12 September 2015, 12:00 | #81 | |
Registered User
Join Date: Nov 2011
Location: Nuernberg
Posts: 818
|
Quote:
|
|
12 September 2015, 14:57 | #82 | |||
Natteravn
Join Date: Nov 2009
Location: Herford / Germany
Posts: 2,539
|
Quote:
Quote:
The minimum requirements will be so high that the group of potential customers becomes really small. And why should they run a badly performing 3D game on their expensive Amigas when any 20-Euro Raspberry Pi can do it better? You can't sell that. Retro-computing is the key. I think many people will buy for games which feel and play exactly like the Amiga classics of their youth and run on their old hardware. So you must support the A500. Quote:
|
|||
12 September 2015, 14:58 | #83 |
Registered User
Join Date: Feb 2015
Location: Sweden
Age: 50
Posts: 2,977
|
|
12 September 2015, 15:02 | #84 |
AmigaMan
Join Date: Oct 2012
Location: Castro Urdiales/Spain
Posts: 763
|
It runs in 68k but needs gfx card.
Perhaps in the future It can offer an AGA plugin but doubt It will be fast enought. |
12 September 2015, 15:20 | #85 | |
Registered User
Join Date: Nov 2011
Location: Nuernberg
Posts: 818
|
Quote:
I normally use it on emulation not real hardware. AGA is promised and it has to be seen how it runs on a A1200 as example. I think it will become more interesting when fast FPGA based accellerators are available. And yes handcoded and optimized assembler games might be faster but if you want to reach developers outside the community I fear you need support of commercial platforms. So to me Hollywood seems to be the only realistic option. And regarding assembler developers, how many are left? I contacted a number of former devs that I found in internet but I found noone really being interested to return. Some have even died, most who are still active develop for smartphones. And in any case all are getting old. That is not offering real future. Last edited by TCD; 12 September 2015 at 16:29. Reason: Back-to-back posts merged. |
|
12 September 2015, 15:31 | #86 | |
Registered User
Join Date: Feb 2015
Location: Sweden
Age: 50
Posts: 2,977
|
Quote:
|
|
12 September 2015, 15:34 | #87 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
|
12 September 2015, 15:38 | #88 |
Registered User
Join Date: Nov 2011
Location: Nuernberg
Posts: 818
|
|
12 September 2015, 15:39 | #89 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
i'm currently training Alex up as well
|
12 September 2015, 15:47 | #90 |
Registered User
Join Date: Nov 2011
Location: Nuernberg
Posts: 818
|
I could not even mention 5 (besides you). Outside the community nobody used asm anymore (except perhaps for specific tasks). And even if somebody would know asm, you need lots of experience with amiga hardware when directly using it. I do not think there are many left.
|
12 September 2015, 15:49 | #91 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
Slipstream demo group is just trying to break back into Amiga again. There are also several members here who use asm.
|
12 September 2015, 15:59 | #92 |
Total Chaos forever!
Join Date: Aug 2007
Location: Waterville, MN, USA
Age: 49
Posts: 2,200
|
Re:Assembly coding on 68k
The compilers on Amiga and AROS for 68k suck rotten eggs. They generate terrible code, in other words. I write code in C and AmigaE but still there's no way to get sensible results without inline Assembly. If writing chipset specific code, the portions that happen in the main loop should be hand-tuned even if the first draft was in a high level language. Re:Hollywood 6.0 I have Hollywood 6.0 but I wouldn't think of using it for a 68k Amiga because it compiles just-in-time. If I were going to target 68k, I'd require ahead-of-time (load-time on user's machine) or static (on developer's machine) compilation or some combination of the two. |
12 September 2015, 16:07 | #93 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
about that, could you tell me about the specific sorts of foolishness the C compilers produce?
|
12 September 2015, 16:34 | #94 | |
Code Kitten
Join Date: Aug 2015
Location: Montreal/Canadia
Age: 52
Posts: 1,178
|
Quote:
The stock kittens have much more historical significance, are more numerous and they are the flag bearers of the Amiga brand, if people want retro they will use these. Boosted kittens fragment the Amiga ecosystem without any advantage other than being interesting on a technical standpoint, it is fine but it cannot sustain game creators reliably alas in my opinion (and obviously everyone is entitled to differ ). |
|
12 September 2015, 16:38 | #95 | |
Total Chaos forever!
Join Date: Aug 2007
Location: Waterville, MN, USA
Age: 49
Posts: 2,200
|
Quote:
|
|
12 September 2015, 16:46 | #96 | |
Registered User
Join Date: Nov 2011
Location: Nuernberg
Posts: 818
|
Quote:
|
|
12 September 2015, 16:51 | #97 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
i wasn't expecting anything quite that insane... thanks for that
|
12 September 2015, 16:51 | #98 | |
Code Kitten
Join Date: Aug 2015
Location: Montreal/Canadia
Age: 52
Posts: 1,178
|
Quote:
[un]linkopcodes for each function call which is both useless and inefficient. The last versions of Lattice did not do it however if I recall correctly. I remember being anxious that it might generate these at the time after I bought it. Gcc did not have this problem however and when I worked a bit with a play version of quake ported by Samuel Devulder on my 1200 in the 90's I recall that the generated assembly code was not too bad. It could be improved obviously but it was far from horrible. Samuel's "popt" peephole optimizer (on Aminet) managed to squeeze a few more cycles out of it but not an enormous amount. Addendum: when I say "far from horrible" I'm not comparing to assembly experts output, but medium range assembly coders. Anyone can outclass a 68k on a few routines given enough time. Last edited by ReadOnlyCat; 12 September 2015 at 16:57. |
|
12 September 2015, 16:55 | #99 |
Glastonbridge Software
Join Date: Jan 2012
Location: Edinburgh/Scotland
Posts: 2,243
|
yes, indeed, LINK/UNLK are quite unneccessary instructions, but in one of the examples in the above link we see even "LINK A5,#0" which is positively certifiable!
it looks like in some cases it is deliberately moving things around for optimisation, but is achieving exactly the opposite of that. or this little gem: Code:
moveq #1,d1 move.l d1,(a0)+ moveq #1,d1 move.l d1,(a0)+ moveq #1,d1 move.l d1,(a0)+ moveq #1,d1 move.l d1,(a0)+ a purely literal rendering of the original source code would surely be better than any of those example outputs... in fact i would rather compile literally and let the programmer in charge of the optimisation. Last edited by Mrs Beanbag; 12 September 2015 at 17:03. |
12 September 2015, 17:03 | #100 | |
Code Kitten
Join Date: Aug 2015
Location: Montreal/Canadia
Age: 52
Posts: 1,178
|
Quote:
https://en.m.wikipedia.org/wiki/Peephole_optimization |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Amiga Indie Hardware Developments | gulliver | support.Hardware | 34 | 02 March 2021 19:41 |
PLAY Margate - Indie and Retro gaming show coming to Margate | Neil79 | Retrogaming General Discussion | 0 | 13 April 2015 19:16 |
RGCD - New Retro/Indie Gaming Blog | Heavy Stylus | Amiga websites reviews | 0 | 05 May 2011 09:56 |
Most "valuable" way to contribute to the Amiga gaming scene today | NewDeli | Nostalgia & memories | 38 | 28 March 2009 15:38 |
Skate or die : AMIGA! | Crown | request.Old Rare Games | 17 | 06 May 2006 16:01 |
|
|