Merger between AmigaOne and Pegasos dev teams

[uncharted]

downix wrote:
@uncharted

Huh?  Where the heck are you getting this stuff on the OS's design?  BOTH OS's are using a sandbox technique here.  The only difference is that one seperates the new MP-enabled kernel from the old non-MP kernel.  This way, should a non-MP app kill itself, you don't loose the entire system, just the non-MP protected section.

The OS 4 feature list it explains that it does not use a Sandbox approach.  I mean before you started slagging it off you did actually READ the information available on it didn't you?

So you are saying that I can write an application for MOS that makes use of all the advanced features of Quark including MP?
That's either great news or total rubbish.

-EDIT-
And MOS has VM.

Available to the applications written for ABox?

[Russ]

Hyperion still made an offer.
Bplan has always wanted to be the only on on the block and Amiga
is in their way. The only problem is that they are
going for Amigans and not the general computer population.
This is too small of a crowd to go for. It only hurts
the community. Amiga people are rabidly loyal.
The board you have now is no longer being made. A
new one is promised real soon. It is funny how Morph OS
crowd hang around Amiga.org so much.

[Kronos]

@Russ

Hyperion made an offer ?
 :roflmao:  :roflmao:

When should that have happened ?
Before 11/01 when Hyperion wasn't in charge of OS4 ?
Or afterwards when it was allready decided that they would go head-on ?

There were talks between bPlan and AInc in spring 01, but looking how things evolved I can
only congratulate bPlan for refusing to sign a deal.

What ever was offered afterwards, wasn't any more honest than bPlan offering/asking
Hyperion to port games for MorphOS (which also happened at that time).

Also, it is Genesi who are cooperating with AROS,Phoenix or external developers
like Titan or e.p.i.c., they are just to smart to tie themselves to incompetent clowns
from snow-city.

Bout cooperation today:

What would AInc/Hyperion have to offer (except maybe the name) that Genesi needs ?

NOTHING !

[meerschaum]

am I the only one who dosent want to see cooperation?... I dont want Amiga.inc and friends  involved with Pegasos/MOS... in my opinon they could only serve to bring down Peg/MOS ...the gain would be one sided only...

[GGS]

@ HMetal

#Any input or test results you have regarding AmiFTPd would be most welcome. Thank-you.

Sorry for the delay. AmiFTPd works quite well with MorphOS. It's only the gui-part that gets some 'locks' for a second or two in certains situations. The TCP-part works without problem.

Also, when talking about AmiFTPd, what I'm missing is that its not possible to create drawers. This should be very useful.

/Gunne

[downix]

uncharted wrote:

downix wrote:
@uncharted

Huh?  Where the heck are you getting this stuff on the OS's design?  BOTH OS's are using a sandbox technique here.  The only difference is that one seperates the new MP-enabled kernel from the old non-MP kernel.  This way, should a non-MP app kill itself, you don't loose the entire system, just the non-MP protected section.

The OS 4 feature list it explains that it does not use a Sandbox approach.  I mean before you started slagging it off you did actually READ the information available on it didn't you?

I read Ben Hermans latest statements on AOS4, and he said clearly that to seperate new-API from old-API they will be using a sandbox.

And, yes, I have read the AOS4 feature list.  I dissected it quite completely, and it meshes with what Mr. Hermans said, a sandbox is needed for future roles.

So you are saying that I can write an application for MOS that makes use of all the advanced features of Quark including MP?
That's either great news or total rubbish.

Nope, but you can't make an application that can use all of the advanced features of ExecSG either.  So no difference here.

-EDIT-
And MOS has VM.

Available to the applications written for ABox?

Yes.  However, like in QNX, with MOS the VM is added using a system module, so apps can only make use of VM if they are designed to.

Having run such a system with QNX, I know very well how nice that is, not having the OS decide when to allocate VM but instead the app decide on exactly how much VM it needs.  Gives me far more control over the applications that way.

-edit-
Got some clarification about MOS's VM system.  It is currently in internal beta, but not a high priority for inclusion into the public release.  

[downix]

@Russ

Made an offer, when?  IIRC, it was Genesi that made the offer, not Hyperion.

[uncharted]

downix wrote:
I read Ben Hermans latest statements on AOS4, and he said clearly that to seperate new-API from old-API they will be using a sandbox.

And, yes, I have read the AOS4 feature list.  I dissected it quite completely, and it meshes with what Mr. Hermans said, a sandbox is needed for future roles.

Can you link to this comment?  I've had a look through all the recent interviews and have had a look through the some comments (it gets boring reading the same stuff over and over) and i couldn't find anything about a sandbox.

If you are refering to what was said about the OS4 sandbox for the mythical "OS 5" then that is a different story and hardly relevant to what we're talking about here.

Nope, but you can't make an application that can use all of the advanced features of ExecSG either.  So no difference here.

This is where I'd disagree as some of the functions are already in place such as resource allocation and limited MP.  Can MOS apps have access to *any* of quark's functions?  I was told (some time ago admittedly) by Ralph that things running in the A Box *Including* PPC native apps could not see Quark at all.

You see, unlike other people, I actually did lots of research into MOS (which wasn't easy as information is rather thin on the ground), I've read nearly every interview, I've actually tried it on Pegasos.  

On this point I'll just have to leave it at I agree with you on some points but not others.

Yes.  However, like in QNX, with MOS the VM is added using a system module, so apps can only make use of VM if they are designed to.

Having run such a system with QNX, I know very well how nice that is, not having the OS decide when to allocate VM but instead the app decide on exactly how much VM it needs.  Gives me far more control over the applications that way.

Cool.  I can see the advantages of that.  I note that AOS can turn off VM (on the fly in some cases) which would be good to.  Personally I've never had a problem with VM and am not so paranoid about having it on or off.

[uncharted]

downix wrote:
@Russ

Made an offer, when?  IIRC, it was Genesi that made the offer, not Hyperion.

I still find it laughable that people actually take Bill's "offer" seriously.  If it was serious surely Hyperion would have recieved thier free board by now.  And how comes (as was originally offered) Ben had to go to Paris to collect it when the other boards were sent to the developers in question?

And the biggest one of all.

Why did Bill need to offer a free board anyway?  Why didn't Genesi just process Hyperion's orginal order for a Pegasos?  All this nonsense about Hyperion not accepting a free board when Genesi wouldn't *sell* them one in the first place!!!

[AmiGR]

Ehm, their original "order" was not exactly an order,
it was a registration at the MORPHOS developer
queue. Importand (for MorphOS) developers, got the
first boards.

BTW, that was before Genesi was founded. It was
actually before their team up with Thendic got public.

[Kronos]

@uncharted

Well those boards were sold by bPlan (a company that didn't really exist anymore in 2002)
and clearly targeted at companies wanting to develop for MorphOS.

So it was clear that Hyperion would not get on of the very few boards produced in late 01.

It's also questionable whether that would have still be a legally binding order ~1 year later
when the board was offered.

The offer was made for one reason:
Before that ben's standard answer to OS on Peg was allways something like
"they won't give us the board we ordered", when in reality everybody knew that it was the
licence-issue that made it impossible.

This PR (or shall I say FUD) campaign was contered with offering the board, which directly
leed to statements like "sending a board is not enough" and "getting a Pegasos is easy"
from the same ben Hermans.

But somehow I don't think the board was what Nate was refering to, but when Genesi
offered to sell Pegasos bundled with OS4, just not with an licence from Ainc or a dongle.

Or maybe he was even refering to an offer made from bPlan in 01, to make Hyperion
port games to MorphOS, just like Titan or e.p.ic.

[uncharted]

@Kronos

:lol: :lol: :lol:

[Kronos]

@uncharted

What ??

[downix]

@uncharted

Egads, you really don't know how OS's work, do you?

[downix]

But somehow I don't think the board was what Nate was refering to, but when Genesi offered to sell Pegasos bundled with OS4, just not with an licence from Ainc or a dongle.

Genesi never offered such a deal, so quit spreading it so thick.

I was refering only to the times a board was offered to Hyperion, and nothing more.

[uncharted]

downix wrote:
@uncharted

Egads, you really don't know how OS's work, do you?

Then explain it to me.

-Edit-

I hope that just wasn't a lame excuse to get around my statements.