jquery-migrate

[fulv]

No description provided.

[fulv]

For the record:

WARN: 'JQMIGRATE: jQuery.fn.load() is deprecated'

This is caused by the jquery.anythingslider plugin, which has jQuery pinned at ">=1.4.2", so it's probably expected to work even with old versions of jQuery.

[fulv]

The warning

WARN: 'JQMIGRATE: $(html) HTML text after last tag is ignored'

was caused by this type of code, which was repeated in multiple places:

var $el = $('<div class="pat-stacks"></div');

Note the missing final angle bracket. I'm not sure if this was intentional, or simply a typo that proliferated due to copy & paste.

Ref: https://github.com/jquery/jquery-migrate/blob/1.x-stable/warnings.md#jqmigrate-html-text-after-last-tag-is-ignored

[fulv]

Now we only have the following warnings left:

WARN: 'JQMIGRATE: jQuery.browser is deprecated'
WARN: 'JQMIGRATE: jQuery.browser is deprecated'
WARN: 'JQMIGRATE: jQuery.fn.load() is deprecated'
WARN: 'JQMIGRATE: jQuery.fn.load() is deprecated'
WARN: 'JQMIGRATE: Can't change the 'type' of an input or button in IE 6/7/8'
WARN: 'JQMIGRATE: Can't change the 'type' of an input or button in IE 6/7/8'
WARN: 'JQMIGRATE: Can't change the 'type' of an input or button in IE 6/7/8'
WARN: 'JQMIGRATE: Can't change the 'type' of an input or button in IE 6/7/8'

The first one is probably going to be tricky.
This can probably be remedied by judicious use of http://api.jquery.com/jQuery.support/, or modernizr.

fn.load(), as mentioned in an earlier comment, is caused by jquery.anythingslider, which is only used by carousel-legacy. Not sure if we should drop this at some point, or build it with our own fork of anythingslider.

The JQMIGRATE: Can't change the 'type' of an input or button in IE 6/7/8 can safely be ignored, in my opinion.

Please evaluate whether to merge this now, to provide a baseline for further upgrade to jQuery 3. Either way, I'm planning on building on top of this branch for further migration work.

[fulv]

I'll leave the tests to fail until I'm done with all the fixes, instead of adding jquery-migrate.

[pilz]

Really nice progress.
Note that we don't really support IE6/7/8/9 and even 10 is out of scope as it is not even supported by MS anymore. I believe to remember that the jquery.browser stuff was to work around some old IE bugs. If that is correct, I would be fine by removing that compatibility code instead of doing a costly replacement of jq.browser.

[fulv]

@pilz good to know, I will keep that in mind.

[fulv]

I figured out the issue:

WARN: 'JQMIGRATE: jQuery.fn.offset() requires an element connected to a document'

It comes from select2: select2/select2#5173

I have not tried different versions of select2 to see whether it goes away, because the above issue is still open, so I guess there is nothing we can do.
From what I can see, there is no other code causing this warning other than select.

FTR, this is where the warning gets triggered: https://github.com/select2/select2/blob/stable/3.5/select2.js#L1386

[fulv]

And this concludes the migration. The remaining warnings are all due to external dependencies, i.e. select2 and anythingslider:

WARN: 'JQMIGRATE: jQuery.fn.bind() is deprecated'                                                                                                                                    
WARN: 'JQMIGRATE: jQuery.fn.bind() is deprecated'                                                                                                                                    
WARN: 'JQMIGRATE: jQuery.fn.unbind() is deprecated'                                                                                                                                  
WARN: 'JQMIGRATE: jQuery.fn.unbind() is deprecated'                                                                                                                                  
WARN: 'JQMIGRATE: jQuery.fn.load() is deprecated'                                                                                                                                    
WARN: 'JQMIGRATE: jQuery.fn.load() is deprecated'                                                                                                                                    
WARN: 'JQMIGRATE: jQuery.fn.offset() requires an element connected to a document'                                                                                                    
WARN: 'JQMIGRATE: jQuery.fn.offset() requires an element connected to a document'                                                                                                   

I verified this by hacking select2 and anythingslider in node_modules, and by doing so I was able to get a totally clean test run.

Next steps:

• Update the dependencies which are mentioned by npm and github as outdated or containing vulnerabilities by uppinning
• Checking tests again
• Creating a test bundle and run our integration tests with the new jquery bundle
• Do manual tests with the new bundle.

[pilz]

Excellent. Can you try to merge current master in? I think it already works with a modern select2, that might fix it. Perhaps there is also a newer version of anythingslider?

https://github.com/patternslib/Patterns/network/dependencies shows 11 vulnerabilities, the next step should be to upgrade them to latest version and check again.

[fulv]

@pilz where do you see those 11 vulnerabilities? When I go to https://github.com/patternslib/Patterns/network/dependencies, I can't find anything related to vulnerabilities.

[fulv]

whoa!

🎆 ⚡️ 😌 😆 💥 👊 👏 👯 💌 💪 👍 😂 😁 😤 🆗 👌 🙆‍♂ 🙆 🤛 👊 ✊ 🤜 🤘 😹 🎉 💣 🎂 🎁 💝 🎈 👑 💎 💴 💶 💷 💵 💳 💰 🎊 🎄 🎅 🤶 🎏 🏁 🌠 🏮 ㊗️ ✳️ 🔝 💟 💲

[pilz]

nice, eh? I do see some random failures on master now which don't occur locally, and may be timing issues. They are always on pat-scroll but the failing test is also changing randomly. I don't consider that a problem, but annoying.

[fulv]

Yes, I know the pat-scroll test failures, I see them randomly on local, too. -- Fulvio

…

On Tue, Apr 30, 2019 at 12:24 PM Alexander Pilz ***@***.***> wrote: nice, eh? I do see some random failures on master now which don't occur locally, and may be timing issues. They are always on pat-scroll but the failing test is also changing randomly. I don't consider that a problem, but annoying. — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#611 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADQPRT2JA6IX26E5ZXJDH3PTAM5HANCNFSM4FLQP57A> .