Hyperion bankrupt?

[matthey]

I just bought a new-ish Camaro and gave my oldest son the 97 Firebird I've been driving forever.

After having the car for two days (tonight) he calls me and says that "there was a loud noise so I pulled over and when I put it in park, it still rolls".

He's broken either a u-joint (I hope) or the drive shaft or something. Not sure which yet.

I just now got it towed back home, so now I know what I'm doing for the weekend.:pint:

The Firebird sounds as broke as Hyperion. Otherwise automatics won't roll in park. You better hope it is something between the tranny and diff or it's going to be expensive. I hope you show your son how to work on it for next time he goes on a a test drive. Make sure he can at least change all the spark plugs. I had a mechanic friend with an earlier LT1 manual Firebird and there was one spark plug that he never changed because he couldn't (from the bottom with several swivels). Maybe GM fixed that problem but then again we are talking about GM. They are to engineering as Hyperion is to software development. See, we can stay on topic .

There is nothing sweeter than the sound of a 427 Chevy engine with mechanical lifters and a tri-barrel carb at idle.

I've heard the rhythmic lope and hollow rumble of exhausts of various sweet sounding V8s as I live in the Midwest (I've even driven a few). However, not many people have heard the sound of an 80ci 2 rotor sequential twin turbocharged Wankel with titanium exhaust easing into WOT at mid-rpm in 3rd gear. First comes the sound of the aluminum intake sucking in air and the rpm wanting to go higher as the sound becomes something like a propellor plane taking off. As I reach WOT, the torque that sneeks up on the passenger can be felt in the lower stomach and some murmer an expletive as they reach for whatever they can hold on to. The exhaust now sounds like a hornets nest trapped in the titanium but smooths out into one deep take-off pitch that is loud but not too loud. Let off and the swush of the blow off valves can be heard. Fun, but now lets try it on some twisty country backroads. My car is a silver base model '93 Mazda RX-7 with bolt-ons. Unlike Hyperion, I enjoy diversity and choice which is a better business strategy than going bankrupt with one processor for a nich upper class market. See, our car conversation is on topic again.

[matthey]

How about a ROM that can find a CDRom to boot so I can ditch my slowly decaying boot floppies?

And has HD support for >4GB hard drives.

A-EON owns Warp3D now.  I know upgrading it for 68k would be something of interest. All of A-EON's new software technologies developed will be back ported to AmigaOS 3 if the hardware can support it.

Did somebody pick up some toys at Hyperion's garage sale or should we say fire sale? It makes me wonder if Hyperion is just a shell to maintain a license agreement for A-Eon, not that I would complain if that was the case. If Trevor is more clever than Hyerion, he would back port Warp3D and Reaction to the 68k and license to users for free to proliferate these APIs, ease porting between AmigaOS 4 and AmigaOS 3 and increase software development. It would sure make our job of getting NetSurf running on AmigaOS 3 easier. I know a thing or two about Warp3D that might be useful too .

[matthey]

I personally would very much like to see Radeon R200 support available for Classic AmigaOS 3 on the Mediator.  If you have a proposal so we can support that, I will be very much interested.

A new AmigaOS 3 (and possibly later AROS) Warp3D Radeon driver would be great but the whole W3D software for AmigaOS 3 needs serious updating. Here is what is needed:

The Avenger (Voodoo 3) driver has a Z-buffer bug which overwrites innocent memory in a big way. The C generated CheckIdle() function is so slow it doesn't work right. The optimization level (NOP instructions in the code which flush the CPU pipeline for example) and floating-point rounding are very inefficient. The 6888x is not suppoted and the 68040 compiled driver traps on the 68060 which is very slow. The AmigaOS3 Avenger driver version lacks support for most Napalm (Voodoo 4-5) enhancements. I have free optimized assembler code available which would help the speed on the 68k.

The Permedia2 driver has a bug with floating point rounding modes which causes distorted or black colors for objects or even the whole screen (bright color values may wrap to dark). This was a compiler problem of GCC not initializing the FPU properly when the Permedia2 code was compiled. Karlos has an updated Permedia2 driver for W3D with more features added.

The 68k Warp3D library itself doesn't have too many bugs but suffers from serious optimization problems. I optimized it in assembler to nearly 1/2 the size with less than 1/2 the size being possible. With all my optimizations, I can get 25fps in GLQuake with 68060@75MHz and more is possible.

I would try to get Alain Thellier to work on an update (with the help and testing of others). He created Wazp3D and knows what he is doing. He could probably even make a software renderer for Warp3D which would be good. I would be willing to lend a hand myself so the same optimization disaster doesn't happen again.

I just thought of something. Up until now we have all assumed that the best scenario of a bankruptsy is that A-Eon buys Hyperion upright in order for AOS4 not to return into A Inc.'s hands. But is that really the best scenario?

Remember, if HE goes down they will take with them a lot of debt. Debts Trevor has to pay off to get the OS. That sounds expensive. But if no one buys the whole company (who would besides Trevor?), the rights to the OS reverts to A Inc., then it must be in McEwans interest to sell the OS. Cloanto already got the original. And since there is only one realistic buyer A Inc. can't ask for too much.

So, instead of paying off a lot of debts, it might be cheaper to negotiate a transfer of IP rights with Amiga Inc. instead.

Thoughts?

The License between Amiga Inc. and Hyperion may no be transferable or at least a buyout would make the situation more complex. A-Eon would like to continue AmigaOS 4 development as it's needed for their hardware. If the debts aren't too large, the easiest solution may be for A-Eon to bail out Hyperion (probably becoming majority shareholder) and Hyperion becomes just a software development house for A-Eon and possibly a legal shell or zombie business. It wouldn't be a bad business arrangement as it could avoid costly legal costs and keep AmigaOS 4 out of the hands of the blood sucking criminals at Amiga Inc. A-Eon seems to be more friendly and open toward AmigaOS 3 which is a good strategy that could benefit AmigaOS 4 also. Honey attracts more Amiga users than vinegar.

[matthey]

Open sourcing it might be nice, but I'd just like to see it being maintained and updated by professionals.

Open or not, unless you have a strong project leader, you're going to get people going off in left field doing crazy additional stuff that drags down the project and breaks compatibility.

I tend to agree as AROS is open source and hasn't replaced 68k AmigaOS 3. Several of the current 68k developers are around and could continue organized development instead of throwing development into a state of chaos and uncertainty. The best would be if the AmigaOS 4 and AmigaOS 3 developers could share code and try to maintain similar APIs where possible. One could argue for bringing AmigaOS 4 to the 68k which is a bit heavy but could be done. AmigaOS 4 needs to move toward 64 bit, memory protection, more security, SMP etc. (workstation, server and business stuff) which is not the greatest fit for the 68k unless we made an ASIC. AmigaOS 3 is probably best as an efficient and compact "fun" but still useful AmigaOS for small computers, old and retro computers, hobbyists, electronic gadgets, netbooks and maybe some embedded purposes. If both groups of developers were allowed to create freely but share code, I think there would be mutual benefits like faster development and bug fixing.

In order to open source Amiga OS 3.x one would need the source.

Check you history; The entire source code for 3.x does not exist. It went missing somewhere in the Commodore labyrinth. This was one of the issues faced by the early AmigaOS 4 team. There was no "gold" code stored in a source repository. Bits and pieces here and there. Decompile this and reverse engineer that.

Most of it exists and is available. ThoR and Olaf "olsen" Barthel are active on the forum here. ThoR claims he knows how to contact Heinze Wrobel also. There is the AROS code to look at if anything is missing. We need to either get AmigaOS 3 back or fix AROS up for the 68k.

[matthey]

>The best would be if the AmigaOS 4 and AmigaOS 3 developers could share code and try to maintain similar APIs where possible.

Yes, and with this end in mind I was going to port OS4 piechart.gadget to OS3, however the developer refused to release source code for that purpose and was very insulting. What a prick. With that kind of attitude from the OS4 community why should I or any other OS3 developer bother to support OS4-specific features in the future?

Because we don't want to sink to the level of these kinds of people and we don't want to punish innocent users for the actions of some misguided developer. People like that reap what they sow. He might just fall off his pedestal when the high and mighty he worships goes bankrupt and he finds himself in the same predicament as us 68k AmigaOS users .