Oolite Bulletins

If you really want power, I think you can flash the firmware on a gtx 295... smokin.

Good morning

is that to run on a mac (not pc)(or both?)

local graphics joint offering "EVGA PCIe gtx285 1024mb for Mac" deal (on octo-core)

or "Leadtek PCIe Quadro FX4800 1.5GB MAC" -- no 5800 but I read (?review) both have 4GB

local graphics joint probably not crooks, of course..

So the 295 is just two gtx 260s in SLI, which are supported in snow leopard. So you should, without too much work, be able to make a 295 run on a mac pro.

I think.

Without any firmware changes, the gtx 295 will run on windoze. But as you posted in the mac forum, I assume you wanted to know about macs.

local graphics joint offering "EVGA PCIe gtx285 1024mb for Mac" deal (on octo-core)

or "Leadtek PCIe Quadro FX4800 1.5GB MAC" -- no 5800 but I read (?review) both have 4GB

I don't really understand what you mean.. It says for the 285 1024 mb (1 gb) and for the fx4800 1.5 gb, so neither of them have 4 gb. The choice of the two cards depends on what you plan to use it for primarily. IIRC, the gtx 285 is a gaming card, while the fx4800 is more for graphic design work.

Edit: SL does add support for the ATI Radeon HD 4870 and 4850, which are beasts of cards themselves, although widely considered inferior to the GTX 295.

Something completely different:

Today I finally received my copy of Snow Leopard and just finished upgrading. As a rather new Mac OS X user I'm really impressed how smooth the upgrade went.

And most important of all: Oolite still works in 10.6! ... not that I ever had a doubt

Today’s revision log:
Now it’s time to find some Grand Central Dispatch sample code. :-)

I kind of hate the way Mac's have gone the entire route of if your using a high OS you've got to buy all your software again and the old stuff won't work.

Or if your running an older system then sorry your poor (you may be creative) but your not rich enough to deserve an OS.

They've done it twice now, removal of support for 9 applications (software) and the intel switch (hardware). Basically I find it quite ironic that you could play elite practically on a calculator but now compatibility is forcing creative peoples hands when it comes to who is "allowed" to play the game.

It's not the peoples fault it's apples fault for screwing it's following over time and time again. Can we please have an archive of last version oolite. So poor people who can't afford a new system (or just bloody mindedly don't want to change) can continue to enjoy this game too?

I find rampant capitalism extremely damaging to computing. I feel like hacking the whole bloody thing out and installing linux.

I dont think that is entirely fair, when I moved to NT4.0 I had all kinds of stuff that didnt work, and I still have a handful of other windows and dos games that dont play on xp, even in compatability mode, or dosbox, or anything else I have found.

True, apple have moved hardware, and backward compatability is not guaranteed, but I dont see Windows as significantly better, and there are various things like rosetta that have been used to minimise the pain.

And if you want support for legacy hardware, I really dont think you can fault apple over microsoft. You may not be able to run SL on a ppc machine, but I doubt you can run Vista on a similar ages x86 machine either -- generally osx is considered to be more usable on old hardware than windows, ymmv

EDIT: I dont think this is capitalism in action, these changes are technically driven not idealogical.

I mean, it does add a certain amount of options for the devs for them to ditch PPC support with SL, ya know?
As the 360 uses a PPC processor, do you think it would be possible to run OSX 10.5 on it? Or does that fall into the category of hackintosh - do not discuss?

And you might be able to run elite on a calculator, I'm sure you could run it on the TiNspires they want us to buy for math this year...

Well, a 360 isnt apple labelled, so technically, if it were possible then I guess it would be a "hackintosh" so it is technically verboten.

That said, I dont think you are talking about illegally running osx, I take the question to be more whimsy, so as long as we dont get into it too far, I guess I can answer.

No, you cannot run OSX on a 360, not with any normal definition of possible. I also think it is unlikely that you can run osx nor indeed whatever funky windows version 360 has on IBM p-series servers which are power based.

wackyman465 wrote:

I mean, it does add a certain amount of options for the devs for them to ditch PPC support with SL, ay know?

Yes. It gives them the option to work on tasks other than ensuring everything works on PowerPC, which benefits the large majority of customers who are using Intel systems. Programmer time isn’t unlimited (and this generally can’t be solved by throwing more programmers at a problem, even if you have unlimited supplies of money).

True, but there might be the odd PPC/Linux user around (not me)...

Let us not forget that the resale is the corner stone of any business model. Why sell someone something once when you can sell it to them 3 times.

Backward compatibility costs money and yields little profits. Of course this leads to a difficult truth for many programmers to swallow. Purposefully bad inefficient, inelegant but future profitable software.

Maybe those good old boys at apple and Microsoft aren't as dumb as they look. Maybe it's like that by design.

ClymAngus wrote:

Of course this leads to a difficult truth for many programmers to swallow. Purposefully bad inefficient, inelegant but future profitable software.

I tend towards cynicism myself, but all the Apple programmers I know to any extent are passionate and driven people who definitely don’t think that way. Each OS release is better than the last, not because the last was bad but because people who are very good at what they do have put a year or more of concentrated effort into making it better. It’s a huge project, and new techniques and new challenges arise all the time.

The hardware keeps changing, too, and so do the techniques required to make use of it. When OS X was new, multi-processor systems were rare; all Intel-based Macs have at least two cores. When OS X was new, windowing systems were software-based; now it would be crazy not to design for graphics hardware. Both these transitions were on the horizon, and the system was designed to be ready for them, but adapting still took a lot of work, and software that makes good use of newer systems was unusable on older systems. (Software that didn’t is inherently obsolete.)

The current big obsolescence threshold is PowerPC, and the next will be 32-bit Intel systems. Apple certainly didn’t switch processors just to make life difficult for users; it was a great risk for them. But IBM just wasn’t delivering the goods (and still hasn’t). If the switch hadn’t been made, the platform would be dead. The 64-bit transition is also necessary, but in most ways less dramatic.

The 64-bit transition also provided the opportunity to clean up old cruft, since there’s no need for backwards compatibility. This means that code written specifically for 64-bit systems will be faster and easier to write. The alternative isn’t to make 32-bit code faster and easier to write, it’s to make 64-bit code unnecessarily bad. This benefits no-one.

You can use a ten-year-old computer quite effectively. But you won’t enjoy it, because even if it’s the best of what ten years ago had to offer, the fact is that modern systems – both hardware and software – are immensely better, as the result of driven, passionate people doing hard work.

To be fair, I should mention that there are driven, passionate people at Microsoft, too. The main problem seems to be that they get mired in an inflexible bureaucratic structure until their brains turn to mush.

ClymAngus wrote:

Of course this leads to a difficult truth for many programmers to swallow. Purposefully bad inefficient, inelegant but future profitable software.

I think this is hugely unfair for a number of reasons.

Firstly, I think it is outright untrue that any code is purposefully bad. I have seen (and myself written) plenty of bad code, for a number of reasons, time, experience, legacy integration for example, but never because the author wanted, or was instructed to write bad code. It is easy to laugh at microsoft, as an example, but do you seriously think they have project managers walking round the cubicles screaming at their programmers to make sure windows is a dog, and punishing them for fixing bugs?

It is rarely directly a programmers fault that features, such as backward compatabilty, are missing or not perfect, this is typically a business or project management decision. And it is a justified business decision. Think how many different macs there are going back into history, hardware and software, think of the number of combinations that you need to test and support, think of the effort you are spending to ensure not only your code, but the code of your customers (if you are talking about xcode) which could be spent on making things more efficient, or adding features.

Consider oolite. We could abandon cross platform and backward compatability, use the latest and greatest libraries, and the game would almost certainly be "better". Alternatively, we could decide not to take advantage of new graphics card features and have a less beautiful game, but more compatability. Where do you draw the line? There is no perfect answer, you make decisions, you balance, and you do what you think is best overall. In fact that is exactly what is being discussed, the oolite dev team are moving to a newer, better xcode and that has limited support for older OSes. Swings and roundabouts. The dev team will find their job easier, be more motivated (less frustrated) and get more done. Balancing that, people with 5+ year old gear may find they cannot run the latest version. It may be possible for someone to set up a non xcode build for older osx, just like the ones for windows and linux, but again its a compromise, a dev spends some time doing that instead of making the game better. Is it better for oolite to run on xyz, or to have new feature abc? For you, compatability is key, but then, if you dont want the new features, you can stick with the existing version.

I also dont think its hard to swallow at all, most of the coders I know are in an almost constant state of frustration that they do not have enough time to deliver to the best of their abilities, and that they must deliver cool feature X that an exec wants, instead of making improvements that are not on the press release.

Cmdr James wrote:

ClymAngus wrote:

Of course this leads to a difficult truth for many programmers to swallow. Purposefully bad inefficient, inelegant but future profitable software.

....
Firstly, I think it is outright untrue that any code is purposefully bad. ......, but never because the author wanted, or was instructed to write bad code.

Probably true in 99% of the cases but I know of one occasion of adding deliberate bugs. In 1999 a friend was working on the famous millennium bug in administrative software. They needed a new release to be sold in 2000. But in that software the VAT percentage was hard coded and they already knew the percentage would change in 2001. Wow, new update to be sold in 2001 because the government changed the vat rate. He as programmer was not allowed to make it a user changeable variable. To me it sounded as a big rip-off at that time (and still does).