@OlafS25

Odyssey is much newer than OWB, has more features and loads pages better. And it sees more updates (though the last update has been almost a year ago, but the newest version will be ported to AOS 4.1 within the next months).

Last edited by Vistaus on 19-Oct-2015 at 12:15 PM.

_________________
Proud user of AmigaOS 4.1 on an AmigaONE 500. This is the first Amiga I've ever had so I don't know all the ins and outs of AmigaOS yet, so I'm sorry if I'm asking noob questions and stuff.

@Vistaus

Timberwolf in its unfinished state as it is now has no hardware acceleration whatsoever that is why Odyssey is so much faster because it is hardware accelerated.

_________________
Amiga A1200T - TF1260 - R9200 - Indivision AGA MK3
Amiga A500 - PiStorm EMU68
Pegasos 2 G4 - AmigaOS 4.1 FE / MorphOS 3.16

@Fairdinkem

personally i dont think the sources in question are of any value anymore, they are simply too far behind. you might as well be starting from the scratch. so none stops you to appoint another bounty. and keep in mind that a port in its complexity needs to be maintained. me and others, we have mentioned it repeatedly. but be prepared whom you assign the task to. otherwise whatever goes wrong it will again be blamed on community. and it wont be entirely wrong. because if you insist on putting your bet with the wrong people, no matter all caveats, then its your responsibility.

@OlafS25

I could be wrong but the reason I understand that development has stopped with regards to Odyssey for AmigaOS is there is something unavailable for Kas1e to bring it up to parity with the MorphOS port. But I could be mistaken with this.

_________________
Amiga A1200T - TF1260 - R9200 - Indivision AGA MK3
Amiga A500 - PiStorm EMU68
Pegasos 2 G4 - AmigaOS 4.1 FE / MorphOS 3.16

@Vistaus

really? if you say that

Aros port is based on April sources, deadwood is working one year on it. It is not easy to port it to PPC, because of that even MorphOS lags now

@wawa

Judging by the response in this thread no one sees value in Firefox so a bounty seems useless, I guess my Pegasos2 will continue to collect dust.

_________________
Amiga A1200T - TF1260 - R9200 - Indivision AGA MK3
Amiga A500 - PiStorm EMU68
Pegasos 2 G4 - AmigaOS 4.1 FE / MorphOS 3.16

@Fairdinkem

it has to do with OWB being for X86 as I understand it, it is not simple recompiling it for PPC (besides differences between the different amiga platforms). One of the reason why even the MorphOS version has stopped at the moment.

I cannot find the thread on morphzone right now... Fab explained it there

@Fairdinkem

People will say another "Firefox" bounty? A hard time to explain. And then as i wrote... who will update it regularly? OWB port is really open source now so different camps are working on it and using the sources. Firefox would be ported one time, bounty fullfilled and then? Another bounty every couple of months for updating?

@OlafS25

here it is...

Fab:
Updating WebKit is quite some work considering my free time, but the real issue preventing a new release is that there is a blocking endian issue in JavascriptCore: it just doesn't work at all on bigendian/32bits architectures (all of them), and they just don't care at all...

My latest build, with broken javascriptcore, is based on a webkit from may 2015.

@OlafS25

Well all I can say is if the AmigaOS, MorphOS and Aros community/developers gathered together to port Firefox which is a maintained open source browser the Amiga community could work together to maintain the updates for the port. the reality is each of the factions developers have different priorities/focuses and time constraints that such an undertaking will never happen.

It made for an interesting chat anyhow.

_________________
Amiga A1200T - TF1260 - R9200 - Indivision AGA MK3
Amiga A500 - PiStorm EMU68
Pegasos 2 G4 - AmigaOS 4.1 FE / MorphOS 3.16

@Fairdinkem

it comes down to find developers who do the work. As the situation is right now I would say OWB offers best chances, I do not see enough interest for another big browser and no developers who would care for it. OWB port is main browser on MorphOS and Aros and most people I think are happy with it. Everybody is free to make a bounty but I doubt that many people would give money and developers taking the huge task.

@Fairdinkem

Quote:

Fairdinkem wrote: @OlafS25 Well all I can say is if the AmigaOS, MorphOS and Aros community/developers gathered together to port Firefox which is a maintained open source browser the Amiga community could work together to maintain the updates for the port. the reality is each of the factions developers have different priorities/focuses and time constraints that such an undertaking will never happen. It made for an interesting chat anyhow.

odyssey development is shared between the platforms to certain extent (not to forget that its thanks to fab in first place), morphos, aros, and os4 have more less up to date implementations. its what has been achieved thanks to work and inspiration of kas1e and what i have been advocating all along as the subject of timberwolf came up.

now if the os4 developers do not want to share the sources, they have been actually paid for, you cant really go around blaming others for lack of cooperation.

@Fairdinkem

Quote:

Fairdinkem wrote: @duga When I write OWB I mean Odyssey 1.23 it is what I used when I last booted my Pegasos2

But it's not the same. Both exists for AmigaOS 4.1 so it's very confusing to call Odyssey OWB.

@Fairdinkem

Quote:

Fairdinkem wrote: Now that we have heard from Steven Solie that Timberwolf is held hostage due to community members making threats to the Friedens it seems that Timberwolf will never eventuate or ever become open source. Do we think it is possible that we can generate funds to have another port of Firefox to AmigOS that will be open source, I sincerely hope so because I would donate?

I don't think that there's any problem. Since the Friedens don't want to complete Timberwolf, they give back the money that they already got, and the backers can then decide to finance another FireFox port, or another browser (e.g.: enforcing OWB, for example).

Looking at some comments in German I wonder, how SMP could be done in AmigaOS4 - it seems to me, that full SMP may be not so easy.

So what solutions are thinkable?
(links to older discussions are wellcome)

Well, how abot this: Either old compatibility and no SMP or NG-AmigaOS with SMP.

It's quite simple: The NG-AmigaOS will use a new and sophisticated lock-method. Classic Forbid() or Disable() are in the SMP mode prohibited and programs, who still use them, will be forwarded to the Grim Reaper. The user can choose at startup whether he likes to use SMP or the classic Forbid().

Now it's not so easy to develop a stable NG-AmigaOS with SMP but so what? Modern single core is fast enough for AmigaOS and not to forget at least DDR2 SDRAM and fast SSDs. Multicore is desirable and until the NG-AmigaOS with SMP is ready, one could use Linux.

Perhaps, there could be an comparatively easy way, to use other cores from the main classic (single core) AmigaOS4 as slave for some tasks. How? Well, how about this: a special program could pack the slave executable with all dependencies, reserve (lock) some RAM and discspace and move the package to the next best core, where it could run happily all alone. Obviously there shoud be a way to terminate the slave task from the master core.

There could be a first primitive way but in the future small tasks could be very quickly queued and distributed to a free core (there could be an option to automatic execute the job an the main master core after a timeout because of too busy slave cores). Data exchange goes simple over (at first prereserved) RAM. At first there should be only read rights for other resources (like HD, screen, and so on), so there should be no lock problems. (all imho)

Is it possible that newer Power Architecture specifications are not optimal?

Well, because of Nintendo mit Gewinn – Absatz der Wii U zieht etwas an I have a look at Wii U and quite to my surprise the Wii U uses a kind of modernized G3 PowerPC ("... i.e. PowerPC 750 based, but enhanced with larger and faster caches and multiprocessor support. ..." Espresso).

(maybe partly not reliable) Specifications of the Espresso microprocessor:
- Out-of-order execution PowerPC based cores
- 45 nanometer process technology
- Die size: 4.74 mm × 5.85 mm = 27.73 mm2
- IBM silicon on insulator (SOI) technology
- Backward compatible with the Broadway and Gekko processors
- Broadway-based core architecture
- Three cores at 1.243125 GHz
- Symmetric multiprocessing with MESI/MERSI support
- Each core can output up to 4 instructions per clock using superscalar parallelism.
- 32-bit integer unit
- 64-bit floating-point (or 2× 32-bit SIMD, often found under the denomination "paired singles")
- A total of 3 MB of Level 2 cache in an unusual configuration: Core 0: 512 KB, core 1: 2 MB, core 2: 512 KB
Espresso (microprocessor)

"... The console contains 2 GB of DDR3 system memory consisting of four 512 MB (4 Gb) DRAM chips with a maximum bandwidth of 12.8 GB/s. This is 20 times the amount found in the Wii. Of this, 1 GB is reserved for the operating system and is unavailable to games. The memory architecture allows the CPU and GPU to access both the main DDR3 memory pool and the eDRAM cache memory pool on the GPU, removing the need for separate, dedicated memory pools. The console includes either an 8 GB (Basic) or 32 GB (Premium (WW) / Deluxe (NA)) internal eMMC flash memory, expandable via SD memory cards up to 32 GB and USB external hard disk drives up to 2 TB. ..." Wii U

How good is the newest Power Architecture?

How should a universal SoC for industry, military, PC, notebooks, smartphones be designed ideally? Well, for sure there have to be more than one version, perhaps from 1W to 25W TDP, from 1 to 4 or more cores, different caches and so on and the number of ports for USB, SATA, LAN and so on could vary from 0 to 4. One must develop a universal modular design and remember KISS. Ideally the layout is mostly done by a computerprogram, where only the desirable features (how much cores, cache, TDP, ...) have to be chosen beforehand. Don't laugh, this is possible.

The hardware is one thing, another is the software: the world needs a perfect operating system, a modular, secure, reliable, efficient RTOS and this OS has to be open source, because who else should own it if not the people, everyone?

That's the destination. Well, AmigaOS seems to be a little off this track ... but it has to be perfectly clear, that a kind of perfect OS has to come and that would obviously be a problem for not so good operating systems. In the future practically no one would use a stupid inferior OS, if he could use a better one.

AmigaOS has to get on track or AmigaOS will kinda get lost forever. So what must be done? I think, open source could be a good idea, if not the only possible solution.

Obviously some freaks couldn't do this (a perfect OS) alone ... so find help worldwide, think of Asia, India, Russia. Promote the idea, try to convince government, military and industry that an open source OS could be the winner for all. There are trillions (meaning here 10^12, in German Billionen) Dollars out there in the world, there are mighty states, who just don't want to dependent on Microsoft and Intel any more, there are billions (10^9) of people and Billion Dollars to earn.

What's to do? Somehow a convincing team has to be gathered and then it is not only possible but likely that supporters spend much (or at least enough) money for the project. The question is: what has Amiga to do with this? Well, what? It's up to you.

Maybe it is not so wise to firstly develop small programs/applications instead (Call for Amiga Developers)- it would be better, to provide a good way to use Linux programs. The main task is to develop the operating system ... and the main problem is that AmigaOS isn't open source. (all imho)

@KingKong

" it would be better, to provide a good way to use Linux programs"

why not simply use Linux instead?

Quote:

KingKong asked: what has Amiga to do with this?

Absolutely nothing.

@KingKong

>Well, how abot this: Either old compatibility and no SMP or NG-AmigaOS with SMP.

To get robust SMP in this lifetime I still vote for the faster route for multicore future, breaking compatibility & dualbooting 4.1 vs 4.2 when needed.

(+ in 100% 4.1 compatible release, developers might be able to use some performance of those other cores via OpenCL kind of solution ... In too high simplification, OpenCL can be a little like PowerUP/WarpUP NG
+ if there would be enough resources to build virtualization support for AOS4.2+, we could "partition" some extra core for 4.1 apps etc)

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

@OlafS25

Quote:

OlafS25 wrote: @KingKong " it would be better, to provide a good way to use Linux programs" why not simply use Linux instead?

For the same reason that linux runs wine and people don't say: "why not simply use Windows instead"

_________________
No PowerPC, No Fun
Make Amiga Great Again