Page demos problems in Firefox 44

[pocesar]

When using the first input in the page https://leaverou.github.io/awesomplete/ it shows as this (double tooltip):

then after selecting the first item:

and down with the combobox dropdown, the toggle is misaligned

[vlazar]

Awesomplete was included twice by mistake. This created 2 completer instances. Fixed in 6c20ebe

[vlazar]

Closing this one since the first bug is fixed and second one is in separate issue now #16838

[pocesar]

👍

[TxHawks]

This brings up an issue though.

Should Awesomplete check if an input is already initialized before
initializing?

Could be done by checking the Awesomplete.all array

On Sun, Jan 31, 2016, 12:34 Vladislav Zarakovsky [email protected]
wrote:

Closing this one since the first bug is fixed and second one is in
separate issue now #16838
#16838

—
Reply to this email directly or view it on GitHub
#16836 (comment)
.

[vlazar]

Yes.

[LeaVerou]

I’m not so sure. It introduces a performance overhead on every use for little benefit (since including it twice is essentially the author's bug).
I'm not necessarily saying no, I'm saying "yes" is not that obvious to me.

[vlazar]

Not necessarily performance overhead. Including script twice is an obvious error. Creating new instance for input, which already has awesomplete maybe not.

Anyways, won't implement if it adds significant performance overhead or complicates code to much.

[TxHawks]

I think the easiest way to do this is to check is input is the same as
one of the inputs in _.all just before element creation.

I don't think the performance overhead should be significant, as everything
is already cached, so there is no unneeded DOM querying.

I think the check can probably be done pretty neatly with _.all.reduce()

On Mon, Feb 1, 2016, 18:50 Vladislav Zarakovsky [email protected]
wrote:

Not necessarily performance overhead. Including script twice is an obvious
error. Creating new instance for input, which already has awesomplete maybe
not.

Anyways, won't implement if it adds significant performance overhead or
complicates code to much.

—
Reply to this email directly or view it on GitHub
#16836 (comment)
.

[LeaVerou]

Why reduce and not indexOf?

[ghost]

Fuck you! Stop emailing me

Thank you,
Stacey

On Feb 4, 2016, at 7:23 PM, Lea Verou [email protected] wrote:

Why reduce and not indexOf?

—
Reply to this email directly or view it on GitHub.

[LeaVerou]

LOL, Stacey, nobody is emailing you. You (or someone using your email) has subscribed to this conversation. Just press unsubscribe.

[ghost]

I would have 3,823 emails ago. There is no unsubscribe. And thank you for not taking my message personally. Apple should allow blocking of email rather than just junk

Thank you,
Stacey

On Feb 4, 2016, at 7:25 PM, Lea Verou [email protected] wrote:

LOL, Stacey, nobody is emailing you. You (or someone using your email) has subscribed to this conversation. Just press unsubscribe.

—
Reply to this email directly or view it on GitHub.

[LeaVerou]

The notifications you are getting are automated for every message posted on this conversation. It's not an email conversation, the emails are just notifications that you can reply to. The actual conversation is on the website. Note the "View it on Github" link under every message? That takes you to the website.
However, to unsubscribe, I imagine you need to log in, so if you didn't register, you need to figure out who did using your email.

[ClausMandal]

Lea you radiate professionalism left and right! Well done.

[TxHawks]

You could either use reduce or all.map().indexOf().

You can't use indexOf() directly, because Awesomplete.all is an array of objects, each with a property of input, which is the HTML element

[LeaVerou]

Yes, that's what I meant. I think it would be more readable than reduce.

[TxHawks]

Sure thing. Want me to submit a PR?

[LeaVerou]

Eh, why not. :)

[vlazar]

The short answer would be - tests. I need to figure out a clean way to test Awesomplete's addition via script tag to the page. This is pretty much the only part not covered by tests yet.