Lazarus 1.6 - 1st Release Candidate

[georgebou]

Good to hear it guys.
Keep up good work

[Jurassic Pork]

hello,
thanks for your work ! 
i have a big problem when i try to install the packages easydockmanager and easydockmanagerdesgn

a window with an access violation  message appears when lazarus try to restart 

Lazarus 1.6 RC1 windows 32      on windows 7     french lang choosen

i can't start lazarus  any more.

[regs]

and same problem as with typhon - can't switch into app if integrated designer is active. will only switch if source is active.
no matter if clicking taskbar or using alt-tab. at least on windows.

[Ondrej Pokorny]

i have a big problem when i try to install the packages easydockmanager and easydockmanagerdesgn

easydockmanagerdsgn isn't maintained anymore and cannot be used in the IDE any more.

[Jurassic Pork]

hello,
thanks ondrejpokorny !
i made a big mistake      , i wanted to use anchordockingdsgn not easydockmgrdsgn and easydockmgr   
anchordockingdsgn is OK   

maybe package easydockmgrdsgn should be remove from the list of installable packages of lazarus.

[Ondrej Pokorny]

maybe package easydockmgrdsgn should be remove from the list of installable packages of lazarus.

Yes, good point. I'll open a thread about it in mailing list.

[SnoopyDog]

Is this thread the right place to report about bugs and/or problems?

[Ondrej Pokorny]

Is this thread the right place to report about bugs and/or problems?

IMO yes, you can ask here first but if problems/bugs are confirmed, you'll then have to report them in the bug tracker as well.

[abonic]

Paint bug in DBGrid while editing the cell

OS: Windows 10

Steps to reproduce:

Press F2 in some cell in DBGrid control. Text shifts for pixel or two. Left and top border of the cell are blue (highlight color). Look really unprofessional.

[Nick]

Please, anybody, check equality of fpc.exe in lazarus \fpc\ folder and in fpc release.
Seems that they are different. And many other files too.

1) platform/OS?

2) They may be build with different optimization or similar. So they may not be the same binary. They are (or should be) from the 3.0 tag on the fpc svn. So any source file included should be the same as in the fpc release.

1. win7, win32
2. I don't know how they are builded. But most of exe files in "fpc" folder of released lazarus and released fpc are different.
And moreover, they produces different executables (I try empty project with one form).
So, what's the difference?

[lazie]

TDBGRID is not triggering all ONSELECTEDITOR-events

Got a problem with this new version. - I am using TDBGRID and its ONSELECTEDITOR-event to use TEDIT as a replacement for the grids cell-editor. - Its working fine with prior versions (The position of the editor-rect ist given from - DbGrid.SelectedFieldRect).

The NEW VERSION does NOT fire the ONSELECTEDITOR-event when entering a cell by KEYBOARD. But everything ist allright when entering a cell by MOUSE.
When a cell is entered by keyboard the onselecteditor event is ONLY triggered - after one has CHANGED the cell-text (thats not enough ;-)).

With Lazarus 1.4.4 its allright:
ONSELECTEDITOR is fired - EACH TIME when EDITORMODE is going to true (one is editing).

Where should the bug being reported ?
I am using Debian Linux, ZEOSDBO-7.1.4.

[Martin_fr]

2) They may be build with different optimization or similar. So they may not be the same binary. They are (or should be) from the 3.0 tag on the fpc svn. So any source file included should be the same as in the fpc release.

1. win7, win32
2. I don't know how they are builded. But most of exe files in "fpc" folder of released lazarus and released fpc are different.
And moreover, they produces different executables (I try empty project with one form).
So, what's the difference?

I don't know what the default settings are used for the installer released by the fpc team, so I can't say.

The build for the Lazarus install is done by lazarus\tools\install\win\build-fpc.bat (I have no idea who wrote this, its been around a long time). I also don't know why a special build script was needed, but maybe there were some differences intended.

My first guess would be to look at the OPT=
Any option that is passed in, if it is not used by default, is likely to cause a difference.
-Ur only marks ppu as release (fpc will not rebuild, even if src is edited) So no diff expected
-CX -XX is smartlinking. That may well cause some differences.

Yet that happens after the compiler was build. So I don't know what affects the compiler. It may be that OPT (that I do not know) were passed when the fpc team did the build.

---
Also check the fpc.cfg

--

EDIT:
http://lists.freepascal.org/pipermail/fpc-devel/2015-December/036438.html
http://lists.freepascal.org/pipermail/fpc-devel/2015-December/036439.html

> What options are specified for the release by fpc?

None. The top level Makefile automatically adds RELEASE=1, which enables
a default set of options built into all Makefiles.

I am not at current going to search for it myself. But the next step is to read the makefile and see what those options are.

[regs]

After setting docking and setting some most used panels, at one sudden lauch i got dozens, if not not hundreds, of WMSize loop detected - Widgetset does not like LCL bounds or sends unneeded wmsize messages error for nearly every single IDE component. BoundsRealized were identical to NewBoundsRealized.

And it wouldn't restore maximized with Win32.

IDE Builder doesn't take into account LCL widget type selected in Configure Build Lazarus window and keeps building with Win32 widgetset. Putting LCLqt or LCLgtk2 into defines makes it work. With qt it seems restores into maximized mode.

AnchorDocking wouldn't remember tab position.

[bill0287]

This database (tdbf) bug still seems to be unfixed since 1.2.6. How long before it gets into the codeline?

http://bugs.freepascal.org/view.php?id=27959

[Nick]

2) They may be build with different optimization or similar. So they may not be the same binary. They are (or should be) from the 3.0 tag on the fpc svn. So any source file included should be the same as in the fpc release.

1. win7, win32
2. I don't know how they are builded. But most of exe files in "fpc" folder of released lazarus and released fpc are different.
And moreover, they produces different executables (I try empty project with one form).
So, what's the difference?

I don't know what the default settings are used for the installer released by the fpc team, so I can't say.

The build for the Lazarus install is done by lazarus\tools\install\win\build-fpc.bat (I have no idea who wrote this, its been around a long time). I also don't know why a special build script was needed, but maybe there were some differences intended.

My first guess would be to look at the OPT=
Any option that is passed in, if it is not used by default, is likely to cause a difference.
-Ur only marks ppu as release (fpc will not rebuild, even if src is edited) So no diff expected
-CX -XX is smartlinking. That may well cause some differences.

Yet that happens after the compiler was build. So I don't know what affects the compiler. It may be that OPT (that I do not know) were passed when the fpc team did the build.

---
Also check the fpc.cfg

--

EDIT:
http://lists.freepascal.org/pipermail/fpc-devel/2015-December/036438.html
http://lists.freepascal.org/pipermail/fpc-devel/2015-December/036439.html

> What options are specified for the release by fpc?

None. The top level Makefile automatically adds RELEASE=1, which enables
a default set of options built into all Makefiles.

I am not at current going to search for it myself. But the next step is to read the makefile and see what those options are.

fpc.cfg is the same.
When I change path to fpc.exe from D:\Develop\Lazarus\fpc\3.0.0\bin\i386-win32\fpc.exe to C:\FPC\3.0.0\bin\i386-win32\fpc.exe in IDE settings (without any other changes), then compiler produces different exe (I press Shift-F9).
There are 3 bytes different:
0014CF71: 31 32
0014CF73: 31 30
0014CF74: 36 37

When I set path back to D:\Develop\Lazarus\fpc\3.0.0\bin\i386-win32\fpc.exe and press Shift-F9, then this 3 bytes reverts back:
0014CF71: 32 31
0014CF73: 30 31
0014CF74: 37 36