 'Read-Write' groups in webadmin. See https://trac.openmicroscopy.org/ome/ticket/10751

To test:

• Log in as Admin, go to webadmin, create a Group...
• See that there is now a 4th "Read-Write" option for group perms.
• Create a group with these perms.
• Test new group behaves / appears as expected in web (or insight)
• Edit the new group (check that perms are "Read-Write" before editing)
• Edit permissions and check updated OK.

[joshmoore]

NB: I'd propose this is one for a "space", i.e. something it's good to get in and start testing, but which realistically might not make it into 5.1.2 based on other constraints (like testing).

[joshmoore]

@pwalczysko suggests that this should be marketed properly since this is a special type of group where there needs to be a lot of trust. cc: @hflynn

[joshmoore]

General feeling that we should likely show a warning to sysadmins when creating such a group. cc: @jburel & @dominikl for insight; @sbesson for CLI.

[mtbc]

Most likely server bug in r-w group is deleting others' data (e.g., annotation on image that is being deleted) in obviously wrong cases. (Less clear cases can probably wait for larger permissions review.)

[hflynn]

I've stuck a card on the docs board for this - https://trello.com/c/dUmuMrkk/325-read-write-groups

[will-moore]

Added a warning on selection of "Read-Write" group, using the same 'WARNING' dialog as when you select "Private" group. It's consistent but if this is too aggressive we can think again....?

[mtbc]

I like it. Better warn ahead of time, as it's not like our deletes have an undo.

[hflynn]

I agree, better a slightly aggressive warning message than angry users who have lost their data!

[pwalczysko]

@will-moore : I suppose the warning will pop up even when just upgrading the permissions of an existing group (in some respect this is the more dangerous workflow)

[will-moore]

Certainly that last commit shouldn't have caused any Travis failures.
Travis errors are:

[nomemorytestng] PASSED: testUsage on null(ome.services.graphs.GraphSpecUnitTest)
[nomemorytestng] FAILED: testCreateJPEG
[nomemorytestng] org.testng.internal.thread.ThreadTimeoutException: Method org.testng.internal.TestNGMethod.testCreateJPEG() didn't finish within the time-out 5000
[nomemorytestng]    at java.awt.geom.Area.pathToCurves(Area.java:193)
[nomemorytestng]    at java.awt.geom.Area.<init>(Area.java:126)

and

[nomemorytestng] 2015-05-04 18:03:35,590 WARN  [n.sf.ehcache.config.ConfigurationFactory] (      main) No configuration found. Configuring ehcache from ehcache-failsafe.xml  found in the classpath: jar:file:/home/travis/build/openmicroscopy/openmicroscopy/lib/cache/net.sf.ehcache/ehcache-core/jars/ehcache-core-2.1.0.jar!/ehcache-failsafe.xml
[nomemorytestng] 2015-05-04 18:03:46,093 ERROR [ o.services.sessions.stats.MethodCounter] (calStats()) ome.services.messages.stats.ObjectsReadStatsMessage[source=ome.services.sessions.stats.MethodCounter@243a0c0b] produced an error: java.lang.NullPointerException
[nomemorytestng] 2015-05-04 18:03:46,109 ERROR [ome.services.db.SelfCorrectingDataSource] (oRetries()) Failed to acquire connection after retries=0
[nomemorytestng] ome.server.utests.SelfCorrectingDatabaseUnitTest$MySQLException: null
[nomemorytestng]    at org.jmock.core.stub.ThrowStub.invoke(Unknown Source) ~[jmock-1.0.1.jar:na]

seem unrelated to this PR.

[joshmoore]

Kicked travis.

[atarkowska]

All seems to be working fine. No issues with adding RW group and updating users. Deletion works as expected.

Just noticed one issue with deleting Tag

<li class="jstree-last jstree-open" rel="tag-locked" id="tag-803"><ins class="jstree-icon">&nbsp;</ins><a href="#" class="jstree-clicked"><ins class="jstree-icon">&nbsp;</ins>by member</a><ul style="">
        <li class="jstree-last jstree-leaf" rel="dataset" id="dataset-154"><ins class="jstree-icon">&nbsp;</ins>
            <a href="#" class=""><ins class="jstree-icon">&nbsp;</ins>test</a>
        </li>
    </ul></li>

Cannot delete someone else tag because is locked "tag-locked" even Trash button is enabled. Data manager should be reviewed and fixed in the separate PR. This is good to go

[will-moore]

Made a note at https://trello.com/c/ScEErwT9/222-5-1-2-follow-up to fix Delete tag-locked.

[jburel]

Since we turn on that functionality in a point release we should complete it on this PR. (release in few weeks), the "follow-up" PR might not be ready on time and we won't have the functionality completed.

[will-moore]

@jburel Sorry - not sure I understand your last comment. Is there more that I need to do with this PR? Do you mean the delete of 'tag-locked'?

[jburel]

Discussion on RW (not only web related) scheduled on Wednesday morning when everybody is back.

[hflynn]

I'm going to make a start on the permissions docs BTW, I want to get this documented before I go on holiday next week.

[will-moore]

Disabled "Read-Write" for group owners:

[gusferguson]

@will-moore

Tested with trout merge user-6 and user-3.

The warning is showing up as expected for admin.
The grey-out of Read-write for group owners and tooltip explanation are good.

Good to merge.

[joshmoore]

re: #3783 (comment) -- what's the status of the delete tag?

[will-moore]

@joshmoore Handle delete of 'tag-locked' is on my todo list at https://trello.com/c/ScEErwT9/222-5-1-2-follow-up.

[will-moore]

@joshmoore I'm assuming that this is awaiting a final decision to "support read-write for 5.1.2" and that there's nothing else I actually need to do on this PR?

[jburel]

• Delete tag PR has been opened see Delete Tags #3827
• Minor follow-up for Read write #3811. This will be done in a follow-up PR (post 5.1.2)

When all the PR are ready, we should be good to go. gh-3827 will be reviewed again tomorrow

[joshmoore]

Merging for 5.1.2.