@NutsAboutAmiga

Quote:

Nightly build will always be unstable

Yes, by definition, but how does the stability/usability of the PPC build compare to the stability/usability of the X86_64 build? Is AROS on NG currently viable as a daily system? This is what I'm not sure about.

_________________
"Unix is supposed to fix that." -- Jay Miner

@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @TRIPOS It's not that long ago I read up some stuff from ANN.lu, its kind finny in retrospect, the cooperation between MorphOS and Hyperion broke down, because Ben trusted the kernel design to Hyperion, and MorphOS team had a different vison of how the kernel should look like. At the time there was amitalon and it restricted by Linux, Hyperion did not wont that, they did not get the idea of hypervisor, micro kernel design like quark was supposed to be, and frankly it divided the community into two as well. Anyway Amiga community had lots history from Atari / Amiga food wars before that its carried on a tradition I guess.

I remember the old times pretty well. Early releases of MorphOS did not include AmigaOS 68k parts, but MorphOS was not complete. You needed some 68k parts to actually run MorphOS. But they were never _distributed_ with MorphOS. In the early days it was no issue anyway, as the target hardware were Amigas souped up with P5 ppc cards, hence AmigaOS was legaly present there anyway.
The 1.0 release (I still have the CD here) was complete and needed no 3rd party additions to operate.
Ben Hermans (and Steffen Häusser) either mixed up things or wanted to provide FUD.
It was disgusting!

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@amigadave

That is, essentially, my sentiment as well. Hyperion was cavalier about making enemies and it's high time they make some corrections - even amends - or accept the consequences of their mistakes and move on.

@outrun1978

Wow. You're a special kind of stupid arent you?

EVERY SINGLE THING you criticized about MorphOS also applies to OS4.
One is a well coded, efficient, more advanced AmigaOS implementation, the other is OS4.
Why would one use OS4 when they can use MOS?
MOS for fun, and Linux for something serious.

This is painfully obvious to those whose shoe size isnt a great number than their IQ

But you have such a 1/2 track mind that you appear to be blind to what is clear to a half witted monkey.

You're the sort of person that velcro was invented for. Those that find laces too complicated an idea.

@fishy_fis

Quote:

One is a well coded, efficient, more advanced AmigaOS implementation, the other is OS4. Why would one use OS4 when they can use MOS?

Because the IP IS important. No one outside our community knows or cares about MorphOS period. Sad but true!

While I admire MorphOS I would rather mess around with my Classic and save up for a Tabor than mess up my iBook G4 to 'try out' MorphOS. My iBook G4 works very well as a old scanner terminal for an old OS9 era scanner that works fine under the Classic Emulation layer on OSX Tiger. Why would I mess that up? I can also play The Settlers 2 under Mac OSX!

What exactly would I do with MorphOS? Play Wipeout 2097? It wouldn't even have The Prodigy on the soundtrack! My Classic allows me to use Deluxe Paint, PageStream, Photogenics and DrawStudio. What benefits would a MorphOS machine bring? Video? Internet? Who cares? The Apple AirPort card in the iBook G4 is on the fritz and crashes the machine when enabled so I keep it disabled and use it for the scanner! It's a solid machine but why mess with it if it's already failing in some areas?

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@Thread

And what about getting back to topic (and ignore the usual trolls)?

_________________
PowerMac G5 2,7Ghz - 2GB - Radeon 9650 - MorphOS 3.14
AmigaONE X1000, 2GB, Sapphire Radeon HD 7700
FPGA Replay + DB 68060 at 85Mhz

@BigD

Sorry but kernel in AmigaOS4 is not AmigaOS3.x related, just some reimplementation like in MorphOS. AmigaOS4.x has a name which is important but the more MorphOS will be advanced (currently is advanced in the most things except support for newest Radeons) than AmigaOS4.x the more hardliners will jump into blue train.

and:
1) Pagestream and Photogenics works very well on MorphOS, much better than on BVision+040, i didn't tried on AmigaOS4 but probably it is true also for OS4
2) Settlers2 can be played on Basilisk2 (both on MorphOS and AmigaOS4)
3) Which Scanner do You have? It may be supported by current MorphOS

@K-L

That’s practically impossible in amongst some of these drama queens. They say only ostriches bury their heads in the sand but it’s clear these trolls do just the same.

Regardless I am off back to play some GNGeo, in glorious HD, with joystick support on my X5000 because at least that works properly on AmigaOS and didn’t under MOS.

After I might just go and play Wipeout 3 via FPSE afterwards, that too runs on AmigaOS compared to MOS which just locks up mid game after being assured by MOS users that it ran better in their platform. I can also use my analogue stick too because guess what on AmigaOS that too works.

_________________
Amigaone X5000/20 4GB Radeon RX 550 Polaris 12 AmigaOS4.1 Final Edition Update 1
Amiga 1200 Workbench 3.1.4
Amiga CD32

@K-L

I can bring discussion back to topic, but .. not in positive side sadly. I watched both 2 videos (so 2 hours), and for first, what amaze me is how one can talk 2 hours about things which can be said in just dunno .. 5 minutes ? But probably its just intersting to discuss even simple things for long, just to have some nice mood :)

Now to details and facts . As i expected, in general all what is done NOW , is:

1. Done all preparation work to actual doing work. I.e. just put things in one place, making sure they all compiles and nothing broken. Nothing too fancy of course, as kernels compiles even before, of course, how else they will be released :)

2. There is some team, who can or can not work on it. How fast, how much they will works no one say, and everyhting can be the same as usuall : team here and no work done by any of reassons. That not unusuall (expectually in amiga land, where everyone loose motivation and keep busy with other things). Sure names in the Team looks pretty good, but, who will say that Thomas will now spend all his time on it ? Who can hope that Fredirk will spend all his time on it ? Who can hope that Olaf Barthel, which isn't _that_ active will like drop all his deal and working on it ?

See, just a good names together know that there will be good to have something done. Nothing done at moment.

3. As i expect , nothing currently done. Maybe just something about DMA support in kernel (x1000 one already done, as some hack in graphics.library, so maybe it will be better when will be all in kernel). But in general NOW there is only roadmap. And that damn have needs with bananas and phanfares spread it all over the net ?


4. That post on Amiwest site are veeeery wrong. I am not about that "aeon" have rights but Trevor, but about everything. I.e. all those 4gb bariers, multicore, "3 more weeks" whatever else is writen , that is pure "we wish to have". There was writen that "kernel now is blazing fast", that also untrue : its just some optimised functions was added to TABOR version of kernel, so to make it faster (there i can mistake things, but seems it like this), not that all kernels now start to be 50% faster as one may have thing from that amiwest's posts.

Most of those things will be never done in usuable for amigaos4 users form. The only thing which probabaly can be done its that DMA (the call it GART) support in kernel (its not that GART which need it for gfx drivers, that another one, just in kernel).

Usage of multicore can be done probabaly in some primitive form (as it all discussed everywhere in last year). Breaking of 2gb (actually 1.8gb) barier can be probabaly done, but again, maybe with bank-swithing , which mean that again its of no use until someone will not start use it, and rewrite os componets for it (will never happens).

In other words as i expect, its uzuall amiga-buzz. There is nothing really done currently. Nothing to really betatest. Nothing to really release with benchmarks. Its usuall "amiwest" news , much of which we have in last 10 years, 90% of which never ever matearilized.

Sadly, but expected.

Of course some work probabaly being done in some alpha tests. But that didnt' mean it all will have final phase and being bug free.

In other words .. I never understand news about news. Or news about what will be maybe done someday. One actually should deliver firstly some beta version of things in discussion, and only then starting all those bananas.

I read everywhere now how those fans spreading "amiga future starts here!" , damn, it make me cry :) There is nothing new done. Just a team who will (in hope) works on some new version of kernel, where they will try to make all this they wish. How , when, how it will be, when released : that is take my words another 3-5-10 years.

Last edited by kas1e on 25-Oct-2019 at 02:18 PM.

_________________
Join us to improve dopus5!
zerohero's mirror of os4/os3 crosscompiler suites

"Settlers2 can be played on Basilisk2 (both on MorphOS and AmigaOS4)"

Came to my mind... sorry about the side step ... but is there any sane way to run PPC Mac SW via Amiga(like)OS?
(eg https://www.macintoshrepository.org/3516-simcity-3000 )

And is there any modern browser for 68k linux (to be run in 68k emulation on Amiga(like)OS)?


*******************
"I watched both 2 videos (so 2 hours), and for first, what amaze me is how one can talk 2 hours about things which can be said in just dunno .. 5 minutes ?"

+1

"Usage of multicore can be done probabaly in some primitive form (as it all discussed everywhere in last year)."

Any form would be a nice starting point.

Would be interesting to see a workshop/brainstorming to be done about how to proceed, so that there will be good enough solution.
(if needed, shovel PPC legacy into a sandbox etc.)

" Breaking of 2gb (actually 1.8gb) barier can be probabaly done, but again, maybe with bank-swithing , which mean that again its of no use until someone will not start use it, and rewrite os componets for it (will never happens)."

Extra ram as swap space + RAM disk would be a start.

Would be interesting to see a workshop/brainstorming to be done about how to proceed, so that there will be good enough solution.
(if needed, shovel PPC legacy into a sandbox etc.)


Currently multicore + more ram etc. are being implemented multiple times (AOS+Morph+AROS), what a waste of developer resources!

Last edited by KimmoK on 25-Oct-2019 at 02:23 PM.
Last edited by KimmoK on 25-Oct-2019 at 02:14 PM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@kas1e

Quote:

I read everywhere now how those fans spreading "amiga future starts here!" , damn, it make me cry :) There is nothing new done. Just a team who will (in hope) works on some new version of kernel, where they will try to make all this they wish. How , when, how it will be, when released : that is take my words another 3-5-10 years.

I can't get excited about kernels. If the promised SMT stuff speeds up actual programs then great otherwise: Where's the software?

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@BigD

Amigaos have no big needs for real to support more than 1 cpu now. There is a lot of areas need updates and improvements, and all software we have fits very fine to one core with no problems. Its just marketing or non-understanding of technical details, that any kind of multicore will help _that_ much as one hope.

But instead things like DMA in kernel for sure need it for a long time for all platforms (expectaully for x5000 where is none). Faster booting also for sure have needs for a looooong time. I do not remember how it done on morphos, but it was always faster on loading on about 50%, and if i remember right they have load compression version to the ram from start, then unpack and continue from ram. And have ability to use 4gb instead of 2gb of ram also can be cool , if only it possible to do "malloc(2gb)", without bank switching as it done now with extendented memory which no one except Andy with Sketchlock is using. But as we on 31 bits, i somehow thing 2gb will be our limit forever. I was once port some thing, which want more than 1gb of memory at once (malloc just 1gb), and then once i undertstand how hard is to add that extendent memory support to, i give right away :)

And in all of this everyone talk about multicore, which really, really, FAAAAR away for reall needs in compare with unresolved issues we have.

_________________
Join us to improve dopus5!
zerohero's mirror of os4/os3 crosscompiler suites

@BigD

Quote:

I can also play The Settlers 2 under Mac OSX!

A good game! I play Widelands (a copy, more or less) on linux.

_________________
"Unix is supposed to fix that." -- Jay Miner

@Kas1e

Thanks for this summary.

So, as said ssolie : there is nothing to be seen. Move along.

Too bad !

_________________
PowerMac G5 2,7Ghz - 2GB - Radeon 9650 - MorphOS 3.14
AmigaONE X1000, 2GB, Sapphire Radeon HD 7700
FPGA Replay + DB 68060 at 85Mhz

@kas1e

I really do not understand what sense s super-modern kernel (base) would make if the house (OS) running on it is a old castle and the architects cannot transform it in a modern house (because they have no sources)

@OlafS25

Thats what I'm thinking too, what is needed is 64bit xkernel (hypervisor level0), SMP/MP everything modern, then put ExecSG inside a sandbox, move critical drivers to level 0, and provide sandbox for 64bit and SMP compatille software. this I belive is the only way to move forward without building some kind of hackory.

Making radical changes to ExecSG means breaking a lot of software, and if you are not in control of OS components like GUI, and other stuff, how can you make the changes without screwing it up.

Any we want our modern software not to be limited by 31bit limits, and other issues that classic software requires, so there has to be a two worlds system.

Last edited by NutsAboutAmiga on 25-Oct-2019 at 05:11 PM.
Last edited by NutsAboutAmiga on 25-Oct-2019 at 04:38 PM.

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

@outrun1978

Quote:

outrun1978 wrote: @K-L That’s practically impossible in amongst some of these drama queens. They say only ostriches bury their heads in the sand but it’s clear these trolls do just the same.

The thread went off-topic in post #37. Your post. You did it, you are the troll.


@KimmoK

If you want to be Amiga binary-compatible, then there are certain rules that could never be broken. Like 31-bit memory addressing (which also means it can’t be expanded beyond that through “virtual memory” or swapping) of unified unprotected memory, single core processing, big endianness, etc.

If you don’t care about the compatibility, then it’s fairly easy to implement all those things and go little endian as well on a real, contemporary platform (meaning: ARMv8 or AMD64).

But you can’t have both.

@NutsAboutAmiga

Quote:

NutsAboutAmiga wrote: @OlafS25 Thats what I'm thinking too, what is needed is 64bit xkernel (hypervisor level0), SMP/MP everything modern, the put ExecSG inside a sandbox, move critical drivers to level 0, and provide sandbox for 64bit and SMP compatille software. this I belive is the only way to move forward without building some kind of hackory.

If only someone had thought about this before?


Quote:

Making radical changes to ExecSG means breaking a lot of software, and if you are not in control of OS components like GUI, and other stuff, how can you make the changes without screwing it up.

Indeed!

@BigD

If you don't know what you would use MorphSO for- why saving up for a Tabor then?
I mean en gros the situation of usability and applications is pretty similar for MorphOS and OS4.

And: To play around with MorphOS on the ibook you don't need to hack/mess that much. A little typing exercise on the Open firmware and you can boot from usb w(o touchig the original installation. If you happen to somehow like it, there are ways to shrinken the oS X partitition and parallely install MorphOS.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@kas1e

People have SSD hard drives these days, you most likely won't gain a lot by doing that. But I guess what you want do is load in a large file, instead of lots small files. I guess can just link all modules into one, and load it as is, and just relocate it in memory using ELF header, maybe simple compression to reduce the file size. Anyway the biggest delay on X1000 is not the kernel, its CFE that has to scan all unnecessary USB ports. (it's not CFE's fault, actually I like CFE, but if be great if there was option to see what was going on insted of the graphic at boot.)

The biggest delays at boot is in the Firmware, that has to emulate x86 bios calls, to initialize the graphics, then its scanning all the other hardware (USB ports, PCIe) , then its scanning the SATA ports, this part can be slow if you have slow DVD or something is not connected to port, it think something should be connected. Also at boot CPU is not running at full speed, that happens when CFE console is ready.

Last edited by NutsAboutAmiga on 25-Oct-2019 at 05:52 PM.
Last edited by NutsAboutAmiga on 25-Oct-2019 at 05:46 PM.
Last edited by NutsAboutAmiga on 25-Oct-2019 at 05:43 PM.
Last edited by NutsAboutAmiga on 25-Oct-2019 at 05:42 PM.

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