If people want to simply explore what this project is about, but they are not completely clear on what a FAS name is, or they know what FAS is but don't know any packager FAS name, they simply can't enter and see at least some example output. That's why I included some example links in the readme in #41. But that doesn't help people who just see the homepage and not the readme (for example people who stumble to the homepage from https://fedoraproject.org/wiki/QA:Tools or from https://apps.fedoraproject.org/ - in the future). Or perhaps they are a packager who owns just a single package and their output doesn't illustrate what the dashboard is capable of.
What about including some example names that people can input (or make them clickable) to create a good idea of what the packager dashboard has to offer? One name for a reasonably short output of a few packages and updates, one name for a long output of possibly everything that the dashboard can display. Something like that?
Consider including groups. They may also serve as good entry level pages for volunteers.
While I absolutely understand the thought behind this, the actual *One name for a reasonably short output of a few packages and updates, one name for a long output of possibly everything that the dashboard can display. * is not necessarily a straightforward task. Especially since the underlying data change in time, and what was "short output" last month can easily be "nothing" or "long AF" in future. So selecting a couple of static "names" probably won't cut it.
If it the intent was "showing the capabilities" it might be better (while still a tradeoff, and I'm not sure a good one) to have an "example" username, for which the backend would just respond with static pre-populated data. The problem here is, that we'd need to make sure the "example" data are in sync (format and capabilities-vise) with what the backend responds with in reality.
So maybe it would be best to just select a -sig for the "example page" since there is a good chance of it always having "some data to be shown". Still not sure about a good way of making the easonably short output of a few packages and updates happen on real-world data though :(
Any bright ideas?
I think you're overthinking this, Josef :-) I really imagined just picking a few packagers names who are likely to have a small/medium/large output displayed, and mentioning them as a hint/example on the homepage. Like what I did in readme (but that was just an immediate 1-minute idea, perhaps we can actually spend 10-15 minutes going through different packagers or groups and finding some well suited ones). Yes, their content will change over time, but if someone has 5 packages, it's unlikely that he will have 50 packages in a few months. The same situation is when somebody has 100 packages already. The example output might not always show all the stuff we wanted, it might be shorter or longer than a few months back, but that's good enough in my eyes (and an improvement over the existing state & it can be trivially implemented)! Also, that's the point of having several example names, if one of those packagers retires all his packages and his output is suddenly blank, well, there are still more examples you can click on, no big deal. And we can trivially adjust it any time.
We've settled on something like this:
which will lead to a detailed page containing links to some pre-selected dashboards, and some tips/explanations on using the UI.
Good enough for you @kparal?
Well I didn't see the detailed page. But it surely sounds like an improvement over the existing state :-) :thumbsup:
This is deployed for a while now, closing. Feel free to reopen/comment if you find any issues or typos.
Metadata Update from @lbrabec: - Issue status updated to: Closed (was: Open)
This issue has been migrated to Fedora Forge: https://forge.fedoraproject.org/apps/packager_dashboard/issues/43
Please continue any further discussion there.