> On Tue, Mar 18, 2008 at 4:46 AM, Cyril Plisko <cyril.plisko at mountall.com> > wrote: > >> [snip] >> >> Going with Apple is swapping one obsolete platform for another. >> It doesn't matter how cheap you can pick the kit up from ebay _today_. >> In some time in the future we will again find ourselves with something >> that runs on quickly diminishing hardware and a need to pick something >> else. >> > > > I'm confused. It seems like many are under the impression that this port > will only be done for only one platform, and that we need to debate which > platform makes the best business decision long term.
Thus far, the reality is that the port work has all been on one platform. The vision was never restricted to just one. It is open source and you can go any direction that you want. > To my mind, this "port" will have to be done many times. Probably. Maybe. It would be better to go 64-bit sooner than later but that will only happen with millions of dollars of serious R&D by both hardware and software engineers. What we have right now is a 32-bit single processor port on a very specific hardware platform. > But the work that Brian, Guy, Tom > (and others) have done has largely been to just get PPC working, period, > no? >From power on to single user prompt on a serial console. Yep. Along with some basic tools like maybe GCC and such. That was the plan. > I'm working under the assumption that as long as there are enough basic > resources to develop on (including documentation or code to > reverse-engineer), any platform is as good as the next because we'll > /learn/. And the next hardware port will be easier because we'll have > already have 1 platform under our belt. Sounds great in theory. I agree. > Personally, I'm just going to go ahead and get as far as I can get (which > may not be much due to my lack of time and hardware experience at this > level) with the PowerMac I already have. If I can get to a self hosting > state, I'll try to tackle other platforms. Perhaps by that time the magic > will happen and a sponsor will show up with more "relevant" hardware. There is no such thing as "magic". Nothing will happen at all, unless we choose to take action and then actually do something. In my case I am talking with IBM people. > Meanwhile, PowerMac/Linux/FreeBSD hackers everywhere will be enthused about > a port of OpenSolaris. Maybe. Maybe not. In a perfect utopia we would have software engineers and computer scientists working towards the betterment of mankind as well as possibly solving some interesing math problems.[1] The reality is that we have a world fighting over how to produce cheap stuff even more cheaply with slave labour and sell it to the unwashed masses with software that doesn't work and updates that make it worse all while charging you money at every step.[2] Car manufacturers are even worse maggots. That is the real world. Sorry. No magic. > I'd rather be using the same platform as the rest of > you, but I know what my ability is to purchase more hardware for this > personal endeavor - $0. What if it were hosted elsewhere? Afterall, you are only ever going to see a serial console anyways. > To me, what's perhaps equally as important as the platform choice is the > knowledge/experience aspect. I'm going to make an effort to get involved in > the PPC communities for Linux and FreeBSD and see if there are ways to > leverage their knowledge and experience. I'm somewhat sceptical about the > ongoing involvement from the current brain trust (the former Sun Labs > folks). Are they still around? Sure they are. Possibly lurking. Possibly busy doing other things. Tom Riddle and Guy Shaw could probably change the face of the PPC world all by themselves if we raise $1 million for them. > Will they continue to be involved? If > they've moved on to other projects or companies, are they in the > same boat as the rest of us -- needing a new platform and hardware > to port on? Q1 : Will they continue to be involved? Q2 : If they've moved on to other projects or companies ... Q3 : are they in the same boat as the rest of us My Answers : Maybe, Yes and probably. In that order. > It occurs to me too that even if I go with the Mac -- which at this point, I > believe would be documented well enough because of the existence proofs in > abundance -- how easy will it be to do low level bring up without things > like JTAG/hardware debugging? Probably near impossible because of the totaly lack of observability. Even the memory map will be different and have you considered the cache issues? > Would another EFiKA run be more beneficial > because of the additional debugging tools that would bring? I don't know. At $100 a unit ( or whatever ) at least we would have access to all the docs and specs. It is something as opposed to nothing. Yes, when it comes to the 32-bit PPC world I am grasping at straws. > This the answer > to this last question is the only thing that'd sway me from using a > PowerMac. > > I guess Cyril's speculation make sense. Until money shows up and the > requirement for a viable and healthy and future-proof platform becomes a > necessity, this will have to be a hacker effort/itch scratching. It started as a hacker effort in 2005. Full circle here we are again. :-) > I'm > tempted to suggest that SMI could just be hatching plans for a PPC > port with IBM behind closed doors which would make all this moot, > but that'd just be mean. Always ask the question "where is the money?" Once you know that, you need to follow it. Is there money in a PPC edition of Solaris/OpenSolaris today? Has the landscape shifted from what it was back in 1994 when IBM was selling a PPC version of the PS/2 and Sun was shipping Solaris 2.5.1 ( 2.6 ? ) on the RS/6000 gear? If there is no money in it then there is no business case for it. Plain and simple. Open source is not all about some utopian vision of free wheeling and fancy free give aways. It is business. - Dennis Clarke [1] http://mathworld.wolfram.com/RiemannHypothesis.html [2] http://blogs.sun.com/jmcp/entry/microsoft_now_doing_nightly_builds
