32915 - need to insert "prv:" before milestone # in user agent string

Status: VERIFIED INVALID

(Core :: Networking, defect, P3)

Description

[ekrock's old account (dead)]

Per the recent consensus on netscape.public.mozilla.netlib, the Mozilla user 
agent specification will be updated to recommend inserting the string literal 
"prv:" (standing for Pre Release Version) immediately before the VersionModifier 
string (with no intervening space). 

e.g. We must change from this:
Mozilla/5.0 (Windows; N; Windows NT 5.0; en-US; m14) Netscape6/6.0b1

... to this:
Mozilla/5.0 (Windows; N; Windows NT 5.0; en-US; prv:m14) Netscape6/6.0b1

The purpose of this change is to make the VersionModifier substring 
algorithmically findable within the user agent string.

Jud, could you please do this before beta2 at the latest? Thanks!

Comment 1

[Warren Harris]

Moving to M17 (also beta2).

Comment 2

[leger]

ekrock, can you give more detail to PDT on WHY this is a beta stopper?  Thanks!

Comment 3

[ekrock's old account (dead)]

We can ship beta2 without this *if* PDT agrees to accept the implementing fix 
thereafter. Getting this feature in prior to FCS is CRITICAL to enabling 
interoperability of web content and web apps across all browsers/set top 
boxes/devices that embed Gecko, which is strategically crucial.

But I heard verbally that somebody (valeski?) has the fix for this ready anyway. 
If that's true, could the person with the fix just check it in prior to 5/16? 
Thanks!

Comment 4

[leger]

Reassigning to leaf per chofman.  Putting on [nsbeta2-] radar. Not critical to 
beta2. 

Comment 5

[leger]

Update, per PDT - marking INVALID 

ekrock - if we have user agent string request that we REALLY think that we need, 
assign those bugs to leaf.

Comment 6

[ekrock's old account (dead)]

REOPENing. INVALID is not the correct resolution for this report. 

(1) This change is intended to bring the Netscape 6 client into compliance with 
the mozilla.org user agent specification. It would be most unfortunate if the 
Netscape 6 client were not fully compliant with the UA specification we're 
trying to establish for all Mozilla-based clients.

(2) It's such a trivially easy change to make that we're at risk of spending 
more time debating whether we should make the change than it would take to make 
the change and be done with it. (Remember Jim Barksdale's rules for snakes? ;-> 
)

cc:ing Mitchell. Mitchell, any progress on getting the Mozilla UA spec doc 
updated? It's hard for me to make a strong case for complying with a document 
that isn't up to date yet. ;->

Reopening bug, marking FUTURE for now due to PDT's lack of interest in assigning 
Netscape resources to make this change, reassigning to leaf.

Comment 7

[ekrock's old account (dead)]

David, did you by any chance (a) fix this bug to bring our UA string into 
compliance with the spec, or (b) change the UA draft spec in a way that makes 
this bug INVALID now?

Comment 8

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

The spec no longer has this because we don't do it.  So it's invalid according
to the current spec:

http://www.mozilla.org/build/revised-user-agent-strings.html

Comment 9

[ekrock's old account (dead)]

INVALID. If we don't do this in the first release, might as well not bother. The 
Gecko datestamp is the most valuable piece of information and tells you what you 
really need to know anyway. Thanks David!

Comment 10

[Keyser Sose]

No longer valid according to Mozilla specs. Verifiying.