How does the Netscape team figure into this?  For instance, as part of 
the next big release, they're actively working on a quicksearch for the 
thread pane, and there is a little debate going on about whether the 
toolbar should get a grippy.  (bug 103734).  Would they have to take 
this to newsgroups before a decision is made?  If they went ahead with 
it, would mozilla the organization "reject" a new feature?  What is 
their stance?

Gervase Markham wrote:

> This is the third draft of a document which attempts to define how 
> changes can be made to the UI of Mozilla.
> 
> Feedback is welcome in netscape.public.mozilla.ui.
> 
> Gerv
> 
> <snip>
>       Mozilla UI Design and Implementation Process - Draft 3
>       ------------------------------------------------------
> 
> This document defines how significant changes to the Mozilla UI are 
> made. It is not supposed to cover trivial things, like fixing typos or 
> mis-sized dialogs.
> 
> This process is based on specs. A spec defines exactly what a piece of 
> UI should look like and act like. It provides a way of defining whether 
> UI is correct or not - "does it meet the spec?" In general, specs should 
> be for an individual dialog or window; in the larger cases such as the 
> main Navigator UI, they may be for a subset, such as "Navigator window 
> toolbars". The central repository of UI specs will be located at
> http://www.mozilla.org/projects/ui/specs
> If there is no spec for a piece of UI in the repository, the spec is "as 
> it appears and works at the moment."
> 
> This process is overseen by a person who in the following document will 
> be referred to as the Module Owner. If there is an appointed Mozilla UI 
> Module Owner for the relevant piece of UI, then this role is filled by 
> that person. If not, then the role is filled by the peers in that module 
> in cooperation. The job of this person is to achieve closure by pulling 
> together ideas from discussion to produce the spec that will be 
> implemented. They are expected to act according to the guidelines set 
> out in the Module Owners document, and to take input from the Module 
> Owner of the code which implements the UI. Module Owners are also 
> expected to coordinate with each other to attempt to ensure as much 
> cross-app consistency as possible in the product. There will certainly 
> be cases where Module Owners make changes with a less formal design 
> process, but significant changes should happen as follows.
> 
> Anyone, including the Module Owner, planning/proposing a significant 
> change should file a bug in the User Interface: Design Feedback 
> component noting the UI proposal (e.g., "Proposed [X] Implementation"), 
> and leave the Default Owner and QA Contact as is. The person proposing 
> the change should also post the proposed specs at the central repository 
> as soon as practical. The posted spec should be clearly marked as "Draft 
> - not to be implemented." They should then make an announcement with the 
> following headers:
> 
> Newsgroup:   netscape.public.mozilla.ui
> Newsgroup:   netscape.public.mozilla.<area>
> Newsgroup:   netscape.public.mozilla.ui.announce
> Followup-To: netscape.public.mozilla.ui
> Followup-To: netscape.public.mozilla.<area>
> 
> where <area> is e.g. browser or mail-news or editor. (The ui.announce 
> newsgroup will be created.)
> 
> Discussion will then ensue. After a period, the Module Owner will call a 
> halt, and the spec proposer will revise the spec based on feedback, and 
> the expressed wish of the Module Owner. Specifically, the Module Owner 
> may call a halt to the discussion if:
> - discussion has died out, or
> - no new points have been made in the past few days, or
> - a reasonable period of time has elapsed.
> The Module Owner needs to sign off on the spec for it to be considered 
> final. The Module Owner is requested to articulate the reasoning behind 
> their decision on particularly controversial issues.
> 
> The final spec produced must be immediately implementable - that is to 
> say, it should not contain UI for features which are not yet written, 
> unless either those features are due to be implemented at the same time 
> by the person implementing the UI, or there is a strong expectation that 
> they will be finished by the time the UI is implemented. If anyone 
> wishes to post an "ideal" UI spec which does not meet this restriction 
> to the repository, they may do so, but it should be clearly marked as 
> "Ideal - not to be implemented."
> 
> This spec is then considered canonical, and should be marked as 
> "Current", until further changes are proposed and it goes through this 
> process again. From this point on there will be no significant further 
> review of the content of the UI, and the Module Owner is free to ignore 
> further comment. This does not restrict the Module Owner's right to 
> change the specs, but Module Owners are requested not to change the 
> spec, particularly on issues which have proved controversial, without a 
> further discussion period.
> 
> This spec can then be implemented by anyone. Implementors are warned 
> that, if they begin before a spec is finalised, they may have to make 
> changes. The existence of patches should have no effect on the UI 
> decisions taken. The patch will go through code review and code 
> super-review in the normal way. There will be a requirement for UI 
> review (uir) from one of the peers in that module - this review is 
> limited to answering the question "Does it match the spec?" and so 
> should not be burdensome. uir= can be given by the reviewer or 
> super-reviewer if they are a Peer for the relevant UI module. If they 
> feel strongly that a change should be made which was not discussed in 
> the discussion period, they should consult the Module Owner.
> 
> There will be hard cases, where a group believes a suggested change is a 
> bad idea and should not be implemented by anyone, and others strongly 
> disagree. In such cases we expect open and honest communications without 
> personal attacks. This means actually listening and considering 
> alternative points of view. As in all things, [EMAIL PROTECTED] are the 
> ultimate arbitrators.
> 


Reply via email to