Welcome

aha · 5544

aha

  • Junior Member
  • **
    • Posts: 66
    • Karma: +41/-0
on: November 11, 2018, 11:13:32 AM
https://blog.alb42.de/2018/11/11/hex2-0-2/

Calculator Hex2

edit: moderator moved all reactions from the sticky welcome topic into this separate thread.
« Last Edit: November 15, 2018, 06:00:36 PM by magorium »



magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #1 on: November 13, 2018, 08:35:47 AM
Thank you aha... i seem to have missed that the calculator was updated.


nikos

  • Senior Member
  • ****
    • Posts: 374
    • Karma: +71/-3
    • aspireos
Reply #2 on: November 13, 2018, 08:40:29 AM
Whow. That looks real nice. Thx :)


aha

  • Junior Member
  • **
    • Posts: 66
    • Karma: +41/-0
Reply #3 on: November 13, 2018, 11:26:16 AM



magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #4 on: November 13, 2018, 01:01:25 PM
@aha,

Hehe... that one came in earlier at my u-tube feed...

Be sure to check out the software and give the developer some feedback ... that's the only way that things are able to improve (if necessary).

Thanks for posting aha.

edit PS: don't miss the updated MUIPlot.
« Last Edit: November 14, 2018, 03:09:27 PM by magorium »



jacko

  • Newbie
  • *
    • Posts: 36
    • Karma: +1/-0
Reply #5 on: November 13, 2018, 02:58:51 PM
Interesting programs. Reminds me a little bit of school.

Wasn't alb42 around in Aros-Exec, in the past?



magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #6 on: November 13, 2018, 04:07:40 PM
Wasn't alb42 around in Aros-Exec, in the past?
Yes, he was  :)


jacko

  • Newbie
  • *
    • Posts: 36
    • Karma: +1/-0
Reply #7 on: November 15, 2018, 02:57:30 AM
so he is not anymore?
I remember he usually posted his new programs here.



magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #8 on: November 15, 2018, 02:08:19 PM
so he is not anymore?
I have no clue whatsoever, so Your guess is as good as mine  :)

What i do know is that neither his usual handle or name is being listed as registered user on this new forum (you can check the memberlist yourself once logged in)

Quote
I remember he usually posted his new programs here.
On the old forums, yes. Here on amigalife i haven't seen any sign of life of ALB42 so far.

It is heartwarming to see how you seem to miss ALB... and i can't blame you as he's a great person...  :D

You can always contact him either by using his website alb42.de or by using the freepascal.org (lazarus) forums (he does have an account overthere).
« Last Edit: November 15, 2018, 02:13:02 PM by magorium »



Jackokring

  • Newbie
  • *
    • Posts: 13
    • Karma: +0/-0
Reply #9 on: November 15, 2018, 02:54:11 PM
Is code page 437 (classic IBM) available as a font or does the IDE render gfx?



magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #10 on: November 15, 2018, 05:16:06 PM
Is code page 437 (classic IBM) available as a font
i'm sure it is...

... but i am assuming that you wish to use the box drawing characters in a shell window. If so then that is not a very good idea as a) the shell font is a system wide (user) setting and b) a shell window on amiga/amiga is (by comparison) dead slow. Amiga is originally a graphical OS... the shell is build on top of that.

Quote
or does the IDE render gfx?
FP ide does indeed render the graphics into a intuition Window. One of the amicommon include files contains a (open source) vga BIOS font.

FP IDE (on amiga and alike OS) uses a virtual textmode buffer (with dimensions the size of the intuition expressed in a number of horizontal/vertical character count) that gets (re)rendered to the graphical intuition window when required.


Chain-Q

  • Newbie
  • *
    • Posts: 1
    • Karma: +0/-0
Reply #11 on: January 28, 2019, 05:45:33 AM
Err, not sure what the heck is going on with the new AROS Exec, or if this is even the right topic to post to, but anyway. Just want to share this news. After the FPC MorphOS version through yesterday night, the AROS version - at least on i386 for now - also gained proper support for the lineinfo unit, so address-to-sourceline mapping on runtime errors, software exceptions (i.e. induced from Pascal code) and heaptraces is now possible.

See here:
https://twitter.com/chainq/status/1089735617863471107

What's special about AROS and MorphOS (and OS4) is, that instead of just using the STAB section values in the debug binary as on most other systems, you have to relocate the STAB values on-the-fly, while searching for the address mapping in the debug info, otherwise it cannot work. So now I added support for these ELF relocations. And of course PowerPC and i386 flavors are slightly different, but the differences are not so big. Available in FPC 3.3.1 trunk right now, and in Alb42's nightly builds probably from tomorrow.

Stealth Ranger - Amiga demoscene and Free Pascal superhero


magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #12 on: January 31, 2019, 05:20:45 AM
Hi Chain-Q.

Welcome to the new aros-exec and thank you for your post (it seems to have skipped my rss feed and just noticed it).

It is ok to have posted here. I had the sticky thread locked in order to reserve it for news posts like the one you just posted. I see if i can add it to the sticky thread.

Just want to share this news. After the FPC MorphOS version through yesterday night, the AROS version - at least on i386 for now - also gained proper support for the lineinfo unit, so address-to-sourceline mapping on runtime errors, software exceptions (i.e. induced from Pascal code) and heaptraces is now possible.

See here:
https://twitter.com/chainq/status/1089735617863471107

What's special about AROS and MorphOS (and OS4) is, that instead of just using the STAB section values in the debug binary as on most other systems, you have to relocate the STAB values on-the-fly, while searching for the address mapping in the debug info, otherwise it cannot work. So now I added support for these ELF relocations. And of course PowerPC and i386 flavors are slightly different, but the differences are not so big. Available in FPC 3.3.1 trunk right now, and in Alb42's nightly builds probably from tomorrow.

Splendid news and thank you for the implementation. I have many questions about it as i took a stab at it myself but i refrain from doing so because i have a more pressing matter (and looking at your implementation would probably answer many of my questions as well):

Do you happen to know if FPC 3.2 release will support cross-compiling from arm to i386.... ?

right now it doesn't because i386 soft float is not implemented for ARM. Not a big showstopper but i got 68k an ppc cross-compilers working with vlink and vasm.

Second question would be to ask what is required to get line support added for 68k ? hunk debug is not supported atm, i presume ?



aGGreSSor

  • Member
  • ***
    • Posts: 184
    • Karma: +25/-0
    • russian transit
Reply #13 on: December 03, 2020, 07:31:38 AM
right now it doesn't because i386 soft float is not implemented for ARM. Not a big showstopper but i got 68k an ppc cross-compilers working with vlink and vasm.
I don't know who and what will support, but I builded this vasm and if you used it, you can probably test it (here)..


magorium

  • Moderator
  • Legendary Member
  • *****
    • Posts: 632
    • Karma: +62/-0
  • Convicted non contributor
Reply #14 on: December 03, 2020, 05:57:19 PM
I don't know who and what will support, but I builded this vasm and if you used it, you can probably test it (here)..
Thank you aGGreSSor, although i have those compiled as well (see my github. i should really update those  ??? ).

In this day and age i compile vlink/vasm etc. on-the-fly with some script that downloads all necessary sources and builds the FPC compiler as well.

The actual issue is that the Free Pascal compiler itself is unable to cross compile from arm (or any other non x86 platform) to i386 because of the fact that the i386 version of the Free Pascal Compiler uses the FPU floating point, and none of the the other platforms are currently able to 'emulate' that correctly until the Free Pascal softfloat unit is completely implemented/supported for the host architecture.

It is not a very big deal, other then annoying because i had intended to use a raspberry-pi for building things automatically for me whenever the sources are updated (targets AROS-i386, AROS-arm, MorphOS-ppc, OS4-ppc and 68k) and this is currently not possible because of the mentioned missing softfloat support. So atm i have to use my x86-64bit rig to do that.

That said, i haven't looked at the softfloat unit and its status for almost a year now, so it might be that things have improved over time.... (* see edit)

Thank you for the effort though. It is appreciated :thumbsup:


edit: *, just checked in the bugreport. Unfortunately there seem no progress  :'( 13 years and counting.... must be a record somewhere  ;D
« Last Edit: December 03, 2020, 06:26:33 PM by magorium »