Click Here
home features news forums classifieds faqs links search
6071 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
51 crawler(s) on-line.
 49 guest(s) on-line.
 1 member(s) on-line.


 Hammer

You are an anonymous user.
Register Now!
 Hammer:  8 secs ago
 DiscreetFX:  5 mins ago
 bhabbott:  9 mins ago
 gonegahgah:  11 mins ago
 Hypex:  27 mins ago
 MEGA_RJ_MICAL:  38 mins ago
 matthey:  2 hrs 1 min ago
 Gunnar:  2 hrs 5 mins ago
 agami:  4 hrs 21 mins ago
 OlafS25:  4 hrs 33 mins ago

/  Forum Index
   /  Classic Amiga Hardware
      /  News about Vampire and Apollo
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 | 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | 89 | 90 | 91 | 92 | 93 | 94 | 95 | 96 | 97 | 98 | 99 | 100 | 101 | 102 | 103 Next Page )
PosterThread
Kremlar 
Re: News about Vampire and Apollo
Posted on 21-Jun-2019 18:24:23
#1761 ]
Regular Member
Joined: 12-Aug-2010
Posts: 108
From: Milford, MA

@kolla

Quote:
See? I am not the one "demanding" 100% compatibility here - this _claim_ is coming from Gunnar himself - and this claim is yet to be proven correct.


I don't think you understand the word "goal".

 Status: Offline
Profile     Report this post  
Lou 
Re: News about Vampire and Apollo
Posted on 21-Jun-2019 20:38:53
#1762 ]
Elite Member
Joined: 2-Nov-2004
Posts: 4169
From: Rhode Island

@Overflow

Quote:

Overflow wrote:
@Lou

Being a non-technical guy, nor a developer;

But wont having more chips add complexity in development, in hardware, core development and production line cost/overhead?

It's an FPGA! What extra cost?
What is more complex than an entire cpu redesign?
Wasn't adding a single AMMX unit already complex - what would be hard about moving it onto the custom chipset so that it had DMA and could have 16, 32 or up to 256 units working at once on a chunk of chipram while the cpu ran actual game code?

These "demos" are nice but the Quake demos I saw had ridiculous screen tearing and simplistic lighting. Look at the PS1 port of Quake 2.
https://www.youtube.com/watch?v=ObPWiqbMp30
Considering the system was memory-limited and had to load often, it looks MILES ahead of any Vampire demo despite the clockspeed, memory size and memory bandwidth advantage of the Vampire. The PSX/PS1 was a 34Mhz machine with 2MB of RAM and 1 MB of VRAM (aka chipram in Amiga-land).

Last edited by Lou on 21-Jun-2019 at 08:46 PM.

 Status: Offline
Profile     Report this post  
ne_one 
Re: News about Vampire and Apollo
Posted on 21-Jun-2019 20:40:22
#1763 ]
Cult Member
Joined: 13-Jun-2005
Posts: 905
From: Unknown

@Birbo

Quote:
This was really disappointing to me. Sill, I will probably buy a Vampire Standalone, because I think it's an amazing piece of hardware and I like the idea behind. But I will probably not develop anything for it.


Which begs the question: you abandoned a potentially rewarding project based on an unsubstantiated action in a forum?

This traditionally happened with fora and is now prevalent with social media. But ultimately, why let the actions of others dissuade you from doing something when you know going in that there is a high probability of exposing yourself to idiocy?

 Status: Offline
Profile     Report this post  
ppcamiga1 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 4:56:38
#1764 ]
Cult Member
Joined: 23-Aug-2015
Posts: 762
From: Unknown

On vampire core 2.11 when one use calculator 1.3 * 5 will be 6.5 ?
Not 6.49999999 like on vampire core 2.8 from march 2018 ?

Ok. It is good to know that finally after so many years vampire fpu is compatible.

Still MMU, 3D and speed left.

MPU without software in real file equals to no MMU at all.

Some extremely dumb powerpc haters may think it is nothing, but for any sane member of amiga community it is obvoius that MMU provide big help for developers.

3D. Vampire still not get PS1 quality and performance.

MMU, 3D nad performance are good reasons to buy Amiga NG hardware.

Comparing to rpi is stupid beacuse rpi do not provide binary and source compatibility to old Amiga Os software.

Last edited by ppcamiga1 on 22-Jun-2019 at 05:06 AM.

 Status: Offline
Profile     Report this post  
megol 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 9:31:14
#1765 ]
Regular Member
Joined: 17-Mar-2008
Posts: 355
From: Unknown

@ferrels

Quote:

ferrels wrote:
@megol
Quote:
Well he isn't too positive but often correct. I'd take correct over positive.


You have an odd perspective on correct. For months Kolla attacked the Apollo Team relentlessly over the lack of an MMU and FPU even after Gunnar announced that an FPU with 100% compatibility was the goal. This hasn't changed. And there is an MMU. Gunnar just hasn't released the info on how to take advantage of it, probably just to spite Kolla for his months-long attacks.

I agree with describing it as relentless attacks even to the level of derailing threads but the questions or statements are correct (AFAIK).

A MMU not compatible with the 68040 or 68060 is of very little use outside the Amiga market, that incompatible MMU not being documented is of even less use.

Quote:

As for demos, many of them won't run on stock or accelerated Amigas for various reasons, most of them hardware related and they will never run on every Amiga. I'd wager that the Vampire runs more Amiga demos than most stock or accelerated Amigas, simply because the team still actively develops the Apollo core and responds when someone finds a demo that doesn't work. You won't see that happen for classic, stock or accelerated Amigas unless you're under the delusion that Hyperion will come to the rescue or that Commodore was just resurrected.

This is true, don't know if the solution used on the Vampire works well in general* but as you say ensuring compatibility is a huge plus for running old hardware hitting games. Active development in general also helps of course.

(* actually don't know how it works at all and details sadly tend to be secretive)

Quote:

As for me attacking anyone in a personal manner, that's just bull. I don't know what Kolla looks like, what he smells like, or anything else about him that's personal in nature nor did I refer to anything about his person. I referenced how he was banned at the Apollo Forums. Kolla was banned at least twice from the Apollo forums, first when using his Kolla profile for his trolling and then again after he registered a shill profile when his Kolla account was permanently banned. Since when has stating a fact become a personal attack? There's quite a long thread about it over at the Apollo forum where he and Gunnar go back and forth over his misuse/abuse of accounts and constant trolling. He's also been threatened numerous times over at EAB for the same behavior. The ax he has to grind with Gunnar and the Apollo Team goes far beyond personal. Whenever anyone from any forum discusses the Vampire, he inevitably appears and starts attacking by parroting the same comments over and over like a broken record.


You accused him of often being banned, ban evasion, being on an anti-Vampire crusade, having obsessive compulsive disorder, and stalking. None of which is a factual arguments or having any bearing on him being wrong = ad hominem.

The only things that can be seen as anything but a personal attack is the claim that striving for 100% is retarding development (could be true) and that the current FPU can run "nearly every demo" (not relevant). That doesn't make sense though as the claim of 100% compatibility came from Gunnar and came before the fixes to some demos that required more precision than the old FPU, precision that a 100% compatible FPU would have had from the beginning.

@Kremlar
Quote:

Kremlar wrote:
I don't think you understand the word "goal".

I think you don't understand the word "is".

Last edited by megol on 22-Jun-2019 at 09:36 AM.
Last edited by megol on 22-Jun-2019 at 09:35 AM.

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 15:23:32
#1766 ]
Elite Member
Joined: 9-Jun-2004
Posts: 12791
From: Norway

@ppcamiga1

Quote:
but for any sane member of amiga community it is obvious that MMU provide big help for developers.


but not all developers need it, many program languages work on higher level with lots of checking to check valid inputs, don't poke into memory mutch.

Like Amos, Arexx, phyton, ruby, php, javascript.


Quote:
Comparing to rpi is stupid beacuse rpi do not provide binary and source compatibility to old Amiga Os software.


there is version of AROS that works on ARM, so it can be more or less compared, not sure if the drivers are written yet, or has to run hosted on Linux.

Last edited by NutsAboutAmiga on 22-Jun-2019 at 03:28 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
Lou 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 16:37:42
#1767 ]
Elite Member
Joined: 2-Nov-2004
Posts: 4169
From: Rhode Island

@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote:
@ppcamiga1


Quote:
Comparing to rpi is stupid beacuse rpi do not provide binary and source compatibility to old Amiga Os software.


there is version of AROS that works on ARM, so it can be more or less compared, not sure if the drivers are written yet, or has to run hosted on Linux.

He's just mad that the AROS-driven Pi-miga destroys the price/performance of the UNIVERSE!!!
...and BEYOND!!!

Last edited by Lou on 22-Jun-2019 at 04:38 PM.

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 17:26:51
#1768 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@NutsAboutAmiga


Quote:
but not all developers need it, many program languages work on higher level with lots of checking to check valid inputs, don't poke into memory mutch. Like Amos, Arexx, phyton, ruby, php, javascript.


You left out the primary, lower-level languages of C and C++. I've been a C/C++ developer since the late 1980's and started out on the Amiga. On the Amiga, an MMU was/is essentially useless for development and well over 90% of all Amigas never had an MMU anyway. Most developers and publishers targeted the lowest common denominator in order to maximize profits, and for Amigas, that was a stock A500 which didn't have an MMU.

It seems that those who are crying the loudest about the Vampire's lack of a compatible MMU are the ones who know least the very purpose of said MMU. The purpose of an MMU is to manage virtual memory, something which classic Amigas and AmigaOS never had, even to this day. Some enterprising hacks were written to simulate virtual memory using the MMU as well as some hacks to assist some developers who had poor programming habits, such as overwriting areas of memory that had already been allocated for use, either by the OS or by other apps. Overwriting such areas caused an unstable system at best and a full system crash at worst.

And these days, anyone doing development for classic Amigas uses a PC and a cross-compiler such as GCC or AmiDevC++ or they develop under WinUAE using a period C/C++ compiler such as SAS/Lattice and then test their code on a real Amiga. So the argument that the Vampire is somehow flawed or less of an Amiga than an original, classic Amiga because it lacks an MMU is simply ludicrous.

About two years ago I undertook the challenge of updating the PLPLOT plotting library from version 2.6 to version 5.0.1 for classic Amigas. I decided to use a real, classic Amiga with period development tools from the early 1990's: PLPLOT 5.0.1

I conducted my development on an A1200 which had been upgraded many years ago with a 68030 board and Lattice/SAS C version 6. The MMU wasn't used at all during any of the development. Any system crashes were minimal but development was extremely slow due to the speed of such an ancient system. I won't be repeating any such challenges due to the amount of time wasted by using such a slow system but will instead use AmiDevC++ on a PC and have the resulting binaries stored in a folder that is shared with WinUAE for testing. Using real, classic hardware for development is simply too slow to be practical these days.

So you have to wonder why some people here are crying so loudly about the Vampire not having an MMU when the very Amiga they have at home on their desks don't have an MMU either.

 Status: Offline
Profile     Report this post  
kolla 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 20:43:41
#1769 ]
Elite Member
Joined: 20-Aug-2003
Posts: 2852
From: Trondheim, Norway

@ferrels

The only Amiga systems I have that have no MMU are the FPGA based ones and CDTV, everything else has at least 030 upwards (and no EC). You are correct that most users didn't have, nor need, MMU on Amiga, as most users just used their Amiga systems for games. But for development on the platform, it was rather essential. But then again, these days development for Amiga does not happen on Amiga anymore, but rather in emulators, WinUAE, Qemu, Vamos etc.

Anyways, you may not have noticed, but my "whining" regarding the AC68080 was not so much about the lack of it, as it was about it being hidden, and about the constant claim that "nobody needs it", yet it is there in the core, hidden and (publicly) undocumented, used to do various tricks behind the scenes on the vampire cards. Also, I pretty much let it rest as it becamr clear that the core would be "vampire only", the FPU is much more critical for "amiga use" (though "nobody needs it (except for Quake) was the argument there as well).

Btw, what was the outcome of this?
http://apollo-core.com/knowledge.php?b=4¬e=11498

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

 Status: Offline
Profile     Report this post  
Overflow 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 20:59:42
#1770 ]
Super Member
Joined: 12-Jun-2012
Posts: 1628
From: Norway

@kolla

Since Im not a developer, Im asking that you and others elaborate exactly why, when and for what the MMU is needed in development.

Ferrels, as a dev, said he didnt utilize it.

What is he missing?

Be spesific please, or else it becomes a thread of pages discussing abstracts.

@ferrels

With the increase in both screen resolutions and processing power vs your old a1200, wont using a Vampire be less painful for development?

I realise a properly set up cross-compiler system is much more userfriendly.

Last edited by Overflow on 22-Jun-2019 at 09:19 PM.

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 22:31:48
#1771 ]
Elite Member
Joined: 9-Jun-2004
Posts: 12791
From: Norway

@Overflow

Quote:
I realise a properly set up cross-compiler system is much more user friendly.


I don't think it is, I like writing my C/C++ programs on AmigaOS4.1, but AmigaOS4.1 has some what limited memory protection, so MMU is great for that.

AmigaOS3.x does not have any memory protection built in but there are tools you can have running in the background to help find bugs in programs.

But the memory protection on AmigaOS does not detect a lot of the issues, (some times bugs show up in wrong place) so some times, you need to test the code on more secure operating system like Linux or Windows. or you can pull out all the hair on your head, in frustration.

I often test just code snippets in visual studio, often it's not required to test everything or compile everything under Windows.

Last edited by NutsAboutAmiga on 22-Jun-2019 at 11:40 PM.
Last edited by NutsAboutAmiga on 22-Jun-2019 at 10:33 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
hth313 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 22:38:34
#1772 ]
Regular Member
Joined: 29-May-2018
Posts: 159
From: Delta, Canada

Quote:

Overflow wrote:

Since Im not a developer, Im asking that you and others elaborate exactly why, when and for what the MMU is needed in development.

Ferrels, as a dev, said he didnt utilize it.

What is he missing?

Be spesific please, or else it becomes a thread of pages discussing abstracts.


I am a developer and bought an A3000 (new) at the time and the MMU was main reason for that purchase.

The MMU helped to run various tools like MungWall and Enforcer (if I remember the names right) which can help detecting certain hard to find bugs in the code. Typically related to illegal use of memory you did not properly allocate or dropped and user anyway.

So yes, I used it and it helped me as a developer.

We had a somewhat similar tool a couple year later. It relied on modifying compiler output (object files) on a C compiler for HP-UX workstations (Purify I think it was called). It was expensive and did a similar job but with much more friendly output. It was totally worth it. We found several hard to find problems with it.

The attitude ferrels shows with "some hacks to assist some developers who had poor programming habits" makes me cringe. A developer with such attitude would fail any job interview with me. A developer should use every tool available.

On the other accounts with the cross compiling and such, ferrels is spot on. And yes, they were hacks, but very useful hacks.

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 23:33:43
#1773 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@Overflow



Quote:
With the increase in both screen resolutions and processing power vs your old a1200, wont using a Vampire be less painful for development? I realise a properly set up cross-compiler system is much more userfriendly.


That's a very good question. Only slightly less painful and not powerful/fast enough to justify moving my development to the Vampire from my PC. The Vampire, even under the best of conditions runs on par with a 100 Mhz Pentium from 1995 (188 MIPS), so it would still be painfully slow for compiling code. Also keep in mind that native compilers for classic Amigas are stuck in the past, as are any debugging tools created for them, with no way to move forward with modern features. They are closed-source and abandoned, making them nearly impossible to update. Classic Amiga compiler development/improvement essentially ceased when Commodore died. There also aren't any native, modern and comprehensive IDE's for C/C++ that run on classic Amigas. Such isn't the case when it comes to cross-compilers on Windows or Linux. The 68K branch of GCC is still actively developed....as active as one can expect for a dead CPU architecture anyway. There are also very good IDE's and debuggers available on the PC that simply don't exist for an Amiga or a Vampire. The IDE and debugger for Amiga Lattice/SAS C/C++ is from around 1992-94 and these days I wouldn't even call it an IDE. It's really just a poor editor and the debugger is extremely poor when compared to the debuggers that exist now for modern cross-compilers.

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 22-Jun-2019 23:39:18
#1774 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@hth313

Quote:
The attitude ferrels shows with "some hacks to assist some developers who had poor programming habits" makes me cringe. A developer with such attitude would fail any job interview with me. A developer should use every tool available.


Don't worry. The chances of me ever being interviewed by you are zero. I'm the owner of a software development firm in the field of terrestrial LIDAR mapping/cartography. Here's my company: LIDAR Widgets

And I don't use every tool available. I use the best and most relevant ones. That's a huge difference. MungWall and Enforcer are no longer relevant when using a modern compiler/cross-compiler with a modern debugger and modern IDE.

And yes, tools such as Mungwall and Enforcer are hacks...hacks because they bypassed the OS since AmigaOS doesn't/didn't natively support an MMU. And you yourself admit that after a couple years you moved development over to an HP-UX Workstation with a cross-compiler.....These days, using MungWall and Enforcer for development on a classic system for classic systems when there are better, more relevant tools (I.E. a PC with a cross-compiler) available is simply absurd unless you are trying to make a statement about masochism.

Last edited by ferrels on 23-Jun-2019 at 12:01 AM.
Last edited by ferrels on 22-Jun-2019 at 11:59 PM.
Last edited by ferrels on 22-Jun-2019 at 11:57 PM.

 Status: Offline
Profile     Report this post  
kolla 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 4:50:42
#1775 ]
Elite Member
Joined: 20-Aug-2003
Posts: 2852
From: Trondheim, Norway

And besides, the apollo core has an MPU which is widely supported by all relevant development tools by now... right? And where is LIDAR for Apollo Core? Oh yes, that's right...

Cross-compilation can be tricky, and sometimes not even an option. It is often safer and easier to use native build environments under emulation, and even on right architecture with a more capable OS. Today, most compilers are natively developed on free *nix platforms such as Linux and BSDs, which is why any new 68k also would benefit greatly from being capable of running such operating systems (and it would make many Linux and BSD developers happy to have a new, more modern and fast 68k). To build a new CPU architecture, like the AC 68080 in many ways is, without _any_ consultation or cooperation with compiler developers at all, is borderline to pointless.

Last edited by kolla on 23-Jun-2019 at 05:08 AM.
Last edited by kolla on 23-Jun-2019 at 05:04 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 5:32:10
#1776 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@kolla

Quote:
And besides, the apollo core has an MPU which is widely supported by all relevant development tools by now... right? And where is LIDAR for Apollo Core? Oh yes, that's right... Cross-compilation can be tricky, and sometimes not even an option. It is often safer and easier to use native build environments under emulation, and even on right architecture with a more capable OS. Today, most compilers are natively developed on free *nix platforms such as Linux and BSDs, which is why any new 68k also would benefit greatly from being capable of running such operating systems (and it would make many Linux and BSD developers happy to have a new, more modern and fast 68k). To build a new CPU architecture, like the AC 68080 in many ways is, without _any_ consultation or cooperation with compiler developers at all, is borderline to pointless.


Now you're just being an @ss, as usual...the Kolla we've all come to know and love. Why would anyone want to process LIDAR data on a Vampire or any other Amiga for that matter? The data sets are so large that they require 64-bit addressing and Amigas are so slow that we'd be waiting for hours instead of seconds, even on a PPC equipped Amiga. And why would I care if the Vampire or any classic Amiga has an MMU or not when a cross-compiler on Windows, Linux or Mac is a much better option?

As for a classic Amiga being a safer build environment, emulated or not, you're out of your mind. The Amiga suffers from memory fragmentation problems which makes it an extremely poor choice for native development. Even in the 1990's, any serious programmers switched to cross-compilers on other platforms. The coding cycle of write/edit, compile, test, came to a screeching halt rather quickly due to the Amiga's memory limitations and memory fragmentation. I recall having several megabytes or free RAM, but no contiguous blocks large enough to continue testing even after just one or two test cycles. And I haven't even touched on the stability issues tied to this fragmentation. The fragmentation required constant reboots to free up enough contiguous RAM to continue development which was a death knell for any programmer's productivity.

So explain to all of us here why cross-compilation isn't even an option in some cases....I'm extremely curious to hear the scenario you've dreamed up because I can't think of a single situation where it ISN'T an option. So do enlighten us.

Last edited by ferrels on 23-Jun-2019 at 05:41 AM.
Last edited by ferrels on 23-Jun-2019 at 05:37 AM.
Last edited by ferrels on 23-Jun-2019 at 05:36 AM.
Last edited by ferrels on 23-Jun-2019 at 05:32 AM.

 Status: Offline
Profile     Report this post  
Overflow 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 5:39:32
#1777 ]
Super Member
Joined: 12-Jun-2012
Posts: 1628
From: Norway

@kolla, hth313, ferrels

Ok, I havent made any bones about I often find HOW you present critique is quite important, with regards to how said feedback is recived.
I remember how harsh Daniel was with Dark whatever his name was, and he seemed to just check out after trying to make a new high level language for game development and demoplatform.
Daniel did go into detail, but he was also extremely direct regarding fundemental flaws, and Dark seemed alltogether too sensitive. So it all escalated badly.


@ferrels

I understand what you are saying (as far as a non-developer can claim understanding) regarding bad habits and not to say the least of outdated tools.
Then the question becomes; with updated versions of the tools that hth313 refrences to, wouldnt there be any use for them?

I had a extremely limited exposioure to Assembly (Devpac and AsmOne) in the 90s, where I dabbled with democoding, and when I where expanding the code beyond the starter lessons and examples I was reading, I kept crashing the system when I made bad mistakes.
Coming from the C64, rebooting of a HD powered Amiga felt instant almost, but in todays world, running programs in a Virtual box or having the program crash without taking the system down with it is how I expect it to behave, since it does save time.

Are there tools/compilers/setups that allows you to happily do mistakes, while not taking the system down with it?

hth seems to be alluding to having tools that allows you to learn/debug quicker, hopefully weeding those bad habits out quicker?

@kolla

Using the lack of support for MPU from relevant development tools for the Apollo isnt much of an argument since even the tools being refrenced seems to be very outdated .. or?
Yes, the tools where available when the Amiga was in its "prime", so in that regard they are "up to date", but not really :)

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 6:11:33
#1778 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@Overflow

Quote:
I understand what you are saying (as far as a non-developer can claim understanding) regarding bad habits and not to say the least of outdated tools. Then the question becomes; with updated versions of the tools that hth313 refrences to, wouldnt there be any use for them? I had a extremely limited exposioure to Assembly (Devpac and AsmOne) in the 90s, where I dabbled with democoding, and when I where expanding the code beyond the starter lessons and examples I was reading, I kept crashing the system when I made bad mistakes. Coming from the C64, rebooting of a HD powered Amiga felt instant almost, but in todays world, running programs in a Virtual box or having the program crash without taking the system down with it is how I expect it to behave, since it does save time. Are there tools/compilers/setups that allows you to happily do mistakes, while not taking the system down with it? hth seems to be alluding to having tools that allows you to learn/debug quicker, hopefully weeding those bad habits out quicker?


That's the problem. The Amiga native tools (C/C++ compilers) used for 68K development are stuck in time. They are closed source and have been abandoned by their devlopers/owners, so there is no way to bring them up to modern standards and features. The Apollo Team was/is working to update one of the assembly language compilers, but that only helps a very limited number of assembly coders....not C/C++ coders. Using the debugger for Lattice/SAS C is a painful experience and it doesn't even tie into its own IDE as the IDE for Lattice/SAS C is just very poor editor. Modern debuggers allow for much more, such as setting break points and code profiling and are integrated with their IDE's. Modern IDE's also provide syntax highlighting and will even highlight errors prior to compilation or debugging. That would have been amazing to have back in 1985 on the Amiga. This type of functionality and integration are non-existent on classic Amiga C/C++ dev tools.

And by using a C/C++ cross-compiler on Windows, Mac or Linux, you don't risk crashing your development system. You do risk crashing the target system during testing, whether that system is WinUAE or real hardware, but your development cycle continues without any fear of data loss or drive corruption. Modern cross-compilers also implement modern C/C++ standards too, such as C99 and have modern IDE's and debuggers. The coding cycle is also much faster on a modern system hosting a cross compiler. BTW, C99 is ancient but it's modern when compared to anything classic Amiga.

I'm not saying that the old native tools are worthless I'm saying there are better choices for anyone who wants to do serious development. For example, AmiDevC++ is an amazing cross-compiler that runs on Windows. It has a great IDE and debugger and it truly lets me write code once, and then compile it for a classic 68K Amiga, PPC Amiga, or for an AROS x86 system. I just select the target system and hit the compile button and it generates the binary. It uses GCC for compilation which is still in development, as are the debugging tools. So the old native tools for use on a native system are still a choice, but I'm at a loss as to why anyone would want to use them on real native, or emulated hardware, unless they enjoy doing things very slowly and with lots of crashes and reboots.

Last edited by ferrels on 23-Jun-2019 at 06:27 AM.
Last edited by ferrels on 23-Jun-2019 at 06:21 AM.
Last edited by ferrels on 23-Jun-2019 at 06:13 AM.

 Status: Offline
Profile     Report this post  
Overflow 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 6:26:07
#1779 ]
Super Member
Joined: 12-Jun-2012
Posts: 1628
From: Norway

@ferrels

Thanks for the explainations

I read your comments above regarding memory fragmations/blocks, but out of curiosity;

Are there any way updated tools would get around these issues, or is this a fundemental OS memory handling issue?

Or, with your best judgement/guesstimate, would the MMU/MPU of the Apollo be able to get around this problem, if AROS/AOS ...sorts out its memory management issues?

Just about now, Im sure you are facepalming, and thinking "this is why I say you should crosscompile, cause the hypotetical scenario isnt on the horizon anytime soon! FFS!" etc.

But Im just asking questions, since I sense that you and others seems to have some common ground, but the tone of the debate usually creates a fog of war, that spirals out of control during text communications.

Last edited by Overflow on 23-Jun-2019 at 06:27 AM.

 Status: Offline
Profile     Report this post  
ferrels 
Re: News about Vampire and Apollo
Posted on 23-Jun-2019 6:59:07
#1780 ]
Cult Member
Joined: 20-Oct-2005
Posts: 922
From: Arizona

@Overflow

Quote:
I read your comments above regarding memory fragmations/blocks, but out of curiosity; Are there any way updated tools would get around these issues, or is this a fundemental OS memory handling issue? Or, with your best judgement/guesstimate, would the MMU/MPU of the Apollo be able to get around this problem, if AROS/AOS ...sorts out its memory management issues? Just about now, Im sure you are facepalming, and thinking "this is why I say you should crosscompile, cause the hypotetical scenario isnt on the horizon anytime soon! FFS!" etc. But Im just asking questions, since I sense that you and others seems to have some common ground, but the tone of the debate usually creates a fog of war, that spirals out of control during text communications.


The memory fragmentation problem is a limitation of classic AmigaOS. Ideally, AmigaOS needs a complete overhaul that correctly implements virtual memory and garbage collection, but the overhaul would likely break a lot of apps. The sources for AmigaOS are in such a legal shambles that there will never be a re-write of AmigaOS that properly implements virtual memory and garbage collection so you're left with 2 choices. Develop for AmigaOS using classic tools on real or emulated hardware (painfully slow and crash prone), or use a cross compiler on Windows, Linux or Mac. On real or emulated Amiga hardware, MungWall can assist with the memory fragmentation issue if you chose to use classic tools for development, but such development can be slow and frustrating even if you use emulated hardware on a fast PC. Even MungWall won't stop all the stability and memory issues that crop up during testing and you risk losing your code or corrupting your drive on your real or emulated Amiga.

By using a cross compiler you gain development speed thru the use of better tools and a faster CPU without the worry that you're going to crash your dev system and lose your code. You might crash the target during testing, but that's a small price to pay when compared to the alternative.

Currently I don't think AROS supports any form of memory protection, but I'm just a layman on anything AROS related so that may have changed. I would refer you to this page but I'm unsure of how current or accurate it is: AROS MMUHID

So to answer your question, no, even using the MMU on the Vampire won't alleviate the problems we've been discussing as the OS is the problem, not the presence nor the lack of an MMU/MPU.

Last edited by ferrels on 23-Jun-2019 at 07:04 AM.

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 | 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | 89 | 90 | 91 | 92 | 93 | 94 | 95 | 96 | 97 | 98 | 99 | 100 | 101 | 102 | 103 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle