Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread LacaK

Hi,
I have question indirectly related to this subject ;-)
If there are fixed some bugs in trunk (2.7.1 now) can we expect that 
they will be backported in some of next stable release 2.6 (2.6.2, 2.6.3)?

If yes , are there any rules which bugs are / can be / will be backported ?
(Now I am not speaking about new features added, but only about bugs fixed)
Depends it on fact how complex / risky is fix or must be backporting 
explicitly asked by sb ?

Thanks
-Laco.
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Hans-Peter Diettrich

Paul Ishenin schrieb:

02.05.12 18:18, Graeme Geldenhuys wrote:


So instead of jumping around with various delphi versions (a bit of D7
and a bit of 2009 etc), maybe start from the oldest delphi version
(eg: D7) and move towards the newest?


Maybe you can teach us how to do this by sending appropriate patches? We 
will sit then and criticize them.


You cannot expect *users* to supply patches, it's already work enough to 
only supply test programs which allow to reproduce failures. FPC should 
not be patchwork, this could be achieved even without core developers.


DoDi

___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Hans-Peter Diettrich

Tomas Hajny schrieb:

On Wed, May 2, 2012 12:27, Hans-Peter Diettrich wrote:

Marco van de Voort schrieb:

In our previous episode, Hans-Peter Diettrich said:

The solution is so easy: don't mark it as deprecated.

I also have a clear opinion about Delphi compatibility: Every FPC
version must be compatible to a Delphi version. A mix of incompatible
features from different Delphi versions is useless.

We don't have packages. That is what, D3? functionality, so we should
now
strip down FPC to D3 level?

There exist *limitations* due to the FPC multi-platform approach, of
course.


We cannot claim full compatibility to any Delphi version and we cannot
provide complete list of incompatibilities between any particular FPC
version and any particular Delphi version. If you're able to create
something like that, you're welcome to document and maintain it (e.g. in
our Wiki).


This would require a list of FPC features, which could be compared to a 
list of Delphi features.



As far as what is more advantageous for users or not - your statement is
apparently based on certain assumptions which are probably not valid for
all FPC users and probably not even most FPC users. In particular, your
statement is fully correct for users extensively using almost all features
provided in a particular Delphi version and always rewriting their own
previous code (based on language constructs and units delivered with
earlier Delphi versions) to the newest set.


I had in mind users which must maintain legacy software, written for a 
specific Delphi version. Or for newbies, which have a working Delphi 
project and want to check whether it would work with FPC/Lazarus, too.


Some picky Delphi users jump in whenever I suggest to have a look at 
FPC/Lazarus, in a Delphi forum, and report that they downloaded the 
latest release and tried to run a simple test project, which failed 
miserably.


Other people observe a strange behaviour in an FPC/Lazarus project, and 
write a test program to find out what Delphi does. They are not really 
happy when either Delphi or FPC refuse to even *build* that project, due 
to arbitrary incompatibilities.


DoDi

___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Graeme Geldenhuys
On 2 May 2012 15:59, Paul Ishenin  wrote:
> ...
> Details about these new features can be found at
> http://wiki.freepascal.org/FPC_New_Features_2.6.0


Thanks Paul. I didn't know (or forgot) about these links. This would
be helpful in creating a wiki matrix page of FPC vs Delphi features.
If I get the Delphi versions wrong, maybe somebody with more knowledge
of recent Delphi versions could correct those for me.


-- 
Regards,
  - Graeme -


___
fpGUI - a cross-platform Free Pascal GUI toolkit
http://fpgui.sourceforge.net
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Marcos Douglas
On Wed, May 2, 2012 at 11:47 AM, Martin Schreiber  wrote:
> On 02.05.2012 15:41, Marcos Douglas wrote:
>>
>>> This last one is bad advice, this code will break as soon as they switch to
>>> 2.6.3. Which, presumably, eventually they will.
>>>
>>> If you really want to avoid the messages, it is better to use TBookmark,
>>> GetBookmarkData and FreeBookmark.
>>> That will not generate warnings and will continue to work with 2.6.3.
>> Well, that was the way I chose.
>>
> Please don't forget to call FreeBookmark() in a try ... finally block
> otherwise there will be memory leaks in case of exceptions.

Yes, I do this always.
This is not the best way, I agree, but works.

>>> Or you can simply ignore the warnings, which is by far the easiest option.
>>> I can't believe people are making such fuss over a couple of warnings.
>>>
>>> Michael.
>> The problem, for me, would be break the sources in production.
>> If GetBookmarkData and FreeBookmark will continue work so, that is
>> what I will use.
>>
> You could define a "bookmarkty" alias yourself if Lazarus does not
> provide it.

This is one idea, yes. But I will wait the finish of this history and
what will be decided.

Marcos Douglas
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Trolling about: Breaking change in FPC 2.6.1

2012-05-02 Thread Mattias Gaertner

Giuliano Colla  hat am 2. Mai 2012 um 17:21
geschrieben:

> Mattias Gaertner ha scritto:
>[...]
> > This is starting to become off-topic.
>[...]
> Breaking compatibility is something very similar to leaving an old lady
> in the middle of the road.


Now it's off topic.




> Whenever people agree with me, I always feel I must be wrong (O. Wilde)



Since you have this signature for a long time, I assume you agree with it.


I agree with Martin that it is unusual to mark something deprecated in a fixes
branch.



About trolling: Provoking Graeme has never helped.

Mattias
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


[fpc-devel] Trolling about: Breaking change in FPC 2.6.1

2012-05-02 Thread Giuliano Colla

Mattias Gaertner ha scritto:

On Wed, 02 May 2012 18:49:13 +0800
Paul Ishenin  wrote:


02.05.12 18:18, Graeme Geldenhuys wrote:


So instead of jumping around with various delphi versions (a bit of D7
and a bit of 2009 etc), maybe start from the oldest delphi version
(eg: D7) and move towards the newest?
Maybe you can teach us how to do this by sending appropriate patches? We 
will sit then and criticize them.


If you don't like the dish - cook yourself.


Please don't feed the trolls. This is starting to become off-topic.


IMHO a more open attitude towards criticism, instead of just labeling it 
as "trolling" would bring big benefits to everyone.


I'm perfectly aware that we're all in debt to voluntary work created by 
good willing people. Nonetheless, whenever you undertake a task, be it 
for free or for a fee, you assume responsibilities for what you're 
doing, because there are people relying on you.


If I undertake the task to help old ladies to cross the road, every old 
lady I help should be reasonably thankful for that.
But if I leave an old lady in the middle of the road, to help another 
one, I don't believe that the first lady will be thankful. She may even, 
and rightly, use strong words against me.


Breaking compatibility is something very similar to leaving an old lady 
in the middle of the road.


If one knows how to help old ladies then he's welcome to do it. But if 
he happens to leave them in the middle of the road, then he'd better 
think twice before saying "If you don't like my way, then cross by 
yourself" to the lady standing in the middle of the highway and shouting 
between cars and buses!


One of the reasons of the average good quality of open source, is that 
it takes advantage of user feedback. Better remember it, from time to time.


--
Giuliano Colla

Whenever people agree with me, I always feel I must be wrong (O. Wilde)
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Martin Schreiber
On 02.05.2012 10:47, michael.vancann...@wisa.be wrote:
>
>
> Or you can simply ignore the warnings, which is by far the easiest
> option.
> I can't believe people are making such fuss over a couple of warnings.
>
MSEide+MSEgui compiles without FPC warnings and notes. I assume other
FPC users also make code without warnings and notes. While waiting for
compilation we check the compiler output in message window. If there are
warnings from your deprecated statement we have to check every time if
it is a real warning or a "FPC-Delphi-Future" warning which can't be
fixed without using try-finally blocks which should later be removed
with FPC trunk because they are redundant thanks to the managed TBytes type.

Martin
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Martin Schreiber
On 02.05.2012 15:41, Marcos Douglas wrote:
>
>> This last one is bad advice, this code will break as soon as they switch to
>> 2.6.3. Which, presumably, eventually they will.
>>
>> If you really want to avoid the messages, it is better to use TBookmark,
>> GetBookmarkData and FreeBookmark.
>> That will not generate warnings and will continue to work with 2.6.3.
> Well, that was the way I chose.
>
Please don't forget to call FreeBookmark() in a try ... finally block
otherwise there will be memory leaks in case of exceptions.
>> Or you can simply ignore the warnings, which is by far the easiest option.
>> I can't believe people are making such fuss over a couple of warnings.
>>
>> Michael.
> The problem, for me, would be break the sources in production.
> If GetBookmarkData and FreeBookmark will continue work so, that is
> what I will use.
>
You could define a "bookmarkty" alias yourself if Lazarus does not
provide it.

Martin
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Paul Ishenin

02.05.12 21:32, Graeme Geldenhuys wrote:


Maybe the FPC core team will be so kind as to create a new "FPC
features since xxx" page on the wiki (similar to what I have done for
Lazarus).


From the anouncement of FPC 2.6.0 which was posted to this mail list:

Changes that may break backwards compatibility are documented at:
http://wiki.freepascal.org/User_Changes_2.6.0
...
Details about these new features can be found at
http://wiki.freepascal.org/FPC_New_Features_2.6.0

Best regards,
Paul Ishenin
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Marcos Douglas
On Wed, May 2, 2012 at 5:47 AM,   wrote:
>
>
> On Wed, 2 May 2012, Martin Schreiber wrote:
>
>> On 01.05.2012 17:37, Michael Van Canneyt wrote:

 As written before, in MSEgui I'll define a "bookmarkty" type, so
 MSEgui users
 have "bookmarkty" in order to avoid the warning. FPC and Lazarus
 probably
 can't do the same because of Delphi compatibility. Suggestion:
 remove "deprecated" from TBookmarkStr in fixes_2_6.
>>>
>>>
>>>
>>> Well, I think it is better to warn people of a coming change.
>>>
>>> So I have added a description that warns people that it will disappear
>>> in 2.6.3.
>>>
>> For users who do not want to see the "deprecated" warnings in fixes_2_6 do
>>
>> In MSEide+MSEgui:
>>
>> Use "bookmarkty" instead of "TBookmarkStr" for bookmark variables.
>>
>> In Lazarus:
>>
>> Use "string" instead of "TBookmarkStr" for bookmark variables.
>
>
> This last one is bad advice, this code will break as soon as they switch to
> 2.6.3. Which, presumably, eventually they will.
>
> If you really want to avoid the messages, it is better to use TBookmark,
> GetBookmarkData and FreeBookmark.
> That will not generate warnings and will continue to work with 2.6.3.

Well, that was the way I chose.

> Or you can simply ignore the warnings, which is by far the easiest option.
> I can't believe people are making such fuss over a couple of warnings.
>
> Michael.

The problem, for me, would be break the sources in production.
If GetBookmarkData and FreeBookmark will continue work so, that is
what I will use.

Marcos Douglas
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Graeme Geldenhuys
On 2 May 2012 13:55, Mattias Gaertner  wrote:
>
> Please don't feed the trolls.


By no means did I mean to troll. It was a legit statement, and
something that confuses the hell out of FPC developers like myself.

For example:  tiOPF branched off version 3 so as to support Delph
2009+ support including Unicode (though no delphi unicode features
have been added to tiOPF v3 yet). With the mixed bag of Delphi
features in Free Pascal, I have no idea when we will be able to add
FPC support to tiOPF v3??? A brief attempt showed that FPC 2.6.0 was
not able to work, and I have no idea what Delphi features are
implemented in FPC Trunk.

Maybe the FPC core team will be so kind as to create a new "FPC
features since xxx" page on the wiki (similar to what I have done for
Lazarus). And on that page have a table or roadmap indicating what
Delphi features are supported in what FPC versions. That would give
developers like myself, and other commercial entities a clear
indication of how "compatible" FPC is with Delphi and if dual-compiler
support is possible in our products.

@Paul & Mattias
I have no idea why you guys think this is trolling. From my point of
view [in our company], I need to show a business case for spending my
company time working on open source projects especially
dual-compiler (Delphi and FPC) projects like tiOPF that require a lot
more testing. And no, tiOPF is not the only project like this that our
company works on. And again, I don't think I am the only developer
with this problem either. I'm pretty sure other companies would
appreciate such clear and transparent information from the FPC core
team, after all, nobody else is more qualified to know what FPC does
and doesn't support.

@Everybody else
I'm perfectly fine with Michael's solution to this message thread. I
don't mind deprecated compiler warnings at all - this gives me enough
warning to update my affected source code before the next stable
release. But I wasn't appreciative of Macro's idea of simply breaking
a stable branch without warning (thus I raised my concern). I'm glad
this problem is solved though.

@Marco
By the looks of recent events, maybe you advice is indeed correct.
Maybe my idea of only using the latest released FPC version is too a
narrow minded view. Maybe I should test FPC Trunk every few weeks to
keep more up to date with FPC developments, and smooth out any
possible compatibility problems in our code in preparation of new FPC
releases.


-- 
Regards,
  - Graeme -


___
fpGUI - a cross-platform Free Pascal GUI toolkit
http://fpgui.sourceforge.net
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Tomas Hajny
On Wed, May 2, 2012 12:27, Hans-Peter Diettrich wrote:
> Marco van de Voort schrieb:
>> In our previous episode, Hans-Peter Diettrich said:
>>> The solution is so easy: don't mark it as deprecated.
>>>
>>> I also have a clear opinion about Delphi compatibility: Every FPC
>>> version must be compatible to a Delphi version. A mix of incompatible
>>> features from different Delphi versions is useless.
>>
>> We don't have packages. That is what, D3? functionality, so we should
>> now
>> strip down FPC to D3 level?
>
> There exist *limitations* due to the FPC multi-platform approach, of
> course.

We cannot claim full compatibility to any Delphi version and we cannot
provide complete list of incompatibilities between any particular FPC
version and any particular Delphi version. If you're able to create
something like that, you're welcome to document and maintain it (e.g. in
our Wiki).

As far as what is more advantageous for users or not - your statement is
apparently based on certain assumptions which are probably not valid for
all FPC users and probably not even most FPC users. In particular, your
statement is fully correct for users extensively using almost all features
provided in a particular Delphi version and always rewriting their own
previous code (based on language constructs and units delivered with
earlier Delphi versions) to the newest set. In reality, users typically
combine code created for earlier versions with selected parts available in
newer versions depending on their needs and priorities. Users like this
will probably appreciate availability of a certain very new Delphi feature
regardless from the fact that certain older Delphi feature may not be
supported in FPC yet. For one, they may not use the older feature at all.
Even if they do, users like this still need to change less code on their
side if the newer feature is supported in FPC.

Even when talking about incompatibilities between two particular Delphi
versions, they may be less or more relevant/important for a particular
user depending on his needs. Some code may be perfectly shared between
pre-Unicode and Unicode based Delphi RTL without any changes and some code
needs to be rewritten almost completely. In summary, I personally believe
that majority of our users would not benefit from your proposal.
Nevertheless, you're free to create a special branches selectively merging
changes from FPC trunk based on their compatibility with specific Delphi
versions. Something like that may be useful for people creating components
and units targetting both FPC and Delphi users, but not necessarily for
others. In any case, Delphi releases do not drive FPC releases and I
personally believe that it is right that way.

Tomas


___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Mattias Gaertner
On Wed, 02 May 2012 18:49:13 +0800
Paul Ishenin  wrote:

> 02.05.12 18:18, Graeme Geldenhuys wrote:
> 
> > So instead of jumping around with various delphi versions (a bit of D7
> > and a bit of 2009 etc), maybe start from the oldest delphi version
> > (eg: D7) and move towards the newest?
> 
> Maybe you can teach us how to do this by sending appropriate patches? We 
> will sit then and criticize them.
> 
> If you don't like the dish - cook yourself.

Please don't feed the trolls. This is starting to become off-topic.

Mattias
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Paul Ishenin

02.05.12 18:18, Graeme Geldenhuys wrote:


So instead of jumping around with various delphi versions (a bit of D7
and a bit of 2009 etc), maybe start from the oldest delphi version
(eg: D7) and move towards the newest?


Maybe you can teach us how to do this by sending appropriate patches? We 
will sit then and criticize them.


If you don't like the dish - cook yourself.

Best regards,
Paul Ishenin
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Graeme Geldenhuys
On 2 May 2012 09:00, Hans-Peter Diettrich  wrote:
> To the user this means "not compatible with D7 nor D2009" :-[

+1000

-- 
Regards,
  - Graeme -


___
fpGUI - a cross-platform Free Pascal GUI toolkit
http://fpgui.sourceforge.net
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Graeme Geldenhuys
On 1 May 2012 23:48, Hans-Peter Diettrich  wrote:
> Simply specify the Delphi version, to which FPC 2.6.1 should be compatible,
> then the rest is clear.


In a way I agree here... It is damn confusing when just saying "delphi
compatible". Delphi is NOT a product that is standing still any more,
so FPC should start saying which Delphi version they are compatible
with - otherwise the statement is just plain vague and helps nobody.
This fact of "unknown to which delphi version fpc is compatible with"
is causing a maintenance nightmare in some of the dual-compiler
(Delphi and FPC) open source projects I maintain. I'm pretty sure I'm
not alone in this situation too.

So instead of jumping around with various delphi versions (a bit of D7
and a bit of 2009 etc), maybe start from the oldest delphi version
(eg: D7) and move towards the newest?

-- 
Regards,
  - Graeme -


___
fpGUI - a cross-platform Free Pascal GUI toolkit
http://fpgui.sourceforge.net
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Hans-Peter Diettrich

Marco van de Voort schrieb:

In our previous episode, Hans-Peter Diettrich said:

The solution is so easy: don't mark it as deprecated.

I also have a clear opinion about Delphi compatibility: Every FPC
version must be compatible to a Delphi version. A mix of incompatible
features from different Delphi versions is useless.


We don't have packages. That is what, D3? functionality, so we should now
strip down FPC to D3 level?


There exist *limitations* due to the FPC multi-platform approach, of course.

DoDi

___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Martin Schreiber
 On 02.05.2012 10:47, michael.vancann...@wisa.be wrote:
>> For users who do not want to see the "deprecated" warnings in
>> fixes_2_6 do
>>
>> In MSEide+MSEgui:
>>
>> Use "bookmarkty" instead of "TBookmarkStr" for bookmark variables.
>>
>> In Lazarus:
>>
>> Use "string" instead of "TBookmarkStr" for bookmark variables.
>
>
> This last one is bad advice, this code will break as soon as they
> switch to 2.6.3. Which, presumably, eventually they will. 

That's the idea. When one switches to trunk the compiler will stop at
the wrong types and one can change the types possibly in conditional
defines if one needs FPC 2.6.0 compatibility.
Or maybe Lazarus can define "bookmarkty" too for those who don't care
about Delphi compatibility?

Martin
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread michael . vancanneyt



On Wed, 2 May 2012, Martin Schreiber wrote:


On 01.05.2012 17:37, Michael Van Canneyt wrote:

As written before, in MSEgui I'll define a "bookmarkty" type, so
MSEgui users
have "bookmarkty" in order to avoid the warning. FPC and Lazarus
probably
can't do the same because of Delphi compatibility. Suggestion:
remove "deprecated" from TBookmarkStr in fixes_2_6.



Well, I think it is better to warn people of a coming change.

So I have added a description that warns people that it will disappear
in 2.6.3.


For users who do not want to see the "deprecated" warnings in fixes_2_6 do

In MSEide+MSEgui:

Use "bookmarkty" instead of "TBookmarkStr" for bookmark variables.

In Lazarus:

Use "string" instead of "TBookmarkStr" for bookmark variables.


This last one is bad advice, this code will break as soon as they switch 
to 2.6.3. Which, presumably, eventually they will.


If you really want to avoid the messages, it is better to use 
TBookmark, GetBookmarkData and FreeBookmark.

That will not generate warnings and will continue to work with 2.6.3.

Or you can simply ignore the warnings, which is by far the easiest option.
I can't believe people are making such fuss over a couple of warnings.

Michael.
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Martin Schreiber
 On 01.05.2012 17:37, Michael Van Canneyt wrote:
>> As written before, in MSEgui I'll define a "bookmarkty" type, so
>> MSEgui users
>> have "bookmarkty" in order to avoid the warning. FPC and Lazarus
>> probably
>> can't do the same because of Delphi compatibility. Suggestion:
>> remove "deprecated" from TBookmarkStr in fixes_2_6.
>
>
> Well, I think it is better to warn people of a coming change.
>
> So I have added a description that warns people that it will disappear
> in 2.6.3.
>
For users who do not want to see the "deprecated" warnings in fixes_2_6 do

In MSEide+MSEgui:

Use "bookmarkty" instead of "TBookmarkStr" for bookmark variables.

In Lazarus:

Use "string" instead of "TBookmarkStr" for bookmark variables.

Martin

___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Marco van de Voort
In our previous episode, Hans-Peter Diettrich said:
> The solution is so easy: don't mark it as deprecated.
> 
> I also have a clear opinion about Delphi compatibility: Every FPC
> version must be compatible to a Delphi version. A mix of incompatible
> features from different Delphi versions is useless.

We don't have packages. That is what, D3? functionality, so we should now
strip down FPC to D3 level?

Lazarus likewise, due to MDI?
___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel


Re: [fpc-devel] Breaking change in FPC 2.6.1

2012-05-02 Thread Paul Ishenin

02.05.2012 15:00, Hans-Peter Diettrich wrote:

To the user this means "not compatible with D7 nor D2009" :-[


It is both compatible with D7 and some D2009 features.

Best regards,
Paul Ishenin.

___
fpc-devel maillist  -  fpc-devel@lists.freepascal.org
http://lists.freepascal.org/mailman/listinfo/fpc-devel