[Architect guide]

Hi,

I have asking others communities and they have said me that ROO is not a good choice for BIG and COMPLEX project...That it is good to start the project (Maeven..), and then to separate ourself to ROO and to manage component individually.

Last but not least, it is said that we can use one server for the search (solr), one server for the repository (Alfresco :) ) and so on...But if I want a cheap infrastructure to save money, is it possible to have all in one server ?

Tks,

Quote:

---

Originally Posted by Brad Murray

I'm unaware of any such technologies in Spring itself.

For your versioning and CMS though, I'll still echo what other people are saying and grab Alfresco. You can have the open source version for free, and it has a lot of nice restful APIs that you can use to integrate it with your app.

> Ok, so does it mean that I can dl the code of Alfresco, then do my own application with it, and I can do my business with my soft (is it legal) ?

Architecturally, what you want is not one app that has all of these functionalities, but a web-app acting as the integration point for different technologies, often running in seperate apps in your domain.

My own project (large public sector CMS) has both Alfresco and Solr running on dedicated servers, but the end-users do not interact with them, our CMS and integration applications do. You are leveraging thousands of man-hours that has gone into getting those applications as good as they are in their domain.

> I totally agree, and I would like to know more about it, about the architecture, how proceding... I do not know...do you have any website to recommend me ?


Honestly though, I do think it would take more than a month to become proficient in working with all of the techs you could combine with Roo. Roo itself does not dictate any long term design choices though. Every single part of what a default Roo application uses can be pulled out, its just a matter of getting to know where it puts things.

---

> So you are saying, that ROO is not a good choice for large project ? But I do not understand why ?

Tkx,

Roo + hard work = success (any project)

Tossing questions here and there to see what sticks to the wall; probably not.


Thank you
jD

Ok, I will stop asking questions because, you already give me a hand and I have a better understood of my situation. So, can you (ultimately), confirm me that 'it is not stupid' (loss of time and energy) to develop my own application (used by my startup and important for my business) with ROO (I need the following functionalities) :

// common to others CMS features

* RBAC (user's permissions)
* OpenID
* XSS SECURITY
* Versionning + repository
* search
* managing taxonomy
* simple workflow (edit, validate, publish)
* forum
* SEO

// and this is more or less the original part I want

* own user inteface (WIZIWIG) to create the articles
* multilinguistic backend and frontend
* manage multi site + auto generate squeletor
* own way to manage adverts (DFP's API) + accounting
* internal communication tool (gmail ???)
* special Google Map interaction
* will do my own dashboard to have an overview of the production / business / marketing indicator


Most of people recommand to use an existing CMS because it has been done by a lot of experts during many years...but I do not want a CMS that easily allow other person to deploy their website but just in order to let my contributors to work with my application. DIY I am afraid that I will encounter a lot of bug and headaches, but most of all, I am afraid to have an unsecure application. Remember that I would like to be on the GAE. Tks a lot.