google-code-export / ords

Automatically exported from code.google.com/p/ords
1 stars 0 forks source link

Dataset metadata can't be edited #519

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago
What steps will reproduce the problem?
1. Go to your list of saved datasets and think 'I'd like to change the 
description of this one'
2. Click on the dataset title
3. Find that instead of taking you to a page where you can edit the dataset 
details, you go to the query page instead

What is the expected output? What do you see instead?

In the list of project datasets there should be a 'run query' button next to 
each dataset, which will take you to the pre-populated querying page (as 
clicking on the dataset name does at present).

Clicking on a dataset name in the list of project datasets should instead bring 
up a form where you can edit the dataset details (provided you are an owner, 
admit, or contributor).

Please use labels and text to provide additional information.

Original issue reported on code.google.com by jajwil...@gmail.com on 12 Aug 2014 at 8:32

GoogleCodeExporter commented 9 years ago

Original comment by jajwil...@gmail.com on 12 Aug 2014 at 3:11

GoogleCodeExporter commented 9 years ago

Original comment by jajwil...@gmail.com on 18 Aug 2014 at 1:04

GoogleCodeExporter commented 9 years ago
I've just seen that you've moved this to 1.0.5, but I was already working on it 
and its now done.

Meriel, users can now edit datasets by going to the "View previously saved 
datasets" page and clicking the edit button - there's also a button for "Run 
query" in place of the linked view name and the delete link here is now a 
button too.  The form now also has proper validation and error messages.

The only thing that's not editable is whether it's a static dataset, since this 
may be problematic due to the way static datasets are stored.  If we people to 
be able to do this, I'll need to discuss it with Dave first.

Original comment by marxjohn...@gmail.com on 18 Aug 2014 at 3:21

GoogleCodeExporter commented 9 years ago
This sounds good - except that to keep the labelling consistent, the button 
should probably say 'View dataset' rather than 'Run query': I've been trying to 
keep a distinction between datasets. (I can change the button label as long as 
it has a text string that isn't being used somewhere else - I can't currently 
check whether this is the case, as the changes don't yet seem to be showing up 
in dev.)

Original comment by meriel.p...@gmail.com on 18 Aug 2014 at 3:39

GoogleCodeExporter commented 9 years ago
Hmm, well clicking the button has the same effect as the "run query" button on 
the query builder page (shows you the query builder page with the query 
executed), but I can certainly change it.

Just so I can try and keep it too, when you say you're keeping a distinction 
between datasets and queries, what's the distinction you're keeping?

Original comment by marxjohn...@gmail.com on 18 Aug 2014 at 3:52

GoogleCodeExporter commented 9 years ago
A query is a question you ask of your database using SQL. A dataset is a 
pre-saved set of query results.

In practice, yes, you view a non-static dataset by re-running the query that 
was used to create it. But it seems to me worth maintaining a conceptual 
distinction between the two. There are two ways to view a dataset: via a 
project you're a member of, or via the ORDS public project listing. If you go 
via the latter, you don't see the SQL query box or the query itself (meaning 
members of the public who aren't very familiar with databases may not even 
realize there's a query involved). When you view a dataset from within a 
project, it strikes me you're looking at the same sort of thing (though with 
the ability to see the generating query and tweak it if you so desire), so it 
seems to make sense for the button labelling to be the same.

Where static datasets are concerned, it seems (to me, anyway) slightly 
counterintuitive to describe what's happening as running a query - because the 
version of the database that was queried to produce the dataset may not exist 
any more. (I assume that's the case, anyway, and that ORDS is just saving the 
results, not the whole database they were taken from?)

Does the above make sense to people who aren't me?

Original comment by meriel.p...@gmail.com on 18 Aug 2014 at 4:48