Feasibility Assertion - FiveQt

[Pritpal Bedi]

Hello Everybody

Off late I am compelled to think seriously about porting Five* library constructs to HbQt, but...

Rationale behind this initiative:

At the begining of HbQt evolution, I wanted a proven application framework which would go
side-by-side and also become test-bed for HbQt, and tilted towards XBase++ because of certain
reasons I explained many times elsewhere. I succeeded to a large extent but for no-use. Nobody
from XBase++ community participated and gave feedback etc. The initiative turned out to be
completely useless. The 2nd on that priority list was FiveWin but I could not ever get inner support
to go this way because I have never been a user of FiveWin.

Having said that, I am now convinced that porting Five* to HbQt will address much wider audience
then one could expect because 1) FiveWin consumes same compiler, 2) FiveWin community is
already immune to GUI, and 3) FiveWin community is extremely expressive and broad.

But...

Before starting in this direction I want to build a consensus about its viability.
I do not want to face humiliation of non-acceptability of efforts as the only reward coming
my way will be the satisfaction that my efforts will be useful to many of my fellow Harboureans.

So, here are the doubts I confront, please reply...

1. Does FiveWin developers think it right to have this solution ? [ Extremely Important ]
2. Does FiveWin community wants this to happen ? Your day-to-day support will be vital. [ Very Important ]

I have more questions but for now the reply to above two will build the foundation for
setting-up my mind in this direction.

DISCLAIMER: I do not intend to have any financial returns out of this venture. The only reward
I do expect that you will try to use this port on your existing applications.


Pritpal Bedi
a student of software analysis & concepts

[elvira]

Dear Pritpal,

That sounds very exciting!!.

What OS will support FiveQT?.

I think FiveQT will extend your apps to Mac and Linux with same sintax as Fivewin

So, instead of using those commands:

Code: Select all  Expand view

LOCAL oWnd

   oWnd := QMainWindow()
   oWnd:setWindowTitle( "Title of the Window" )
   oWnd:resize( 300, 200 )

   oWnd:show()
   QApplication():exec()

   RETURN

 

We will use Fivewin´s sintax?:

Code: Select all  Expand view

unction Main()

   local oWnd

   DEFINE WINDOW oWnd FROM 3, 6 TO 20, 70 ;
      TITLE "Welcome to Fivewin!!!"

   @ 2, 2 SAY "Hello world!"
   
   ACTIVATE WINDOW oWnd
 

[TimStone]

Pritpal,

I'm not sure what this project would include. I do have a few thoughts for the discussion:

Antonio has long been a supporter of Harbour and encouraged us to use Harbour with our projects
Some of us now have Windows programs running that are built with Harbour / FWH / and Visual Studio. I am now able to build a very large and complex application with VS 2013.
Anything that more closely links the capabilities of Harbour and FWH benefit us all.

With that said, many people still depend on Borland or xHarbour for their projects with FWH. What they have built is large and complex so getting them to move to a more unified platform may be difficult.

People working together is always a benefit. I will be interested in seeing Antonio's thoughts.

Tim

[Antonio Linares]

I fully respect Pritpal's ideas, but I don't share them at all.

I don't see any point at all to use QT, when we can directly talk to the Windows API and get the fastest, self-contained EXEs that don't need any intermediate layer at all.

If we already have the best speed, simpler code and direct communication to the Windows API, why should we want to put something in between our apps and the Windows API ?

To me, Pritpal keeps dreaming about a unique solution for all of us, and that is something that doesn't even exist in nature: Software evolution means diversity, exactly the same way that nature evolves, and not the oposite way. Look at the way that software evolved along the years: diversity and natural selection (the best tools survive) exactly the same way mother nature evolves. So that idea is simply against natural evolution. It is like dreaming of a unique language to be used (see the reality: more and more languages are appearing constantly).

And what tool has proved to survive so well along all these years for all of us ? FiveWin

[Antonio Linares]

More over, I already published here some code to emulate FiveWin using QT (it seems as Pritpal is not aware of this) and nobody paid attention to it at all.

Just download QT, install it, try it, and you will get back to FiveWin is less than one hour. Why should we want a heavier and slower app that what we already have ?

Pritpal already tried this with his hb_ide. Who is using hb_ide here (that I promoted here on this forums, and even I provided feedback to Pritpal) ?

I would say that nobody at all

[elvira]

Antonio,

And could be FiveQT used in other OS rather than Windows?.

Of course, the best for Windows is Fivewin!!!.

But what about Mac, Android, iPhone, etc?.

[Antonio Linares]

I even documented it here:

http://wiki.fivetechsoft.com/doku.php?id=fwqt_english

It was also published in these forums, (I don't find them now), and got no interest at all.

[Antonio Linares]

Here are my posts in these forums about it:

viewtopic.php?f=6&t=24375

[Antonio Linares]

Elvira,

But what about Mac, Android, iPhone, etc?

We have been providing FiveLinux, FiveMac, FiveDroid, FivePhone, etc. since many years ago, and the interest about them it has been really low, what made me understand along these years, that Windows (like it or not) owns the 95% or more of the bussiness applications market, and the other platforms have very little interest for bussiness applications.

Linux users want everything for free (I am talking generally), Mac is not used for bussiness apps (most apps are multimedia, but no bussiness ones), and regarding the iPhone and Android, to me the apps should be developed using Apple and Google official development tools and using other languages is to go against the flow, because they are so different in the way that they are built, that it has very little sense to use a different approach (please review my FiveDroid development, how complex it is: https://code.google.com/p/fivedroid/wiki/GUI_editing_with_Android_and_Harbour)

In all those products, I never wanted to use any intermediate layer, because nothing is as fast and gives you such level of control as talking directly to each GUI API.

As an example look at Java apps: they tried for many years to be a standard solution for all the different operating systems. If you review a Java app, most of the time they look really ugly, because they don't use each operating system own GUI controls. So it is a layer on top of the APIs and thats quite absurd to me as we simply don't need it.

[Antonio Linares]

To me, Pritpal should open a forum of his own, and promote there his QT ideas instead of coming here to try to convince FWH users to support him, and confusing FWH users.

And I have nothing personal against Pritpal, to me its simply that he keeps trying the same for some years to no avail...

Pritpal, if you are so happy and convinced about QT, why do you need to come to these forums trying to get followers ? I don't go to your forums (not sure if you have them) trying to get FWH users...

[TimStone]

I really don't know anything about QT. My comments were simply posted to say that all work which enhances the FWH / Harbour / MSVC integration is good.

There have been so many enhancements over the years and they give us many strong capabilites. I am grateful for them.

My focus is business applications. I spent a lot of time looking at tablets, and other operating systems. I agree with Antonio's analysis. If someone has an Apple environment, they can run a Windows application using Parallells. For the next few years I do not see Apple moving into the business environment because they simply do not have the server technology to handle the constant demands of record locking and unlocking while data is added, edited, or removed. Tablets are really too small for a fully integrated, comprehensive business application. Apple and Android devices simply do not communicate well with intricate data sources. They are designed for simple tasks, document reading, and perhaps updating of a few fields.

My primary application has a section to maintain workorders. The main screen uses tabed dialogs, each with a browser and data editing fields. There are about 300 data fields, with 6 browsers, and 15 ancillary popup tasks that are themselves complex. There is no way this could be handled on a tablet screen. I've tried. I can run it on my Surface Pro 3, but the screens are small so the result is a potential for a much higher user error rate.

With FWH, Harbour / MSVC and Windows, my clients have their needs met completely. With other resources, that cannot happen. My point is that if we know what we do well, and focus on continuing to enhance that, we don't have to try and build systems on all platforms. It is just not practical.

Tim


Sent from my iPad using Tapatalk HD

[Antonio Linares]

Tim,

+1

[hmpaquito]

Hello Pritpal,

Today fivewinners day for developing for Win32 API, FWH is the best tool.

Where accommodates your FiveQt then? I see if it will allow to create programs for Android. Market for Android is bigger than win32 market. Android applications are here and YES are for bussiness. There is who wants to deny the reality.

Fivedroid from Fivetech was only a taste test, it was never something solid. Maybe soon ... I was not know it. Qt has many rich controls, perhaps not known to many Win32 developers.

There is talk of layers ... Parallels on Mac is THE BIG LAYER, or not? Today, on the computer are all layers. Harbour vm is a layer on C language. Java is a layer on OS. . .Net is a layer on Net machine. Fwh is a layer on Win32 API and OS. All layers are. Layers are not important if they run fast and transparent.

My 2 cents.

[Antonio Linares]

Paco,

Thats exactly the reason why I always install Windows on Apple natively (using bootcamp) and never used an intermediate layer. I don't like Parallels at all

I have used VirtualBox sometimes but just to do something quite specific and thats all. If you can have the best speed, why would you accept to use a slower solution ?

Applications speed is very important. We have seen slow apps as dogs developed in VB, Java, .NET, etc. Look at a Harbour + FWH speed, thats real fast

Recently I did a large demo for a software development company, and when they saw Harbour + FWH apps speed they were really impressed.

Thats my advise, but of course, everyone is free to choose his own path and start building slow and fat apps... (I will not do it)

[elvira]

Hello,

The only purpouse I see interesting for FiveQT is in Android, iPhone, Mac, etc, but not in Windows, as Fivewin is great and very fast.

Or, to give a solution, Pritpal could join Fivetech team to develop FiveMac, FiveLinux, etc?. He is a master too in Harbour and in GUIs and his tallent would be useful for our community.

What do you think?.