@Another User said:

The fact is that with the Mac Pro machines, they're already best set up for samples and mixing by the large bus architecture, large shared cpu caches, fast double bus ram, and the abilit to put four large fast sata drives right inside the machine - it doesn't really have anything that you don't need for samples and mixing.

I agree, but they may be overkill for slave machines. That or the next machine is the one I'll use as my main one when all the software is ready.

I got on really well with my atari, and the fairlight was special.

sigh...........

However, i'm in Daryl's fortress here as far as brand loyalty. I don't have any.
It's the goods that count. (I wonder in the new Nuendo 4 will have a decent score component. Or if the new cubase 4 score editor is finally half decent. Sigh..........)

Seriously now, i'm wondering if we could run a DAW on an entry level CRAY, with their custom built integrated multi processor setup, the Gazillion gigs of ram, and raid array raptors.

You know, like a fairlight, with a supercharger.

http://www.cray.com/products/xd1/index.html">http://www.cray.com/products/xd1/index.html


Sort of puts the 'friendly hardware joust' about performance in perspective.

Regards,

Alex.

just to weigh in here... I love my Mac, much of time, I hate my PC, much of the time. They both work, much of the time, and they both do what I need them to do, much of the time.

Having the MIR running in a dedicated box makes a great deal of sense, to me. Yes, it can be a pain to get multi-machine setps working smoothly, but it's unavoidable, generally speaking. What I personally think would be ideal, though, would be for MIR to be offered as a standalone hardware box. Yes, it would be expensive, but it would be soooo simple! Or, perhaps it could be sold pre-configured in a dedicated Linux-based machine, with some OS setup that's not user-configurable... so we can't f**k it up!

J.

@jbm said:

Having the MIR running in a dedicated box makes a great deal of sense, to me. Yes, it can be a pain to get multi-machine setps working smoothly, but it's unavoidable, generally speaking. What I personally think would be ideal, though, would be for MIR to be offered as a standalone hardware box. Yes, it would be expensive, but it would be soooo simple! Or, perhaps it could be sold pre-configured in a dedicated Linux-based machine, with some OS setup that's not user-configurable... so we can't f**k it up!

J.

I agree, but apparently that is not what VSL wants... [:(]

DG

DG: I disagree it is about the hardware as well as the software, not just the packaging too although that is important.

But you're right the whole argument is annoying so lets agree to disagree and acknowledge that we're none the worse for it ! [*-)] [:D]

Nick: Saying that Mac Pro's are too much as a slave machine is, sorry, but kind of stating the obvious.

All the best,
Miklos.

JBM: That's a good idea however my point earlier in the discussion was this: why buy 16 cpu's when you only need 8. Or why buy 32 when you only need 16 - whatever is needed for MIR. Somebody like me who can't afford to buy more gear than I need - it would be great to be able to run MIR on my existing computer and not have to buy the hardware again. Of course it would always be better in a luxury world (for me anyway) to have a stand alone box then I could mix live! awesome, but until that day comes at least it's good for those on a budget to have the *option* of using their existing machine to run the program and when they make more money, yes, buy a stand alone box.

Miklos.

Miklos, let's trace the thread.

I said the advantage to Windows machines is that you can put together hardware that does just what you need. Your answer was that the Mac Pro is perfect for streaming samples.

So I responded with the bleeding obvious.

I have a feeling that we're having two different conversations that only overlap occasionally. [:)]

The mac pro IS good for streaming samples - it is not good as a farm machine. That is what is obvious.

Miklos.

Okay, I missed something. Why is that obvious?

Sorry Nick, both of those points seem to me to be obvious - the mac pro is very good (if not perfect) for samples use, but as a farm, the large case, extra drives and ram capabilities not to mention cpu are obviously to me anyway, overkill for a farm machine - better would be 3 or 4 minis perhaps with a mac pro as your sequencer/mixer? one possibility anyway. So both statements are fairly obvious to my mind. If you are not a Mac user or up to scratch on thier hardware, I can understand in that case how it would not be obvious at all - if you quoted me dell model names I would be lost - but I was assuming the specs were obvious, perhaps my mistake.

Miklos.

OMG - what have i done ... [:P]
anyway ... just to keep some things in the right order - one can discuss who invented LCD monitors, but for sure it has not been apple, AFAIK all (at least earlier) LCD monitors from apple starting with the 15" studio display 1998 have been produced by LG - the same company which produced my alltime favourite SGI cinema display, the SW1600 (note the specs! - the only LCD-display i have ever seen with an integrated colour calibration and 110 dpi)

what i don't like with the newer apple cinema displays is more caused by the operating system resp. aqua - the smoothed contours, i always have the feeling the picture is unsharp and unfortunately it cannot be switched off.

as an early bird and enthusiastic BSD user (where the roots of mach, openstep, nextstep and at last OS X have to be found) i think i may comment on obvious inconveniances. i have never seen corrupted file permissions in BSD, i have seen them sometimes in windows, but i have never seen them happen so regulary with a simple install of an application like i had to notice it when installing eg. logic on osx - there _has_ in fact some improvement to be done.

the new macPros are - sorry to say that - just another, though powerful, PC with an intel-board and i wrote in an earlier post that IMO this was the really clever strategy. on one hand windows users can run XP on the macPro and on the other hand apple has to develop their further operating systems just for one platform. i'm pretty sure it will not take too long until we see OSX for *ordinary* PCs so opening the market from two sides for apple.

one of the major advantages of BSD is that it runs on almost every platform you can think of - i ran it on ALPHA(†), MIPS(†), PPC(†?), X86, X64 and i think it was the most important piece of open source software ever released. a lot of webservers, the majority of mailservers and AFAIK all DNS-servers run BSD because of its performance, reliability and stability - not all of this attributes have found its way into OSX so far.
what i'm missing most (especially regarding sample streaming) is a configurable pagefile - why let an operating system (released for generic use of a wide range of applications) decide if and how much virtual memory to use if samples cannot even use any? eg. one of the basic rules for good performance is to have the pagefile on another disk than operating system and applications.

hopefully you can notice i do see the advantages, but also just don't conceal the inconveniances. all this is only my personal view only, of course.
christian

christian: of course apple didn't invent lcd's I didnt' say that... [*-)] point taken on the sgi - good screen, but for the time the apple displays have been consistently of the best quality and offered much larger screens that most of the competition: but I digress....

me likey da aqua

[[:|]]

permissions corruption: we are agree don that point - it should not happen and should be fixed, but is this not nit picking?? [:O]

fourth point: AHA [I] so they're not that different from PCs [I] [I] [I] so you *could* adapt MIR to run on it! as paralell software? or mac osx native. Even if you have political issues with Apple [8-)] which I, ahem, respect... [[;)]]

Your fifth paragraph only strengthens the above argument, if you really think Apple with release osx for ordinary machines (he he) then all the more reason to develop for it.

I agree that mac osx should be more tried and true to the original philosophies of bsd and practicalities that flow from that. and I agree about the pagefile, that is a pain, and should be fixed... however, the near future shows that the usage or ram in the next 12 months will grow to a theoretically enourmous number, and practically will only be limited by how much ram you can cram into a machine - lets assume 32 Gb by 12 months from now (optimistic perhaps) - even 16Gb is capable now but for software limitations - isn't that all that would be needed? Or am I technically missing what you said?

Miklos.

@mpower88 said:

Even if you have political issues with Apple [8-)] which I, ahem, respect... [[;)]]

With all respect, I didn't read the slightest political issues in cm's posts, but a lot in yours.

I don't deny my political leanings, but I have prejudices on both sides. However Mathis the plot runs deeper and thicker than you know! Based on reading other posts of Christians in other threads and here I have made an inference led into an assumption, which I hasten to add does not necessarily make it any less true or relevant! [:)]

@Another User said:

permissions corruption: we are agree don that point - it ... but is this not nit picking?

an example why i can't consider such issues not as *nit picking*: if you try to extract an archive (eg. the compressed EXS samples from a pro edition DVD) with UnRarX on a machine with corrupt file permissions you get messages like file not found, format error, unexpected quit, sometimes actually nothing. this comes up without having the corruption related to the files from UnRarX or the DVD of course and disappears as soon as you repair permission.
you don't want to know how many hours, days, weeks of user support such a tiny issue can generate ... and of course no one blames OSX, everyone is furious about the *poor behaviour of VSL libraries* - too bad they have that size.

just check out the ratio of support requests for XP and OSX to understand why i'm alittle bit *picky* about fulfilling basic requirements for an OS. unfortunately this is backed up by a marketing campaign which tries to present OSX as the ever and world's best, unfailing, easy-to-use OS when it is not.
i understand and am aware why: because it's a modern multiple purpose and highly complex system - and it would be no shame to admit this instead of reiterating to communicate one would not need to know anything about computers to use them. this is, especially for *NIX, not true.

my example supporting this point of view: how many OSX users know that program code for running processes stays in the memory even if the respective file has been updated - and is it noticeably mentioned somewhere? it is not. the results can be disastrous.
now one can ask either for the technical understanding or instruct to reboot the computer before and after updating something - both contradicts the original and still published apple philosophy leaving us with the decision to choose the less worse ...
christian

Christian you make good points and although I understand what you are saying I didn't know that until you just said it. However, my simple point, while I reserve that I respect your own experienced final judgement on it is just that I think if you weigh up everything on all sides of the argument, the playing field is *more or less* level now compared to before - that's what I meant by nit picking, not necessarily that particular issue of permissions repair, I probably should have said it better - sorry for that. What I'm saying is that I think things are different than when the original decision to only develop for PC was made - (at which time it made more sense) - now however all I'm saying is that perhaps a review of that decision is warranted and my main argument for such a review is the fact that it would be great for a lot of people to only have to purchase one box of cpu's instead of two. - And by the way thank you for your patience and time in discussing this, I really appreciate it.

Miklos.

thanks for reading and considering my arguments, miklos. to have everything well-balanced i could now write about some things i don't like in windows and probably such a posting would easily become far longer than the ones above [:P]

but back to the topic (at least i think it has been the topic ...): the MIR development started (fortunately) on windows and will be continued for this platform. when all needed requisites will work together as expected we will have some hard times to find the best possible machines - and be assured there will be strong system requirements and only a limited number of supported components (you see parallels to available apple hardware? - intended!)

if we somwhen later have both feet firmly on the ground again and the developers returned from their well-earned vacancies i'm sure someone will check out if a) all components are crossplatform b) all parts of code are portable c) resources are available for a possible OSX version - and this is _no_ promise, this is just to state it is TBD and the anwer can finally be yes or no.
christian

OK - - so now all we need to know is a likely timeframe for the appearance of MIR and whether it is 32 or 64 bit and whether it will run on Windows XP or will require Vista (and which version of XP or Vista it will need) and whether it is going to need 2 or 4 or 8 or 16 processors and whether the processors need to be Intel or AMD and whether it is going to need 2 or 4 or 8 or 16 or 32 or 64 or 128GB of RAM and what sound cards it will be compatible with and what specs the soundcards need to have (is 96KHz enough or will we need 192Khz?), etc. Just a little info so we can start budgeting now! [:D]

miklos, I've been working on Macs every day all day long since 1985. Literally.

Since you're obviously irritated that I would point out something so obvious, I want to reiterate that I only pointed it out because what you wrote implied that you didn't understand it! Oy veh!

And now let me move on to the next point (which I've made before in other threads): I personally don't think the Mac Mini is a very good slave machine, for several reasons: 1. only 2GB of RAM; 2. only 2GB of RAM; 3. only 2GB of RAM; 4. only two outputs unless you use a FW interface, and since the machine only holds 2GB of RAM, you need more machines than necessary and therefore more FW interfaces; the other problem with the built-in interface is that optical formats are not ideal for carrying clock (*very* jittery), and there's no other way of clocking the machine other than via its optical input; 5. it comes with a sissy hard drive, so you have to use FW...in which case you have a drive and a FW interace on the FW bus, and I'm not crazy about that idea; 6. only 2GB of RAM.

I may yet buy a Mac Mini to have an Intel machine for testing software and other things, but I wouldn't buy several of them as slaves. On the other hand, space and noise aren't issues for me, since my machines are on the other side of a wall in the garage.

I forgot to add the subtext, which is the most important part of it all: much as I prefer to use Macs (only because I'm more comfortable with them, not because I necessarily think they're better), I think you can put together a Windows machine with equivalent components to the ones in the Mac Minis and solve all the problems I just mentioned.