New controls for FWH

[hag]

Otto:

you asked great questions but before I give you my response let me give you my perspective.
I have one application that has been in the marketplace for 20 years. I'm not a developer I've only
the one application. It is sold to single users.

Presently competitiors have developed a web-based applications that I compete with.
Even though in my opinion a Web based application is no better than they single-user
application web based is the trend. We can develop beautiful, graphically enhanced applications but if the
user wants a web-based application the sale will be lost.

I believe there are two ways to provide a web-based application using fivewin. One is to have a stand alone
application on a server such as Citrix. Provide the user access through a website and they would have a
web-based application even though they are using a stand alone application on a Citrix type server.
To the user it would be Internet-based accessible from any computer anywhere. To the developer the only issue
would be getting it on an appropriate server at a reasonable cost.

You could literally carve out a number of folders on Citrix server and have the application provided
to many users with each user having access to their stand alone app in their folder on the server.
This might be a reasonable temporary solution to having web-based apps. I am investigating it for my program
but so far it is too expensive. Citrix which is an ideal server for this charges a monthly fee per user which
would drive the cost of my program significantly higher.

But we're still looking for a way to accomplish this.

The second method would be to have sufficient programming options to convert fivewin applications directly to
Internet-based applications. I can't say too much about this because I am not familiar enough with
web-based apps to have any idea how something like this might be accomplished.

[Otto]

Hello Harvey,
thank you for your response.

In AUSTRIA they are celebrating 20 years internet these days.
I remember something around 1993? when I started my business I organized an “Intenet” demonstration to make my clients familiar with the internet. And we spoke that in future all will be in the net. But I feel many things are still offline or partly offline.

For example we sell a guest registration application where we post data against a XML server. Although there exists a free web interface with online registration form an with own online database space many prefer offline solutions. I don’t know the reason I only can think of data security and speed.

>Presently competitiors have developed a web-based applications that I compete with.
This is exactly what’s happening here too.

I see the internet more a versatile DEVICE for example a display, a keyboard a data input terminal etc. Sure we need to interface with all that.


Best regards,
Otto

[TimStone]

Otto,

I think the biggest reason people are not happy with web based applications is security. Putting a lot of information into the "cloud" is only good if you trust whoever is hosting the service. Trust is hard these days.

Tim

[xProgrammer]

As most of you know I use FiveLinux rather than FiveWin but have been following this with some interest.

It's easy to propose a FiveNet or FiveWeb or whatever but I think you need to be a little more specific about what such a product would be. It would also be important to realise that Antonio is only (super?) human and if a great deal of work is required that represents so much less time available for FiveWin development. It would be essential also that a sufficient number of users are going to use the product so that it is worthwhile for him to continue actively developing and supporting any such product and it gets up to production quality rapidly enough.

I also saw a request for a FiveLinux that was 100% compatible with FiveWin. I think that that would only be achievable if you fairly radically restructured FiveWin (which I don't think many FiveWin users would appreciate). Five Win is structured around native Windows calls. FiveLinux is based on GTK (which has a Windows version) but its structure is based on FiveWin (in order to give it a fair degree of compatibility with FiveWin) but that does place limitations on FiveLinux.

The real strength of FiveWin is Antonio's support and commitment to ongoing development. I think history is showing that he can't put the level of effort that he would like into his products if there are too many of them.

Just my thoughts
Regards
xProgrammer

[MdaSolution]

Antonio,
What do you mean with FiveWeb?

[frose]

I fully agree with xProgrammer.

If there is a strong customer wish for a browser/web based application (FiveWeb) then my recommendation is to choose Eclipse, PHP, Ajax, GWT, etc.

For the future of FiveWin I’d like to have more sophisticated design patterns similar to them offered from http://www.extjs.com with an add on for scheduling http://ext-scheduler.com or http://www.treegrid.com for Java/JavaScript.

There are a lot of reasons why I’m still developing with (x)Harbour/FiveWin:

No need of browser/web based application
Perhaps performance advantages
A huge existing code base
ADS with free dbf-tables, PHP requires an ADS dictionary .adt
Etc.

[Antonio Linares]

The way I see a "FiveWeb" solution is as a powerful and simple way to quickly develop a database application for the web.

We can't take a desktop application (built with FWH) as it is and "port" it to the web. For such purpose there are already solutions as VNC, Citrix, LogMeIn, remote desktops, etc. We are not going to reinvent the wheel

If we think about porting the PRG source code from the desktop to the web we face a major trouble: A desktop application is a self contained solution (unless it uses a client/server model) meanwhile the web uses the client/server model. Most of the processes happen in the server, where the databases are located, though some other processes occur in the client side (input analysis, formatting, painting style, hyperlinks selection, cache, etc.)

These two separated processes require different tools to be used:

In the client side the internet browsers allow some specific languages and some plugins that extend the functionality of the browsers. Painting format and styles are defined by CSS files. Client side language scripts like javascript are used to validate user input, make some local processes and to invoke remote actions.

In the server side some tools have become standards, widely used and available from many hosting companies: PHP, Python, Java, ASP, etc. The major difference here is between the free ones and the commercial ones. The more populars and widely available are the free ones. These languages have the mission to organize and send the users requests to the databases and to return the result to the queries as HTML pages that will be delivered to the web browsers.

Is there a way to automatically "split" the PRG source code and to know what parts should be translated to the client side, as javascript, etc. and what parts should go to the server side, as PHP, etc ? Sadly, no

We have to face the reality: We need to develop a new application for the web. Even if we want to offer a similar functionality as our desktop based solution, we have seen that the web architecture forces us to use a different model.

And here it is where FiveWeb fits: A collection of already implemented and tested solutions to mimic the functionality of a desktop application as much as possible. Is it a PRG based solution ? No. Does it automatically translate my PRG ? No. Will I have to learn some html, javascript and php and similars ? yes. But don't be scared, the same way as FiveWin eased the transition from MsDos to Windows, FiveWeb will do its best to easy the pain of this new transition

The target of FiveWeb is to provide you working skeletons ready to be assembled into a powerful web application. We save you the lots of R+D hours that you need to become an expert on web development. You can quickly deliver to your customer and in the way you will become a web applications master the same way as you already mastered Windows programming meanwhile you were delivering your FiveWin based applications

[Antonio Linares]

I would also like to comment that FiveTech is not just me

FiveTech actually is a team formed by great experts that most of you already know very well: Daniel, Rao, Enrico, Charly, Anser and me. Besides the lot of contributors that we have found along all this time that we are together, great masters too, like Manuel Mercado, Ramón Avendaño, Alfredo Arteaga, etc. a very large list!

And of course, this great community where there are great friends and great programmers. If I try to mention all of them then I would have to write here most of all your names

FiveWin is not the key, all of us are the real power behind FiveWin

[frose]

Antonio,

thank you for the clarification.
So FiveWeb is on my wish list too!

Do you have a basic orientation when FiveWeb will be released?

[Bayron]

So FiveWeb is on my wish list too!

I want one too.

I guess the more people confirming interest, the easer is going to be for Antonio to prospect profitability and apply more resources and investments to this project...

[Richard Chidiak]

Same for me , i am candidate for a fiveweb also

[lailton.webmaster]

Antonio,

xWeb

Please See:
http://www.4shared.com/file/6roKuXMq/xweb.html

Its to use xharbour with CGI work fine !

think that is very easy change to add fivewin comandas ( and inside make code html. )

[Marco Turco]

Hi Fivewinners,
I read all suggestions,
most requests are related to graphic controls, improve the printer preview and solution for web app.
but i would like to put to your attention some aspects you didn't consider:

1) Windows graphics controls. I was one of the more "aggressive" FWH developer that asked Antonio to develop better graphics controls for fivewin (tOutlook, tTitle etc.). Antonio has made a really good work and we can make just NOW really competitive app on the graphics side.

2) Printer preview/Report. In my opionion it has no-sense to ask Antonio to spend time to improve the standard FWH preview/report. We can easly use Fast Report from NOW that is integrated in full with FWH, it is used by thousands of developer in varios languages (Delphi, VC etc.), it has an integrated preview with export facilities in Word, Excel, OpenOffice, XML,PDF and various graphic formats.

3) Web applications. Interesting but this is a solution that is not interesting to a wide number of end-user at least for the next few years.

I speak as a software developer that has about 5,600 customers that use products developed with FWH - still thanks to Antonio,
IN MY OPINION WE MUST CONCENTRATE OUR ATTENTION ON THE DATABASE MANAGEMENT AT THIS MOMENT !!!!

The DBF database is nice but - unfortunately - it is a tecnology that in my and - more important - in the market opinion has no future !!
Now the develop of new device (Ipad, Palmtop, SmartPhone...) require to connect different device to PC and vice-versa,
objectively we can't make this using DBF in a simply way.

We urgently need FWH embedded class to manage DIRECTLY and in easly way SQL database like Sqlite (it is coming a standard database for stand-alone/small network) and Firebird (for larger network/web solution) without the use of ODBC, ADO and other interfaces !!!
That's all. Any opinion appreciated.

[xProgrammer]

Hi Marco

"IN MY OPINION WE MUST CONCENTRATE OUR ATTENTION ON THE DATABASE MANAGEMENT AT THIS MOMENT !!!!"

What about a client-server database written in [x]Harbour with underlying dbfs. You can do it on any platform that supports [x]Harbour and IP sockets (hbtip library)?
It can understand and directly update [x]Harbour objects. Doesn't even need any code written in C. Otherwise you are traversing [x]Harbour code, a binding from [x]Harbour to database client software (usually a mixture of [x]harbour code and C code) and the client software itself and you will have to map the returned data to your object's properties on return. And SQL isn't exactly standardised (a standard exists but vendors don't fully comply).

At least that has been my solution.

Regards
xProgrammer

[George]

Antonio,
I agree 100% with Marco:

We urgently need FWH embedded class to manage DIRECTLY and in easly way SQL database like Sqlite (it is coming a standard database for stand-alone/small network) and Firebird (for larger network/web solution) without the use of ODBC, ADO and other interfaces !!!
That's all. Any opinion appreciated.

It's critical to our future, as [x]Harbour + FWH programmers, moving ASAP from DBF to a more standardized database.

Regards,

George