I don't know if it's an undefined behavior or a bug.
The environment is fpc 2.6-bugfix branch x86-64 linux.
See the details in the attached code
program t2;
{$mode objfpc}
uses sysUtils;
type
ip= interface ['{f0cea771-a3f1-4223-bea2-c46c5f0f61fa}']
end;
iq= interfac
On 22 Aug 12, at 19:03, Michael Van Canneyt wrote:
Hi Michael,
> As you know, FPC uses fpmake to compile everything in packages.
>
> As I've been steadily improving fcl-passrc for fpdoc,
> (it's now almost on par with the compiler, barring errors and assembler
> statetements),
> I created a
On 22/08/2012 18:08, Michael Van Canneyt wrote:
Hi,
for those that do unit testing, I committed a small tool "pas2ut" which
analyses a unit, and creates a unit with a number of (empty) test
cases. Obviously, the unit should be compileable (well, at least it
must be
parseable).
I just updat
On 22 August 2012 18:47, Daniel Gaspary wrote:
> The obvious example: insert calls to logging procedures at the enter
> and exit of each method, procedure or function of a unit.
This is already possible for some time. Take a look at the
'fpprofiler' project. I have a version of it that I worked
On Wed, Aug 22, 2012 at 2:08 PM, Michael Van Canneyt
> I would appreaciate feedback and
suggestions.
I believe could be interesting to have a similar tool to generate
"annotations" to a unit(or set of units).
The obvious example: insert calls to logging procedures at the enter
and exit of each m
Hi,
for those that do unit testing, I committed a small tool "pas2ut" which
analyses a unit, and creates a unit with a number of (empty) test cases.
Obviously, the unit should be compileable (well, at least it must be
parseable).
The output is very configurable, and should cover a lot of case
Hi,
As you know, FPC uses fpmake to compile everything in packages.
As I've been steadily improving fcl-passrc for fpdoc,
(it's now almost on par with the compiler, barring errors and assembler statetements),
I created a small tool pas2fpm which takes as input a bunch of unit names,
and outpu
On Wed, 22 Aug 2012, Ludo Brands wrote:
There are more problems: the library name is not
cross-platform, so the project will no longer be
cross-platform. Which is a far more serious problem, IMHO.
(and one which would exist even if I had followed your
suggestion of implementing the property in
On Wed, Aug 22, 2012 at 11:05 AM, Sven Barth
wrote:
> Am 22.08.2012 12:37, schrieb Timothy Madden:
>
>> On 08/22/2012 09:57 AM, Marco van de Voort wrote:
>>>
>>> In our previous episode, Marcos Douglas said:
>>
>> [...]
>>>
>>>
> Requiring all functions to be allways fully qualified will make
On Wed, Aug 22, 2012 at 9:22 AM, Jonas Maebe wrote:
>
> Jürgen Hestermann wrote on Wed, 22 Aug 2012:
>
>
>> Am 2012-08-22 13:45, schrieb Lukasz Sokol:
>>>
>>> Recompilation when one of units/libraries was upgraded, almost always
>>> implies
>>> rebuild. Who on Earth is so trigger-happy to upgrade,
leledumbo wrote on Wed, 22 Aug 2012:
I've created a .so from a unit by compiling it to .ppu and .o and then
executes ppumove on the .ppu. Now I've got the .so and .ppl (extension
specified with -e), how can I tell fpc to link against them instead of the
.ppu and .o? I've tried with -XD with no
I've created a .so from a unit by compiling it to .ppu and .o and then
executes ppumove on the .ppu. Now I've got the .so and .ppl (extension
specified with -e), how can I tell fpc to link against them instead of the
.ppu and .o? I've tried with -XD with no success.
--
View this message in conte
Am 22.08.2012 13:47, schrieb OBones:
Jonas Maebe wrote:
OBones wrote on Wed, 22 Aug 2012:
Olivier SANNIER wrote:
Does any of you have any suggestion as to explain this behavior, and
best of
all, how to fix it?
I went further and changed the code this so that I can get better
details as to w
On Wed, Aug 22, 2012 at 3:57 AM, Marco van de Voort wrote:
> In our previous episode, Marcos Douglas said:
>> Yes, we know too... but how prevent this using third-party libs?
>> For example, what do you think if I have a project that have a file
>> called lnet.pas? Can I use this identifier?
>
> A
Am 22.08.2012 12:37, schrieb Timothy Madden:
On 08/22/2012 09:57 AM, Marco van de Voort wrote:
In our previous episode, Marcos Douglas said:
[...]
Requiring all functions to be allways fully qualified will make the
most commom case worse than current implementation just to be better
for the
> There are more problems: the library name is not
> cross-platform, so the project will no longer be
> cross-platform. Which is a far more serious problem, IMHO.
> (and one which would exist even if I had followed your
> suggestion of implementing the property in TSQLConnection).
>
That is no
Jonas Maebe wrote:
OBones wrote on Wed, 22 Aug 2012:
Jonas Maebe wrote:
You can try debugging syswin32_i386_exception_handler in
rtl/win32/system.pp to see what exception Win32 reports instead.
Thanks for the confirmation as I was heading that way when I posted
the message, without much ho
Jürgen Hestermann wrote on Wed, 22 Aug 2012:
Am 2012-08-22 13:45, schrieb Lukasz Sokol:
Recompilation when one of units/libraries was upgraded, almost
always implies
rebuild. Who on Earth is so trigger-happy to upgrade, rebuild and
not test for
regressions? This is the only sane way to get the
OBones wrote on Wed, 22 Aug 2012:
Jonas Maebe wrote:
You can try debugging syswin32_i386_exception_handler in
rtl/win32/system.pp to see what exception Win32 reports instead.
Thanks for the confirmation as I was heading that way when I posted
the message, without much hope.
But now that I
Am 2012-08-22 13:45, schrieb Lukasz Sokol:
> Recompilation when one of units/libraries was upgraded, almost always implies
> rebuild. Who on Earth is so trigger-happy to upgrade, rebuild and not test for
> regressions? This is the only sane way to get the process going - document it
> *as MML said
On Wed, 22 Aug 2012, Ludo Brands wrote:
(In the IDE: drop, set connector type from picklist, maybe
change library name, set to enabled).
I'm having some problems using this in the IDE. The picklist is empty.
Probably something missing in
procedure TSQLDBConnectorTypePropertyEditor.GetValues
> Also, if I understand the mode of operation correctly, the
> TSQLDBLibraryLoader, when enabled, loads the library before
> the TxxxConnection tries to load it. But how can you be sure
> TSQLDBLibraryLoader is streamed before a TxxxConnection that
> has connected set to true
>
Previous messag
Jonas Maebe wrote:
OBones wrote on Wed, 22 Aug 2012:
Olivier SANNIER wrote:
Does any of you have any suggestion as to explain this behavior, and
best of
all, how to fix it?
I went further and changed the code this so that I can get better
details as to where the exception is coming from:
On 21/08/2012 17:43, Timothy Madden wrote:
> On 08/21/2012 03:10 PM, Jorge Aldo G. de F. Junior wrote:
>> "With no error messages, or even with no changes to the program since 1
>> and a half year in the repository, the scientific calculations are now
>> all blown up, and program outputs only error
On Wed, 22 Aug 2012, Ludo Brands wrote:
After recent discussions about which library to load when
loading DB client libraries, I changed the default names for
Firebird, MySQL, Postgres.
That was well in time ;) Thanks.
Another little bug and patch in http://bugs.freepascal.org/view.php?id
OBones wrote on Wed, 22 Aug 2012:
Olivier SANNIER wrote:
Does any of you have any suggestion as to explain this behavior, and best of
all, how to fix it?
I went further and changed the code this so that I can get better
details as to where the exception is coming from:
EControlC is the tra
Olivier SANNIER wrote:
Does any of you have any suggestion as to explain this behavior, and best of
all, how to fix it?
I went further and changed the code this so that I can get better
details as to where the exception is coming from:
program test;
uses
sysutils,
classes;
function GetMX
> (In the IDE: drop, set connector type from picklist, maybe
> change library name, set to enabled).
>
I'm having some problems using this in the IDE. The picklist is empty.
Probably something missing in
procedure TSQLDBConnectorTypePropertyEditor.GetValues(
Proc: TGetStrProc);
Var
L : TSt
On 08/22/2012 11:30 AM, Sven Barth wrote:
> Am 21.08.2012 21:42, schrieb Jorge Aldo G. de F. Junior:
>> Instead of implementing the half-assed C++ namespace model, just add
>> to the compiler a warning when it detects that there is a collision in
>> the current scope (two functions with the same pa
On 08/22/2012 09:57 AM, Marco van de Voort wrote:
> In our previous episode, Marcos Douglas said:
[...]
>
>>> Requiring all functions to be allways fully qualified will make the
>>> most commom case worse than current implementation just to be better
>>> for the uncommon case. This is a step back,
>
> After recent discussions about which library to load when
> loading DB client libraries, I changed the default names for
> Firebird, MySQL, Postgres.
>
That was well in time ;) Thanks.
Another little bug and patch in http://bugs.freepascal.org/view.php?id=22691
Ludo
Just saw this commit fly by:
r22165 | michael | 2012-08-21 21:43:26 +0200 (di, 21 aug 2012) | 1 line
* pas2ut, initial version (create unit tests from pascal unit)
That looks very interesting! Thanks for this, I'll look into it.
Regards,
Reinier
___
f
Am 21.08.2012 21:42, schrieb Jorge Aldo G. de F. Junior:
Instead of implementing the half-assed C++ namespace model, just add
to the compiler a warning when it detects that there is a collision in
the current scope (two functions with the same parameters from
different units that can be called fr
Michael Van Canneyt wrote:
Hi,
After recent discussions about which library to load when loading DB client
libraries, I changed the default names for Firebird, MySQL, Postgres.
araminta: >./loadlibdemo list
Available connection types:
Firebird, Default library name: libfbclient.so.2.5
MySQL 4.0
34 matches
Mail list logo
