Depends on D24477

Depends on D24478

Hi Victoria,

In case you have some time for remote debugging, we are trying to add a feature that definitely needs UX guidance.
You can read https://phabricator.services.mozilla.com/D24478#718989 for more context.

When a user clicks on the "Connect" button in the Sidebar of about:debugging, we will try to connect to the remote Firefox. By default, a connection prompt will be displayed on the device, and if the user doesn't accept the prompt, the connection will never succeed.

If the connection has not resolved after some time, we want to show a notification in about:debugging "Connection still pending, there is probably a prompt you need to accept on the target browser". Do you have any suggestion on what would be a good way to show this information to the user?

Edit: adding a screenshot of the current patch below for reference.

Depends on D25032

Depends on D25033

Hi, since I have made a second prototype, I attach the screenshot.
If I get agree from all, I'll add a test and fix a test failure of browser/base/content/test/static/browser_misused_characters_in_strings.js which :flod told me at slack.
Thanks!

I really like this second prototype, the flow feels very nice! My only remaining concern is the size of the messages. Also knowing that when the close icon lands, we will lose 32px more on the right of the container :)

Let's try shorter texts:

• for the error message, what about simply "Connection failed" for now?
• for the warning message, it's difficult because we want to convey more information. "Connection still pending, check messages on your device" or "Connection still pending, please check your device"? Other suggestions welcome :)

In follow ups I think we could discuss about reducing the font size for those messages, maybe adding the technical error as a collapsible content in the error message.

I would really like to get Victoria's feedback on this one. Maybe we can land a first version without localization while we wait for UX input? Will review the code now.

"Connection still pending, check messages on your device" or "Connection still pending, please check your device"

Well the message might also be displayed on remote desktop browsers... so device is misleading "Connection still pending, check for messages on the target browser" ?

Daisuke I found an issue with the current patch queue, can you verify the following STRs:

Prerequisite

• add a network location that will not match anything (eg "blablabla:1234")
• connect a phone via USB, start a debuggable with connection prompt enabled

STRs:

• open about:debugging, enable USB debugging if needed
• in the sidebar, click on connect for blablabla:1234
• (it should show the error message)
• in the sidebar, click on connect for the USB runtime

ER: it should show the warning after 1s
AR: it shows the a connection failed error, same message as for the network runtime. If you click on connect again then you get a connection prompt on the device and the flow resumes as expected.

Maybe this is specific to my local setup, can you verify?

Depends on D25033

(In reply to Julian Descottes [:jdescottes] from comment #13)

Daisuke I found an issue with the current patch queue, can you verify the following STRs:

Prerequisite

• add a network location that will not match anything (eg "blablabla:1234")
• connect a phone via USB, start a debuggable with connection prompt enabled

STRs:

• open about:debugging, enable USB debugging if needed
• in the sidebar, click on connect for blablabla:1234
• (it should show the error message)
• in the sidebar, click on connect for the USB runtime

ER: it should show the warning after 1s
AR: it shows the a connection failed error, same message as for the network runtime. If you click on connect again then you get a connection prompt on the device and the flow resumes as expected.

Maybe this is specific to my local setup, can you verify?

Thanks Julian!
Yes, I could reproduce this issue. (However, it seems the caused error is happening even without my patch.)
I have investigated a bit, the error occurs from https://searchfox.org/mozilla-central/source/devtools/shared/security/socket.js#328
So, we might need to investigate more around https://searchfox.org/mozilla-central/source/devtools/shared/security/socket.js#253 ?
I want to fix this in followup since I am not sure how long time will it take to resolve.

Sure, can you file a follow up bug?

https://hg.mozilla.org/mozilla-central/rev/e72444917015
https://hg.mozilla.org/mozilla-central/rev/67c4c04fe27c
https://hg.mozilla.org/mozilla-central/rev/1fc7cb841027
https://hg.mozilla.org/mozilla-central/rev/3737b3b3acb0

The version of this I saw in Nightly, with shorter error message, looks great :). Small spacing tweaks recommendations in the other bug.