Open Bug 444399 Opened 16 years ago Updated 2 years ago

Add Hotmail mailto protocol handler

Categories

(Firefox :: File Handling, defect)

3.0 Branch
defect

Tracking

()

People

(Reporter: mic, Unassigned)

References

(Blocks 1 open bug)

Details

Please add Microsoft Live Mail Mailto Protocol Handler to Firefox 3.0.x for en-US. Also please add to 3.0.x for a set of locales tbd (will be added as soon as they are confirmed). 

AFAIK Hotmail protocol has been tested and works

the url should be:
http://mail.live.com/?rru=compose%3Fto%3D%s<http://mail.live.com/?rru=compose%3Fto%3D%25s

It should be branded "Hotmail" in all locales in which it will ship
I have tested and I don't believe the URL here is valid. The mailto always redirects to the welcome screen of the hotmail account, not the compose screen. Can we get confirmation from MS that this is the correct string?
Can you try http://mail.live.com/?rru=compose&to=%s ? That's unescaping a few things.
Sorry, should have mentioned I tried both. Same results. Who performed initial testing, do we know? I may be doing something incorrectly.
Clint- do you know if we covered hotmail in any QA testing or your testing specifically?  Can we add this to your protocol handling tests?
Tim - just for a little extra info, I've been using Clint's tests for these (the mailto URLs, not the URI registration). I've spoken with Clint recently about it (Jun 30), and he had not done testing as we didn't have the new URI at that point.
just heard from our contact at Microsoft I had the wrong url :( here's the right one

http://mail.live.com/?rru=compose%3Fto%3D%s
That's the same URL :)
(In reply to comment #2)
> Can you try http://mail.live.com/?rru=compose&to=%s ? That's unescaping a few
> things.

%3f is also "?", so the ampersand makes more sense. I'll try that.
With Version 13.1 (Build 0124.0611) of Hotmail, I get the same results with
http://mail.live.com/?rru=compose%3Fto%3D%s (which is direct from MS, and loads as http://mail.live.com/?rru=compose?to=%s)

I get redirected to http://by126w.bay126.mail.live.com/mail/TodayLight.aspx?errorloading=true&n=200586 when I use the mailto URI.

Mic, can we get confirmation that the server/cluster I'm getting directed to has the update applied?





(In reply to comment #4)
> Clint- do you know if we covered hotmail in any QA testing or your testing
> specifically?  Can we add this to your protocol handling tests?
> 
I really hate to pile on here, but as Kev said, we did not do any testing with the official Microsoft Hotmail URI.  We did have some unofficial tests that used it back in the beginning of the 3.x project, but those were very unofficial and were more focused on testing our implementation of the protocol handling api (as per WHATWG spec) and less on the actual functionality of the mailto redirect.  

For the official version, we should be testing the functionality of the mailto URI spec using some common mailto style links and ensure that the organization providing that support (aka Microsoft in this example) has decent support.  

I have a list of mailto and webcal links for this type of testing available here: http://people.mozilla.org/~ctalbert/test-protocol-links.html

i've asked our contact at Microsoft to comment in this bug or send me more information by email (of course as soon as I have any updates by email, I will post here)
(In reply to comment #0)
> Please add Microsoft Live Mail Mailto Protocol Handler to Firefox 3.0.x for
> en-US. Also please add to 3.0.x for a set of locales tbd (will be added as soon
> as they are confirmed). 
> 
> AFAIK Hotmail protocol has been tested and works

OK, but do you know that Firefox user cannot create a new Live Mail account? ("We're temporarily unable to check ID availability. Please try again.")

It is not a temporary issue. In Opera Browser it works fine.

Tested on Mozilla/5.0 (X11; U; Linux i686; pl-PL; rv:1.9) Gecko/2008061015 Firefox/3.0

Are you sure that you want to add Live Mail?
(In reply to comment #12)
> OK, but do you know that Firefox user cannot create a new Live Mail account?
> ("We're temporarily unable to check ID availability. Please try again.")
> 
> It is not a temporary issue. In Opera Browser it works fine.

I also cannot creat new Japanese Hotmail accunt with Firefox.
To test mailto: protocol, I made an accunt with Safari.

I don't like supporting webmail servese who don't support creating new account with Firefox. But for users who already use Hotmail, better to have.
I'm not sure if we should or should not include webmail service who don't support Firefox.

# I think this decision (if support service who don't support Fx) should be
# global decision, not locale only one.
# Once we decided to support Hotmail, we Japanse also support it.
dynamis upon testing has also found that this does not currently support BCC parameter - I will ask Microsoft to comment
per comment #14, this is not a block ship IMHO but a nice to have. logged here to capture comments and feature upgrade from Microsoft
per #14, Bcc: should not be supported. From the RFC:

"Note that some headers are inherently unsafe to include in a message generated from a URL. For example, headers such as "From:", "Bcc:", and so on, should never be interpreted from a URL. In general, the fewer headers interpreted from the URL, the less likely it is that a sending agent will create an unsafe message."

kev
Blocks: 442892
(In reply to comment #16)
> per #14, Bcc: should not be supported. From the RFC:

I see, then Hotmail is fine and Gmail who support BCC should be fixed.
Thanks.
Yup. I'll also ask ctalbert to update test cases to include "bad" headers such as Bcc: and From: that should be ignored. Thanks for pointing that one out.
Added tests 9 and 10 for this: http://people.mozilla.org/~ctalbert/test-protocol-links.html.  Let me know if you need further changes.
we need a patch for this to land for upcoming release. 
Depends on: 451386
No longer blocks: 442892
Are we going to have this landed for Firefox 3.1 or Firefox 3.0.next?
.next releases are stability and security updates, and as such shouldn't introduce major changes to the user experience of the browser. Adding a new mailto: handler would be such a change in my opinion, so I think we should target 3.1 here (and beta3 more specifically).

That said, we still don't know if the URL we have is the right one and the handler has not been tested. I'll try to contact Microsoft and get a confirmation on the URL to help us move forward here.
Alright, I was able to test the handler using the following URL:
http://mail.live.com/?rru=compose%3Fto%3D%s

I made the results available in a tracking spreadsheet:
http://spreadsheets.google.com/pub?key=pdvm49_Er3oUaB-_uml7DCg

It seems that on 3.1 builds, the textarea with the message's body is not active, which makes it impossible to write anything (and, obviously, makes a couple of tests fail). I tested on two profiles. In 3.0.x the textarea works fine. 

The error console reports an error: c._oldWidth is undefined. Here's the code snippet featuring this property:

var c = a.CSSStyleDeclaration;
// [cut]
c._oldHeight = c.prototype.__lookupSetter__("height");
c.prototype.__defineSetter__("height",
function(a) {
  c._oldHeight.call(this, a);
  b()
});
c._oldWidth = c.prototype.__lookupSetter__("width");
c.prototype.__defineSetter__("width",
function(a) {
  c._oldWidth.call(this, a);
  b()
})

I can't really tell where the problem is, but right now, it makes it impossible for us to ship 3.1 with the handler. Clint, can you help me get to the bottom of this? If not, who else can?
Are you maybe hitting bug 462428?
That's possible. I tested a build with the patch from bug 462428 and I was able to view messages. While writing a new message, however, the body was still inactive.

Adding dependency, changing the summary.
Depends on: 462428
Summary: add Hotmail to 3.0.x mailto protocol handlers → Add Hotmail mailto protocol handler
On 3.0.x, the only failing test is test 6 ("Use a mailto and a to header - all addresses should be on To line").

The spec at http://www.faqs.org/rfcs/rfc2368.html says:

"Also note that it is legal to specify both "to" and an "hname" whose
value is "to". That is,
  mailto:addr1%2C%20addr2
is equivalent to
  mailto:?to=addr1%2C%20addr2
is equivalent to
  mailto:addr1?to=addr2
"

However, in the last case, Hotmail will only parse addr2, ignoring addr1.
(In reply to comment #26)
> On 3.0.x, the only failing test is test 6 ("Use a mailto and a to header - all
> addresses should be on To line").
> 
> The spec at http://www.faqs.org/rfcs/rfc2368.html says:
> 
> "Also note that it is legal to specify both "to" and an "hname" whose
> value is "to". That is,
>   mailto:addr1%2C%20addr2
> is equivalent to
>   mailto:?to=addr1%2C%20addr2
> 
Stas, I added these extra two tests to the existing repertoire under test 6 on the testing page.
Thanks Clint. 

Andy Chin from Microsoft told me that the next release of Hotmail including the fix for test case 6 should roll out by mid-February. I'll do the testing again once it's out. Hopefully we can take this in for the RC.
Blocks: 451386
No longer depends on: 451386
Is no one interested in doing this anymore?

1. The site name is now Outlook.com
2. The URL is now HTTPS: https://mail.live.com/?rru=compose&to=%s
3. The test page at comment 19 is gone.
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.