Simone,

Wherever you can, I recommend going with the flow. Some important questions
to illustrate what I mean, though these are not the only ones:

1. Do you have lots of already-existing databases for which you just want to
create web front ends, or will you be creating mostly new databases for each
project?

In general, the more you're able to start from scratch and accept a
framework's defaults and preferred way of doing things, the better. Rails,
for example, is well-known for this. Its creators consider it a feature, and
advertise Rails as "opinionated software". This truly can be a nice feature
if your opinions match the framework's opinions. Rails and most other
frameworks build in lots of short-cuts and conveniences for doing things in
default ways. Most will allow you to defy their opinions to some extent, but
the more you do, the more pain you'll feel.

This is especially true when it comes to database design. Other project
requirements may go against a framework's flow, too, but databases seem to
be where most people feel the most pain. Many frameworks require you to
design database schemata in very particular, restrictive ways. Not all of
them do, though. For example, the Catalyst framework (Perl-based, not Ruby
or PHP) was designed to be completely database-agnostic from the beginning.
In contrast, though I understand that Rails has been getting much better in
this respect recently, it's not what it was designed for. And since Rails
has gotten so much hype, many other frameworks have followed a similar
opinionated approach. I highly recommend looking for database-agnosticism if
you're putting web front ends on already-existing databases.

Disclaimer: I'm a former contributor to Catalyst, so I'm biased. It's also
been a year or two since I did a lot of comparison research on web
frameworks.

2. What tools do your developers already know? What tools do they want to
use?

You said that you're already using Groovy and Grails. If your developers
already know those tools, why not just use them? The more a tool lets
developers use skills they already have, and the more they want to use that
tool, the more likely the project is to succeed, all other things being
roughly equal.

HTH!

David

--
David Naughton
University of Minnesota Libraries
[log in to unmask]
5-0473

On Wed, Sep 15, 2010 at 3:33 PM, Simone Q. Vuong <[log in to unmask]> wrote:

>  Absolutely, Dave.
>
> *PROJECTS
>
> *We are a research center for *Environmental Health Sciences* under the *School
> of Public Health*. The projects are mostly secure, web-based forms and
> reports as well as some client applications.  Ideally, we would like to have
> all of our applications web-based. Examples of common projects we have are:
>
>
>    - generation of letters, forms, etc for mass mailing
>     - dynamic reports in which one can enter criteria for selection (e.g.,
>    date range, month, year, group, etc.) and select multiple sorting
>     - LOTS of data entry applications
>
> We're starting to create open-source applications that can be shared with
> other study centers.
>
> *DATABASE
>
> *We use both MySQL and MSSQL, depending on the study although we would
> like to move to MySQL for all.
>
> *USERS
>
> *We're almost done with converting to Active Directory(AD) so that would
> be our means of authenticating users. Everyone who needs access our secure
> site -- that's most of our users -- must have an AD account.
>
> *OTHERS*
>
> Also, we've recently been building web applications in Groovy and Grails.
>
> Does this help?
>
> David Peterson wrote:
>
> Hi Simone:
>
> Could you be a bit more specific with the scope of your project? The
> framework to chose would depend upon what exactly you are trying to
> accomplish.
>
> Thanks,
>
> Dave
>
> On Wed, Sep 15, 2010 at 2:35 PM, Simone Q. Vuong <[log in to unmask]> wrote:
>
>> Has anyone explored using *PHP* with a framework -- I'm considering the *
>> Yii* framework for PHP -- and/or *Ruby on Rails*?
>>
>> Could you tell me:
>>
>>    - what you are using it for,
>>     - your pros and cons for using either,
>>     - easy of use,
>>     - the availability of resources for learning and trouble-shooting,
>>     - references you highly recommend, and
>>    - anything else you find relevant.
>>
>>
>> Thank you kindly.
>>
>> Simone Vuong
>> Health Studies Section
>>
>>
>
>
> --
> David Peterson
> IT Professional
> Office of Institutional Research
> University of Minnesota
> http://www.oir.umn.edu
>
> 612-625-0914
> [log in to unmask]
>
>