@# Quotes DB     useful, funny, interesting





Google
 
Web www.quotesdb.info
Undernet  |  EFnet  |  Quakenet  |  Freenode  |  Dalnet  |  Ircnet  |  Galaxynet
Page: 1 2 3 4 5 6



Comments:

<0> that's cool
<1> you wrote a ppc emulator OS?
<2> sup folks
<3> yeah, that's what it is, essentially
<0> and it's fast. your dynamic translator works pretty well
<4> PearPC.
<3> yo geist, sup?
<2> just got home
<3> cool, cool
<1> pretty cool
<3> yeah, i'm hoping to get within 80% efficiency
<5> eieio: Hey.
<4> whats ur emulator called?
<1> monarc
<5> eieio: That's very cool dude. I was just thinking about this today, just a virtualization engine
<5> eieio: Do you have the code up somewhere?



<5> monar = eieio's thing?
<1> sbahra: he is selling the code for $2000, about the price of a mac :)
<5> heh
<5> That is a really nice approach though.
<5> It makes a lot of things rather trivial too (process migration, saving state, etc...).
<1> did you see the video?
<5> Yes.
<5> $2000 is under-priced though. :-/
<1> heh
<5> eieio: You should just open-source it. :-P
<4> <3 mplayer aalib output
<3> heh, i don't have the code up anywhere right now. there are some important elements that may need some additional protection
<0> do i recall correctly that apple is coming up with a system similar to that, to allow old mac programs to run on the new x86 macs?
<6> you mean, rosetta?
<5> They already have this.
<0> ah
<5> and universal binaries (which is ugly IMHO :-/).
<0> thanks, i don't use a mac, i don't follow it
<3> rosetta, based on transetta stuff
<3> transmetta*
<0> really?
<3> er
<6> right, licensed by apple for lots of $$$
<3> transmeta*
<7> oh, really?
<7> so that's what happened to Transmeta
<3> yeah, but it's only an application level translator that uses static recompilation
<0> ya
<3> and from the Adobe Photoshop tests, it isn't really acceptable for a lot of larger applications
<0> heh
<0> i have no idea what apple is playing at, jumping architectures like that. is the ppc really that far behind now?
<3> well, from what i know it's mostly because of IBM not wanting to keep at the technology contest with Intel of the personal computing market
<0> ah, okay
<0> i find it ironic that apple is moving to x86 while microsoft is making a ppc-based machine
<3> yeah, that is kind of funny
<1> apple shouldve just used power5 processors :)
<3> the deal actually happened with MS a long time ago, around 2001 IIRC
<0> hehe
<0> the ppc has ended up being a popular ip core lately
<0> it's popping up in all sorts of strange places
<0> like on xilinx fpga's
<3> IBM had nintendo, apple, and ms in contracts at the same time. that was a big hush hush
<3> yeah, the PPC really has been a good product for IBM
<0> ibm is really trying to take some ground from arm now
<3> seems like they are doing well at it
<0> indeed
<3> wb nn
<3> hmm, i suppose the broadcom translator will be easier with some additions to the MMU ***ist. perhaps i should just wait for those first.
<3> it's certainly fortunate that apple chose to support some of the more common chips sets. less driver work for me :)
<0> hehe just using a pci-to-pci translator directly?
<3> yep
<0> common bus architectures really helped portability, i imagine
<0> are you planning to do hardware architecture translation then?
<0> or is that what you're talking about? making one peripheral look like another?
<3> umm, the translators allow the emulator to treat the real devices as devices controlled by the virtual bus
<0> oh i see.
<3> in some cases it requires some knowledge of the way the device operates, but often it's much less effort to build the interface than to make a full driver
<0> ah, so you need to do translator modules for the devices
<0> how about providing virtual devices through translation from a fully-driven hardware?
<3> an added advantage is that the coupling between the client OS and the device is much tighter



<0> yep
<0> i believe that
<0> i suppose there's not much point in doing that sort of emulation, now that i think about it. unless there's a problem getting drivers in the os
<3> yeah, i use emulated hardware to host driven hardware also
<0> ah okay
<0> cool
<0> awesome :)
<3> some ATA controllers/drives don't support bus master dma, for instance, and the OS X drivers require it, so i have to emulate in that case
<0> ahh
<3> so i've been wondering about the whole legality issue. from what i understand, emulation has a certain amount of legal precident in favor of it being fair use
<3> wb nn
<4> tx
<4> had some issues with my power supply emitting sparks
<3> ooo, that's a bad thing when that happens
<4> Sometimes
<4> this time it was my lights broken
<2> ooga booga
<8> Mikaku,
<9> mur:
<8> Mikaku:
<9>
<8>
<8>
<9>
<8>
<8> cya now
<8> i go as i explained above
<8> cya
<9> cya ;)
<10> morning
<2> morn
<4> hmm
<2> indeed you do
<4> geist: vomit is my kernel :)
<11> hi
<12> hello
<11> What do you think about the following address space management technology (for ppc64):
<11> every process has ***ociated a list of segments
<13> you're asking one who stole the name of a famous mathematician
<11> there is one large hash page table
<11> Mathis: this is to everyone
<13> it does tell you that the one has no fantasy to create an own nickname
<11> mappings are inserted into the hpt as necessary (i.e. all mappings for kernel and all processes, which do not collide if they all have different segments ***igned)
<11> slb[0] = kernel segment
<11> on process swith, flush slb except segment 0
<10> yo ness
<11> in slb miss handler, test if the requested segment is in the current process' list
<11> hm, that's bull****
<11> forget it
<12> Mathis: was that comment a reference to a lack of knowledge?
<13> *beep*beep*beep*
<14> guildo hat euch lieb
<12> heh, i am not offended, though judging by appearance is usually unsuccessful
<12> also, i am unsure if i shall keep this handle
<12> i can not find a new handle
<12> my previous one was annoying
<12> i had it since i was 12, though it no longer applies
<13> was it j35u5 ?
<12> no
<12> it was LinuxFreak
<12> i wanted a name that expressed my strong interest in mathematics, thus, i took Fibonacci, and placed Phi in front, Phi as in 1.618033
<12> thus i have the Fibonacci series and the number Phi in one name
<13> you are soooooo uuuber-coool
<12> erm
<13> you are my gaaawd
<13> *bend*bend*bend*
<10> Mathis: just stop bitching around ...
<9> http://www.caballe.cat/media/2006/08/2463_startrek-bluescreen.gif
<9> (1.4MB size)
<15> link doesntwork
<9> hmmm


Name:

Comments:

Please enter the result of the sum 63 + 46 (to avoid spam):






Return to #osdev
or
Go to some related logs:

#css
Failed to import module 'portage_db_cdb.database gentoo
xmms idjc
Can not call constructor php
python telnetlib cntrl
#gentoo
undefined symbol: cairo_xlib_surface_create_for_bitmap
gentoo eix messed up
jigajoule
launch+win+arp



Home  |  disclaimer  |  contact  |  submit quotes