React UI: Select time range with mouse drag

[LeviHarrison]

This PR adds support to select the time range with a mouse drag.

time-select.mp4

Fixes #1237

[LeviHarrison]

I added the v7.5.8 distinction as that was the last version that was Apache 2 licensed and the one I got this file from.

[LeviHarrison]

I copied the header from another file and I just noticed it says CONTAINS FIXES FOR DISPLAYING NULL VALUES IN STACKED GRAPHS, which has nothing to do with this plugin, but I checked and all the other plugins have it. Maybe we should change that?

[roidelapluie]

We should update the license in other files

[LeviHarrison]

@juliusv Do you know which one this line actually pertains to?

[juliusv]

Nope! So #6603 just copied Grafana's vendored flot version into Prometheus. The reason for this was that the normal Flot version didn't handle null values in stacked graphs correctly, but Grafana's fork did, because they did something to it to fix it. So I guess one would have to do some archaeological digging in Grafana's Flot fork itself to try and find out what they changed, but not sure how easy it is. Also, who knows, maybe upstream Flot has been fixed wrt. stacked null values in the meantime.

[LeviHarrison]

I'll remove that line from this new file because we definitely know it has nothing to do with the fix. Doing updates on the other files/maybe using the new version of upstream Flot is a separate PR.

[LeviHarrison]

It may also be good to add the v7.5.8 tag to the other links, especially the ones referring to the Apache 2 license that is now not an Apache 2 license.

[roidelapluie]

Would it be relevant to unzoom on double click ? (unzoom with the double-clicked point as center)?

[roidelapluie]

Note: This is awesome :)

[juliusv]

Would it be relevant to unzoom on double click ? (unzoom with the double-clicked point as center)?

So far I am happy with just pressing the "back" button on my mouse, but not sure. Is double click intuitive for unzooming?

[LeviHarrison]

I think that double-click is too subtle an action. I'm sure many accidentally double-click on a regular basis, especially with a click-heavy element such as the graph.

[RichiH]

Nice feature!

Adding /* SPDX-License-Identifier: Apache-2.0 */ to jquery.flot.selection.js and noting where you copied the file from might be the best course of action.

I have adding SPDX Ids on my bucket list; this would should make licence checking easier going forward. Unclear if I will find time to apply them all to the old branches as well.

[roidelapluie]

Nice feature!

Adding /* SPDX-License-Identifier: Apache-2.0 */ to jquery.flot.selection.js and noting where you copied the file from might be the best course of action.

I have adding SPDX Ids on my bucket list; this would should make licence checking easier going forward. Unclear if I will find time to apply them all to the old branches as well.

Instead of doing it yourself, why not creating an issue ? The community might help here. That's a lot of repos.

[RichiH]

I fear that outside of you and me, most people wouldn't care. The community interest is so low that I needed to start https://github.com/RichiH/spdx_helper myself as there was nothing to automate addition (it needs to detect and exclude vendored files before I can make the first production runs).

[LeviHarrison]

If there's a consensus, I'd be happy to start the process here. Just to be clear, it would look something like this:

/*
SPDX-License-Identifier: Apache-2
Source: https://github.com/grafana/grafana/tree/v7.5.8/public/vendor/flot/jquery.flot.selection.js
*/

[roidelapluie]

We still need the apache header:

https://www.fosslife.org/how-apply-license-your-open-source-project

[LeviHarrison]

/*
Copyright 2015 Grafana Labs
SPDX-License-Identifier: Apache-2
Source: https://github.com/grafana/grafana/tree/v7.5.8/public/vendor/flot/jquery.flot.selection.js

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/

Ok here's a revised version. I garnered the copyright directly from the license. It is worth mentioning that I had to make some small modifications to get it to work, I'm not sure if that changes anything with the license stuff.

prometheus/web/ui/react-app/src/vendor/flot/jquery.flot.selection.js

Line 199 in c3a54d1

//anyUsed = false;

prometheus/web/ui/react-app/src/vendor/flot/jquery.flot.selection.js

Line 412 in c3a54d1

})(window.jQuery);

[RichiH]

@roidelapluie I disagree with pulling the full boilerplate in. But we're also at risk of bikeshedding

@LeviHarrison Please bear with us while we figure this out. It's not fair to you that need to follow our discussion

[roidelapluie]

@roidelapluie I disagree with pulling the full boilerplate in. But we're also at risk of bikeshedding

VM is way more competent than me in this area, I trust her.

[RichiH]

It's somewhat hard to argue with "someone else is right"; more of a monologue than a conversation.

The problem I see is that we would be saving literally the same information twice; the inherent risk is that this can diverge over time. A license blurb is also more legally binding than an SPDX header, as such the header becomes useless and you need to parse the file with brittle heuristics again. The law systems I am familiar with (Germany, USA) both rely on intent. Putting a SPDX header in makes the intent clear. As a court would always try to find out and follow intent, the SPDX header itself should(TM) be enough.

Along similar lines, claiming that Grafana Labs holds all the copyright seems wrong; https://github.com/grafana/grafana/blob/v7.5.8/public/vendor/flot/jquery.flot.selection.js clearly states that initial copyright was "Copyright (c) 2007-2013 IOLA and Ole Laursen." Having done this sleuthwork helping Debian's ftp-master ages ago, I know how endless this digging can be. A relative pointer to where the file was copied from initially seems to be enough for someone to dig if they needed to dig.

All that being said, I don't think this discussion should block the merge. For the reasons stated above I strongly prefer the short form proposed in #8977 (comment) but don't want to block either.

[RichiH]

Deliberately in a second comment: "Back" makes more sense to me than a double-click. In more complex drill-downs it can become unclear what the double-click would jump back to whereas atomic back steps would be explicit and predictable.

[roidelapluie]

double click is not the same as back, double click is zoom out. this is what grafana is doing.

[roidelapluie]

e.g. you do not need to zoom first to double click, you can do it directly.

[LeviHarrison]

I believe zoom out is range control, equivalent to entering a time range into the box on the graph controls, is it not? If so, I think it's unrelated to this and should be considered as a separate feature.

[roidelapluie]

Let's unblock this

[juliusv]

👍