TOPIC:

How close to integrate the Pandroid to CodeTyphon?
What prepares or planning CT about apk JVM FPC?

Hello Sternas,

Thanks for the quick response, problem solved! It was indeed a naming conflict.

I pre pended the package name ending with a underscore to the bmp names (should have done that when i ported the package from Delphi7 as standard practice) and made a .res instead of lrs (deprecated) of it to be future proof.

Maybe useful for others, i extended my fpc.cfg to crosscompile to arm (Beaglebone and RPi) under linux from the IDE by conditionally setting the searchpath for libraries.

# searchpath for libraries
#IFDEF CPUARM
#IFDEF Raspbian2
-Fl/usr/local/codetyphon/binLibraries/arm-linux-Raspbian2
#ESLSE
-Fl/usr/local/codetyphon/binLibraries/arm-linux-Beagleboard
#ENDIF
#ELSE
-Fl/usr/lib64
-Fl/usr/lib/x86_64-linux-gnu
-Fl/usr/lib/gcc/x86_64-linux-gnu/4.9
-Fl/usr/lib/gcc/x86_64-linux-gnu/4.8
-Fl/usr/lib/gcc/x86_64-linux-gnu/4.7
-Fl/usr/lib/gcc/x86_64-linux-gnu/4.6
-Fl/usr/lib/gcc/x86_64-linux-gnu/4.2.3
-Fl/usr/lib64/gcc/x86_64-unknown-linux-gnu/4.7.4
-Fl/usr/lib/gcc/x86_64-redhat-linux/4.8.2
-Fl/usr/lib64/gcc/x86_64-suse-linux/4.8/
-Fl/usr/lib/gcc/x86_64-redhat-linux/4.8.3
#ENDIF

Since i mostly cross compile for the Beaglebone i opted for a extra ifdef for RPi. For cross compiling to RPI just add -dRaspbian2 to custom options in your rpi target build mode.

Again thanks!

Thanks Sir
have fun...

Avra wrote:

---

Please add mORMot. All the FPC and Lazarus RTTI integration is already done at newpascal.org

+1

Vitaly wrote:

Avra wrote:

---

Please add mORMot. All the FPC and Lazarus RTTI integration is already done at newpascal.org

+1

Mormot is already good documented but has no ready support for Linux and needs Delphi partially (See Synopse-Homepage). tiOPF is really working fine on Windows, Mac and Linux with FPC but is less documented.

I don't know about Linux support, because I use win and my target is windows server (what is this? synopse.info/files/html/Synopse%20mORMot...201.18.html#TITL_142)

Delphi-needing: it' s true, I use delphi, but only for generating missing interface RTTI (delphi 6 totally enough). In other cases, like classic mORMot client-server, I don't need delphi at all.

tiOPF obviously good, but (as I understood) it is only ORM without any web-services abilities for example.
I work with mORMot (with CT) for less than month and didn't work with tiOPF at all. So, I admit that I can be wrong. Still I don't see how tiOPF can replace mORMot for me.

You're right, there is some linux support and I will test it soon. I think if mORMot is working with linux nativly it is the better solution. Also the documentation in mORMot is mutch better and complete. What I can't see is how to generate the RTTI interface for nativ Linux, when I have to use Delphi for that.
Thank you for the above link - I hadn't seen this.
There are web-services also in tiOPF but there is no json support in th moment (its in development)
I use tiOPF for some weeks now and I did not know all functions for now. The documentation is incomplete but there are many demo projects.

Bad News from Synopse on 2016-02-11:

'Sadly, we do not have full support of all FPC versions.
We would like perhaps to validate a particular version of FPC...

Any version of Delphi, since Delphi 6 would work just fine with mORMot.

Honestly, I did not have time lately to focus back on support of latest versions of FPC...
Since we do not use FPC here on production, and are implementing features on proprietary mORMot-based code using Delphi and (old) CrossKylix,'



There are warnings to not use mORMot on FPC for production purpose. (But many did this without any issues !)
There are also massive problems with string constants in FPC with mORMot. (may be it's a fpc problem in codepages)

(See mORMot forums)

If all problems are solved mORMot will be the better solution I think.

Why not implement both tiOPF and mORMot? Its not a great thing to do this ...

Klaus Riesterer wrote: If all problems are solved mORMot will be the better solution I think.

Why not implement both tiOPF and mORMot? Its not a great thing to do this ...

That's why I posted "+1" on mORMot proposal I used CT for years (instead of Lazarus) and became too accustomed to get everything I need out of box at any time. [Many thanks to CT Team by the way, many!]
Now I use mORMot with CodeTyphon. There are some troubles, but they are all "overcomeable".
Therefore I'm sure that CT Team will solve all that and future problems while implementing mORMot. Of course, if they decide to do it

You are right! I'm totally for implementing every single qualified library in CT. It will be only better for everybody.
For example, I didn't ever used even half of components, introduced in CodeTyphon. But I'm pleased to have an ability to use them, if I get such a need.

Vitaly,

Your information on mORMot is outdated. It supports and follows the FPC trunk. There is a contributor (Alf) who helped the Linux and Android development greatly. There are people actively developing Linux servers with FPC, me included. Delphi starter is available for free now. It is just one extra step to generate the necessary RTTI for interfaces. You can create a batch file for that. No biggy.

Leslie wrote: Vitaly,

Your information on mORMot is outdated. It supports and follows the FPC trunk. There is a contributor (Alf) who helped the Linux and Android development greatly. There are people actively developing Linux servers with FPC, me included. Delphi starter is available for free now. It is just one extra step to generate the necessary RTTI for interfaces. You can create a batch file for that. No biggy.

With some help in other forums I get mORMot running under Linux now. May be that my informations are outdated, but they are from official source.

To use Delphi in any form is not what I will do, either payed nor free. I have licences for Delphi as I used it for many years, but I will not do a new Setup on Windows - I use Linux.

Looking at the sample codes in the mORMot demo folder I decided to use tiOPF. Its - in my opinion - the better, more clear and structured solution and it is highly integrated in Lazarus, especially with fpGUI.
Please don't missunderstand, mORMot is also great and I will follow it's future! May be some day ...

Leslie wrote: Vitaly,

Your information on mORMot is outdated. It supports and follows the FPC trunk. There is a contributor (Alf) who helped the Linux and Android development greatly. There are people actively developing Linux servers with FPC, me included. Delphi starter is available for free now. It is just one extra step to generate the necessary RTTI for interfaces. You can create a batch file for that. No biggy.

I didn't say that mORMot doesn't support fpc. Where did you see that?
In fact I wrote that I successfully use it with FPC (CT) in my current projects.

And I said I have mORMot running with fpc on Linux now ...

Guys, I am sorry, did not mean to offend anyone. I have not read the whole topic, but the over all impression was that mORMot with FPC is not ready for Linux.

Leslie wrote: Guys, I am sorry, did not mean to offend anyone. I have not read the whole topic, but the over all impression was that mORMot with FPC is not ready for Linux.

That's ok
Let me summarize our discussion: mORMot is good, worth and ready for implementing in CT build
I believe everybody will agree with that.

We release Lab CodeTyphon ver 5.90 revision 005850 (RC)
This version has a lot of new things and many fixes/improvements
More at first post

Please, test this LAB CT release and report/suggest here

-FIX aarch64 (arm64) CrossBuild problems



Distributor ID: Debian 64bit
Description: Debian GNU/Linux 8.5 (jessie)
Release: 8.5
Codename: jessie



CT 5.9 64bits
[ERROR]: FPC64 Cross Element aarch64-android Param:xxxx NOT Exists
[ERROR]: FPC64 Cross Element x86_64-android Param:xxxx NOT Exists




Distributor ID: Debian 32bit
Description: Debian GNU/Linux 8.5 (jessie)
Release: 8.5
Codename: jessie

CT 5.9 32bits
[ERROR]: FPC64 Cross Element aarch64-android Param:xxxx NOT Exists
[ERROR]: FPC64 Cross Element x86_64-android Param:xxxx NOT Exists

Sir, just a moment to boot Debian...
testing...

1)The problem is that FreePascal NOT support
aarch64-android and
x86_64-android yet..

FPC SVN Android possible targets CPU are: arm, i386 jvm and mipsel

2) The Lab CT 5.9 rev 005850, crossbuild scripts are OK
and give this error on aarch64-android

Which is correct, for 1) reason

so, we must wait .... for aarch64-android and x86_64-android

3) aarch64-Linux work OK

it's true

Problem is
aarch64-android and
x86_64-android

Hello sir,

I just tested 02-11-2016 LAB CT 5.9.0 Revision 005850 [RC] on Ubuntu 16.04 x86-64 xfce and the source editor page remains invisible on the editor tab when the embedded form designer is disabled.

Steps to reproduce, in te ide menu go to "Tools > Options > General Extra" and disable "Use Embedded Form Designer" and restart the ide.

This happens in both docked and undocked modus. Switching between form and editor with F12 (classic key map) does focus the code editor tab but the editor remains invisible. In 04-10-2016 LAB CT 5.9.0 Revision 005840 [BETA 2] version the code editor is visible with the embedded form designer disabled.

* I can't use the embedded form designer because of custom mdi form hackery under gtk2...