Re: [lazarus] Default target OS/CPU for projects ?

[Michael Van Canneyt]

On Sat, 26 Jan 2008, Marc Weustink wrote:

> Michael Van Canneyt wrote:
> > Hello,
> > 
> > How does the IDE determine the target OS/CPU if none is specified for a
> > project ?
> > 
> > Apparently, it takes the same OS/CPU combination as the one the IDE itself
> > was compiled with.
> > 
> > In my opinion, this is wrong. It should see what compiler was selected,
> > and use the OS/CPU combination that this compiler reports.
> > 
> > Specifically:
> > 
> > - I use the i386/Linux IDE.
> > - No target OS/CPU has been specified.
> > - I select the ppcx64 compiler in the Environment options.
> > - The IDE constructs all paths with units/i386-linux.
> > - Recompile IDE.
> > 
> > Of course, compilation fails. All goes well for the LCL/IDE
> > because the makefile detects that it needs x86_64-linux, but
> > compiling the packages fails, because the IDE constructs all paths with
> > 386-linux...
> > 
> > The same is true for any other project as well, of course.
> > 
> > I think that this behaviour should be changed to the same behaviour as used
> > by the makefile: detecting the default
> > OS/CPU of the used compiler.
> > 
> > Or that at least there should be an option to allow the choice between:
> > - Always use IDE CPU/OS combination
> > - Detect compiler CPU/OS and use that.
> > 
> > Opinions, comments ?
> 
> IIRC, this was a todo to base CPU/OS on selected compiler.
> (I've the same issues when compiling on win32/64 for wince/arm)

Good, this todo has my vote  :-)
 
> OTOH, (I've only heard this, not tried) if you set fpc as compiler the correct
> flags are passed, st fpc can choose the correct compiler

I tried this too. That it chooses the 'correct compiler' is correct.
But the unit paths are still wrong, because the IDE constructs those...

Michael.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Felipe Monteiro de Carvalho]

I just tested the patch and it doesn't fixes selecting utf-8 words on
synedit. This doesn't mean it isn't on the right direction. I don't
know what is missing, as from the description I would think this
should start working.

-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Felipe Monteiro de Carvalho]

Another test line:

  Caption := 'éé';

Behaves like before patching.

-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

I tested with this line (utf-8 encoded), specifically the last word:

  Application.Title:='Minha Aplicação';

Double clicking on the left part selects "Aplica". Clicking on the
accented part does nothing and clicking on "o" selects only "o" and
puts the carret at the end of the line.

Lazarus is compiled with build+clean and with the option -dWindowsUnicodeSupport

thanks,
  


What function does "double clicking" call? It might be possible I missed 
some. I'll test it on my "spellchecker" to see if it selects a whole 
word and if so then I'll try double-clicking. Thanks, I'll report soon..


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Felipe Monteiro de Carvalho]

Oh, and another detail. I am not 100% sure, but I think that changes
should be around ifdef SYN_LAZARUS

thanks,
-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

I tested with this line (utf-8 encoded), specifically the last word:

  Application.Title:='Minha Aplicação';

Double clicking on the left part selects "Aplica". Clicking on the
accented part does nothing and clicking on "o" selects only "o" and
puts the carret at the end of the line.

Lazarus is compiled with build+clean and with the option -dWindowsUnicodeSupport

thanks,
  
I missed some function. If you use GetWordAtRowCol or 
GetWordBoundsAtRowCol or
NextWordPos functions, you get the whole word/boundaries. Doubleclick in 
synedit uses SetWordBlock (in lazarus/non-lineselect case) which I 
didn't look at (and it seems to try some ugly utf-8 conversion which I 
guess didn't work)


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

I just tested the patch and it doesn't fixes selecting utf-8 words on
synedit. This doesn't mean it isn't on the right direction. I don't
know what is missing, as from the description I would think this
should start working.

  
What words did you test? Can you send the test sample to me? I tested on 
some slovak accented (eg: 2byte chars) words and it worked perfectly.


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: synedit patch from ales

[Felipe Monteiro de Carvalho]

On Jan 27, 2008 11:33 AM, Ales Katona <[EMAIL PROTECTED]> wrote:
> What words did you test? Can you send the test sample to me? I tested on
> some slovak accented (eg: 2byte chars) words and it worked perfectly.

I tested with this line (utf-8 encoded), specifically the last word:

  Application.Title:='Minha Aplicação';

Double clicking on the left part selects "Aplica". Clicking on the
accented part does nothing and clicking on "o" selects only "o" and
puts the carret at the end of the line.

Lazarus is compiled with build+clean and with the option -dWindowsUnicodeSupport

thanks,
-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Component

[Michael Van Canneyt]

On Sat, 26 Jan 2008, [EMAIL PROTECTED] wrote:

> That's it! http://wiki.lazarus.freepascal.org/Lazarus_Packages should be
> enhanced...

I did this, added some stuff.

Michael.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

[lazarus] Why do Delphi users hate Lazarus so much?

[John Stoneham]

Well, I guess the term "Delphi users" is a little broad, but it seems
that every post that mentions Lazarus on
borland.public.delphi.non-technical gets BLASTED by Delphi fanboys. I
mean, it's really amazing to see the outright hostility towards a free
software package that might very well save Pascal as a language in the
future (since it sure seems that CodeGear and Borland are heading down
the drain pretty fast these days). Some of these guys even start
entire threads with subjects like "Quit talking about Lazarus here" or
"If you think you need cross-platform you don't need Delphi" or some
other nonsense. Why do they hate Lazarus so much??

As a (former) Delphi user, I just can't understand this. Delphi 7 was
the last version I purchased -- which was some 6 or 7 years ago -- and
I refused to purchase a newer version because they refused to fix a
very serious bug in D7* (here's a link to one of my complaints about
it on the newsgroup, where I describe it in more detail than I care to
here: http://tinyurl.com/2ws5gw ). Of course, they *claimed* it was
fixed -- in fact, the QS bug listing showed it as closed and fixed
because it was fixed in D8. Well it may have been fixed in D8, but I
didn't buy D8, I bought D7. They really expected us to pay for an
upgrade to get this bug-fix! That was the last straw for me, and I
swore I'd never purchase another Delphi product.

Anyway, the point I'm trying to make is that I see Lazarus as
"salvation" for the Delphi community, not something to be reviled and
hated. So what if it's not as polished as Delphi! At least the FPC and
Lazarus teams respond to bug reports!! And if it's something I really
need fixed, I can always do it myself because it's open source.

I just hope the Lazarus community isn't discouraged by those negative comments.

-- John

[*] It was actually two bugs. The main one was integer math
optimization. But when they fixed that bug they broke Cardinal
multiplication. So I had a choice, to use optimized integer math and
making sure I didn't use any Cardinals, or not worry about the
Cardinals and stick with un-optimized integer math. In my opinion,
this was completely unacceptable, since I was developing a simulation
package that made heavy use of both Cardinals and integer math, and it
needed to be optimized.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[Florian Klaempfl]

John Stoneham schrieb:
> Well, I guess the term "Delphi users" is a little broad, but it seems
> that every post that mentions Lazarus on
> borland.public.delphi.non-technical gets BLASTED by Delphi fanboys.

90% of the posts in b.p.d.non-tech would be qualified as troll posts on
other mailings lists or in other newsgroups so I wouldn't care about it :)

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

[lazarus] New synedit patch

[Ales Katona]

This is a revised version of my synedit patch which adds support for 
"doubleclick" selection of utf-8 text as well as puts all the changes in 
{$IFDEF SYN_LAZARUS}.


To sum it up:
1. it inverts parsing logic to use Highlighter.WordBlockChars + 
TSynWhiteChars (or TSynWordBlockChars + TSynWhiteChars if there's no 
highlighter), thus allowing utf-8 words to be parsed as long as the 
block chars don't contain chars > 127.


2. it fixes the doubleclick selection by said logic + fixing final caret 
position based on true length of given line if UseUTF8 is true.


Ales
Index: components/synedit/synedittypes.pp
===
--- components/synedit/synedittypes.pp	(revision 13890)
+++ components/synedit/synedittypes.pp	(working copy)
@@ -44,6 +44,7 @@
 const
   TSynSpecialChars = ['À'..'Ö', 'Ø'..'ö', 'ø'..'ÿ'];
   TSynValidStringChars = ['_', '0'..'9', 'A'..'Z', 'a'..'z'] + TSynSpecialChars;
+  TSynWhiteChars = [' ', #9];
   TSynWordBreakChars = ['.', ',', ';', ':', '"', , '!', '?', '[', ']', '(',
 ')', '{', '}', '^', '-', '=', '+', '*', '/', '\', '|'];
 
Index: components/synedit/synedit.pp
===
--- components/synedit/synedit.pp	(revision 13890)
+++ components/synedit/synedit.pp	(working copy)
@@ -6453,9 +6453,6 @@
   Runner: TPoint;
   TempString: string;
   IdChars: TSynIdentChars;
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  BufChars: array of PChar;
-{$ENDIF}
 begin
   { Value is the position of the Carat in bytes }
   Value.x := MinMax(Value.x, 1, fMaxLeftChar);
@@ -6463,53 +6460,36 @@
   TempString := Lines[Value.Y - 1];
   if TempString = '' then exit;
   // Click on right side of text
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  UTF8ToArrayOfUTF8Char(PChar(TempString), Length(TempString), BufChars);
-  if Length(BufChars) < Value.X then Value.X := Length(BufChars);
-{$ELSE}
   if Length(TempString) < Value.X then Value.X := Length(TempString);
-{$ENDIF}
+
   Runner := Value;
-  if fHighlighter <> nil then
+  if Assigned(fHighlighter) then
+{$IFDEF SYN_LAZARUS}
+IdChars := [#1..#255] - (fHighlighter.WordBreakChars + TSynWhiteChars)
+{$ELSE}
 IdChars := fHighlighter.IdentChars
+{$ENDIF}
   else
+{$IFDEF SYN_LAZARUS}
+IDchars := [#1..#255] - (TSynWordBreakChars + TSynWhiteChars);
+{$ELSE}
 IDchars := [#33..#255];
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  if not (BufChars[Runner.X - 1]^ in IdChars) then begin
-{$ELSE}
+{$ENDIF}
   if not (TempString[Runner.X] in IdChars) then begin
-{$ENDIF}
 // no word under cursor and next char right is not start of a word
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-if (Runner.X > 1) and (not (BufChars[Runner.X - 1]^ in IdChars)) then begin
-{$ELSE}
 if (Runner.X > 1) and (not (TempString[Runner.X] in IdChars)) then begin
-{$ENDIF}
   // find end of word on the left side
   while Runner.X > 1 do begin
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-if (BufChars[Runner.X - 1]^ in IdChars) then break;
-{$ELSE}
 if (TempString[Runner.X] in IdChars) then break;
-{$ENDIF}
 Dec(Runner.X);
   end;
 end;
 // no word on the left side, so look to the right side
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-if not (BufChars[Runner.X - 1]^ in IdChars) then begin
-{$ELSE}
 if not (TempString[Runner.X] in IdChars) then begin
-{$ENDIF}
   Runner := Value;
   while (Runner.X < fMaxLeftChar)
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  {$IFDEF FPC} and (Runner.X < length(BufChars)){$ENDIF} do begin
-if (BufChars[Runner.X - 1]^ in IdChars) then break;
-{$ELSE}
   {$IFDEF FPC} and (Runner.X < length(TempString)){$ENDIF} do begin
 if (TempString[Runner.X] in IdChars) then break;
-{$ENDIF}
 Inc(Runner.X);
   end;
   if Runner.X > fMaxLeftChar then
@@ -6518,11 +6498,7 @@
 Value := Runner;
   end;
   while Runner.X > 0 do begin
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-if not (BufChars[Runner.X - 1]^ in IdChars) then break;
-{$ELSE}
 if not (TempString[Runner.X] in IdChars) then break;
-{$ENDIF}
 Dec(Runner.X);
   end;
   Inc(Runner.X);
@@ -6530,24 +6506,21 @@
   fBlockBegin := Runner;
   Runner := Value;
   while (Runner.X < fMaxLeftChar)
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  {$IFDEF FPC} and (Runner.X <= length(BufChars)){$ENDIF} do begin
-if not (BufChars[Runner.X - 1]^ in IdChars) then break;
-{$ELSE}
   {$IFDEF FPC} and (Runner.X <= length(TempString)){$ENDIF} do begin
 if not (TempString[Runner.X] in IdChars) then break;
-{$ENDIF}
 Inc(Runner.X);
   end;
   if Runner.X > fMaxLeftChar then Runner.X := fMaxLeftChar;
   fBlockEnd := Runner;
 // set caret to the end of selected block
   CaretXY := Runner;
+  {$IFDEF SYN_LAZARUS}
+  if UseUTF8 then begin
+CaretX := CaretX - Max(0, (Length(TempString) - UTF8Length(TempString)));
+  end;
+  {$ENDIF}
   InvalidateLine(Value.Y);
   StatusChanged([scSelection]);
-{$IFDEF NEW_UTF8_SETWORDBLOCK}
-  FreeArrayOfUTF8

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[Gabor Boros]

I Delphi user and I love Lazaros (and FPC). :-)
See the newsgoup name. Contains delphi and borland. That is a newsgroup 
on Borland's own newsserver about Delphi.


Gabor

John Stoneham írta:

borland.public.delphi.non-technical


_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[John Stoneham]

On Jan 27, 2008 7:05 AM,  <[EMAIL PROTECTED]> wrote:
>
> BTW:
> the cardinal mul/div overflow bug still exists in Delphi 2007 ;-)
>

Wow, that's weird! They fixed it in the 7.1 update, but that update
broke the integer optimization, which in my opinion was just as
serious a bug. I guess they re-broke the overflow bug in a later
"fix"! Whew, I'm glad I got out of that terrible loop.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[user1]

BTW:
the cardinal mul/div overflow bug still exists in Delphi 2007 ;-)

John Stoneham wrote:

Well, I guess the term "Delphi users" is a little broad, but it seems
that every post that mentions Lazarus on
borland.public.delphi.non-technical gets BLASTED by Delphi fanboys. I
mean, it's really amazing to see the outright hostility towards a free
software package that might very well save Pascal as a language in the
future (since it sure seems that CodeGear and Borland are heading down
the drain pretty fast these days). Some of these guys even start
entire threads with subjects like "Quit talking about Lazarus here" or
"If you think you need cross-platform you don't need Delphi" or some
other nonsense. Why do they hate Lazarus so much??

As a (former) Delphi user, I just can't understand this. Delphi 7 was
the last version I purchased -- which was some 6 or 7 years ago -- and
I refused to purchase a newer version because they refused to fix a
very serious bug in D7* (here's a link to one of my complaints about
it on the newsgroup, where I describe it in more detail than I care to
here: http://tinyurl.com/2ws5gw ). Of course, they *claimed* it was
fixed -- in fact, the QS bug listing showed it as closed and fixed
because it was fixed in D8. Well it may have been fixed in D8, but I
didn't buy D8, I bought D7. They really expected us to pay for an
upgrade to get this bug-fix! That was the last straw for me, and I
swore I'd never purchase another Delphi product.

Anyway, the point I'm trying to make is that I see Lazarus as
"salvation" for the Delphi community, not something to be reviled and
hated. So what if it's not as polished as Delphi! At least the FPC and
Lazarus teams respond to bug reports!! And if it's something I really
need fixed, I can always do it myself because it's open source.

I just hope the Lazarus community isn't discouraged by those negative comments.

-- John

[*] It was actually two bugs. The main one was integer math
optimization. But when they fixed that bug they broke Cardinal
multiplication. So I had a choice, to use optimized integer math and
making sure I didn't use any Cardinals, or not worry about the
Cardinals and stick with un-optimized integer math. In my opinion,
this was completely unacceptable, since I was developing a simulation
package that made heavy use of both Cardinals and integer math, and it
needed to be optimized.

_
 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] New synedit patch

[Felipe Monteiro de Carvalho]

Thanks, applied as it's much better then the current behavior.

However, it still isn't perfect. On:

  Caption := 'éé';
  Edit1.Text := 'éé';

If you double click "Caption" the cursor ends up 1 char before "C"

If you double click "Text" it goes between "t" and "1"

thanks,
-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Paul Ishenin]

Felipe Monteiro de Carvalho wrote:

If you double click "Text" it goes between "t" and "1"
  

Sorry, where is "1" in word "Text" :) ?

Best regards,
Paul Ishenin.

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

Thanks, applied as it's much better then the current behavior.

However, it still isn't perfect. On:

  Caption := 'éé';
  Edit1.Text := 'éé';

If you double click "Caption" the cursor ends up 1 char before "C"

If you double click "Text" it goes between "t" and "1"

thanks,
  
You're right, this is a logical problem on my end. I need to know how 
many utf8 chars are BEFORE the current word, not on the whole line. I'll 
try to fix this.


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

[lazarus] DBImage patch

[Michael Van Canneyt]

Hi,

2 issues with TDBImage:

First, TDBImage does not handle the case of a NULL blob very well.
When a blob field is NULL, CreateBlobStream can return Nil.
The Nil case is not handled correctly by TDBImage.

I submitted a bug report (10711) with a patch attached,
please commit it.

Secondly:
In general, the TDBImage should be more flexible about the kind
of data it expects: it now requires that the first part of the 
blob stream is an ansistring with the extension of the image. 
This is of course not true unless the image was saved by TDBImage 
itself, but is not correct for other, legacy, data. For instance,
the Borland TDataset writes a TGraphicHeader record to the stream
before the actual image data (guess how I found out).

I think that this can be easily fixed by adding an event handler
'OnGetImageFormat' to TDBImage, which, if set, should return this 
ansistring; if it is not set, the current behaviour remains. 
Similarly, an 'OnSetImageFormat' will write the type to the 
stream (if a write is needed at all).

I think this is a simple change, and if the Lazarus devels agree,
I can create a patch for it.

Michael.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

Thanks, applied as it's much better then the current behavior.

However, it still isn't perfect. On:

  Caption := 'éé';
  Edit1.Text := 'éé';

If you double click "Caption" the cursor ends up 1 char before "C"

If you double click "Text" it goes between "t" and "1"

thanks,
  

This patch fixed the oversight :)

Ales
Index: components/synedit/synedit.pp
===
--- components/synedit/synedit.pp	(revision 13892)
+++ components/synedit/synedit.pp	(working copy)
@@ -6516,6 +6516,7 @@
   CaretXY := Runner;
   {$IFDEF SYN_LAZARUS}
   if UseUTF8 then begin
+Delete(TempString, fBlockEnd.X, Length(TempString));
 CaretX := CaretX - Max(0, (Length(TempString) - UTF8Length(TempString)));
   end;
   {$ENDIF}

Re: [lazarus] New synedit patch

[Felipe Monteiro de Carvalho]

On Jan 27, 2008 3:48 PM, Paul Ishenin <[EMAIL PROTECTED]> wrote:
> Felipe Monteiro de Carvalho wrote:
> > If you double click "Text" it goes between "t" and "1"
> >
> Sorry, where is "1" in word "Text" :) ?

It's on the word on the left of it: Edit1.Text

The idea is that you click one word and the caret goes to the other
word on the left.

-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Felipe Monteiro de Carvalho]

Thanks applied. I did some testing and it looks perfect now =)

thanks again,
-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Ales Katona]

Felipe Monteiro de Carvalho  wrote / napísal(a):

On Jan 27, 2008 3:48 PM, Paul Ishenin <[EMAIL PROTECTED]> wrote:
  

Felipe Monteiro de Carvalho wrote:


If you double click "Text" it goes between "t" and "1"

  

Sorry, where is "1" in word "Text" :) ?



It's on the word on the left of it: Edit1.Text

The idea is that you click one word and the caret goes to the other
word on the left.

  
If you apply my last patch (see a few mails earlier, synedit3.patch), 
you'll fix this :)


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Felipe Monteiro de Carvalho]

Ok, so we are now reduced to zero known regressions introduced by
unicode win32 Lazarus.

Just thougth I could use this opportunity to encourage people to test it =)

thanks,
-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] DBImage patch

[Joost van der Sluis]

Op zondag 27-01-2008 om 16:18 uur [tijdzone +0100], schreef Michael Van
Canneyt:
> Secondly:
> In general, the TDBImage should be more flexible about the kind
> of data it expects: it now requires that the first part of the 
> blob stream is an ansistring with the extension of the image. 
> This is of course not true unless the image was saved by TDBImage 
> itself, but is not correct for other, legacy, data. For instance,
> the Borland TDataset writes a TGraphicHeader record to the stream
> before the actual image data (guess how I found out).

I wrote the current behavior, because it didn't work at all. So I came
up with this idea, since i wasn't aware of the precise header that
Delphi uses.

> I think that this can be easily fixed by adding an event handler
> 'OnGetImageFormat' to TDBImage, which, if set, should return this 
> ansistring; if it is not set, the current behaviour remains. 
> Similarly, an 'OnSetImageFormat' will write the type to the 
> stream (if a write is needed at all).
> 
> I think this is a simple change, and if the Lazarus devels agree,
> I can create a patch for it.

Go ahead offcourse although you should ask a lazarus-developer to commit
it. ;)

Joost.

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

[lazarus] LazDoc for preparing documentation files

[Chris Kirkpatrick]

I have been trying out the new functionality of LazDoc in the IDE, on 
which Mattias Gaertner has been working very hard.


Although it is still incomplete, and has been the subject of several bug 
reports, it has a number of useful features. One of these is the 
facility to 'copy from inherited', ie if a property has been described 
for an ancestor, the description can be pasted into the descendant's 
documentation.


I have recently done this for the properties of TGroupBox and TComboBox 
in StdCtrls.xml, pasting in the descriptions from TCustomGroupBox and 
TCustomComboBox (or earlier ancestors such as TControl or TWinControl if 
the property is described there).


It is very easy to use, but I believe we should not routinely use it 
until we are very sure the description in the ancestor class is correct 
and complete, because if we decide that the ancestor's description isn't 
right and we have already made the copies, we will have to retrace our 
steps and change all the descendants' properties as well.



I should therefore welcome some quality control on the work I (and 
others) have done in writing descriptions for the components in the LCL. 
I started with TControl and other primitive components in Controls.xml, 
and worked through Graphics, Forms, Dialogs, and then into the 
components that are actually visible in the IDE, in StdCtrls, ExtCtrls 
and ComCtrls as well as Menus, and have recently started on Grids, 
DBCtrls and DBGrids.


I tried to write the descriptions for the ancestors such as 
TCustomXx before describing TXx. In general I have not provided 
detailed descriptions for the child classes if the properties have been 
described in the ancestors; the on-line help and LazDoc hinting facility 
can trace the description back to the ancestor. We should wait until the 
descriptions for the primitive classes are complete and correct before 
pasting the descriptions into the visible components. (This will 
ultimately be needed for completion of the LCL manual, to make it like 
the RTL manual. At present, if a property is not described explicitly 
for a component, it doesn't appear in the HTML documentation files).


I have also been concentrating on Public or Published properties, and 
recognise that there is still a lot to be done on the Protected properties.


All comments are most welcome.

Regards - Chris

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] LazDoc for preparing documentation files

[Florian Klaempfl]

Chris Kirkpatrick schrieb:
> All comments are most welcome.

Just curious, are the docs as pdf available somewhere or how can I build
them?

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] LazDoc for preparing documentation files

[Chris Kirkpatrick]

No pdf at present;

html docs are built regularly and latest snapshot is at
ftp://ftp.hu.freepascal.org/pub/lazarus/

HTML is also visible on-line at http://lazarus-ccr.sourceforge.net/docs/lcl/

I don't think we are ready for PDF yet - it would be full of holes!
Regards - Chris

Florian Klaempfl wrote:

Chris Kirkpatrick schrieb:
  

All comments are most welcome.



Just curious, are the docs as pdf available somewhere or how can I build
them?

_
 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

[lazarus] TBitmap.LoadFromStream Fails in Lazarus but works fine in Delphi...

[Dominique Louis]

Hi all,
  I'm using the following function to retrieve a Bitmap from a memory 
stream.


function MyStream.ReadBitmap : TBitmap;
// Reads a bitmap from the stream
var
  sss : TMemoryStream;
begin
  Result := nil;
  if not ReadBoolean then
Exit;
  Result := TBitmap.Create;
  sss := ReadStreamC;
  try
Result.LoadFromStream( sss );
  except
Result.Free;
Result := nil;
  end;
  sss.Free;
end;

ReadStreamC simply creates a TMemoryStream populates itself from an 
internal stream and returns the results. From what I can tell the 
TMemoryStream is fine, but the Result.LoadFromStream() call always 
throws an exception. Strangely, while running exactly the same code with 
exactly the same data file using Delphi, works fine. I am also using 
Delphi compatibility mode with Lazarus ( if that makes any difference ).


Is anyone aware of this problem? Any Suggestions?


Thanks,


Dominique.

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[zaher dirkey]

I Delphi programmer too (10 years) , and i like FPC/Lazarus too match, than
Delphi.

This day (before i see this comments), my co worker ask me if some day we
see some people try to make fight Lazarus, i have no answer for him, and i
hope no.
(OK bad english :P )

On Jan 27, 2008 3:22 PM, John Stoneham <[EMAIL PROTECTED]> wrote:

> On Jan 27, 2008 7:05 AM,  <[EMAIL PROTECTED]> wrote:
> >
> > BTW:
> > the cardinal mul/div overflow bug still exists in Delphi 2007 ;-)
> >
>
> Wow, that's weird! They fixed it in the 7.1 update, but that update
> broke the integer optimization, which in my opinion was just as
> serious a bug. I guess they re-broke the overflow bug in a later
> "fix"! Whew, I'm glad I got out of that terrible loop.
>
> _
> To unsubscribe: mail [EMAIL PROTECTED] with
>"unsubscribe" as the Subject
>   archives at http://www.lazarus.freepascal.org/mailarchives
>



-- 
Zaher Dirkey

Re: [lazarus] Mac OS X : TBitBtn Transparency

[Dominique Louis]

Tom Gregorovic wrote:

It used to work, but after bitmap rewrite (which moved transparency in
TBitmap to mask) it was broken. I hope, I have fixed it in latest SVN.


Hi Tom,
  I totally missed this update, I'm downloading the latest snapshot so 
will let you know if it works for me.


Thanks,


Dominique.

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Mac OS X : TBitBtn Transparency

[Dominique Louis]

Hi Tom,
  I just checked and transparency does indeed work correctly in Carbon 
in the latest Snapshot(27/01/2008). Great work!


Dominique.

Tom Gregorovic wrote:

On Jan 11, 2008 10:09 AM, Dominique Louis
<[EMAIL PROTECTED]> wrote:

Hi Paul,
   Does this mean that TBitBtn under Mac OS X does not support
transparency for Bitmaps only via PNGs?

I just searched through the code and the Mac OS X Function
TButtonGlyph.Draw() in buttonglyph.inc is a bit sparse and there is no
code there makes use of the passed in Transparency. I also noticed in
TCustomBitBtn.ActionChange/CopyImage it has a // TODO : Transparency
comment.

So I'm guessing it's not done yet.



It used to work, but after bitmap rewrite (which moved transparency in
TBitmap to mask) it was broken. I hope, I have fixed it in latest SVN.

Tom


_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] TBitmap.LoadFromStream Fails in Lazarus but works fine in Delphi...

[zaher dirkey]

try
sss.Seek(0, soFromBeginning); //<---
Result.LoadFromStream( sss );

On Jan 27, 2008 8:45 PM, Dominique Louis <[EMAIL PROTECTED]>
wrote:

> Hi all,
>   I'm using the following function to retrieve a Bitmap from a memory
> stream.
>
> function MyStream.ReadBitmap : TBitmap;
> // Reads a bitmap from the stream
> var
>   sss : TMemoryStream;
> begin
>   Result := nil;
>   if not ReadBoolean then
> Exit;
>   Result := TBitmap.Create;
>   sss := ReadStreamC;
>   try
> Result.LoadFromStream( sss );
>   except
> Result.Free;
> Result := nil;
>   end;
>   sss.Free;
> end;
>
> ReadStreamC simply creates a TMemoryStream populates itself from an
> internal stream and returns the results. From what I can tell the
> TMemoryStream is fine, but the Result.LoadFromStream() call always
> throws an exception. Strangely, while running exactly the same code with
> exactly the same data file using Delphi, works fine. I am also using
> Delphi compatibility mode with Lazarus ( if that makes any difference ).
>
> Is anyone aware of this problem? Any Suggestions?
>
>
> Thanks,
>
>
> Dominique.
>
> _
> To unsubscribe: mail [EMAIL PROTECTED] with
>"unsubscribe" as the Subject
>   archives at http://www.lazarus.freepascal.org/mailarchives
>



-- 
Zaher Dirkey

Re: [lazarus] New synedit patch

[zaher dirkey]

How we can test it?, i use compile from SVN i will use Arabic language in my
units (strings).
Is we can use it also for WinCE?

On Jan 27, 2008 6:20 PM, Felipe Monteiro de Carvalho <
[EMAIL PROTECTED]> wrote:

> Ok, so we are now reduced to zero known regressions introduced by
> unicode win32 Lazarus.
>
> Just thougth I could use this opportunity to encourage people to test it
> =)
>
> thanks,
> --
> Felipe Monteiro de Carvalho
>
> _
> To unsubscribe: mail [EMAIL PROTECTED] with
>"unsubscribe" as the Subject
>   archives at http://www.lazarus.freepascal.org/mailarchives
>



-- 
Zaher Dirkey

Re: [lazarus] New synedit patch

[Ales Katona]

zaher dirkey  wrote / napísal(a):
How we can test it?, i use compile from SVN i will use Arabic language 
in my units (strings).

Is we can use it also for WinCE?

On Jan 27, 2008 6:20 PM, Felipe Monteiro de Carvalho 
<[EMAIL PROTECTED] 
> wrote:


Ok, so we are now reduced to zero known regressions introduced by
unicode win32 Lazarus.

Just thougth I could use this opportunity to encourage people to
test it =)

thanks,
--
Felipe Monteiro de Carvalho

_
To unsubscribe: mail [EMAIL PROTECTED]
 with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives




--
Zaher Dirkey 


Just doubleclick a word with special (non latin1) chars, and see if it's 
properly selected. Or try other functions like SynEdit.NextWordPos etc.


Ales

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] New synedit patch

[Felipe Monteiro de Carvalho]

On Jan 27, 2008 10:43 PM, zaher dirkey <[EMAIL PROTECTED]> wrote:
> How we can test it?, i use compile from SVN i will use Arabic language in my
> units (strings).

You need to recompile with unicode enabled:

http://wiki.lazarus.freepascal.org/LCL_Unicode_Support#Compiling_LCL-Win32_with_Unicode

Then you can write unicode text to the code editor and the object
inspector. Legacy code needs to be converted to utf-8

> Is we can use it also for WinCE?

Not yet. I hope to implement it in the next month(s).

-- 
Felipe Monteiro de Carvalho

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[Marius]

I just hope the Lazarus community isn't discouraged by those negative comments.


;-) Everything there is being bashed. In my opinion nothing in that 
forum should ever be taken to serious (including the teamb or 
codegear-team msgs).


Its all free advertisement in the end (that included the bad msg).

_
To unsubscribe: mail [EMAIL PROTECTED] with
   "unsubscribe" as the Subject
  archives at http://www.lazarus.freepascal.org/mailarchives

[lazarus]Have Utilities of parse local html files to Dom?

[wfifi]

hi,all
i write a program need get some content in html files.
Have Utilities of parse local html files to Dom?

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Re: [lazarus] Why do Delphi users hate Lazarus so much?

[Graeme Geldenhuys]

On 27/01/2008, John Stoneham <[EMAIL PROTECTED]> wrote:
> Well, I guess the term "Delphi users" is a little broad, but it seems
> that every post that mentions Lazarus on
> borland.public.delphi.non-technical gets BLASTED by Delphi fanboys. I

Take all those comments with a pinch of salt.  ;-)  I believe they can
do and say what they want, it's a Delphi newsgroup after all. BUT
every time they mention the word Lazarus, they are making our product
know.  Somebody might start wondering, what is this "Lazarus" that
everybody is going on about  Free advertising!  :-)  Ask any
person in advertising, there is no such thing a good or bad
advertising!  Any publicity is good publicity.

Regards,
  - Graeme -


___
fpGUI - a cross-platform Free Pascal GUI toolkit
http://opensoft.homeip.net/fpgui/

_
 To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives