Common Lisp for Mac OS X
in [Lisp]
|
From: Harald Hanche-Olsen on 2 Jun 2010 10:36 + Tim Bradshaw <tfb(a)tfeb.org>: > On 2010-05-29 18:33:58 +0100, Harald Hanche-Olsen said: > >> Appl's X11 support after the release of Leopard was a disaster. > > I think you mean "X11 was a disaster" don't you? No, I don't. I am well aware that many people think so, but X11 has served me well over the years, warts and all. I am not really qualified to speculate on whether or how a networking window system could have been done better, though. -- * Harald Hanche-Olsen <URL:http://www.math.ntnu.no/~hanche/> - It is undesirable to believe a proposition when there is no ground whatsoever for supposing it is true. -- Bertrand Russell
From: Andrew Reilly on 2 Jun 2010 20:40 On Wed, 02 Jun 2010 10:36:47 -0400, Harald Hanche-Olsen wrote: > + Tim Bradshaw <tfb(a)tfeb.org>: >> I think you mean "X11 was a disaster" don't you? > > No, I don't. I am well aware that many people think so, but X11 has > served me well over the years, warts and all. I am not really qualified > to speculate on whether or how a networking window system could have > been done better, though. I think that there are quite a few people who don't think that the "network" part is necessary, given that most machines capable of running X (well) are also capable of running the whole GUI, and that the partition should be elsewhere. This was the original premise of Java, as it grew out of experience with NeWS (which was a PostScript based GUI that Sun played with for a while --- I can't remember if you could run NeWS apps over a network, but I can't think of any reason why not.) The only other networked window system that I've heard of was that of plan9. I've never used it, but I do like the model, as I've read about it. The difference from X is that it ran the window manager on the terminal, and every window was an independent communications link over a tty mux (pty-style) to some server process on the (remote) host. I don't know the details of the graphics language that was spoken over these channels, though. Thinking about it now, it could very easily have been Display PostScript, and that would have been very fine indeed... Cheers, -- Andrew
From: Tim Bradshaw on 3 Jun 2010 03:14 On 2010-06-03 01:40:17 +0100, Andrew Reilly said: > I think that there are quite a few people who don't think that the > "network" part is necessary, given that most machines capable of running > X (well) are also capable of running the whole GUI, and that the > partition should be elsewhere. This was the original premise of Java, as > it grew out of experience with NeWS (which was a PostScript based GUI > that Sun played with for a while --- I can't remember if you could run > NeWS apps over a network, but I can't think of any reason why not.) I'm fairly sure NeWS stood for *network* extensible window system, and yes, it could run things over a network. It was a good idea, but it was slow on anything you could afford and Sun cocked it up in the way they cocked up almost everything. Obviously when X was designed, a window system that would run over a network was pretty useful, because if you had a 3/50 on your desk (which would run X11 OK after the release that improved performance - R4 maybe?), but there were much bigger systems in the basement, you didn't want to run too much on the 3/50. Nowadays it's still important, but in a way X doesn't provide. Obviously the Linux fanboys don't see it as important because they live in a world where compute power is cheap and it's easy to provide on your desk. (But no one actually cares what they think.) But think about a bank or something: they have maybe a few tens of thousands of desktops, all of which have operating systems which are busily persistently caching data. This is a hideous security & compliance nightmare for them, so there's basically nothing they'd like more than to put all of that in some central secure location and remove any persistent storage from desktops. As an added bonus they then get rid of a lot of maintenance costs. But what they need, which X can't do, is to be able to move the whole desktop to another screen, so people can move around. Of course there are systems which can do this, they're just not X.
From: Pascal J. Bourguignon on 3 Jun 2010 04:56 Tim Bradshaw <tfb(a)tfeb.org> writes: > Nowadays it's still important, but in a way X doesn't provide. > Obviously the Linux fanboys don't see it as important because they > live in a world where compute power is cheap and it's easy to provide > on your desk. (But no one actually cares what they think.) But think > about a bank or something: they have maybe a few tens of thousands of > desktops, all of which have operating systems which are busily > persistently caching data. This is a hideous security & compliance > nightmare for them, so there's basically nothing they'd like more than > to put all of that in some central secure location and remove any > persistent storage from desktops. As an added bonus they then get rid > of a lot of maintenance costs. But what they need, which X can't do, > is to be able to move the whole desktop to another screen, so people > can move around. Of course there are systems which can do this, > they're just not X. This is possible, using xmove. -- __Pascal Bourguignon__ http://www.informatimago.com/
From: Tim Bradshaw on 3 Jun 2010 05:12
On 2010-06-03 09:56:00 +0100, Pascal J. Bourguignon said: > This is possible, using xmove. Which is basically a proxy X server, right? So yes, you can do it by virtualising the server, which is how all these things work. I guess you could argue that virtualising the X server is better than virtualising the framebuffer (I think technically it might be, but ractically it's not). |