This is becoming confusing. Is necessary a clear vision to move this forward

[mamcx]

After read http://groups.google.com/group/three20/browse_thread/thread/8cf56f453c5fbc45,
I start to see a lot of mis-direction in the whole management of the
project.

This are, by far, all sites that pretend to be the source of engaging
of the community:

- http://groups.google.com/group/three20 (this).
- http://three20.info/ - The (i think) main website. Look like very
dead
- http://three20.com/ - Who knows
- http://twitter.com/three20 - Also dead?
- http://three20.stackexchange.com/ - No clear indication if will die
or what
- http://area51.stackexchange.com/proposals/12785/three20 - Another
stackoverflow Han Solo attempt, who knows.
- http://stackoverflow.com/questions/tagged/three20 - The original
stack.

This are *7* websites. The first is the only active enough.

I have managed 3 open source projects before, and is truly demanding
support any infrastructure beside the code itself.

IMHO, this is a complete waste of energy and polarization for a simple
developer library. I don't think more large libraries have so big
diverse sources.

Also, taking in account the current issues and the lack of new commits
I wonder if this is part of the problem.

So, my proposal is stick to max. 2 sites and get rid of everything
else. I don't care which one of the stack* version survive, but man,
only one!


Google groups is a chore, but at least is active. If the plan is move
to anything else,ok. But move. And close this. No get in the
indecision of what to do and which get.

Also, who is now the leader of this project? What happened to the
commits? Require help? Who knows? Need money? What we can do?

I don't care what options be choose. I only want a single vision. No
more diversification of energy.

[Prime31]

Totally agree. The state of Three20 is essentially nil. The last
major update (or even minor for that matter) was so long ago I don't
even remember. It's either time to create a new head somewhere with
some active maintainers or kill the project completely.

[Ulon]

Hi there,

I don't see anything wrong in people starting websites to talk/discuss
about a given technology, in this case the Three20 library. It's
perfectly healthy, in my opinion. Some of them will eventually die
while others will grow.

Btw, you forgot the main "hub" of Three20, where the code lies:
http://github.com/facebook/three20


On Oct 4, 10:02 pm, mamcx <ma...@elmalabarista.com> wrote:
> After readhttp://groups.google.com/group/three20/browse_thread/thread/8cf56f453...,

> I start to see a lot of mis-direction in the whole management of the
> project.
>
> This are, by far, all sites that pretend to be the source of engaging
> of the community:
>
> -http://groups.google.com/group/three20(this).

> -http://three20.info/- The (i think) main website. Look like very

> dead
> -http://three20.com/- Who knows
> -http://twitter.com/three20- Also dead?

> -http://three20.stackexchange.com/- No clear indication if will die
> or what
> -http://area51.stackexchange.com/proposals/12785/three20- Another

> stackoverflow Han Solo attempt, who knows.

> -http://stackoverflow.com/questions/tagged/three20- The original

[mamcx]

> I don't see anything wrong in people starting websites to talk/discuss
> about a given technology, in this case the Three20 library. It's
> perfectly healthy, in my opinion. Some of them will eventually die
> while others will grow.

Thats fine, but the "normal" way is use personals blogs, youtube,
twitter, etc (personal places) to talk about that stuff, and still
kept a central place.

Spin-offs regularly happen when:

1- The user base become very large. Think jquery/django/ruby
2- Localized info (others languages)
3- New, more or less successfully branches

This is not the thing I see. What I see is that each X time a new site
is created with the intention of replace this, then become dead.

And the "plan" is grab the users that participate here and move to
that site, and for people that look core developers *. Fine with me.
But then the "old" is kept. So, fragmentation.. and this is not
android!

A more centralized effort is more important, IMHO, until massive
adoption & maturity happen.

> Btw, you forgot the main "hub" of Three20, where the code lies:http://github.com/facebook/three20

Well, that is the exactly one that is *truly* necessary ;)


* That is my perception, I don't have a clue who is part or not of the
dev team...

[Drew McAuliffe]

I agree that there is a great deal of fragmentation, far out of line with the size of the project. I have been an advocate of a centralized source of information about the project and am concerned that there is still not a real "home" site for the project. There is a lot of stuff in this project that needs documentation and right now it's scattered across way too many different sites. It might be healthy that there's a good deal of discussion about the project, but it's not healthy that there's no central "home". And I also worry that there is little movement on the commit side. I don't think I know who's in charge anymore, and that might be my fault, but it looks like I'm not alone. I find myself less and less inclined to use the library due to the scattered nature of the project's management; why tie myself to it for certain things if I can't be sure that it will be managed over the long haul? The recent issues around 3.2/4.0 builds were a perfect example of why my faith has been somewhat tested.
Drew

> --
> You received this message because you are subscribed to the Google Groups "Three20" group.
> To post to this group, send email to thr...@googlegroups.com.
> To unsubscribe from this group, send email to three20+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/three20?hl=en.
>

[elpuerco63]

This is a good discussion bringing home the reason why I basically shy away from such frameworks.

Unless it is maintained and documented it becomes to confusing and risky to use.

I agree there should be a single official central home with true commit otherwise it will just die.

I have no intention of using what has the potential to be a seriously cool framework under the current setup as down the road it will cause me headaches!

ep

[Michael Grinich]

Just a quick update:  Jeff and I are working together to get much of this sorted out. We're working to get a proper site up-and-running, but we need your help.

Like others mentioned, there's no central source for information right now. A lot of tutorials and guides exist on the web, but they're not organized. That's where you come in. 

I've created a repo on GitHub to act as a sketchbook for articles, tutorials, and documentation. http://github.com/grinich/three20-articles

Please fork it and help add content. Even if you just want to write a couple paragraphs, combined together it will be a huge step forward. Blog posts, "cookbook recipes", getting started guides for specific pieces: all of these would be great. This content will then be used when we launch the new hub of Three20. If you want this framework to survive, then I encourage you to contribute! 

Also, http://three20.stackexchange.com/ and http://scope.three20.info/ are on the way out. If there is any knowledge there which you would like to preserve, please write an article about it!  Use the main StackOverflow site with a Three20 tag for Q/A from now on. http://stackoverflow.com/questions/tagged/three20

Feel free to email me directly with questions.

Michael

[mamcx]

> Just a quick update:  Jeff and I are working together to get much of this
> sorted out. We're working to get a proper site up-and-running, but we need
> your help.

Glad to know!


> Also,http://three20.stackexchange.com/andhttp://scope.three20.info/are

> on the way out. If there is any knowledge there which you would like to
> preserve, please write an article about it!  Use the main StackOverflow site
> with a Three20 tag for Q/A from now on.http://stackoverflow.com/questions/tagged/three20

Thats fine. Will be good put a deprecated warning to both sites?

[Jeff Verkoeyen]

I'll put the deprecation warning up there within this week.

Cheers,
- Jeff

--

[Robert Mathews]

Will you be grabbing the google group pages, especially the recipes/cookbook?

Otherwise I see the notice from Google that they are going away.

And, thanks for bringing this together.

[Laurent C]

Good to see that jeff is still around. I was wondering about the
future of this framework.

> > three20+u...@googlegroups.com<three20%2Bunsu...@googlegroups.com >
> > .

[William Lachance]

2010/10/5 Michael Grinich <mgri...@gmail.com>:

> Just a quick update:  Jeff and I are working together to get much of this
> sorted out. We're working to get a proper site up-and-running, but we need
> your help.

What about the library itself? The last commit to the facebook
repository was August 20th (and that was only a documentation update).
Meanwhile, there's a million seperate forks out there with useful bug
fixes and changes (mine is at http://github.com/wlach/three20 in the
redirectresponses branch).

From my perspective, it seems like facebook isn't really providing the
resources to properly maintain the library and move it forward. While
I'm totally grateful for Joe/Jeff's work (and fb's sponsorship
thereof), is it perhaps time to get another
person/company/organization to step into that role? I currently lack
the resources to take the lead on this, but would be willing to help
out if someone did want to do this (I use three20 very regularly in
the course of my work).

Finally, not to complain too much, but I'm totally in agreement that
the constant churn of "official" three20 groups/websites has been very
confusing. Given this, what's the justification for yet another
website? What was wrong with three20.info, which has its source on
github?
--
William Lachance
wrl...@gmail.com

[Jeff Verkoeyen]

To be clear, we're not launching another website. We are shutting down the other community sites and focusing efforts entirely on three20.info and the Google Group.

As for the state of the repository, I will be dedicating a concrete amount of time to Three20 each week going forward. I'm quite aware of the number of pull requests and forks at the moment, and I'm hoping that we can start engaging in solid discussions about handling this load going forward.

My time over the last four months has been spent on "vacation" effectively. The library and its community has suffered dearly as a result of this lack of any form of leadership. I began making efforts along the way to try to minimize this by attempting to set up a panel of contributors to the site. Again, due to my effective sign-off from the net, this organization didn't get created as quickly as I'd have liked.

If you'd like a sneak peak on what Michael and I are working on for the new Three20 site design, you can have a look here:

http://new.three20.info/

The goal is to remove all of the cruft from the current Three20.info design and emphasize documentation. The site is also now much, much easier to edit by outside contributors. In fact, you can actually edit articles directly on GitHub as though it was a Wiki.

For example, note the link in the top right:

http://new.three20.info/article/2010-10-06-URL-Based-Navigation

Cheers,
- Jeff

--

[Jeff Verkoeyen]

I rather enjoy the way this article has transitioned to the new site.

http://new.three20.info/article/2010-10-07-Debugging-With-Three20

The new Navi boxes should hopefully catch attention (scroll about halfway down the page to see one).

Cheers,
- Jeff

[David Morford]

Would like to chime in on this discussion with some suggestions. 

- Build issues have affected a lot of folks using Three20. This was largely a result of a change in Xcode 3.2.4 with iOS 4.0 SDK that seems to ignore BASESDK values when set in xcconfig files. I have a radar issue open with Apple on this (rdar://8192536 , http://openradar.appspot.com/8192536). The reason for this change is unclear. It is likely related to either; 1.) Getting developers used to building with the latest SDK or 2.) A temporary change until the iPhone and iPad SDKs are "unified". The later seems more likely considering a bullet point in the last 4.2 beta. Since it is still under NDA, I will leave it at and encourage everyone to read the release notes for the 4.2 SDK Beta 2.

- Jeff based some the build revisions on an early version of a project I have on Github, BuildKit. I think it might be helpful to re-examine the build setup. Beyond the need to force the BASESDK value to be set in each project with each new 4.x SDK update, I have never experienced the kinds of issues I've seen on the mailing list. I  updated the UXKit fork last month to build on the 4.0 SDK and only had to change two xcconfig files and touch the Xcode projects to get them to update the BASESDK value. I've used the configs, scripts and structure with every iOS SDK and they are based on a similar Mac OS X app setup I've used since 2005 with nearly every version of Xcode. My private repository that I use for building my own apps and apps for clients current has over 30 static libraries used by over a a half a dozen apps for clients currently in the AppStore dozens of smaller catalog and example apps. The concepts behind it are stable, reliable and proven.Using it as designed might also enable *much* easier sharing of drop in libraries and extensions. Read up here – http://github.com/davidmorford/BuildKit

- The single most useful aspect of Three20 to me has always been the Navigator. Last month I opened a private rewrite of that class cluster on Github that I started when the iPad SDK beta first shipped. I evolved it over six months time until I felt it was stable enough to use in my own app development. It reduces the included classes to the minimum needed to use Navigator and URLs as a mechanism for architecting iOS applications. It includes the full set of usage patterns for the iPad UI idiom and supports UISplitViewController and MGSplitViewController by Matt Gemmell. The project is aggressively forward facing. It uses LLVM/Clang 1.5 and the modern runtime ABI switch and the implications of that are visible in ivar and synthesize and the overall class design. NavigatorKit is very stable but  is targeting  the 4.2 SDK and forward. I am working on having it ready for that release. The project is setup with BuildKit, includes an example application demonstrating usage patterns of the Navigator in both UI idioms, includes a ready to go Xcode project template and has preliminary (but still not fully tested) support for the 4.2 SDK. More here – http://github.com/davidmorford/NavigatorKit

When thinking about the future features and overall structure and design of Three20, please use and borrow from both of the projects. Making Three20 better will increase the reliability and usefulness of the libraries for developers and help them focus on creating high quality iOS apps for users.

Cheers,

David