Le 04/07/2017 à 16:57, Michael Brohl a écrit :
Hi James,
thanks for your suggestions.
As far as I know, JSF would introduce some new technologies because it relies on beans and JSP's (correct me if I'm wrong). I'm not sure if we want
to go so far.
Facelet is now the recommended technology for JSF https://stackoverflow.com/questions/2095397/what-is-the-difference-between-jsf-servlet-and-jsp and
both are parts of JavaEE.
I agree with Michael and would not like to change OFBiz widgets for JSF. Not that I don't like nor trust JSF (and Oracle, but then a bit less), but
the work is overwhelming and obviously we don't have the resources for that.
I digged a little deeper into the UI stuff, templates and theming and have to correct my summary a bit: I mentioned AngularJS and Bootstrap on the
same level which is like comparing apples and oranges. AngularJS is a client-side JavaScript framework to build single page applications, icluding
his own model-view-controller mechanism while Bootsrap is a CSS framework which provides comprehensive UI elements in a structured way.
I guess that the use of Angular would need a whole lot more changes in OFBiz
than the use of Bootstrap.
So I tend to think that we have to agree on a CSS framework like Bootstrap and rewrite the UI to use the proper CSS classes for this framework. That
would possibly reduce the complexity and makes this statement of mine obsolete:
> - we will need a new approach to be able to "plug in" different UI frameworks. We'll need a UI layer who represents the screen contents in an
abstracted way (possbly an enhanced Freemarker macro library) and make it possible to generate HTML code with the right css attributes for the
target library.
It's maybe too ambitious wanting OFBiz to be able to be used with different frameworks. The Bootstrap CSS world is well documented [1] and there are
a lot of really good looking and functional free templates out there. So if we provide the UI code for it, together with one basic theme, users can
put their own themes on top of it.
Maybe this is a way to come to a competitive UI in a relative short amount of
time. I don't think that we can afford to make this a year-long project.
What do others think?
I agree that using Bootstrap would be a good thing. An alternative is Foundation https://www.keycdn.com/blog/bootstrap-vs-foundation, this could be
possibly discussed.
That's what ilscipio has used, with some success at the UI level I'd say (they now tend to lean to Foundation). Now they derived from OFBiz at other
technology levels (no or less form widgets but more FTL macros, even an API of FTL macros). So I'd try to compare the rest...
I'd also let Angular out of the picture. Some prefer React (initially from FB) and I wonder what those who have used Angular 1 think about Angular 2!
I also remember another Google "attempt": GWT. Are there still people using it with OFBiz? I guess you get my point, trends pass and tools with them...
Jacques
Best regards,
Michael Brohl
ecomify GmbH
www.ecomify.de
[1] https://www.w3schools.com/bootstrap/bootstrap_ref_all_classes.asp
Am 03.07.17 um 15:00 schrieb James Yong:
Hi Michael and all,
We can look into JSF 2.2 as a possible candidate. It is similar to OFBiz Widget
and seems to fit the new requirements described so far in this thread.
Regards,
James Yong
On 2017-07-03 17:42 (+0800), Michael Brohl <michael.br...@ecomify.de> wrote:
Hi Sharan,
thanks for the reminder.
It's fine to have another theme to choose for the "old" UI, I just want
to point out that (in my mind) the new theme/UI initiative goes far
beyond having just another theme on base of the current technological stack:
- new themes should be responsive
- we should be able to use different UI frameworks like Bootstrap and
AngularJS who take care of responsiveness and browser compatibility
- it must be easy for developers to write the screen structure and also
easy for webdesigners to build a good design on base of this
- developers should not care about CSS styles and classes, and
webdesigners should not cara about how the screen snippets are put
together or how the screens get their data.
- we will need a new approach to be able to "plug in" different UI
frameworks. We'll need a UI layer who represents the screen contents in
an abstracted way (possbly an enhanced Freemarker macro library) and
make it possible to generate HTML code with the right css attributes for
the target library.
- a rewrite of the screens will be necessary to make the UI less
cluttered and overloaded. This will require some concepts/design work
beforehand
- there are surely many other possible requirements (I am not a UX or
web design expert)
I appreciate the contribution of the new theme. I am also sure that this
will not solve the challenge to drive OFBiz to another level, UI wise.
Thanks and regards,
Michael Brohl
ecomify GmbH
www.ecomify.de
Am 03.07.17 um 10:52 schrieb Sharan Foga:
Hi All
Don't forget that we also had the offer of a theme from Provolve and
Stannah.
https://issues.apache.org/jira/browse/OFBIZ-6985
This is a theme that they are using at the moment (so it working) and
have said it could be contributed back to the project. If it's only a
case of having someone volunteer to implement it into the trunk then
this could be a way to get a nice theme up and running quickly for us.
Thanks
Sharan
On 03/07/17 10:29, Michael Brohl wrote:
Thanks Nicolas,
is there anything, even work in progress, you are able to share at
the moment?
This way other could chime in and help moving further.
Thanks,
Michael Brohl
ecomify GmbH
www.ecomify.de
Am 03.07.17 um 09:26 schrieb Nicolas Malin:
Hi Michael
Le 02/07/2017 à 20:42, Michael Brohl a écrit :
Hi Julien, all,
I'd like to resurrect this discussion and the activities to improve
the OFBiz user interface. I think we really should put some focused
effort on it if we want OFBiz to be recognized as a modern ERP.
Also, if we imporve the UI, more users and also developers will be
attracted which will be a win for the community and further
development of OFBiz.
Nicolas and others who have started work on this: can you give us
an update about the efforts undertaken and where we stand?
Currently I block on the comon-theme with a good information
propagation. My next step will be create a dedicate object as
referent on widget context, but my works has been disturb with the
framework separation and the git-svn link break.
Instead of continue, I help some people to work on the groovy
mini-lang conversion. I plan to improve a few the groovy DSL and
after I continue the work on commont-theme.
Nicolas
