And bravo on the issue number too
@innovate-invent I have already worked on getting that one sorted (albeit for a different reason) and will continue with the suggestions soon, so that one is already being taken care of (galaxyproject/galaxy#9875). I'm currently waiting for a proper development laptop to arrive so I can continue working on it.
do you mind linking that in the issue?
others are also interested in following the progress
@innovate-invent I'll let you know when the PR is done, I hope to have it in the 21.05 release
as a usability tweak... when I search for a dataset in the multi-history view, I would like to only see those histories where the search has a hit, rather than see a whole lot of empty histories that I need to scroll though. I think this might be a not-too-hard thing to tackled, but can someone point me to where to start in the code?
@pvanheus I haven't actually built the multi-history view yet, but you could try mounting the History/HistoryPanel multiple times. The intent was to separate a history component from the fact that the history being shown is "the current one" so we should be about ready to try and build that view.
As for finding histories that contain a text-match of a contained dataset, that is likely going to require a separate search ajax call which youwould use to determine which histories to try to render. That roughly the functionality you're looking for?
At any particular time a history component doesn't necessarily know the entire contents of all its datasets, because that could be a whole lot of datasets. So you will need to do that filter you described first, then render any matching histories in a loop.
@pvanheus The beta history component is in client/src/components/History, although a somewhat cleaner version is just about to hit dev (galaxyproject/galaxy#10971)
You would probably be interested in the HistoryPanel component. Probably you'd do some kind of history query/filter with the ajax call I mentioned, then try to loop over the results and render a HistoryPanel with each history object? I genuinely haven't tried it yet though
I use Galaxy with IRIDA quite a bit and IRIDA has a way of making many histories... thus the interest.
Hi! This is a question from Jochen @bimbam23 --problems using cvmfs with version 16.07 (!!) https://help.galaxyproject.org/t/toolshed-not-working-anymore-with-old-galaxy/5302 He exhausted the existing solutions (reset/allow cookies) except for upgrading (can't rn). Any other advice?
I know that's an old release, but if this was working and we just broke it with the update we should address that
Does anyone of you have an idea how you can display a collection, or better even, a specific dataset that's part of a collection in a workflow report?
For displaying the collection, I know that there is
history_dataset_collection_display, but where do I get an identifier for it?
I don't know, but can you extract the file as part of the workflow?
Conceptually, I agree with you. It's just that, practically, it forces you to generate a more complex history to get a nicer report.
1 reply
Yeah, that's another good point. It seems the WF editor doesn't let you label non-starred datasets. But without a label you cannot access the dataset in a report.
All this is still kind of acceptable, just making things a bit more complicated than necessary.
Worth an issue?
maybe just a coincidence, not sure, don't want to blame @wm75:matrix.org ;)
socket queue seems ok right now. Shoudl I try to get a pdf?