Just a note on MVC licensing.

MVC is a DSpace project (started by Graham & continued in DSpace Summer 
of Code) and as such is under the same BSD license as all other DSpace 
code: http://www.dspace.org/license/

The MVC codebase is a subproject of DSpace's GitHub Repository:
https://github.com/DSpace/webmvc

All code in the DSpace GitHub repository is under the DSpace BSD License 
(unless specifically stated otherwise)

- Tim

On 7/27/2012 2:17 PM, DSpace @ Lyncode wrote:
> Hi,
>
> sorry for the delay. We were internally discussing your comments (thanks
> for the replies).
> Although SpringUI being built from scratch it could re-use most of the
> 'M' and 'C' of webmvc implementation (licensing issues).
>
> About the templating system, Graham Triggs stated and it's undeniable,
> using Freemarker for that purpose it's fair and almost obvious. So,
> explaining, the decision of using another templating system it's based
> upon the idea of improving the UI changeability. So, the main question
> is all about the templating language and how verbose/easy it is to
> create/modify. Freemarker is a great option and it's highly
> customizable, but the need for costumizations to create some templating
> mechanisms is a negative point within Freemarker. Also this creates an
> oportunity to use another templating system.
>
> Although Freemarker has an active community and is widely used, we are
> not fans of complexity when it refers to templates.
>
> Java templating systems Twig/Django alike:
>
> - jbrackets (http://www.jbrackets.org/)
> - jangod (http://www.jangod.org/)
>
> About using a common UI core (Peter Dietz), we think @mire approach is
> the best one (i'm talking about the discovery module approach example),
> basically, developing small and re-usable business modules, mainly because:
>
> - They are distributable as add-ons (optional modules?)
> - As independents and specific task developed APIs they could/should be
> easily identified and used (we wouldn't need to search in a complex and
> all-in-one API)
> - Easy to incorporate into the DSpace default distribution
> - Easy to maintain
>
> As so, we think that its risky to develop a common UI core, mainly if it
> tries to almost all UI work, it could limit future UI developments, but
> also, it would require much more effort when some UI core API changes
> take place. As Mark Diggory said, it could be interesting if it already
> exists simplifying UI development in some cases, the same could be
> achieved with an "independent small business layer modules" approach.
>
> We do not want to develop another XMLUI/JSPUI similar UI.
>
> PS -  I wasn't able to find licensing information about webmvc.
>
> On 23 July 2012 14:07, Graham Triggs <[email protected]
> <mailto:[email protected]>> wrote:
>
>     Hi,
>
>     As the main protagonist of the WebMVC project, I should probably
>     step in and say something. It's good to see that there are others
>     who wish to develop a UI using the Spring MVC framework, however,
>     there are a few comments I would like to make about this proposal.
>
>     If you look at the project organisation of WebMVC, you'll see that
>     it was always conceived to keep the 'M' and the 'C' separate from
>     the 'V'. Whilst I chose to use Freemarker for the 'V', I always
>     intended that the 'M' & 'C' present a common, consistent layer on
>     which different flavours of 'V' could be implemented. I'm not
>     claiming that it perfectly achieves those aims right now, and there
>     may well be some refactoring required, but the basic intention is there.
>
>     I would be very cautious of another full MVC implementation that did
>     not take into account that the 'V' could be entirely replaced with
>     something else, keeping the 'M' and the 'C'.
>
>     Whilst I share some of your concerns over Cocoon, I personally think
>     the poor state of support, and relatively low level of adoption
>     throughout the Java community is a much bigger cause for concern.
>     The resource requirements are excessive, but it's the health of
>     Cocoon itself which really pushes me to think there should be a
>     compelling alternative available for the community.
>
>     Our existing JSP interface is not a good example of separation, but
>     that doesn't mean JSP as technology does not support MVC. Our
>     existing implementation is a very light, non-framework based and
>     poorly followed through MVC architecture. You can use JSP as the
>     view in an MVC architecture, but what you can't do is prevent people
>     from abusing the architecture by throwing Java into the JSP.
>
>     I've not heard of Twig before, but it seems with good reason - it is
>     a PHP templating language. Looking around, I can't find a Java
>     implementation of it, so I'm struggling to see what it is you are
>     proposing here. Is there a library that I'm not familiar with? Or
>     are you writing an implementation of Twig in Java?
>
>     Either way, I would very, very strongly caution against using an
>     unfamiliar, unproven templating language - regardless of what cool
>     features it may have. That way just leads back into the same
>     problems that we are seeing with the Cocoon framework.
>
>     I chose to use Freemarker specifically because it is very widely
>     used (*), and actively supported. It's quite simple to use (very
>     similar to Velocity), and whilst essentially enforcing that you
>     can't simply dump code into the templates, it's quite extensible in
>     the way that you can use it (for example, being able to use JSP tag
>     libraries if you really need to).
>
>     (* A few of the Java based CMS apps use it, as does Vivo, and some
>     major frameworks like Hibernate).
>
>     With regards inheritance, no you can't inherit a template file in
>     the way that Twig allows. But I'm not convinced how useful that is,
>     and you quickly get into much more complicated templates, rather
>     than the simple HTML + CSS that you are envisioning.
>
>     As it stands, I've put a lot of work into the theming support with
>     the Freemarker UI, and whilst you can't inherit a template, you can
>     inherit a theme (so you could have a base theme for your overall
>     repository, with derived themes for each of your collection, for
>     example). Also, the themes allow you to overwrite any of the base
>     templates. So, your ability to inherit / customize the output
>     depends on how you structure the templates - break them down to
>     enough granularity, and you can effectively get the same inheritance
>     effect.
>
>     One other thing the Freemarker implementation has going for it,
>     there is a lot of support for I18n - you can have replaceable
>     messages within the templates, or you can completely overwrite a
>     template for a given locale / language (potentially very useful in
>     some cases).
>
>     Regards,
>     G
>
>     On 20 July 2012 04:12, DSpace @ Lyncode <[email protected]
>     <mailto:[email protected]>> wrote:
>
>         Hi all,
>
>         some of the current lyncode developments for DSpace, in fact,
>         begin by some specific requirements of the SpringUI, that is, an
>         under development UI for DSpace. As we are willing to develop a
>         robust, wide accepted UI and giving it to the DSpace community
>         as a code contribution, we would like to have, firstly, some
>         feedback from the DSpace devel team.
>
>         Let me invite you to read the following draft:
>         https://wiki.duraspace.org/display/~lyncode/SpringUI
>
>         As you could see at the end of the referred page, our contact
>         with the DSpace community in this specific will start by /"//1.
>         Compile an about page with all necessary info"/. So, is there
>         anything left to say?
>
>         We would be pleased, based on your expertise, to listen to your
>         advices/doubts.
>
>         --
>         Thanks, DSpace @ Lyncode
>         DSpace Department
>         *Lyncode*: Official website <http://www.lyncode.com/>
>
>
>         
> ------------------------------------------------------------------------------
>         Live Security Virtual Conference
>         Exclusive live event will cover all the ways today's security and
>         threat landscape has changed and how IT managers can respond.
>         Discussions
>         will include endpoint security, mobile security and the latest
>         in malware
>         threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>         _______________________________________________
>         Dspace-devel mailing list
>         [email protected]
>         <mailto:[email protected]>
>         https://lists.sourceforge.net/lists/listinfo/dspace-devel
>
>
>
>
>
> --
> Thanks, DSpace @ Lyncode
> DSpace Department
> *Lyncode*: Official website <http://www.lyncode.com/>
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>
>
>
> _______________________________________________
> Dspace-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/dspace-devel
>

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Dspace-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dspace-devel

Reply via email to