Bounty for merging NVU

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

Bounty for merging NVU

jonsmirl@gmail.com
I'd like to propose a discussion item for the next Moz Corp staff
meeting. How about using some of the new found Google revenue to fund
a significant bounty for merging NVU back into the trunk? I believe
having a core editor component as part of Gecko is a key long term
strategy.  This is evidenced by today's Google acquisition of
Writely.,

Boris and I have been exchanging emails on this. NVU hasn't been
merged for two years now. Doing the merge is a lot of work. This is
not a fun and interesting project, it is a whole lot of grunt work and
it is highly unlikely that any one will volunteer to do it. If NVU
continues without being merged it will ultimately diverge so far that
the merge will become impossible. Given the magnitude of the project a
bounty of around $25K seems reasonable.

I'm not interested in pursuing the bounty myself, I am interested in
using an editor component as part of an array of services offer by the
xulrunner platform.

--
Jon Smirl
[hidden email]
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Darin Fisher-2
[I don't speak for staff at mozilla]

One challenge with something like this IMO is getting someone (or some
people) to perform a reliable review on any proposed patch that
implements such a large merge.  Speaking from experience, it is really
hard to get good reviews and to give good reviews on large changes.

-Darin



On 3/10/06, Jon Smirl <[hidden email]> wrote:

> I'd like to propose a discussion item for the next Moz Corp staff
> meeting. How about using some of the new found Google revenue to fund
> a significant bounty for merging NVU back into the trunk? I believe
> having a core editor component as part of Gecko is a key long term
> strategy.  This is evidenced by today's Google acquisition of
> Writely.,
>
> Boris and I have been exchanging emails on this. NVU hasn't been
> merged for two years now. Doing the merge is a lot of work. This is
> not a fun and interesting project, it is a whole lot of grunt work and
> it is highly unlikely that any one will volunteer to do it. If NVU
> continues without being merged it will ultimately diverge so far that
> the merge will become impossible. Given the magnitude of the project a
> bounty of around $25K seems reasonable.
>
> I'm not interested in pursuing the bounty myself, I am interested in
> using an editor component as part of an array of services offer by the
> xulrunner platform.
>
> --
> Jon Smirl
> [hidden email]
> _______________________________________________
> dev-planning mailing list
> [hidden email]
> https://lists.mozilla.org/listinfo/dev-planning
>
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Daniel Glazman-2
In reply to this post by jonsmirl@gmail.com
Darin Fisher wrote:
> [I don't speak for staff at mozilla]
>
> One challenge with something like this IMO is getting someone (or some
> people) to perform a reliable review on any proposed patch that
> implements such a large merge.  Speaking from experience, it is really
> hard to get good reviews and to give good reviews on large changes.

Fortunately, the big changes can be divided in smaller ones but still,
you have a point. The serializer changes for instance are big.

Anyway, where would the $25k come from ? Because with such a funding, I
could hire a mozillian at DI and work part-time with him/her on that.

My only problem here is time, not complexity.

</Daniel>
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Boris Zbarsky
Daniel Glazman wrote:
> Fortunately, the big changes can be divided in smaller ones

That's already a good start!

> Anyway, where would the $25k come from ? Because with such a funding, I
> could hire a mozillian at DI and work part-time with him/her on that.

Jon's suggestion is that the Mozilla Corporation (or Foundation?) consider
funding this.  That's where the "having a core editor component as part of Gecko
is a key long term strategy" part comes in.   ;)

-Boris
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Akkana Peck-3
In reply to this post by Daniel Glazman-2
Daniel Glazman writes:
> Darin Fisher wrote:
> >[I don't speak for staff at mozilla]
> >
> >One challenge with something like this IMO is getting someone (or some
> >people) to perform a reliable review on any proposed patch that
> >implements such a large merge.  Speaking from experience, it is really
> >hard to get good reviews and to give good reviews on large changes.

Reviewing big changes is definitely a big job -- and it's a lot less
fun than writing the changes in the first place. A bounty would
help as much with getting reviewers as with getting the work
done in the first place.

But I suspect there are a lot of people who would like to see this
merge happen, and would be very interested in a project to merge nvu
back into the main tree, especially if it was supported by both sides
(Daniel and MoFo). It might not be that hard to find people to do
the work.

Boris wrotes:
> That's where the "having a core editor component as part of
> Gecko is a key long term strategy" part comes in.   ;)

Is that a key long term Mozilla strategy? If so, it's great to hear!

        ...Akkana
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Boris Zbarsky
In reply to this post by Daniel Glazman-2
Akkana wrote:
> Is that a key long term Mozilla strategy? If so, it's great to hear!

I think given the amount of contentEditable-type stuff out there it should be...

-Boris
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Daniel Glazman-2
In reply to this post by Boris Zbarsky
Boris Zbarsky wrote:

> Jon's suggestion is that the Mozilla Corporation (or Foundation?)
> consider funding this.  That's where the "having a core editor component
> as part of Gecko is a key long term strategy" part comes in.   ;)

Boris, just consider I am thinking out loud here : and if we were to
implement a new html/xhtml serializer and the CSS Editing OM extensions
into an independant component ? Does it make sense ? Is it even
technically feasible ?
Because that could drastically simplify the review process, right ?

</Daniel>
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Boris Zbarsky
Daniel Glazman wrote:
> Boris, just consider I am thinking out loud here : and if we were to
> implement a new html/xhtml serializer and the CSS Editing OM extensions
> into an independant component ?

The serializer may be tough -- we probably have code creating those by CID or
even via direct new calls.  I guess we could switch all callers to contractID,
though.  But we still want the serializer fixes back in trunk, no?

I'd have to know more about the CSS Editing OM changes to comment on them
intelligently...

-Boris
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Brendan Eich
In reply to this post by Daniel Glazman-2
Daniel Glazman wrote:

> Boris Zbarsky wrote:
>
>> Jon's suggestion is that the Mozilla Corporation (or Foundation?)
>> consider funding this.  That's where the "having a core editor
>> component as part of Gecko is a key long term strategy" part comes
>> in.   ;)
>
>
> Boris, just consider I am thinking out loud here : and if we were to
> implement a new html/xhtml serializer and the CSS Editing OM extensions
> into an independant component ? Does it make sense ? Is it even
> technically feasible ?
> Because that could drastically simplify the review process, right ?
>

Boris didn't address the last question, but I am not sure how review
workload is reduced by changing where the big change lands.  We still
want review of new code.  We've had loopholes here before, but I say
never again.

/be
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: Bounty for merging NVU

Boris Zbarsky
Brendan Eich wrote:
> Daniel Glazman wrote:
>> Boris, just consider I am thinking out loud here : and if we were to
>> implement a new html/xhtml serializer and the CSS Editing OM extensions
>> into an independant component ?
...
>> Because that could drastically simplify the review process, right ?
>
> Boris didn't address the last question, but I am not sure how review
> workload is reduced by changing where the big change lands.

I assumed Daniel meant "an independent component shipped by NVU, so it doesn't
need to land in the mozilla.org tree"...

Otherwise, I agree that I don't see a difference in review burden.

That said, I think we want a better serializer in the mozilla.org tree, so I'd
really we actually merged the NVU changes in.

-Boris
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning