533361 - "TLS" should be called "STARTTLS"

Status: RESOLVED FIXED

(Webtools :: ISPDB Server, defect)

Description

[Ben Bucksch (:BenB)]

Socket Type has options "SSL", "TLS" and "none".
But TLS != STARTTLS.

It should be:
"No encryption" (XML: plain)
"SSL/TLS (special port)" (XML: SSL)
"STARTTLS (normal port)" (XML: STARTTLS)

Comment 1

[Ben Bucksch (:BenB)]

* No transfer encryption
No SSL or TLS
XML: "plain"
* SSL/TLS
Connect via SSL.
Needs special SSL port.
XML: "SSL"
* STARTTLS
Use TLS, using STARTTLS.
Uses normal port.
XML: "STARTTLS"

Comment 2

[Ben Bucksch (:BenB)]

While you're at it, you could also make a help text for the "port" field.
Incoming:
"Standard ports:
IMAP: 143 (if STARTTLS or No transfer encryption)
IMAP: 993 (if SSL/TLS)
POP3: 143 (if STARTTLS or No transfer encryption)
POP3: 995 (if SSL/TLS)"
Outgoing:
"Standard ports:
SMTP: 587 (if STARTTLS or No transfer encryption, and if available)
SMTP: 25  (if STARTTLS or No transfer encryption, and if 587 is not available)
SMTP: 465 (if SSL/TLS)"

Comment 3

[Ludovic Hirlimann [:Usul]]

Shouldn't we no push STARTTLS as it's something that was there before the standard TLS was implemented ?

Comment 4

[Ben Bucksch (:BenB)]

I didn't understand what you said, but STARTTLS (if enforced) is as good as SSL/TLS, there's not much of a difference from a high level perspective.

Comment 5

[Ludovic Hirlimann [:Usul]]

One is nicely defined in a standard not the other one.

Comment 6

[Ben Bucksch (:BenB)]

STARTTLS is defined in RFC 2595, RFC 2487 and others, see en.Wikipedia "STARTTLS"

Comment 7

[Blake Winton (:bwinton) (:☕️)]

Attachment: A patch to change the values allowed.

I'm not updating the convert.py script or the serializer, because the code in Thunderbird already expects the new values: http://mxr.mozilla.org/comm-central/source/mailnews/base/prefs/content/accountcreation/readFromXML.js#78 (and line 102).

I also didn't add the help text because I feel that these are terms that the person filling out the form should probably know the definitions of.

Thanks,
Blake.

Comment 8

[Ben Bucksch (:BenB)]

> I'm not updating ... the serializer

Yup, the XML values were/are fine as-is, just the UI was wrong.

> these are terms that the
> person filling out the form should probably know the definitions of.

There's a huge confusion of TLS vs. STARTTLS. I present as evidence: This very bug. The programmer here thought that TLS != SSL, TLS == STARTTLS, when TLS is just the new version of SSL, and STARTTLS a certain way to invoke TLS which allows to upgrade plain connections. The descriptions are there to explain exactly that.

Comment 9

[Blake Winton (:bwinton) (:☕️)]

Attachment: [checked-in] A patch to update autoconfig to change TLS to STARTTLS.

And here is some invalid data in autoconfig.momo.com which I found from this bug.

Thanks,
Blake.

Comment 10

[Ben Bucksch (:BenB)]

Comment on attachment 416633 [details] [diff] [review]
[checked-in] A patch to update autoconfig to change TLS to STARTTLS.

blake said I should r+ this.

Comment 11

[Ben Bucksch (:BenB)]

Comment on attachment 416633 [details] [diff] [review]
[checked-in] A patch to update autoconfig to change TLS to STARTTLS.

Commited, revision 57570

Comment 12

[Ben Bucksch (:BenB)]

Comment on attachment 416629 [details] [diff] [review]
A patch to change the values allowed.

Sorry for the delay!
Looks good to me. r=BenB

Comment 13

[Blake Winton (:bwinton) (:☕️)]

Comment on attachment 416629 [details] [diff] [review]
A patch to change the values allowed.

Hopefully this will be a quick review.  :)

Thanks,
Blake.

Comment 14

[David Ascher (:davida)]

Comment on attachment 416629 [details] [diff] [review]
A patch to change the values allowed.

r=davida, although I feel i should read the spec more.  but I suspect the original bug is mine and due to my lack of understanding of the details, so I have no reason to think that my code was better than this one, while I have good reason to think the inverse.

Comment 15

[Blake Winton (:bwinton) (:☕️)]

Checked in as http://viewvc.svn.mozilla.org/vc?view=revision&revision=58255