Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-30 Thread William Ross
On 30 Jun 2010, at 02:04, Jim Gay wrote:

 On Tue, Jun 29, 2010 at 8:50 PM, Jeff Casimir j...@casimircreative.com 
 wrote:
 I'd like to see radiant follow some of the footsteps from merb and
 have two gems: radiant-core and radiant-more.  The 'radiant' gem would
 snip
 
 To me, radiant-more would be the part that adds snippets, layouts,
 users, image/file management, stylesheet/javascript management, and
 perhaps other things.
 
 Other people might not see it that way, but by far one of the most
 popular questions I get about Radiant has to do with adding it to an
 rails app, rather than shoe-horning a rails app into it. The biggest
 sticking point is the User system.

I completely agree here, except that I think a page-management engine has to be 
able to work with layouts, and I'd like to see versioning in core. Then a -more 
gem could add:

- users
- assets
- snippets and js/css
- sites
- configuration
- dashboard

If it did that in a modular way that allowed areas of functionality to be 
omitted or supplanted:

radiant.handle :images, :with = DragonflyExtension
radiant.handle :authorization, :with = RbacExtension
radiant.handle :dashboard, :with = MyLocalExtension
radiant.unhandle :configuration

then radiant could realistically offer to drop in anywhere and manage pages for 
any app, which would be a splendid thing indeed.

will




Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-30 Thread John W. Long
On Tue, Jun 29, 2010 at 8:50 PM, Jeff Casimir j...@casimircreative.com wrote:
 I'd like to see radiant follow some of the footsteps from merb and
 have two gems: radiant-core and radiant-more.

So one other idea I had on about this is to add the ability to the
extension registry for users to create lists of extensions. If you
couple this with the radiant command you could then do something like
this on install:

radiant -d sqlite3 --extensions jlong:fancypants path/to/radiant/project

Which would install all of the gems for extensions from my
fancypants list in addition to creating and configuring my Radiant
project.

Then add a radiant user in the extension registry and radiant-more becomes:

radiant -d sqlite3 -e radiant:more path/to/radiant/project

--
John Long
http://wiseheartdesign.com
http://recursivecreative.com


Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-29 Thread Jim Gay
I'm glad you brought this up for discussion.

On Tue, Jun 29, 2010 at 11:59 AM, Edmund Haselwanter
edm...@haselwanter.com wrote:
 A question for clarification:

 I like to have options, but sometimes options are a bit confusing.

 On 29 Jun., 06:56, john muhl johnm...@gmail.com wrote:
 Jim (saturnflyer) and i have been working (most of the important work
 is Jim's) on a branch of page_attachments that adds some new interface
 features to the page_attachments extension

 so here is the question on options:

 wouldn't it be wise to concentrate on just one (nifty) asset
 extension?
 I see an asset extension as a vital part of a CMS (= see, it says
 content ;-)

Yes.
Check out the prototype http://github.com/radiant/radiant-prototype

And there are tickets like
http://github.com/radiant/radiant/issues#issue/28 (sns) and
http://github.com/radiant/radiant/issues#issue/92 (paperclipped)


 or the other way round: what would you suggest to use in the future?

 - page_attachements
 - paperclipped (and there are some extensions which extend
 paperclipped)
 - gallery (ok. its just for images ..)
 - MediaMaid

The greatness behind paperclipped is paperclip (not to take away from
all the work that Keith and many others put in to make a really useful
extension). I like page_attachments too and recently did a lot of work
to change it on my fork with John Muhl's help, but for a recent
project I was again looking for an asset manager and even though I did
all that work to update page_attachments (and it still needs more) I
chose paperclipped because of paperclip.
I've not used gallery and have only perused the code of MediaMaid but
plan to look at them more in the future.

I wish there was an easy answer, but keeping things simple is hard
work. One of the tickets I mentioned above suggests pulling in
paperclipped as it is just because so many use it and it would do the
job. But Radiant has gained popularity due to it's simplicity, so
just because isn't good enough.


 and with this comes another question:

 Ok, I know that Radiant is different. And I love it. I recently did
 a
 presentation in my web 2.0 community meeting about radiant. And one
 statement
 burned so hard: No end user will use this!

Bullshit. It's just not true. I've had plenty of non-technical,
non-tech-savvy users enjoying Radiant.
But I will say that it is very easy to look at Radiant, and probably
look at the presentation you put together, and claim that no end user
would use it, but that's mere conjecture.
Take what people say with a grain of salt if they haven't actually
used the software.

That said, there is always room for improvement and there are plenty
of people working on improving it.


 Why do I come up with this? And why now? Because I followed the
 discussion about radiant
 version numbers.

For others, this is a reference to a discussion in the comments an
unrelated commit on github
http://github.com/radiant/radiant/commit/488c2a49c40ced133d3c4abebb66e55fc2df9e6c

Hilariously, it was suggested in the thread to move to the mailing
list for discussion, but nobody did it except of course someone who
wasn't in on the thread.
Thank you Edmund!


 And here goes my suggestion:

 Radiant 1.0 should:

 - be Rails3 based
 - have some core asset handling features (doesn't matter if this is an
 extension)
 - have one WYSIWYG Editor integrated with this assets and other
 ressources (e.g. allow for search pages and insert a link)

see this http://github.com/radiant/radiant/issues/#issue/26 and this
http://github.com/radiant/radiant/issues/#issue/33

 - should define a way on how to do i18n content (not admin area)
 - should have some eyecandy like a HUD were you can search a page
 based e.g. on the title (a nvaigation alternative, if you like)
 - should define a blogging API so that it can be used with some
 standard blogging tools.

I'm very interested in how Rails will lead the way with REST and
http://github.com/caelum/restfulie
There is a Ruby Summer of Code project to fix up ActiveResource to do
REST properly with restfulie.

1.0 is whatever we say it is.
And to quote DHH, I've long been impressed and puzzled by the power
of big version numbers. To open source projects like Ruby on Rails,
it's such a divorced measure of quality of features that I feel we
need to take it's importance down a few notches.

http://www.loudthinking.com/posts/20-dont-overestimate-the-power-of-versions

So 1.0 can come whenever we want. If everyone wants all those things
to be in 1.0, fine, that'll be 1.0 but 1.0 will only arrive when those
things do. That's why I'm in favor of moving to 1.0 with Rails 3 and
calling it a day. 1.0 in particular is so overblown that by the time
Radiant reaches 1.0 there will be plenty of complaints about how it
doesn't have this or that feature which could, of course, easily come
at 1.1, 2.0 or version 999.

Radiant is over 4 years old now and is used in hundreds of production sites.


 And again: I love Radiant. I use it very much. I do 

Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-29 Thread Jim Gay
On Tue, Jun 29, 2010 at 2:31 PM, Marshal Linfoot mlinf...@gmail.com wrote:
 On Tue, Jun 29, 2010 at 1:47 PM, Jim Gay j...@saturnflyer.com wrote:

 I'm glad you brought this up for discussion.
 ...
 Check out the prototype http://github.com/radiant/radiant-prototype

 Very nice! Can't wait to see this in a release soon.
 I also want to say thank you to the developers/contributors to
 the Radiant project for all their excellent work. Radiant has simplified the
 management of my (modest) website and it has been a pleasure to use.
 I'm not a programmer but would like to contribute in some way. What kind of
 help do you need?

Find bugs, request features argue about why things should be added,
changed or removed.

Developers are always good, but the application can really grow when
there is a determined direction. Good arguments based on real-world
usage lead to good features. If you have an opinion, please express it
here and on the github issues, just do so with thought and
well-reasoned arguments. We definitely need more of that.

 --
 marshal




-- 
Jim Gay
Saturn Flyer LLC
http://www.saturnflyer.com
571-403-0338


Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-29 Thread Jim Gay
On Tue, Jun 29, 2010 at 8:50 PM, Jeff Casimir j...@casimircreative.com wrote:
 I'd like to see radiant follow some of the footsteps from merb and
 have two gems: radiant-core and radiant-more.  The 'radiant' gem would
 become alias to radiant-more which would be core plus a few of the
 most common extensions (SnS, some kind of attachments, chronicle?,
 etc).  Now that extensions can be gemified, I think this shouldn't be
 that hard.

 And if someone really wants a stripped down version they can just
 install radiant-core, barebones but fully functional.

Yeah, definitely. This has been talked about and requested by many.

And recently there's been a lot of activity in creating extensions
that manage Layouts from the file system. So I've been thinking about
what I see as the true core of Radiant and that is: Pages.

Snippet and layout mechanisms can be swapped out according to whatever
you like. And you might even have the need for a different type of
User system.

So to me, radiant-core would be the ability to manage pages, and the
extendable UI. Loading of extensions I see as something that ought to
be generic and is going to move into Rails Engines. Stripping it down
to those features means it would be easier to integrate them into
existing rails projects rather than having to turn a rails project
into an extension.

I'm sure there are things I haven't considered, but that's why I've
been working out the ideas on separating the features in
http://github.com/radiant/radiant/issues#issue/23 : so I have real
experience on which I can base the decision-making.

To me, radiant-more would be the part that adds snippets, layouts,
users, image/file management, stylesheet/javascript management, and
perhaps other things.

Other people might not see it that way, but by far one of the most
popular questions I get about Radiant has to do with adding it to an
rails app, rather than shoe-horning a rails app into it. The biggest
sticking point is the User system.

But you're right, with extensions as gems, this is much easier and no
matter what decision is made about what goes into radiant-more, it
will be very easy to make radiant-less, radiant-awesome,
radiant-whatever and write your own custom gem.


 - Jeff

 On Tue, Jun 29, 2010 at 8:24 PM, Jim Gay j...@saturnflyer.com wrote:
 On Tue, Jun 29, 2010 at 2:31 PM, Marshal Linfoot mlinf...@gmail.com wrote:
 On Tue, Jun 29, 2010 at 1:47 PM, Jim Gay j...@saturnflyer.com wrote:

 I'm glad you brought this up for discussion.
 ...
 Check out the prototype http://github.com/radiant/radiant-prototype

 Very nice! Can't wait to see this in a release soon.
 I also want to say thank you to the developers/contributors to
 the Radiant project for all their excellent work. Radiant has simplified the
 management of my (modest) website and it has been a pleasure to use.
 I'm not a programmer but would like to contribute in some way. What kind of
 help do you need?

 Find bugs, request features argue about why things should be added,
 changed or removed.

 Developers are always good, but the application can really grow when
 there is a determined direction. Good arguments based on real-world
 usage lead to good features. If you have an opinion, please express it
 here and on the github issues, just do so with thought and
 well-reasoned arguments. We definitely need more of that.

 --
 marshal




 --
 Jim Gay
 Saturn Flyer LLC
 http://www.saturnflyer.com
 571-403-0338





-- 
Jim Gay
Saturn Flyer LLC
http://www.saturnflyer.com
571-403-0338


Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-29 Thread john muhl
On Tue, Jun 29, 2010 at 7:51 PM, Jim Gay j...@saturnflyer.com wrote:
 On Tue, Jun 29, 2010 at 2:24 PM, Haselwanter Edmund
 edm...@haselwanter.com wrote:

 On 29.06.2010, at 19:47, Jim Gay wrote:

 I'm glad you brought this up for discussion.

 It just felt right after following the github discussion :-)


 On Tue, Jun 29, 2010 at 11:59 AM, Edmund Haselwanter
 edm...@haselwanter.com wrote:
 A question for clarification:

 [ .. snippet .. ]

 wouldn't it be wise to concentrate on just one (nifty) asset
 extension?
 I see an asset extension as a vital part of a CMS (= see, it says
 content ;-)

 Yes.
 Check out the prototype http://github.com/radiant/radiant-prototype

 And there are tickets like
 http://github.com/radiant/radiant/issues#issue/28 (sns) and

 quote jlong

 jlong February 18, 2010 | link
 There's a little more work to do on this in the prototype, mainly the upload
 interface. I believe Chris Parish has had ambitions for simplifying his 
 styles 'n
 scripts extension to serve this purpose. If it is going to make it into 0.9 
 it
 should be implemented as an extension that ships with the core.

 /quote jlong

 That's exactly how I see it too :-)

 SNS needs some more thought (which I happen to be doing
 http://www.saturnflyer.com/blog/jim/2010/06/29/ruby-metaprogramming-is-awesome/
 ) but I see that as a necessity for managing a site (along with
 image/file management of course).

having your css and js in the admin interface is a necessity? i tried
to love sns but never could find any benefit to not having the files
on disk. perhaps the extension has matured since i last looked (an
admittedly long time ago). if anybody has a compelling reason why it's
better to have that stuff in radiant than on disk i'd be happy to give
it another try.

the only reason for having your css/js in radiant i could ever dream
up was being able to stick radius tags in your css or js but that
didn't work with sns...until today. in the past when i ran into a
place where i needed that functionality i would just stick that little
bit of css/js inline in a layout or snippet; i think i needed to do
that once or twice in three years and more sites than i can remember.

 The greatness behind paperclipped is paperclip (not to take away from
 all the work that Keith and many others put in to make a really useful
 extension). I like page_attachments too and recently did a lot of work
 to change it on my fork with John Muhl's help, but for a recent
 project I was again looking for an asset manager and even though I did
 all that work to update page_attachments (and it still needs more) I
 chose paperclipped because of paperclip.

 Yes. You are right. And there are some extensions which add to paperclipped.
 This is some kind of vote of the community, isn't it ;-)

 Yes, but partly because paperclipped had so much more done. Keith did
 a great job creating it and it's feature set blows page_attachments
 out of the water, but as I've heard John Muhl point out, some users
 find the interface (like the assets bucket) confusing and the
 simplicity of a list on a page (a la page_attachments) is very easy to
 understand.

it's true. i've tried paperclipped on three occasions and all three
times it ended up being pulled due to the commoners not really
getting the interface. with p_a they get it almost immediately; it's
like attachments in email and everybody seems to understand that.

i have an extension tied up in a client site that uses carrierwave
http://github.com/jnicklas/carrierwave and provides a similar
experience to page_attachments that i could try to get permission to
extract and open source if there is any interest in yet another file
manager (i can't imagine there would be but who knows) and honestly it
doesn't do anything more interesting than p_a but at least it's not
dependent on attachment_fu :)

 and with this comes another question:

 Ok, I know that Radiant is different. And I love it. I recently did
 a presentation in my web 2.0 community meeting about radiant. And one
 statement burned so hard: No end user will use this!

 Bullshit. It's just not true. I've had plenty of non-technical,
 non-tech-savvy users enjoying Radiant.

 Hm. Its not bullshit. It's how some people see it. I had to fight 20 people
 to suggest that it would be much more easier to educate people to use textile
 than to have a WYSIWYG Editor. I'm a textile fan. It just would be create
 to have a solution for this feature request. but ok. there are some
 extensions out there. must not be in the core neither a core extension

 Seriously!? No end user will use this!
 I don't know who, other than end users, actually use any Radiant
 project I've worked on.

ditto what Jim said. it requires more work on the developer's side but
once you work with enough clients you get a feel for how non-technical
users expect things to work (e.g. no magic page parts, no
`r:blahblah` etc) and radiant's minimal core lets you bend it to
meet those expectations without 

Re: [Radiant] What is Radiant 1.0 and which Asset extension to use, was: page_attachments interface enhancements

2010-06-29 Thread john muhl
On Tue, Jun 29, 2010 at 7:50 PM, Jeff Casimir j...@casimircreative.com wrote:
 I'd like to see radiant follow some of the footsteps from merb and
 have two gems: radiant-core and radiant-more.  The 'radiant' gem would

that sounds like a fine idea; and has been discussed in the past but
it wasn't particularly feasible until recently

 And if someone really wants a stripped down version they can just
 install radiant-core, barebones but fully functional.

radiant-core would be for the hardcore :)

seriously though i think you'd have radiant-blog, radiant-portfolio,
radiant-catalog etc. springing up then which would be a killer feature
in my opinion.