forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <rgard...@apache.org>
Subject Re: [Proposal] Development process and a stable trunk
Date Mon, 29 Aug 2005 10:11:13 GMT
Thorsten Scherler wrote:
> On Sun, 2005-08-28 at 14:19 +1000, David Crossley wrote:
> 
>>Thorsten Scherler wrote:
>>
>>>Before you read this reply, please read again my original reply. 
>>>
>>>Did you read it, ok then go ahead and please be not offended that your
>>>name may not be mentioned here or in the other thread but you actually
>>>contributed to views in any form. That is not my intention. I was
>>>focusing on code for views and the common danger of ignoring threads.
>>
>>First of all, you will need to try very hard to
>>be able to offend me. You were not.
>>
> 
> 
> :) Cheers for telling me this, that is a relief.
> 
> I am lucky that you are know me quite well because you help me from the
> beginning. 

That is true, but the point I was trying to make about offending people 
is that many people do not know us quite so well. We have to be careful 
not to offend newcomers. Did you see the recent thread on the infra@ 
list about netiquette? It asked if there is a problem here in Apache. 
The only conclusion I drew from it was that too many of us have formed a 
little "group" within the community who know each other well and so 
cutting remarks are taken in context of that relationship. These remarks 
often alienate newcomers who do not have the background of the older 
community members.

I only raised the issue to keep us aware of the potential problem. I 
think we all know the intention was not to discredit anyone. But even in 
your response you said something about nobody else has committed code. 
That is also not true, nor is it important since discussion is an 
equally valued part of the community. We need to be careful about 
statements that remove the recognition of the community from people who 
contribute.

There's no need for you to respond, I know you well enough to see it as 
an oversifght. I am raising it as a broader community issue, and I know 
you have the thick skin to cope with any percieved criticism - we all 
know that we write poor emails sometimes, this is a "community awareness 
broadcast". Interestingly, when I wrote the original mail it wasn't 
David or myself that I thought may be offended, but some other newer 
members of the community who have also contributed to forrest:views - 
seems my own mail was a problem in the community sense :-((

..

>>I agree whole-heartedly with your warning about ignoring
>>threads and at the same time i am saying that we need to
>>allow people to particpate in some things and not others.
>>
> 
> 
> Yes, I agree but we need to define core components and this core
> components should be understood and enhanceable from many active PMC
> member. I really do not want to see that we depend on individuals, we
> have do depend on the community.
> 
> That is as well why I think we should rename whiteboard to incubation.
> All components that need more community support should go here. If we
> want to follow Stephano's dreamlist we have to be very clear on the
> community part of components. 

I have no problem renaming the whiteboard. There is sense in your 
proposal. I would recomend starting a new thread saying you are going to
do it unless someone objects.

...

>>If other people helped more with applying patches,
>>then people like me would be relieved and could help
>>more with views development. There is one patch
>>sitting there from a new developer. Who is going to
>>commit it before i get compelled to jump in?
>>
> 
> 
> You are right. That is really a thing that I need working on. Anyway,
> like always said, I see views different and by getting into views I
> understood that this will change the general parts of the project. That
> is why I keep on asking for getting the views integration done.

Just do it (in a branch), I proposed removal of views from whiteboard 
immediately after the 0.7 release. You refused, wanting it to stay in 
whitebaord, you said it wasn't ready. I had the time then, but not now. 
I took that time to build a site using views. That site is now in 
production - in my opinion views is stable enough, it is only 
implementation details that will change.

Don't wait for me (speaking personally) since I am more keen to simplify 
our sitemaps using the locationmap, I think this will make forrest:views 
integration easier. However, I don't know when I will be able to do 
this, other commitments are in the way right now.

> Let me give you an example. The xhtml2 change will force us to rewrite
> the same pipes that we need to change for the views core integration!

So will the locationmap work :-(

> Another point is the integration of the locationmap. Right now it is set
> up but there have to be touched a lot of pipes to really use it, again
> that are nearly the same like for views. Knowing this made me ask
> everybody to get into views.

I'm sorry, it just doesn't owrk that way. Most of us are not here as a 
hobby or a play thing. Most of us use Forrest as part of our jobs. THat 
means we have to focus on the parts that are important to our job 
function. Get views into trunk (you have my +1 for a long time now) and 
it will *force* people to "get into views". With it in whitebaord it 
will only get people who have a specific need for views.

I don't have that use case yet, but I know there are a number of use 
cases on the horizon that would benefit from views in the future. Get it 
in core and I am more likely to move forwards.

In the meantime, any time I find to do none-essential stuff on Forrest 
will be refactoring to us the locationmap. That is *my* itch and I will 
scratch it now that Tim has made it possible for me to do so - if you 
start a branch to integrate views, I'll do the locationmap refactoring 
in their too, as you say this makes sense since they touch the same 
pipelines.

(agreeing naming conventions is the hold up right now, but that can be 
done in the branch before we merge)

>>I agree entirely. We were actually giving other cases
>>in support of that. It is a recognised fact that each
>>area has one or two main developers. It is important that
>>we all do broaden our focus to assist with other areas.

...

> We all have to dive into all core areas of forrest I totally agree. IMO
> if somebody start using views that will be pretty obvious to
> him/her. ;-)

Get views into core then ;-)

>>Then we need to finalise it and do the renaming actions
>>that were discussed. That is the backgound work that
>>needs to happen before the rest of the project can
>>really assist with views.
> 
> 
> Agreed, but just let us get over it and like Dave used to say "more
> code, less talk". 

The code is there!

> I am talking about views, their background and
> concepts since last year, it is all in the archives. I am a wee bit
> tiered to constantly repeating myself.

Defining the correct name is nothing to do with talk. It is about making 
future communications easier by removing as many possible 
misunderstandings as we can. That prevents the need to repeat oneself 
because it makes things easier to understand for those without the 
background you have.

>>The naming is extremely important, well chosen names
>>are very powerful. They can instantaneously convey
>>the whole concept.
>>
> 
> 
> Yes, agreed. I have chosen the names because I thought they will do
> that. I failed and need help on that. ;-)

[1] http://marc.theaimsgroup.com/?t=112276643700001
Re: Defining Views Terminology

>>>Please feel free to propose new naming convention for views and assume
>>>that lazy consensus is in operation from my part. Be sure if I see a
>>>problem or an easier way of doing things that I will speak up. 
>>
>>Please no, we need your input.
> 
> 
> Hmm, actually it is all in the archive. Anyway, I will help out where I
> can.

[1] http://marc.theaimsgroup.com/?t=112276643700001
Re: Defining Views Terminology

>>For me, one of the best things about [1] was that it
>>started to define all the separate pieces of the puzzle
>>and helped me to see more of what "views" are about.
>>However, it is still too clouded.
>>
> 
> 
> jeje
> 
> ok, I see your point. It is dead easy if you have started once. ;-)

Cool, David, thank you for reviving this thread. I took a long time to 
craft that mail to try and bring consensus, I thought it had got lost 
and I had wasted my time. Your persistence has, once again, brought it 
to the surface [privately wishing I had Davids organisational skills]

> Discussions (doco, xhmtl2,...) have to lead to roadmaps and code,
> otherwise they follow the motto "nice having talked about it". See
> Joachim, David et. al. they would welcome a to do list that they can
> follow. Having said this, it is hard to make such lists. ;-)

+1

This brings us right back to where this thread started. Managing the 
development process. See the threads linked in my very first reply. We 
need to decide how to use Jira to create this ToDo list and then we need 
to start actually using it.

Ross

Mime
View raw message