An idea from sun and coming from #1090592: [meta] Use HTML5 data-drupal-* attributes instead of #ID selectors in Drupal.settings the module page is a pretty good use-case.
When checking to enable a module there is a JS check on dependencies and check all unchecked dependencies hence avoiding the confirm step.
The patch is very rough but it works :) any comment on this is welcome. It should be using jquery ui instead of a confirm (or maybe not, which is better for screen-readers?).
I'm not certain i'll be super-fast with 100+ modules enabled, there is a lot of room for perf improvement at this point.
Usability review to comment of whether avoiding the confirm form is beneficial or not.
| Comment | File | Size | Author |
|---|---|---|---|
| #24 | use_data_to_check-1473760-24.patch | 4.26 KB | alansaviolobo |
| #22 | use_data_to_check-1473760-22.patch | 1023 bytes | alansaviolobo |
| #20 | core-js-modules-dependencies-1473760-12.patch | 4.99 KB | capynet |
| #18 | core-js-modules-dependencies-1473760-11.patch | 4.41 KB | capynet |
| #7 | core-js-modules-dependencies-1473760-7.patch | 3.42 KB | nod_ |
Comments
Comment #1
nod_me no right spelling.
Comment #2
nod_forgot this too.
Comment #3
cosmicdreams commented@nod_ I looked into doing something like this a while back as well. But quickly got distracted by all the markup that is injected into the render array for the module page and sought out to search and destroy all instances of from that page.
Find the code that builds up the markup for the module page and you'll see what I mean.
IMO, the addition of data attributes gives us a great opportunity to work on the markup being output here. I think a refactoring of the markup should be apart of this patch.
As a next step, I'll try to put out a patch that shows what I mean.
Comment #4
nod_Comment #5
robloach#1: core-js-modules-dependencies-1473760-1.patch queued for re-testing.
Comment #6
fabianx commentedWow, this would be great to have in ...
Comment #7
nod_reroll for the fun of it :)
Comment #8
wim leersI think this can happen post-feature freeze?
Comment #9
Bojhan commentedI am a little skeptical about this, could this have some screens on the intended flow?
Comment #10
nod_#7: core-js-modules-dependencies-1473760-7.patch queued for re-testing.
Comment #12
capynet commentedRe-rolled.
And added modal with a list of modules being marked.
Comment #13
capynet commentedReplaced $.ui.dialog() by Drupal.dialog();
Comment #14
capynet commentedReplaced wrong library "drupal.dialog.ajax" by "drupal.dialog"
Comment #15
tstoecklerThe patch includes the patch file itself... :-)
Comment #16
capynet commentedFix patch loop o_O
Comment #18
capynet commentedRe-rolled.
Comment #19
nod_the dialog dependency should be in the library definition of drupal.system.modules, not attached with it.
Can you put
gettingDependencies: falseoutside Drupal.behavior? We should only have attach/detach in there.$that.is(':checked')to replace by$that.prop('checked')Drupal.dialog expects a HTMLElement, not a jQuery object.
Can you use the Array version of the button option? (We're told it's the preferred way to declare them http://api.jqueryui.com/dialog/#option-buttons).
Drupal.system is useless after all, I can bet people will confuse it for something it's not. Let's drop it and have the getDeps as a banal function in the closure.
The wording could use a little help. Feels different than the usual confirm step.
Can you also make the click handler a function instead of anon function? and also separate the declaration of the dialog options from the call of Drupal.dialog? There is too much nesting, we need to bring it down.
And as an optimisation for getDeps, we could try to use CSS selectors to do the work, kinda same way it's done in #2180921: [META] Use data attributes rather than classes wherever possible.
Thanks!
Comment #20
capynet commented"The wording could use a little help. Feels different than the usual confirm step."
In fact, I copied the confirm step.
"And as an optimisation for getDeps, we could try to use CSS selectors to do the work..."
Sorry, I dont understand what you mean.
Comment #21
nod_What changed exactly, still seeing extra stuff in Drupal.behaviors, anon functions, not a HTMLElement in Drupal.dialog call and too much nesting.
We can delegate the event as well here.
Comment #22
alansaviolobo commentedreroll
Comment #24
alansaviolobo commentedsorry. wrong patch added earlier.
Comment #29
martin107 commentedwe MUST no longer be modifying *.js files directly
We modify *.es6.js and then transpile as described here :-
https://www.drupal.org/node/2815083
as a ultra trivial point new code in ModulesListForm uses array() when [] is preferred.
Comment #35
nod_hesitating on closing this one. Is there really a UX benefit of doing this?
Comment #41
quietone commentedLet's see what Usability thinks of removing the confirm step.
Comment #42
quietone commentedComment #43
benjifisherUsability review
We discussed this issue at #3485024: Drupal Usability Meeting 2024-11-08. That issue will have a link to a recording of the meeting.
For the record, the attendees at the usability meeting were @benjifisher, @rkoller, and @simohell.
In #35, @nod_ raised the question of whether we should remove the confirmation step. The Usability team thinks we should keep that confirmation.
The cost of the confirmation step is small. Enabling modules is not something you do every day on a site, like editing content. There are some site owners who manage many sites, but they normally enable modules with
drush.There is sometimes a large benefit to the confirmation step. If a site owner wants to evaluate a new module, they may not be aware of its dependencies. (Before we used
composer, this was less true. Site owners had to download all the dependencies individually.) What if one of the dependencies is a module the site owner does not want to install? Perhaps it duplicates or interferes with another module, already installed, or perhaps it has only an alpha release.It can be difficult to uninstall modules. Occasionally, a bug in the module makes it impossible to uninstall from the UI. If there is a long dependency chain, then uninstalling requires several visits to the uninstall page.
We plan to close this issue (won't fix) in two weeks (2024-11-22) unless there is further discussion by then.
If you want more feedback from the usability team, a good way to reach out is in the #ux channel in Slack.
Comment #44
benjifisherIt has been two weeks since my previous comment, so I am going ahead with the plan to close this issue (works as designed).