HMGforum.com

bedipritpal wrote: it seems HMG is still using older copy of hbQT, right ?
Update to the latest before I can provide any help.

Pritpal,

I've delayed the HMG binaries update and source code conversion, since I'm waiting for all structural changes that you are doing to HBQT be finished.

I could be wrong, but reading the HBQT related threads at Harbour developers list, appears to be that there are some things to be changed yet.

But, if the pending HBQT changes will have minimal or no impact in HMG sources, we could start the conversion right now.

Roberto Lopez wrote:Pritpal,

I've delayed the HMG binaries update and source code conversion, since I'm waiting for all structural changes that you are doing to HBQT be finished.

I could be wrong, but reading the HBQT related threads at Harbour developers list, appears to be that there are some things to be changed yet.

But, if the pending HBQT changes will have minimal or no impact in HMG sources, we could start the conversion right now.

It is not wise to postpone the changes with an attitude that let it be finalized first.

The changes, which were done, and which will be done, were never thought of beforehand.
More changes, more complex it will be to fix the bugs, if any comes after.
Change in hbQT structures are still on the cards, but how and what, is not knows.

So my humble advise is, adopt to the current one. You might be having some
regressions which have to be sorted out _now_.

The changes to current structure, from PRG callable point, will be minimal in future.
All will be there on class levels which may not affect external libs.

[quote="mrduckIt's not easy to have hmg ported to the last hbqt... quite a lot of changes must be done, some mechanical, some not...[/quote]

Does it mean HMG should never be get ported and stay with outdated version ?

Roberto Lopez wrote: I've delayed the HMG binaries update and source code conversion, since I'm waiting for all structural changes that you are doing to HBQT be finished.
I could be wrong, but reading the HBQT related threads at Harbour developers list, appears to be that there are some things to be changed yet.
But, if the pending HBQT changes will have minimal or no impact in HMG sources, we could start the conversion right now.

1. Commit all what you have not committed, and all others.
2. Give me write access to your SVN, SF id is "vouchcac".
3. I will adopt it to current hbQT.
4. The process will reformat the whole HMG sources - tabs will be removed, etc.
5. The HMG build process, plus its examples, will have hbIDE compatible .hbp files.
6. HMG will be provided with hbIDE compatible setup ( at least for my workings ).
7. You can continue with batches if that is preferred.

bedipritpal wrote: So my humble advise is, adopt to the current one. You might be having some regressions which have to be sorted out _now_.

The changes to current structure, from PRG callable point, will be minimal in future.

Ok.

Let's do it.

bedipritpal wrote: 1. Commit all what you have not committed, and all others.

Currently only Rathinagiri and me are active and I've committed all pending changes.

So, Rathinagiri, if you have some commit pending, please do it.

bedipritpal wrote: 2. Give me write access to your SVN, SF id is "vouchcac".

Done.

bedipritpal wrote: 3. I will adopt it to current hbQT.

Thanks!

bedipritpal wrote: 4. The process will reformat the whole HMG sources - tabs will be removed, etc.

Ok. That is a really needed thing.

bedipritpal wrote: 5. The HMG build process, plus its examples, will have hbIDE compatible .hbp files.
6. HMG will be provided with hbIDE compatible setup ( at least for my workings ).
7. You can continue with batches if that is preferred.

Ok.

Roberto Lopez wrote: 2. Give me write access to your SVN, SF id is "vouchcac".

Done.

Thanks.

I will begin a few hours later.
So please commit any pending changes.
Do nothing until I am finished level one.

In future:

Every developer of HMG will have to respect
certain formatting ethics, very important, and
you will have to enforce it strictly.
I will layout the rules later.

bedipritpal wrote: I will begin a few hours later.
So please commit any pending changes.
Do nothing until I am finished level one.

Ok.

bedipritpal wrote: In future:

Every developer of HMG will have to respect
certain formatting ethics, very important, and
you will have to enforce it strictly.
I will layout the rules later.

I understand the importance of a good code formatting.

In fact, is difficult to me, even read code that is not properly indented and organized.

But you must understand that many of developers here (including me) are simply xBase application developers doing our best efforts and sometimes without previous experience on open-source/team work.

Since that and considering that we are an independent (little) project, we can have some more lax rules.

As of now I had already committed all my codes and I am just awaiting for the new code.

Regarding formatting, yes we shall obey the rules for the betterment of the project. Thanks for your great concern Pritpal.

Pritpal,

I've built Harbour from current binaries and started to do some testing.

I've found a problem with demos using menupopup class..

The error is:

ERROR HBQT/1001 ARGUMENT ERROR
...
CALLED FROM MENUPOPUP:ADDITEM(61)

The code is:

METHOD AddItem( oItem ) CLASS MENUPOPUP
LOCAL oQAction

If ValType( oItem ) = 'U'
oItem := ::oDefaultMenuItem
EndIf

oQAction := ::oQTObject:addAction() <------- line #61

I hope this info helps you.