Re: [lazarus] how to test my changes in the lcl in a small app?
Alexandre Leclerc napisał(a): I did made some changes to the lcl, but would like to test before posting. But I don't see any effets when compiling my app. There is something I'm missing? do I have to: make clear ? make lcl ? I have open one terminal window and after any changes in LCL do cd /usr/share/lazarus //only firmst time cd lcl //its faster compiling only LCL make and then compiling own application. Its better to start new application: one form with one control to test. Darek _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Who needs Unicode controls.
A.J. Venter wrote: One problem, however, is what to do on Windows 9x and on Gtk 1. We would need to convert utf-8 to iso. But to which iso? A easy solution is to just convert to western european iso. Actually there must be ways of getting the current running environment's preferred character set. Under Linux you could use the LANG environment variable to determine what language the desktop is in, and call the best character set for that, surely the windows API must provide something similar ? GetACP. WideCharToMultiByte also has a 'to ANSI' flag. Micha _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Speedbutton property transparant default value should be false.
Delphi compatibility.Delphi compatability is not the issue here I think. The strange thing is that speedbuttons don't have borders in the designer, but at runtime the borders show when the mouse is over the speedbutton. I think speedbuttons are also used in menus where you don't want to have bordres around them, but when you have a row of speedbuttons somewere else, the user might want to have borders around them.I also noticed that the buttons are transparent even when transparancy is false. I think transparancy somway interferes with the borderstyle of the buttons.Shall I report this as a bug?Marien
Re: [lazarus] how to test my changes in the lcl in a small app?
On Fri, 19 May 2006 08:24:47 +0200 DarekM [EMAIL PROTECTED] wrote: Alexandre Leclerc napisa__(a): I did made some changes to the lcl, but would like to test before posting. But I don't see any effets when compiling my app. There is something I'm missing? do I have to: make clear ? make lcl ? I have open one terminal window and after any changes in LCL do cd /usr/share/lazarus //only firmst time cd lcl //its faster compiling only LCL make and then compiling own application. Its better to start new application: one form with one control to test. Right. Or use the Tools - Configure build lazarus to only build the LCL. At the moment the IDE does not check the LCL for changes, so it the LCL is never recompiled automatically. Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Show hint from code
On Fri, 19 May 2006 02:25:16 +0200 A.J. Venter [EMAIL PROTECTED] wrote: [...] Did you make any changes ? I'll test mine tomorrow (it's very late here) No, I did a svn update, then I went to the command line, and did make clean all Doesn't work here. What doesn't work? Next I recompiled only LCL for Gtk2 (from the IDE), and it worked. My lazarus hangs when I do that, Can you reproduce that? If yes, please start the IDE in gdb and create a backtrace. I have to rerun make, then rebuild with gtk1, gives me enough gtk2 for my apps to be compiled - but it might prevent the GTK2 def from being passed to trayicon ? If so that's a bug either in lazarus or in ttrayicon, I'm not sure. On a hunch I opened the ttrayicon.lpk file, changed it's compiler options to use gtk2, reinstalled, rebuilt wole for gtk2 - same result. Same stack trace as well. So it doesn't matter what the component's compiler options say - which makes sense, it ought to receive it from the parent app. But it also ought to do so regardless of whether the LCL is using the same widget set as the project target. When the IDE compiles itself (LCL + IDE + installed packages) it compiles for the target set in the configure build lazarus dialog (Target cpu, os, widgetset). Even if the project has another target. This has been fixed a few weeks ago, so it does not always work with 0.9.14. I reckon I should fix the lazarus issue first but I am stumped as to what could be causing it. Even a gdb backtrace didn't tell me anything usefull, just that it's pango that crashed. Lazarus's own backtraces are talking bull, complaining about crashes in whichever component was the last to be added, remove the component and it just moves one step back in history - e.g. it doesn't show me what is REALLY wrong. Mattias you wouldn't have any theories perhaps ? If I have any idea where to look for the cause of the bug I would be happy to fix it. Where are the crashes? In the IDE? Which widgetset? Which components installed? Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Speedbutton property transparant default value should be false.
On Fri, 19 May 2006, Marien van Westen wrote: Delphi compatibility. Delphi compatability is not the issue here I think. The strange thing is that speedbuttons don't have borders in the designer, but at runtime the borders show when the mouse is over the speedbutton. This is standard behaviour for a 'flat' look. Michael. _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Speedbutton property transparant default value should be false.
On Fri, 19 May 2006 09:01:50 +0200 Marien van Westen [EMAIL PROTECTED] wrote: Delphi compatibility. Delphi compatability is not the issue here I think. I meant: for Delphi compatibility Transparent = true is the default. The strange thing is that speedbuttons don't have borders in the designer, but at runtime the borders show when the mouse is over the speedbutton. I think speedbuttons are also used in menus where you don't want to have bordres around them, but when you have a row of speedbuttons somewere else, the user might want to have borders around them. I also noticed that the buttons are transparent even when transparancy is false. I think transparancy somway interferes with the borderstyle of the buttons. Shall I report this as a bug? How does it work under Delphi? Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] GTK2 Bug - was Show hint from code
Doesn't work here. What doesn't work? TrayIcon gets compiled with wsgtktrayicon instead of wsgtk2trayicon if the project is set to use gtk2 while the IDE is running gtk1. Next I recompiled only LCL for Gtk2 (from the IDE), and it worked. My lazarus hangs when I do that, Can you reproduce that? If yes, please start the IDE in gdb and create a backtrace. Reproduce ? I cannot get avoid it :) but we are talking about two sepperate bugs here - we shouldn't ASSUME a relationnship Backtrace attached. This happens on every machine I have access to and has been for about 2 weeks now. . I have to rerun make, then rebuild with gtk1, gives me enough gtk2 for my apps to be compiled - but it might prevent the GTK2 def from being passed to trayicon ? If so that's a bug either in lazarus or in ttrayicon, I'm not sure. On a hunch I opened the ttrayicon.lpk file, changed it's compiler options to use gtk2, reinstalled, rebuilt wole for gtk2 - same result. Same stack trace as well. So it doesn't matter what the component's compiler options say - which makes sense, it ought to receive it from the parent app. But it also ought to do so regardless of whether the LCL is using the same widget set as the project target. When the IDE compiles itself (LCL + IDE + installed packages) it compiles for the target set in the configure build lazarus dialog (Target cpu, os, widgetset). Even if the project has another target. This has been fixed a few weeks ago, so it does not always work with 0.9.14. I always run up to date SVN. Either way - this is the other bug - I am happy to work on both but the gtk2 ide bug is much more severe. I reckon I should fix the lazarus issue first but I am stumped as to what could be causing it. Even a gdb backtrace didn't tell me anything usefull, Where are the crashes? In the IDE? Which widgetset? Which components installed? IDE starts up, splash screen shows. Then the lazarus editor window comes up, then the source editor window appears - and then the whole thing freezes solid - have to xkill it. 1) This only happens with GTK2 as the IDE widgetset 2) This happens REGARDLESS of what components are installed - even on a brand new fresh install with no .lazarus in existence yet. I also attach an LDD on the lazarus executable. GTK2 version is: gtk+2-2.8.7 But I have the same problem on other versions both older and newer. Ciao A.J. -- 80% Of a hardware engineer's job is application of the uncertainty principle. 80% of a software engineer's job is pretending this isn't so. A.J. Venter Chief Software Architect OpenLab International http://www.getopenlab.com | +27 82 726 5103 (South Africa) http://www.silentcoder.co.za| +55 118 162 2079 (Brazil) bash-3.00$ gdb ./lazarus GNU gdb 6.3 Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type show copying to see the conditions. There is absolutely no warranty for GDB. Type show warranty for details. This GDB was configured as i486-slackware-linux...Using host libthread_db library /lib/libthread_db.so.1. (gdb) run Starting program: /home/silentcoder/lazarus/lazarus [Thread debugging using libthread_db enabled] [New Thread 16384 (LWP 10199)] (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, (lazarus:10199): Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, TApplication.IconChanged - TODO: convert this message...no implementation in gtk or win32 NOTE: editor options config file not found - using defaults NOTE: codetools config file not found - using defaults NOTE: help options config file not found - using defaults TMainIDE.DoNewProject A TMainIDE.DoNewEditorFile A NewFilename= TPascalParserTool.BuildTree B OnlyIntf=False project1.lpr Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 16384 (LWP 10199)] 0x405f83b2 in memset () from /lib/libc.so.6 (gdb) backtrace #0 0x405f83b2 in memset () from /lib/libc.so.6 #1 0x40838f47 in gray_render_span () from /usr/lib/libfreetype.so.6 #2 0x40839022 in gray_hline () from /usr/lib/libfreetype.so.6 #3 0x408391c0 in gray_sweep () from /usr/lib/libfreetype.so.6 #4 0x408396d7 in gray_convert_glyph () from /usr/lib/libfreetype.so.6 #5 0x408055b0 in FT_Outline_Render () from /usr/lib/libfreetype.so.6 #6 0x40805663 in FT_Outline_Get_Bitmap () from /usr/lib/libfreetype.so.6 #7 0x406e1e9a in cairo_pattern_get_extend () from /usr/lib/libcairo.so.2 #8 0x406e2971 in
Re: [lazarus] GTK2 Bug - was Show hint from code
On Fri, 19 May 2006 11:53:19 +0200 A.J. Venter [EMAIL PROTECTED] wrote: Doesn't work here. What doesn't work? TrayIcon gets compiled with wsgtktrayicon instead of wsgtk2trayicon if the project is set to use gtk2 while the IDE is running gtk1. When the project is set to gtk2 and you compile it, the IDE checks all packages. The package should be compiled with LCLWidgetSet macro set to the value of the project (gtk2). If not, then you spotted a bug. In case of TTrayIcon: This component differs between gtk1 and gtk2. But the output directory is the same. I added the LCLWidgetSet macro to the path of the trayicon package, so it won't be recompiled everytime you switch the widgetset. Next I recompiled only LCL for Gtk2 (from the IDE), and it worked. My lazarus hangs when I do that, Can you reproduce that? If yes, please start the IDE in gdb and create a backtrace. Reproduce ? I cannot get avoid it :) but we are talking about two sepperate bugs here - we shouldn't ASSUME a relationnship Backtrace attached. This happens on every machine I have access to and has been for about 2 weeks now. . Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, This means: your theme uses abilities you don't have. Either install another theme, or install the smooth engine. I have to rerun make, then rebuild with gtk1, gives me enough gtk2 for my apps to be compiled - but it might prevent the GTK2 def from being passed to trayicon ? If so that's a bug either in lazarus or in ttrayicon, I'm not sure. On a hunch I opened the ttrayicon.lpk file, changed it's compiler options to use gtk2, reinstalled, rebuilt wole for gtk2 - same result. Same stack trace as well. So it doesn't matter what the component's compiler options say - which makes sense, it ought to receive it from the parent app. But it also ought to do so regardless of whether the LCL is using the same widget set as the project target. When the IDE compiles itself (LCL + IDE + installed packages) it compiles for the target set in the configure build lazarus dialog (Target cpu, os, widgetset). Even if the project has another target. This has been fixed a few weeks ago, so it does not always work with 0.9.14. I always run up to date SVN. Either way - this is the other bug - I am happy to work on both but the gtk2 ide bug is much more severe. Can you give exact steps how to reproduce that? I reckon I should fix the lazarus issue first but I am stumped as to what could be causing it. Even a gdb backtrace didn't tell me anything usefull, Where are the crashes? In the IDE? Which widgetset? Which components installed? IDE starts up, splash screen shows. Then the lazarus editor window comes up, then the source editor window appears - and then the whole thing freezes solid - have to xkill it. 1) This only happens with GTK2 as the IDE widgetset 2) This happens REGARDLESS of what components are installed - even on a brand new fresh install with no .lazarus in existence yet. I also attach an LDD on the lazarus executable. GTK2 version is: gtk+2-2.8.7 But I have the same problem on other versions both older and newer. Because I don't have the time to debug the gtk2 interface, my only advice is to use gtk1 for the IDE. Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, This means: your theme uses abilities you don't have. Either install another theme, or install the smooth engine. Yep, I know that one, though it should just skip them - either way, I also tested with the gtk2 standard theme - this had no impact. Can you give exact steps how to reproduce that? Exact steps I follow: svn co $LAZARUSSVNURL cd lazarus make ./lazarus Go to Tools|Configure Build Lazarus, click build all, set LCL to gtk2 Click ok. Go to Tools|Build Lazarus -problem occurs on the restart. This happens even on a fresh user with pretty much no settings at all for anything. Because I don't have the time to debug the gtk2 interface, my only advice is to use gtk1 for the IDE. Heh, fair enough - though technically I was offering to debug this one for you - all I need is a hint on where to look. A.J. -- 80% Of a hardware engineer's job is application of the uncertainty principle. 80% of a software engineer's job is pretending this isn't so. A.J. Venter Chief Software Architect OpenLab International http://www.getopenlab.com | +27 82 726 5103 (South Africa) http://www.silentcoder.co.za| +55 118 162 2079 (Brazil) _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
On Fri, 19 May 2006 12:32:13 +0200 A.J. Venter [EMAIL PROTECTED] wrote: Gtk-WARNING **: Unable to locate theme engine in module_path: smooth, This means: your theme uses abilities you don't have. Either install another theme, or install the smooth engine. Yep, I know that one, though it should just skip them - either way, I also tested with the gtk2 standard theme - this had no impact. Can you give exact steps how to reproduce that? Exact steps I follow: svn co $LAZARUSSVNURL cd lazarus make ./lazarus Go to Tools|Configure Build Lazarus, click build all, set LCL to gtk2 Click ok. Go to Tools|Build Lazarus -problem occurs on the restart. This happens even on a fresh user with pretty much no settings at all for anything. Because I don't have the time to debug the gtk2 interface, my only advice is to use gtk1 for the IDE. Heh, fair enough - though technically I was offering to debug this one for you - all I need is a hint on where to look. If I understood correct, you should first fix the gtk warnings about the theme. Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
Heh, fair enough - though technically I was offering to debug this one for you - all I need is a hint on where to look. If I understood correct, you should first fix the gtk warnings about the theme. Well no other GTK apps are affected by the missing bits, nonetheless I had done the test (easy way -just change to the GTK2 standard theme which ships with the library) , with or without the warnings however the bug persists. Even if it was theme related however, it would still constitute a bug - the code ought to fall-back if the theme has a problem. Since I know that it is in pango that the crash occurs, although the stack won't show me the exact place - I can from how much appears on the screen get a fairly safe guess that the crash is being caused by a font-change call from inside TSynEdit -so the question is , how do I determine which one so I can fix the code... A.J. -- 80% Of a hardware engineer's job is application of the uncertainty principle. 80% of a software engineer's job is pretending this isn't so. A.J. Venter Chief Software Architect OpenLab International http://www.getopenlab.com | +27 82 726 5103 (South Africa) http://www.silentcoder.co.za| +55 118 162 2079 (Brazil) _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
On 5/19/06, A.J. Venter [EMAIL PROTECTED] wrote: What doesn't work? TrayIcon gets compiled with wsgtktrayicon instead of wsgtk2trayicon if the project is set to use gtk2 while the IDE is running gtk1. I see now why it works for me, and not for you. I use trayicon directly, from code and I add the directory where it is to my project, so it´s really part of my software. The package needed some fixes Mattias pointed. thanks, -- Felipe Monteiro de Carvalho _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] Detecting user cancel
Is there a way that I can detect if a Lazarus application was cancelled by the windows task manager? For example: Does the Destroy event of the main occurs(And then I can intercept)? I want to detect this situation and do some special processing. Greetings from São Paulo - Brazil Ricardo _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] how to test my changes in the lcl in a small app?
Thanks Darek and Mattias. I just saw that my fix was introducing more problems than solving :) So I'll post an example of my findings if someone has better knowledge of the lcl. -- Alexandre Leclerc _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
On Fri, 19 May 2006 13:03:23 +0200 A.J. Venter [EMAIL PROTECTED] wrote: Heh, fair enough - though technically I was offering to debug this one for you - all I need is a hint on where to look. If I understood correct, you should first fix the gtk warnings about the theme. Well no other GTK apps are affected by the missing bits, nonetheless I had done the test (easy way -just change to the GTK2 standard theme which ships with the library) , with or without the warnings however the bug persists. Even if it was theme related however, it would still constitute a bug - the code ought to fall-back if the theme has a problem. Since I know that it is in pango that the crash occurs, although the stack won't show me the exact place - I can from how much appears on the screen get a fairly safe guess that the crash is being caused by a font-change call from inside TSynEdit -so the question is , how do I determine which one so I can fix the code... The stack is corrupted and the pango use in the gtk2 interface is incomplete. It could be anything. I didn't implement the gtk2 intf pango code, so I don't know, what should already work. I would start some simple gtk2 test apps and see what is working and what not. Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
I see now why it works for me, and not for you. I use trayicon directly, from code and I add the directory where it is to my project, so it´s really part of my software. The package needed some fixes Mattias pointed. Mmm, okay, that is indeed an option - will you be fixing the package ? I am looking at the notification window but it's slow going, I have never tried to do a tform descendent component before. I may have to leave it halfway and just create one built-in for my program with hookups where needed, and then later do an abstract component. A.J. -- 80% Of a hardware engineer's job is application of the uncertainty principle. 80% of a software engineer's job is pretending this isn't so. A.J. Venter Chief Software Architect OpenLab International http://www.getopenlab.com | +27 82 726 5103 (South Africa) http://www.silentcoder.co.za| +55 118 162 2079 (Brazil) _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] Re: Detecting user cancel
On Fri, 19 May 2006 15:00:38 +0330, Arí Ricardo Ody [EMAIL PROTECTED] wrote: Is there a way that I can detect if a Lazarus application was cancelled by the windows task manager? For example: Does the Destroy event of the main occurs(And then I can intercept)? I want to detect this situation and do some special processing. Greetings from S�o Paulo - Brazil Ricardo If you mean cancelled as kill task by windows task manager,no i really doubt that. becouse it is use is to force killling of a task,mostly when program is hung, you can be sure by checking your program with spy++ and see if it recieves any message, also maybe global hooking can do something for you. -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] GTK2 Bug - was Show hint from code
On 5/19/06, A.J. Venter [EMAIL PROTECTED] wrote: I am looking at the notification window but it's slow going, I have never tried to do a tform descendent component before. I may have to leave it halfway and just create one built-in for my program with hookups where needed, and then later do an abstract component. Are you having difficulties with anything specific? It should look like this: unit myunit; type MyClass = class(TCustomForm) private lblTitle: TLabel; lblText: TLabel; imgIcon: TImage; btnX: TButton; public constructor Create(AOwner: TComponent); override; destructor Destroy; override; published // publish some properties, others not end; then on create you have to create the components: constructor TMainWindow.Create(AOwner: TComponent); begin inherited Create(AOwner); BorderStyle := bsNone; lblTitle := TLabel.Create(Self); lblTitle .Left := 0; lblTitle .Top := 0; lblTitle .Width := ?; lblTitle .Height :=?; // Initializes the Glass component lblTitle.Align := alClient; lblTitle.Parent := Self; // Connects the methods to events vGlass.OnClick := HideWindow; end; -- Felipe Monteiro de Carvalho _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
At 08:56 19/5/2006, you wrote: On Fri, 19 May 2006 15:00:38 +0330, Arà Ricardo Ody [EMAIL PROTECTED] wrote: If you mean cancelled as kill task by windows task manager,no i really doubt that. becouse it is use is to force killling of a task,mostly when program is hung, Suppose computer is doing a long processing. Then, someone that can copies some information that is available only during this processing, tries to interrupt the processing. I want to detect this and delete the privileged information(that otherwise would be encripted - sources of a interpreted language). you can be sure by checking your program with spy++ and see if it recieves any message, also maybe global hooking can do something for you. I know nothing of this matter. May you develop, please? Thanks a lot Ricardo _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] Fwd: [MailServer Notification]Attachment Blocking Notification
I received a message from this address that my attached file was not received. So I would like to know if the zip file has actually reached the mailling list. Regards. -- Forwarded message -- From: [EMAIL PROTECTED] [EMAIL PROTECTED] Date: 19.05.2006 08:30 Subject: [MailServer Notification]Attachment Blocking Notification To: [EMAIL PROTECTED] The test_dock.zip has been blocked, and Replace has been taken on 5/19/2006 2:29:30 PM. Message details: Server:MPKWFX01 Sender: [EMAIL PROTECTED]; Recipient:lazarus@miraclec.com; Subject:[lazarus] Docked form focus problem - Finding and example app. Attachment name:test_dock.zip -- Alexandre Leclerc _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Fwd: [MailServer Notification]Attachment Blocking Notification
Yup got the attachment. I also received a bounced message the other day from that address. Graeme. On 5/19/06, Alexandre Leclerc [EMAIL PROTECTED] wrote: I received a message from this address that my attached file was not received. So I would like to know if the zip file has actually reached the mailling list. Regards. -- Forwarded message -- From: [EMAIL PROTECTED] [EMAIL PROTECTED] Date: 19.05.2006 08:30 Subject: [MailServer Notification]Attachment Blocking Notification To: [EMAIL PROTECTED] The test_dock.zip has been blocked, and Replace has been taken on 5/19/2006 2:29:30 PM. Message details: Server:MPKWFX01 Sender: [EMAIL PROTECTED]; Recipient:lazarus@miraclec.com; Subject:[lazarus] Docked form focus problem - Finding and example app. Attachment name:test_dock.zip -- Alexandre Leclerc _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives -- There's no place like 127.0.0.1 _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Fwd: [MailServer Notification]Attachment Blocking Notification
2006/5/19, Vincent Snijders [EMAIL PROTECTED]: Alexandre Leclerc wrote: I received a message from this address that my attached file was not received. So I would like to know if the zip file has actually reached the mailling list. It did, but it did not reach some user. Fine, I'll ignore message from this server then. -- Alexandre Leclerc _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Speedbutton property transparant default value should be false.
On Fri, 19 May 2006 14:39:00 +0200 Marien van Westen [EMAIL PROTECTED] wrote: On 5/19/06, Mattias Gaertner [EMAIL PROTECTED] wrote: On Fri, 19 May 2006 09:01:50 +0200 Marien van Westen [EMAIL PROTECTED] wrote: Delphi compatibility. Delphi compatability is not the issue here I think. I meant: for Delphi compatibility Transparent = true is the default. Ok, yes, you are right The strange thing is that speedbuttons don't have borders in the designer, but at runtime the borders show when the mouse is over the speedbutton. I think speedbuttons are also used in menus where you don't want to have bordres around them, but when you have a row of speedbuttons somewere else, the user might want to have borders around them. I also noticed that the buttons are transparent even when transparancy is false. I think transparancy somway interferes with the borderstyle of the buttons. Shall I report this as a bug? How does it work under Delphi? I Delphi (V6 open edition) the speedbuttons have a border. When I place a speedbutton on a panel I cannot see the text Panel1 through the speedbutton although the transparancy is set true (changing this to false doesn't change this. Delphi bug?). The appearence is the same at design-time and at runtime. Attachment shows how it looks in Delphi Maybe at design time transparency and flat are disabled. Please run the program under Delphi. Mattias _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
On 5/19/06, Arí Ricardo Ody [EMAIL PROTECTED] wrote: Suppose computer is doing a long processing. Then, someone that can copies some information that is available only during this processing, tries to interrupt the processing. I want to detect this and delete the privileged information(that otherwise would be encripted - sources of a interpreted language). If it is Windows only application then it should be receiving some sort of WM message. e.g. when the user logs off WM_QUERYENDSESSION is sent (IIRC). you can be sure by checking your program with spy++ and see if it recieves any message, also maybe global hooking can do something for you. I know nothing of this matter. May you develop, please? Try to find the appropriate WM and then use Windows hooks to insert your code before this message is processed. See SetWindowsHookEx / UnhookWindowsHookEx / CallNextHookEx Win32 API functions. Hope this will help you. If you need a sample code using hooks please send me a private message. _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
On 5/19/06, Arí Ricardo Ody [EMAIL PROTECTED] wrote: Suppose computer is doing a long processing. Then, someone that can copies some information that is available only during this processing, tries to interrupt the processing. I want to detect this and delete the privileged information(that otherwise would be encripted - sources of a interpreted language). Hello, My feeling is that deleting that information when the program is killed would only partially solve the problem , and would only give a false feeling of security. For instance: - since it's a long-running operation, it might be possible to copy the sensitive information before the processing ends ? (we are talking about some temporary files , right ? ) - someone might run the program inside a debugger, and stop the execution of the program just before the privileged information is deleted - giving that person enough time to inspect all the secrets. Perhaps it would be better to keep that information in memory in some kind of buffer ? that memory would be freed after the application is killed ; (again - i'm assuming that you are currently storing the information in some temporary files). Cheers, -- Adrian Maier _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
At 10:15 19/5/2006, you wrote: Hello, My feeling is that deleting that information when the program is killed would only partially solve the problem , and would only give a false feeling of security. For instance: - since it's a long-running operation, it might be possible to copy the sensitive information before the processing ends ? (we are talking about some temporary files , right ? ) - someone might run the program inside a debugger, and stop the execution of the program just before the privileged information is deleted - giving that person enough time to inspect all the secrets. First, thank you for you considerations. I think that people that would handle the application have no expertise to do the things you suggest above. This is my hope. I'll put the sources in a (local) ramdrive. This may difficult these actions. At least for access the information thru the network. Perhaps it would be better to keep that information in memory in some kind of buffer ? that memory would be freed after the application is killed ; (again - i'm assuming that you are currently storing the information in some temporary files). My Lazarus application will manage a list of programs to be converted and will run a tool(once for each file of the list) that uses an interpreter to execute sources that would writed in a ramdrive. I have no way of establish a communications with the interpreter in the way it takes the sources from a buffer of memory and not from a real drive. If you know what I mean... Cheers, -- Adrian Maier _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
On 19/05/06, Arí Ricardo Ody [EMAIL PROTECTED] wrote: My Lazarus application will manage a list of programs to be converted and will run a tool(once for each file of the list) that uses an interpreter to execute sources that would writed in a ramdrive. I have no way of establish a communications with the interpreter in the way it takes the sources from a buffer of memory and not from a real drive. If you know what I mean... Hmm. I see. There is nothing you can do if the interpreter accepts input only from files. It's a shame that you can't pass the sources to the interpreter through pipes. Just curious: what interpreter is it ? -- Adrian Maier _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Focus problem -- where is it handled in the lcl?
Alexandre Leclerc wrote: 2006/5/18, Mattias Gaertner [EMAIL PROTECTED]: On Thu, 18 May 2006 14:08:19 -0400 Alexandre Leclerc [EMAIL PROTECTED] wrote: Ok, can someone point me where the focus events / mouse click events and all that low level stuff for visual components (including forms) are handled in the source files? TControl, TWinControl, TCustomForm and the widgetsets. Just a question like that... what must I understand by widgetsets? The buttons, panels, etc? Alexandre, keep in mind that Lazarus is designed to be cross-platform. Each operating system handles focus a different way, and in Linux, each window manager handles focus a different way making it even more complex. That's why it's separated from the actual component and in the interfaces portion - to abstract the OS specific code from the typical developer's code. With Lazarus designed to compile the same code for Windows, Linux, OSX and more, it's much more complex. -Tony _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
Re: [lazarus] Re: Detecting user cancel
At 10:39 19/5/2006, you wrote: Hmm. I see. There is nothing you can do if the interpreter accepts input only from files. It's a shame that you can't pass the sources to the interpreter through pipes. Just curious: what interpreter is it ? -- Adrian Maier A guy here develop an application using que macro language o Mansfield's Kedit editor(www.kedit.com). The interpreter runs under Kedit. It has a powerful language and you can mix editor commands in the macro programs. But it's a high CPU usage. I wrote a front end in Lazarus that allow the user to select a list of input files. The front end remains starting KEDIT once for each input file. [ ] Ricardo _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] TrayIcon patch
Attached patch fixes compilation of trayicon for GTK2 projects. Ciao A.J. -- there's nothing as inspirational for a hacker as a cat obscuring a bug by sitting in front of the monitor - Boudewijn Rempt A.J. Venter Chief Software Architect OpenLab International www.getopenlab.com www.silentcoder.co.za +27 82 726 5103 trayicongtk2.patch.gz Description: GNU Zip compressed data
Re: [lazarus] Speedbutton property transparant default value should be false.
I Delphi (V6 open edition) the speedbuttons have a border. When I place a speedbutton on a panel I cannot see the text Panel1 through the speedbutton although the transparancy is set true (changing this to false doesn't change this. Delphi bug?). The appearence is the same at design-time and at runtime. Attachment shows how it looks in Delphi Maybe at design time transparency and flat are disabled.Please run the program under Delphi.At run-time the speedbuttons have the same behaviour.I show some screenshots. Note that when transparacy is true and the mouse is over the button (runtime only) the borders are visible (2) otherwise the borders are invisible. I hope those screenshots will help .I only tested this in Windows of course.Greetings, Marien speedbuttons-panel-delphi-runtime.png Description: PNG image speedbuttons-panel-lazarus-designtime-transparacy-false.png Description: PNG image speedbuttons-panel-lazarus-designtime-transparancy-true.png Description: PNG image speedbuttons-panel-lazarus-runtime-transparancy false.png Description: PNG image speedbuttons-panel-lazarus-runtime-transparancy true-1.png Description: PNG image speedbuttons-panel-lazarus-runtime-transparancy true-2.png Description: PNG image
Re: [lazarus] Request for PR noise !
[EMAIL PROTECTED] a écrit : [EMAIL PROTECTED] writes: Florian Klaempfl wrote: Lv wrote: So if you guys want to change the name, where does it leave users like me presenting papers at conferences and citing Lazarus in technical papers as my software of choice? Nice to hear that not all users believe into the imo stupid idea to change the name. I don't think either that the board of directors of lazarus (Michael H., Matthias, Marc, Vincents, Micha et al.) consider a name change :) Just for the record, I also think a name change is uncalled for. A name change would have serious consequences - and we could start promoting everything from zero since the new name will be even less known then Lazarus. The word Lazarus is neither offensive or more ambiguous than other well known projects and companies. One more comment: Cheetah is the simbol of FPC, not lazarus. A cheetah on a column is Lazarus. Now you don't see anyone calling FPC cheetah, nor calling Linux Pinguin. I agree with Andreas (and thank the board of directors and their fine insight in dismissing this idea as being a total waste of time and effort) Which fine insight? Calling it a stupid idea is a fine insight? ;-) The drawbacks from a name change are phenomenal!!! Even without the !!!, it would be exaggerated. 1) Loss of current user base who don't keep up on the mailing list and all of sudden can't find their product under the name they know (and hopefully love) What is your estimate of the average IQ of a current user? :-) 2) Loss of new user base. I have actively promoted Lazarus, whenever asked, as a great tool/environment with my contacts. Don't know how many have looked at it yet, but I'm going to be very embarrassed if they do finally take my advice and check it out, and it's called something different than what I was advertising. That's the contrary, there will be more new users more quickly (assuming the new name is good). Moreover, why all that panic around a simple change of name. That's a current practice. Many software editors make use of temporary names for the beta versions of their softwares. 3) I suspect would require a lot of effort in renaming all sorts of directories/web sites/links, blah blah blah on a global basis Yes, and this is the only valuable reason. A change of name could lead to too much work in rewriting for the Lazarus team. This is a problem because the possibility of a renaming has not been taken in account when starting the project. 4) Same as #3 above on each existing users's personal computer(s) Renaming once for all a few directories is not a problem. All this for some cosmetics? Cosmetic!? Some people are paid fortunes just to search for (and to find) the name of a new product. Are you saying that companies who pay them are wasting their money? Much better that we quit talking about this and get back to making a really great product that will really sell itself based on it's stability and functionality, not on whether it is called S__T or Lazarus. Just my two cents. Let's get back to work making Lazarus GREAT! Yes, and a change of name would be good in order to do that. Btw, why do you present things as if we have to choose between either a good name or a good quality. Asking for both of them, that's not possible? I agree that if a software is useful and of good quality, good name or bad name, it will be used. A bad name cannot kill a good software. But a good name can help it. And Lazarus is indeed very bad. What does this name suggest? Bringing back a death to the life. The first thing it leads to think of is the death. This is totally negative. Lazarus might be an appropriate name for a game where you have to destroy lots of death-alive ones in catacombs but there we are talking about a Pascal programming tool. Note that I don't say that naming the software Lazarus was a stupid idea, no, I try to explain why I think it was a bad idea. Stupid idea is not an explanation, it is at most a peremptory judgement. Personally, and very frankly, I don't care whether the Lazarus team decides to change the name or not. Now, I know and use Lazarus. For me its name has no more importance, I know what is behind it. But this is not the case for those who don't know it yet. mm _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
[lazarus] Anybody know what is wrong here ?
I'm trying to extend my tsimple_html component so it can read from a
tstringlist property instead of a URL (which in many [at least of my]
programs could severely reduce overhead).
So I basically just added a tstringlist property exactly as I have always done
it before (and I compared code with my other components, I cannot find the
difference).
But in this one, the moment you click the property editor for the tstringlist,
it throws an access violation - not a good sign.
Any chance somebody knows what I am doing wrong here ? (code attached)
A.J.
--
there's nothing as inspirational for a hacker as a cat obscuring a bug
by sitting in front of the monitor - Boudewijn Rempt
A.J. Venter
Chief Software Architect
OpenLab International
www.getopenlab.com
www.silentcoder.co.za
+27 82 726 5103
Unit simple_html;
{A simple file based html viewer component based on ippro with image viewing
added}
{$mode objfpc}{$H+}
interface
uses
LResources,Classes, SysUtils, Forms, Controls, Graphics,IPHtml,lazjpeg,util,simple_downloader
,fileUtil,dialogs,extCtrls;
Type
TParseEvent = procedure(Sender: TObject;TempFileName:String) of object;
THTML = class(TIpHtml)
public
property OnGetImageX;
end;
TSimpleHtml = Class(TComponent)
private
FTDIR : String;
FURL : String;
FHTML : THTML;
FLines : TStringlist;
FPanel : TipHTMLPanel;
FParseMe: TParseEvent;
FDownLoader : TSimple_Downloader;
Function FetchURL(URL:String):String;
destructor destroy;
procedure setHTMLSTrings(const Avalue: TStringlist);
Public
constructor Create(AOwner: TComponent);
procedure showHTML;
Procedure HTMLGetImageX(Sender: TIpHtmlNode; Const URL: String; Var Picture: TPicture);
published
property URL : String read FURL write FURL;
property HTMLSTrings : TStringlist read FLines write setHTMLSTrings;
property TempDir : String read FTDir write FTDir;
property Panel: TipHTMLPanel read FPanel write Fpanel;
property OnCanParse: TParseEvent read FParseMe write FParseMe;
property DownLoader : TSimple_Downloader read FDownLoader Write FDownloader;
End;
procedure Register;
Implementation
procedure Register;
begin
RegisterComponents('OLPack',[TSimpleHTML]);
end;
Procedure TSimpleHTML.SetHTMLSTrings(const Avalue: TStringList);
Begin
If AValue Nil then
FLines.Assign(AValue);
end;
Function TSimpleHTML.FetchURL(URL:String):String;
Var
SDIR,Ext: STring;
Begin
If pos('://',URL) = 0 then
FetchURL := URL
else begin
FileUTil.ForceDirectory(FTDIR);
SDIR := '';
Ext := UpperCase(ExtractFileExt(URL));
If (Ext = '.JPG') or
(Ext = '.PNG') or
(Ext = '.GIF') or
(Ext = '.XPM') then
SDIR := FTDIR+pathDelim+ExtractFileName(URL) else
If (length(ExtractFileName(URL)) = 0) or (Length(EXT)=0) then
SDIR := FTDIR+pathdelim+'index.html' else
If pos('.html',ExtractFileName(URL)) = 0 then
SDIR := FTDIR+pathdelim+ExtractFileName(URL)+'.html' else
SDIR := FTDIR+pathdelim+ExtractFileName(URL);
SDIR := SingleSlashes(SDIR);
DownLoader.URL := URL;
Downloader.Download;
DownLoader.Output.SaveToFile(SDIR);
if ((ExtractFileExt(SDIR) = '.html') or (ExtractFileExt(SDIR) = '.htm'))
And (Assigned(FParseMe)) then
FParseMe(Self,SDIR);
FetchURL := SDIR;
end;
end;
Procedure TSimpleHTML.HTMLGetImageX(Sender:TIpHtmlNode; Const URL: String; Var Picture: TPicture);
Var
PicCreated: boolean;
IMGFile:String;
Begin
Try
IF (FileExists(FTDIR+PathDelim+ExtractFileName(URL))) then
IMGFile := FTDIR+PathDelim+ExtractFileName(URL) else
if (pos('://',URL) 0) then
IMGFile := FetchURL(URL) else
IMGFILE := URL;
If FileExists(IMGFIle) Then
Begin
PicCreated := False;
If Picture=Nil Then
Begin
Picture := TPicture.Create;
PicCreated := True;
End;
Picture.LoadFromFile(IMGFile);
End;
Except
If PicCreated Then
Picture.Free;
Picture := Nil;
End;
End;
Constructor TSimpleHTML.Create(AOwner: TComponent);
Begin
FLines := TStringlist.create;
inherited create(AOwner);
End;
Procedure TSimpleHTML.ShowHTML;
Var Path : String;
HTMLStream: TMemoryStream;
I :Integer;
Begin
if length(FTDir) = 0 then
FTDir := SysUtils.GetTempFileName;
If Length(FURL) 0 then
begin
Path := FetchURL(FURL);
HTMLStream := TMemoryStream.Create;
HTMLStream.LoadFromFile(Path);
end else
begin
For I := 0 to FLines.Count - 1 do
HTMLStream.WriteAnsiString(FLines[I]);
end;
FHTML := THTML.Create;
FHTML.OnGetImageX := @HTMLGetImageX;
FHTML.LoadFromStream(HTMLStream);
[lazarus] Unicode - basic logic would be...
Sorry for going further on the subject. I have started an introductory wiki-page that should be a kind of knowledge base for those (if any) trying/thinking about Unicode-enabling win32 interface: http://wiki.lazarus.freepascal.org/index.php/LCL_Unicode_Support Of course, currently there is not exactly much information there. Here follow two concept procedures that just describe a possible logic of what should happen at the boundary between win32 widgetset and WinAPI. Could some core developer take a look and comment, please. Those interested in Unicode are also invited to comment, of course. Regards, Borut var UseUnicode : Boolean; //Global variable, so that application can initialise global behaviour procedure FromWidgetToWinAPI(Text:String); begin if UseUnicode then // implicit statement: Text is UTF-8 encoded begin if RunningOnUnicodeenabledPlatform then Call_Set_API_W(UTF8ToUTF16(Text))//reencode to UTF-16 and call *W else Call_Set_API_A(UTF8ToActiveCodePage(Text)); //reencode to ISO- and call *A (possible information loss) end else // implicit statement: Text is NOT UTF-8 encoded begin Call_Set_API_A(Text); // current case (not meaningful if text is UTF-8 encoded) end; end; function FromWinAPIToWidget : String; var W : WideString; S : String; begin if UseUnicode then // implicit statement: Result should be UTF-8 encoded begin if RunningOnUnicodeenabledPlatform then begin W:=Call_Get_API_W; Result:=UTF16ToUTF8(W);//reencode from UTF-16 to UTF-8 end else begin S:=Call_Get_API_A; Result:=ActiveCodePageToUTF8(S); //reencode from UTF-16 to ISO- (possible information loss) end; end else // implicit statement: Result should NOT be UTF-8 encoded begin Result:=Call_Get_API_A; // current case (not meaningful if Result will later be treated as UTF-8 encoded) end; end; _ To unsubscribe: mail [EMAIL PROTECTED] with unsubscribe as the Subject archives at http://www.lazarus.freepascal.org/mailarchives
