Re: [Oracle Forms replacement] Ways to start this project

[ibenheni]

Hi,

As a developer in the gnu-kopi package (distributed under GPL), i want to introduce to you the kopi suite, not only as a replacement for oracle forms but also as a complete framework to develop enterprise-level database-aware applications.

The kopi suite contains a set of several tools, among which 3 are the most important:
- Kjc: a java compiler written entirely in java (it was the first free java compiler written in java when first released in 1999);
- XKjc: a compiler extending the java language with seamless integration of SQL (even mixing java and SQL in the same _expression_), new primary types (e.g. Date, Time, nullable types) and operator overloading (like in C++);
- VKopi: a textual language to create dynamic forms, reports and printable reports linked to the database in a simple manner (a visual library deals with all swing stuff, another library deals with all database stuff);
- other interesting tools (a java assembler, a java disassembler, ...).

Also, some of the features of the kopi suite:
- applications are generated in JVM bytecode, so they are cross-platform;
- localization is done through simple xml files;
- the kopi framework is based on java, so java classes and libraries can be used;
- interfacing with numerous DBMS using JDBC (e.g. Oracle, PostgreSQL, DB2, MaxDB, MySQL);
- compile-time verification of SQL requests by referring to a file describing the DB schema in SQL;
- multiple triggers can be executed on the form fields (e.g. entering a field, leaving a field, changing the value, ...);
- different commands, modes, acces rights are applied to the fields and blocks in forms;
- a common library containing typical commands, types, forms;
- and many other features.

I want also to mention another open source tool developed by KopiLeft called tanit. It is a tool that translates from PLSQL, Oracle Forms and Reports (version 3) to a kopi application. So, it may be an interesting solution to migrate easily from Oracle.

As a conclusion, i think that the kopi framework can be a stable open source solution for developing database based applications. We have already thousands of kopi forms used by our customers and some of our customers use kopi applications to deal with databases containing millions of records with no problem. So it is not an emerging project but rather a mature one. Nevertheless, there are many facts that prevent kopi from being wide-spread (even though it is is a gnu package) and i think the main fact is the lack of documentation and demonstration.
My suggestion is the following: instead of developing a new replacement for oracle forms from scratch, why not contribute to the improvement of the kopi platform. As an improvement example, we are making effort to make kopi generate web applications in addition to desktop swing based ones (in the fashion of: write once compile for both desktop and web, interesting, isn't it?), and we already have a working beta version but a really bugging one.

We really need the help to take gnu-kopi to the next level.
I am personally going to start documenting the kopi project in order to make it accessible to as many people as possible.
If you have any ideas about how i should start, they are welcome. Should i start by writing a quick start guide, making a video tutorial, publishing screenshots of existing kopi generated applications, ... ???
You are also welcomed to post to our forum (it is still empty, what a shame): questions, criticisms, how to, problems, bug report, suggestions, ...
http://forum.kopileft.com/

Here are the links for the kopi and tanit projects (hosted on sourceforge):
http://sourceforge.net/projects/kopi/
http://sourceforge.net/projects/tanit/

------
regards

Imed Ben Heni
ICT engineer / Java developer
KopiLeft - Tunisia

On 03/06/2012 08:44 PM, Andrew Russell wrote:

Yes absolutely, replacing formss is a massive task.
I am a 20year veteran developing in forms and i have some thoughts.
The corporations that have an exising forms infrastructure. They have been running their forms for years they like the dev speed and the tight linking with the db. They hate the desupport cycle and waiting till oracle end of lifes forms. With no clear migration path that handles the massive codebase they are stuck.
They will not accept an immature project as a replacement so that gives us 5 years or so before we need to consider their needs.
Andrew

On Mar 7, 2012 3:40 AM, <address@hidden> wrote:

On Saturday, March 03, 2012 12:52:33 AM Jaroslaw Staniek wrote:

> On 1 March 2012 14:20, Mihail Daskalov <address@hidden> wrote:

> > Probably <quote>‘Oracle Forms has the ability to use another database

> > because of its transaction and database triggers replacement.’

> >

> > </quote> refers to the old “Oracle Open Client Adapter for ODBC”. You

> > can see this topic

> > https://forums.oracle.com/forums/thread.jspa?threadID=495283

> >

> > Unfortunately it is no longer developed or supported (AFAIK).

> >

> > Oracle Forms (actually Procedure builder component) included a Client

> > implementation of the PLSQL language (yes PLSQL is not only Oracle

> > Server language).

> >

> > The lack of FREE client runtime is major problem with Oracle Forms and

> > Oracle ADF. Once Oracle Forms Runtime Client/Server was free (no cost)

> > for use. Then it was desupported and only Forms Server (iAS Enterprise

> > Edition) was left. Then BC4J and JDeveloper emerged as some

> > alternative, which was later named ADF Swing. It was still client

> > server, but WAS NOT FREE. Now it is also being desupported (in

> > JDeveloper 11.1.1.5 and 11.1.2), with only ADF JSF left, which is not

> > client server.

> >

> > The only commercial alternatives I know for Client/Server programs are

> > Java with Swing and JDBC (possibly there are some frameworks that help

> > a little), and Microsoft .NET with Windows Forms components, or

> > eventually LINQ , Entity Framework and Silvelight (but at the MS sight

> > things are quite fragile in my opinion also).

> >

> > It would be good to have something on which to start building upon a

> > Free toolkit for Client Server programs.

> Hi,

> Thanks for outlining the taste of Oracle's business. Here's my take.

>

> I believe the Free World has no visible interest in a free ~1-to-1

> replacement for such tools (tools that once became popular because of

> their freeware aspect). That is because developing replacements of

> something that changes itself is always a short-term goal and

> potential success is not very rewarding. The Free World seeks for

> stability - products that can be utilized for years, software

> developed on current and near-future requirements.

 

There are other forms products in the world for which a replacement is wanted - on VMS there is DecForms - most often found in the FIMS language .IFDL file.

 

afaik - there is no equivalent forms definition language for ncurses or anything else into which one could readily change this very capable forms definitions (and function exits) into a linkable object on Linux.

 

Agreed taking on all of Oracle Forms is a big task - perhaps this smaller task is a more suitable goal?

 

>

> The trouble for me is that out of all consultants that would be

> natural partners for such a free initiative some wouldn't dream of

> stability/predictability of software components since their success

> lies in their knowledge of tools that are affected by changes. Some

> others are in gray area but they rarely seem to be early adopters what

> makes it hard to gain momentum (users are welcome contributors in

> FOSS).

>

> Conversely, most of the free software enthusiasts are not interested

> in this kind of tools. So unless someone does really solid 'market

> research', I am afraid we're left without any large group of interest

> but with fair amount of wishful thinking. This is why I would suggest

> taking less ambitious challenges so goals can be achieved before

> situation (especially technology) changes again in a painful way.

>

> --

> regards / pozdrawiam, Jaroslaw Staniek

>  http://www.linkedin.com/in/jstaniek

>  Kexi & Calligra (kexi-project.org, identi.ca/kexi, calligra-suite.org)

>  KDE Software Development Platform on MS Windows (windows.kde.org)

>

> _______________________________________________

> Replacementforms-discuss mailing list

> address@hidden

> https://lists.gnu.org/mailman/listinfo/replacementforms-discuss

_______________________________________________
Replacementforms-discuss mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/replacementforms-discuss

_______________________________________________
Replacementforms-discuss mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/replacementforms-discuss