Some additional comments from me…

I would have liked a description of the test to be included in the report. Perhaps something along the line of:

Can an EAI address be created? (relevant components)
Can the application send mail to an EAI Address? (relevant component)
- This will include To;, Cc;, & Bcc
- This will also include a mixture of EAI & non-EAI addresses in the To, cc, & bcc
- This will also include evaluating replies.
Can the application receive mail from an EAI Address (relevant component)
- This will include ability to receive, and how the EAI addresses are displayed.
Are Unicode in passwords accepted an EAI Issue? Are the worth testing while we’re testing other components?
Is Linkification an EAI issue? Is it worth testing while we’re testing other components?
Are the open dots transformed to closed dots
By having these last three items in the body of the report we can easily get decisions from the community.

Now looking at the spreadsheet

MUA

Do we care about Unstructured Headers? Does it make a difference
Do we care about linkification in this exercise?
Do we care about passwords in Unicode in this exercise?
EA-MUA-032 – this is the alias/transformation issue. We want the message sent with an ASCII Alias, we do NOT want a transformation on the fly

I’m not sure that we want to define the failure to support an Open Dot as “failure”. But it should be a warning. This specific issue is under discussion.

Similarly, the whole linkification thing should probably be a warning, not a failure. I don’t think that’s core to the EAI issue. But a topic the UASG does have interest.

I don’t think that there’s anywhere that we want to test the Domain Name separate from an email address. Thus, I don’t think this tab is relevant (and not relevant in text unless we’re testing for linkification)

Do we need a ‘reference’ set of software that we know are EAI ready that we can test against?

When we test, do we need to test multiple MUAs against multiple MTAs, etc?

And, again, is there already existing a test suite of data that is commonly used? Or is this something that we are creating as well?

And, finally, the last part of this contract is to estimate the effort to exercise a pilot evaluation for a single environment. How big do we want this pilot evaluation to be? Just a single instance of each of the component parts or a number larger than 1 to allow some diversity in the testing?

Talk tomorrow (my time)

Don

From: Mark Svancarek (CELA) <marksv@microsoft.com>
Sent: Wednesday, 1 August 2018 6:01 AM
To: Don Hollander <don.hollander@icann.org>; ua-eai@icann.org; ua-discuss@icann.org
Subject: [Ext] RE: Additional Documents for EAI Call later this week

Report looks good. I would like to point out that Microsoft Outlook is another example where the product is available on multiple platforms (Windows, Mac, iOS and Android) and has significant code differences between platforms.

The test suite looks very comprehensive but will take longer to review for accuracy and completeness.

From: UA-discuss <ua-discuss-bounces@icann.org> On Behalf Of Don Hollander
Sent: Monday, July 30, 2018 03:38
To: ua-eai@icann.org; ua-discuss@icann.org
Subject: [UA-discuss] Additional Documents for EAI Call later this week

Please see the latest draft of Phase 1 of the EAI Evaluation Study.

This will be part of the discussion at our EAI Call later this week.

The first is a PDF of the report

The second is a spreadsheet of the test suite

Don

Don Hollander

Secretary General – UASG

Skype: Don_Hollander