Both are fixed. the test program I sent now works with any number of messages. I expect a fully working IDE by the end of the week, thank you.

:-)

Michael.

Cliff Baeseman wrote

I have to figure out how to cycle a compiler on this win box...Oh no... This should be fun...

With the provided gnuutils installed, it should be limited to a 'make cycle'. At least it is on my machine.

Michael.

Thanx Jeff please do It might help me out I am having no luck so far. ;(

Cliff

I have to figure out how to cycle a compiler on this win box...Oh no...

This should be fun...

Cliff

Peter Vreman wrote

The crash is fixed.

Do you mean the crash from the exception error or the multiple messages in a class problem?

Cliff Baeseman wrote

I have to figure out how to cycle a compiler on this win box...Oh no... This should be fun...

Believe me, it is! But I don't know if what I have will help. I have ppc386 in an odd directory compared to probably everone else (\pp\bin\win32) but I'll send you my rebuild.bat if you like.

Jeff.

The crash is fixed.

Peter

Both are fixed. the test program I sent now works with any number of messages.

As does the current Lazarus source. :-) Thank you very much. :-)

I expect a fully working IDE by the end of the week, thank you.

Wellllll....that might be pushing it juuuuuust a little bit. ;-)

However I do expect a very robust set of working components by the end of the week.

We do need to get moving though so we can dampen Borlands spirit. :-) I'd love for use to be able to have a working GUI IDE at the same time that 1.0.0 of FPC is ready.

I do really think we should comment out the ide implementation for a little while.

We should create main form at about 400 X 400 and test the components there.

Button probably already works but we have all of the other stuff to implement. I have to check out the code and see where we are right now...

Cliff

Hi!

First of all...

I've been around here for some weeks now but kept quiet because I wanted a more stable version first. Now things seem to get better and as long as my spare time allows it I'll throw in some comments and code snippets every now and then.

Cliff Baeseman wrote

I do really think we should comment out the ide implementation for a little while.

You're absolutely right.

We should create main form at about 400 X 400 and test the components there.

I'd prefer an approach with a simple separate application for every component. This will allow developers of different components more flexibility, since they can ignore a broken component then.

bye,

Stefan

PS: My first code will be a TGroupBox class. I've already done that but that's about 3-4 weeks ago. I'll bring it up to date in the next 2 days and send patches(*) then.

(*) I'm going to send them to Shane, tell me if I should send them to somebody
    else.

Sending to Shane or myself would be just fine...

Thanx for your help...

Cliff

Compiler/rtl problems could be mailed to the fpc-devel mailinglist. But i'll take a look at it, it looks like a constant ansistring->pchar conversion problem.

Here you go. I hope it helps.

Jeff.

*********** REPLY PARTITION ***********

-------------- next part --------------
A non-text attachment was scrubbed...
Name: bin00001.bin
Type: application/octet-stream
Size: 382 bytes
Desc: "rebuild.bat"
Url : http://localhost/pipermail/lazarus/attachments/19990711/87520c99/bin00001.bin

Does anyone know who did the GTK conversion? The reason I am wondering is that I think I may have found a bug.

I found some method calls that take a Pgchar as input and if you pass a long string to them, they get a bunch of errors and don't work. If you typecast the string to a ShortString, they work perfectly. Yet other procedures that take a Pgchar work with long strings just fine.

For example, here is what I am talking about.

mmiFile := gtk_menu_item_new_with_label ('');
tmp_key :=gtk_label_parse_uline(GTK_LABEL(GTK_BIN(mmiFile).child), '_File');
gtk_widget_add_accelerator(mmiFile, 'activate_item', accel_group,
                      tmp_key, gint(GDK_MOD1_MASK), TGtkAccelFlags(0));

All three of these functions require a Pgchar to be passed in (the params with single quotes). When I had {$H-} at the top of the unit, everything worked fine. As soon as I changed it to {$H+}, the first two functions kicked out a bunch of errors and compiled just fine. They just kicked out all kinds of GTK errors when I ran the program and they didn't work. Then when I changed them to the following, they work perfectly again.

mmiFile := gtk_menu_item_new_with_label (ShortString(''));
tmp_key := gtk_label_parse_uline(GTK_LABEL(GTK_BIN(mmiFile).child),
ShortString('_File'));

Notice the ShortString in them? The third function did not need to be changed at all. It continued to work just fine even with long strings. I can do the typecast to fix it for now, but this seems like it is wrong. I checked the functions in the gtk files and cannot find anything wrong. So I don't know who to report this to. Does anyone else have any ideas how to fix this or who to report it to?

Curtis