If you do have a production application built with ADF 11g can you
please email me at
shay.shmeltzer @ oracle . com
And let me know:
Company name
Technologies used
Two sentences describing the application
Would you be interested in talking to analysts about your ADF
experience
Would you be interested in talking at events about your ADF experience

John brings up some interesting points.

At my current client we're getting incredibly close to production for our
first ADF 11g application, and its a corker at 1 page! (ok, there's a bit of
sophistication in that 1 page app, but still only 1 page).  I'm sure a 1
page app would be a poster child for your needs Shay, it's even got a
spinnng icon! ;-)

In adopting ADF my client has taken the recommendation on board to start
small and build up skillsets, resources and similar.  However like John
indicates, the adoption of ADF is slowed by other peripheral activities,
tools etc.  WLS has certainly been a learning curve.  The introduction and
setup of SVN is another, and its disruption of the current change control
processes.  I've also put a lot of emphasis into teaching teams to do
proof-of-concepts to see how things do and can work, as well as how to write
design documents that assist the ADF development process.  This all takes
time.

There have also been 3 other main limiters in our ability to do more:

1) the organisation has a substantial Forms legacy system + other systems to
maintain; these continously pull team resources away. ADF is still niche,
not core to their dev strategy purely because of the scale of the legacy
system
2) there are virtually zero other ADF resources available internally and
externally in the local market. Java resources are somewhat available, but
to be honest I'm not overly keen to introduce them as they're first reaction
to ADF is usually immature
3) continous issues of Forms is more than capable of providing "good-enough"
solutions in many cases, so where to focus for the next ADF application
without rewriting something Forms is already doing well?

In saying # 3, my client does have a number of upcoming projects which will
lead on from a heady 1 page application, which will be particularly suited
to the web delivery ADF is capable of, and in turn more excitingly tailor
fitted to the new task flow features.  We're hoping for at least 2 pages in
the next app! .... logon, logoff ;-)

I guess why I wanted to post in reflection of John's timely comments is the
adoption of a new technology into an organisation can often be a large
undertaking which has little to do with the actual technology at hand, it's
the peripheral issues that take a lot of time, and often requires somebody
who is keen to see it succeed with a long term outlook.  This is why at
times I've taken pains to indicate to Oracle, that beyond providing demos on
features and similar, you need to assist organisations (for free ;-) in
sloting JDev into the existing internal mechanics of the organisation
(remember my rant about SVN + JDev? - it's not about SVN, it's about changes
to change control procedures that was a pain point for us in adopting JDev)
.... but that's also why this group exists.  In addition given there are now
so many "successful" legacy systems installed unlike 10 years ago, new
technologies need to be incredibly good to dethrown their predecessors in a
zippy fashion, even with a lot of bells and whistles.

Regards,

CM.

Ah, you misunderstand.  No page fragments, just a simple-ish search-result
page, with a bunch of smarts; an autocomplete control (based on Frank's
recent article), business rules around when certain results can be shown and
not, combinations of legal search fields, large amount of logging user
actions for statistical purposes, + minor skin work, + template, but in the
end, just a single page app.  A fair amount of the work for the juniors has
just been learning where and how to do things in the framework particularly
with our business requirement to log the users' actions.  Various parts of
the solution have popped up on my blog (or will do soon).

Certainly not the Queen Mary of ADF apps I'm afraid.

CM.

Hi Chris

I'd like to take the opportunity at ODTUG to sit and talk about JDEV and SVN. As you know, it's my area and I'm even more into the whole ALM/SDLC aspects of team working these days - in fact, if anyone else is interested perhaps we could do it as a group. I want to ensure as far as possible that what your group is working on in terms of methodology and best practice dovetails with our plans for JDeveloper.

I also hope to wow some of you with Team Productivity Center - our move into ALM - that I'll be spreading the word about ;-)

rgds

Susan


Chris Muir wrote:

John brings up some interesting points.

At my current client we're getting incredibly close to production for our first ADF 11g application, and its a corker at 1 page! (ok, there's a bit of sophistication in that 1 page app, but still only 1 page).  I'm sure a 1 page app would be a poster child for your needs Shay, it's even got a spinnng icon! ;-)

In adopting ADF my client has taken the recommendation on board to start small and build up skillsets, resources and similar.  However like John indicates, the adoption of ADF is slowed by other peripheral activities, tools etc.  WLS has certainly been a learning curve.  The introduction and setup of SVN is another, and its disruption of the current change control processes.  I've also put a lot of emphasis into teaching teams to do proof-of-concepts to see how things do and can work, as well as how to write design documents that assist the ADF development process.  This all takes time.

There have also been 3 other main limiters in our ability to do more:

1) the organisation has a substantial Forms legacy system + other systems to maintain; these continously pull team resources away. ADF is still niche, not core to their dev strategy purely because of the scale of the legacy system
2) there are virtually zero other ADF resources available internally and externally in the local market. Java resources are somewhat available, but to be honest I'm not overly keen to introduce them as they're first reaction to ADF is usually immature
3) continous issues of Forms is more than capable of providing "good-enough" solutions in many cases, so where to focus for the next ADF application without rewriting something Forms is already doing well?

In saying # 3, my client does have a number of upcoming projects which will lead on from a heady 1 page application, which will be particularly suited to the web delivery ADF is capable of, and in turn more excitingly tailor fitted to the new task flow features.  We're hoping for at least 2 pages in the next app! .... logon, logoff ;-)

I guess why I wanted to post in reflection of John's timely comments is the adoption of a new technology into an organisation can often be a large undertaking which has little to do with the actual technology at hand, it's the peripheral issues that take a lot of time, and often requires somebody who is keen to see it succeed with a long term outlook.  This is why at times I've taken pains to indicate to Oracle, that beyond providing demos on features and similar, you need to assist organisations (for free ;-) in sloting JDev into the existing internal mechanics of the organisation (remember my rant about SVN + JDev? - it's not about SVN, it's about changes to change control procedures that was a pain point for us in adopting JDev) .... but that's also why this group exists.  In addition given there are now so many "successful" legacy systems installed unlike 10 years ago, new technologies need to be incredibly good to dethrown their predecessors in a zippy fashion, even with a lot of bells and whistles.

Regards,

CM.

susan_duncan.vcf < 1K Download

It's been a huge learning curve . . . ADF Libraries did work for us,
so we have struggle with how to design the apps so that we can reuse
the model. To that end, we are currently struggling with how to get 2
view controllers to work within the same workspace so we can use the
Model that exists there. We file one or two SRs weekly.

Dear Shay,

First  to answer your question:

*Company name :  *I am not allowed to mention the name we are in Iran and
generally we are not allowed to use the american technology but as you now
the Oracle products are the most used products here in Iran and next goes
for Microsoft.
* Technologies used :  *Oracle ADF 11g
* Two sentences describing the application :
 *it is a total solution system for a oil-power plant company with more than
3000 employees the main module is :HR,Document Control,Office
automatin,Contracts& change,.. each of them has about 2-3 sub system.
* Would you be interested in talking to analysts about your ADF
experience:  *yes
* Would you be interested in talking at events about your ADF experience:
*yes

It seems that our company is one of the first company that has a full
application deployed in the production environment with ADF 11g, this system
has about 50 pages/page fragments and more than 100 entity/view/view
link...and other ADF BC components.
the release that we are working is 11.1.1.0.2 (passing the process of
migrating from previous releases), the application is deployed on the
internet too and pls do not ask about the license.. :)
I am really satisfied with this version that was a very big step from 10.3
to 11g.
Also, as I am a developer and consultant of ADF I know 3 other companies
that are working on migrating their applications from form builder to ADF.
we are generating the pages in a team of 5(adf developer) 5 analyse and
designer 1 project manager, 2(quality and test) and generally we are happy
with the good speed of developing and we are mainly disappointing about:

1-many bugs
2-lack of standard for developing and  not a really clear solution for
building a big application more than 100 sub-system with ADF
regarding the standard, mainly the people are worried about how to managed
their application modules and other BC component but in production this is
not a really issue the most issue is that how to develope an application
that is consist of 100 or more subsystem that each subsystem has more than
50-100 pages and each subsystem need about 4-5 application modules!
The system that we are developing is consist of 100 subsystem and we at the
present has developed 3 small-sized subsystems of it,I hope you could guess
my worries.
When the number of application modules and pages are getting big, deploying
is risky and error-proven and the performance would be decreased and because
the number of BC components are large each user would gain large resource so
when the number of user is getting high, the performance is decrease
significantly.

good point that help us to deal with above problems:
1-working on standardized application developing with ADF, having more than
2 (starting from 10g) year of pure developing ADF we reach to a point that
at present we have a really acceptable standards. the best step that we have
taken here is that now we have a template applications that it has all the
necessary resources and patterns needed for developing such as the Task Flow
pattern, Skin, customized Entity and View,customized Error Handling,page
template,sample master-detail page,all the utility and library,sample deploy
project,shared app module,needed page for gateway,1 session bean(user can
add other bean but this one is needed for make the gateway works,...
every sub system is developing with this template and based-on the number of
pages that each sub system would have the team member would be adjusted.
2-each sub system is deployed independed for example we have a
financial.ear, cartable.ear,...
the good point here in contrast to having a big ear file with very larg
entity and view is that:
- in a little change there is no need to redeploy the other application only
the changed application (ear) would be updated.
- developing rate is higher
- maintenance is really simpler
the problem with this design is:
- we need a pattern to handle the link between this subsystem(ear) files,
for this issue we make a pattern that is similar to a gateway, I mean that
when a person from page "a.jsp in cartable.ear" goes to the "b.jsp
financial.ear" with this gateway we do following things, first we invalid
the session in cartable.ear, second we direct the person to the b.jsp in
financial.ear (here a new session in generated) and at last we move the
previous session data that user had in cartable.ear to the new application
financial.ear (not a easy step)

maybe our design is not very good but at present we are generally satisfied
with the application,our application has about 400 user (100 concurrent
user) and they will reach to 800 at the next 4 months.
we are now developing 5 subsystem that would be finished in next 6 months.

I hope that this information is useful and if anybody has a better design or
want to discuss the method that we are adopting for developemnt I would be
really happy to share our idea.

Regards,
Amir

Hi Shay,
MNO is a young software company from Belgrade, Serbia formed in december
2008. Developers that work in MNO have about three years of work experience
in JDeveloper and are former employees of Omnilogika doo, Belgrade. 11g
platform was adopted for development since release of technical preview 1.
Since then our developers worked on projects that are described below.

*Primary Healthcare (General Practice - GP)*
ADF Faces 11g, EJB 3.0, Oracle Toplink, Jasper Reports
Developed with JDeveloper 11g tp3, deployed to OC4j 11g tp3.
Currently in migration to R2 and WebLogic
PHC solution covers four main areas in medical primary healthcare:
1. Individual patient data management (Electronic Patient Record - EPR)
2. Research and epidemiology
3. Creation of the patient's database for the introduction of capitation
financing models with reporting mechanisms
4. PHC management and quality assurance
Software is currently in use in twenty five ambulance stations with about
150 doctors and nurses. Database contains about one hundred and fifty
thousands patient records.
MNO developers participated in all phases of software development cycle.

*Sport Facilities Register*
ADF Faces 11g, EJB 3.0, Eclipselink
Developed with JDeveloper 11g R1, deployed to WebLogic 10.1.3
Encapsulates basic common objects in charge of generic management of sport
facilities.
It will be used by all existing and future applications which are related
with e-government in sports.
Software is currently in testing phase by future users in Serbian Misnistry
of Youth and Sport.
Estimated users more than 1000.
MNO developers participated in all phases of software development cycle with
partner company Belit doo, Belgrade.

*SEPA Direct Debit *
ADF Faces 11g, EJB 3.0
This software module enables bank to participate in SEPA Direct Debit
clearing in Serbia.
Currently in test phase.
MNO developers participated partially in UI development.

Regards
Marko

CM.

Regards,
Jean-Marc