XCode uses distcc for it's distributed compiles; the problem is it's an old version and the patch is no where near applying cleanly to modern distcc. Might be easier to just use something to advertise distcc on rendezvous manually.
The reason it's news is that GNU has no support for Mach-O, Apple has their own assembler and linker. Making them work on linux is a real pain in the ass.
There's this service provided by airlines called "baggage handling". What you do is, you put your comfy show-room-floor shoes (that may or may not contain metal) in your baggage, and... check them in!
You can then safely wear your shoes that don't set things off to the airport. Then (*gasp*) take them out of your baggage when you get to wherever you're going.
That's from the previous (abandoned) port of KDE to Qt/Mac (by one of the guys from trolltech).
It went through bit-rot and pretty much needed to be redone from scratch. This time around the plan is to keep the patches up-to-date until we can get all of the changes back upstream.
Qt/Mac apps generally look like Mac apps, in the case of KDE stuff, so far the style engine is horked, so it's confused as to what to make things look like, so it's pretty ugly.
It's on the todo list. Qt still doesn't directly call native widgets, but it does use the OS's mechanism for drawing, so, in theory, it looks like native apps if Qt doesn't have any bugs.;)
Most of gnome2 (well, 2.0) is in fink, and gnome 2.4.1 is being wrapped up right now. "Unstable" is not for testing, Fink unstable means "it's been tested by the developer and is ready to be supported by him/her." 2.4 is being worked on in "experimental" which is the playground for packagers to work on things without having users complain that [foo] isn't working.
"Fink" has no stance on desktops, but Gnome's packager has historically been insanely busy and didn't have much time to keep up with things very quickly... That's recently changed, Gnome is going to be maintained by a group of people (since it's so big) and things should get updated in a more timely manner.
I've not used it myself, but tonight at our local LUG in Raleigh there's going to be a presentation on White Dwarf Linux. From what I've heard they have a pretty nice embedded linux setup as well as some neat embedded hardware platforms they target out-of-the-box.
As for dports being in panther, a combination of technical issues (dports not being in a state they were willing to support forever) and political issues, it got pulled. I don't know if there are plans to re-add it in a later release.
As for metapkg, the idea of a huge database of packages kind of died out, but we've started trying to fill out the wiki with useful porting info, hopefully in time it will become more useful.
Also, the mailing lists are there for anyone wanting to post problems, but we're getting about 2 e-mails a month, so it's not really happened much.:)
Metpkg was never going to be a new joint packaging system, only a resource for all of the packaging groups to avoid some duplication of work by sharing patches for ported software and compilation tips and such.
Generally things that break like that are ones that use undocumented and unsupported APIs, which usually have big fat "use these at your own risk!" comments in the headers.
OK, not everything, I was generalizing, but the point remains. It would be difficult to find apps that don't use libSystem on Mac OS X. I'm not claiming all apps will get a massive speed boost, only saying that your original suggestion that there would be basically no difference is incorrect in my experience.
I've been running the panther seeds for months and it's noticably faster. Believe it or not. =)
All of the base system libraries that all software depends on have gotten a speed boost. Even for apps that aren't explicitly recompiled, they'll be faster; *all* apps spend a not insignificant amount of their time calling code that comes with the OS.
I've been using the prereleases of panther and it *is* noticably faster, I'd say pretty close to the difference in speed between 10.1 and 10.2.
Under the covers they've done a lot of work to make prebinding work better, for one thing, and gcc3.3 provides vast improvements in powerpc code over gcc3.1 (even apple's modified version).
The difference is, TV ads are my way of "paying" for the programming that I watch.
When a telemarketer calls, he's making me pay with my time, without me saying it's OK. I have never ever bought anything from someone calling me on the phone, and never will. If I did, it would likely be from a company I already trust -- it would be reasonable for me to opt in to the n companies I want to deal with on the phone, but it's rather insane to opt out of the (infinity - n) companies I don't.
Telemarketing is the act of peeing on thousands of people to find the one person who actually enjoys a golden shower.
And I can tell you that if they're polling at 1 a second of *anything*, they don't "know what they're doing". That is complete overkill, there's no way the amount of bandwidth being used for testing is worth the 59-second jump on knowing what went wrong. Humans generally have to react to it, that kind of resolution is just crazy.
Slashdot Mirror
not true, I'm getting it when I search for "kinesis ergo keyboard"
Your TLAs are DOA, why R U not bothering 2 explain WTF U R talking about?
IOW, who/what are SBC and CWA?
...it would be inappropriate to call it the "tongue boy". =)
XCode uses distcc for it's distributed compiles; the problem is it's an old version and the patch is no where near applying cleanly to modern distcc. Might be easier to just use something to advertise distcc on rendezvous manually.
The reason it's news is that GNU has no support for Mach-O, Apple has their own assembler and linker. Making them work on linux is a real pain in the ass.
There's this service provided by airlines called "baggage handling". What you do is, you put your comfy show-room-floor shoes (that may or may not contain metal) in your baggage, and... check them in!
You can then safely wear your shoes that don't set things off to the airport. Then (*gasp*) take them out of your baggage when you get to wherever you're going.
Well then check out the screenshots. ;)
I didn't bother to do anything, I was just happy it ran. Making things pretty happens later.
Just to check, I went into dselect, and lo and behold, KDE, pre-built, as binaries, is available.
Try a "fink info bundle-kde" and look at who the maintainer is. ;)
This port uses Qt/Mac instead of Qt/X11... there's a big difference.
That's from the previous (abandoned) port of KDE to Qt/Mac (by one of the guys from trolltech).
It went through bit-rot and pretty much needed to be redone from scratch. This time around the plan is to keep the patches up-to-date until we can get all of the changes back upstream.
Qt/Mac apps generally look like Mac apps, in the case of KDE stuff, so far the style engine is horked, so it's confused as to what to make things look like, so it's pretty ugly.
It's on the todo list. Qt still doesn't directly call native widgets, but it does use the OS's mechanism for drawing, so, in theory, it looks like native apps if Qt doesn't have any bugs. ;)
Actually, they're self-extracting zip files, last I saw. Unix unzip is capable of skipping past the exe junk and unpacking them.
...to be "spoilers," I'd have to still give a shit. I stopped caring about the new Star Wars movies long ago. :)
Wow, great! Your sample set of 1 clearly disproves this obviously fake affliction!
Most of gnome2 (well, 2.0) is in fink, and gnome 2.4.1 is being wrapped up right now. "Unstable" is not for testing, Fink unstable means "it's been tested by the developer and is ready to be supported by him/her." 2.4 is being worked on in "experimental" which is the playground for packagers to work on things without having users complain that [foo] isn't working.
;)
"Fink" has no stance on desktops, but Gnome's packager has historically been insanely busy and didn't have much time to keep up with things very quickly... That's recently changed, Gnome is going to be maintained by a group of people (since it's so big) and things should get updated in a more timely manner.
No one's out to get you.
I'm with you.
I mean, really, what kind of designer are you when you're trying to make new and better ovals?
I've not used it myself, but tonight at our local LUG in Raleigh there's going to be a presentation on White Dwarf Linux. From what I've heard they have a pretty nice embedded linux setup as well as some neat embedded hardware platforms they target out-of-the-box.
As for dports being in panther, a combination of technical issues (dports not being in a state they were willing to support forever) and political issues, it got pulled. I don't know if there are plans to re-add it in a later release.
:)
As for metapkg, the idea of a huge database of packages kind of died out, but we've started trying to fill out the wiki with useful porting info, hopefully in time it will become more useful.
Also, the mailing lists are there for anyone wanting to post problems, but we're getting about 2 e-mails a month, so it's not really happened much.
Metpkg was never going to be a new joint packaging system, only a resource for all of the packaging groups to avoid some duplication of work by sharing patches for ported software and compilation tips and such.
Exactly.
Generally things that break like that are ones that use undocumented and unsupported APIs, which usually have big fat "use these at your own risk!" comments in the headers.
OK, not everything, I was generalizing, but the point remains. It would be difficult to find apps that don't use libSystem on Mac OS X. I'm not claiming all apps will get a massive speed boost, only saying that your original suggestion that there would be basically no difference is incorrect in my experience.
I've been running the panther seeds for months and it's noticably faster. Believe it or not. =)
All of the base system libraries that all software depends on have gotten a speed boost. Even for apps that aren't explicitly recompiled, they'll be faster; *all* apps spend a not insignificant amount of their time calling code that comes with the OS.
I've been using the prereleases of panther and it *is* noticably faster, I'd say pretty close to the difference in speed between 10.1 and 10.2.
Under the covers they've done a lot of work to make prebinding work better, for one thing, and gcc3.3 provides vast improvements in powerpc code over gcc3.1 (even apple's modified version).
The difference is, TV ads are my way of "paying" for the programming that I watch.
When a telemarketer calls, he's making me pay with my time, without me saying it's OK. I have never ever bought anything from someone calling me on the phone, and never will. If I did, it would likely be from a company I already trust -- it would be reasonable for me to opt in to the n companies I want to deal with on the phone, but it's rather insane to opt out of the (infinity - n) companies I don't.
Telemarketing is the act of peeing on thousands of people to find the one person who actually enjoys a golden shower.
As far as I understand it, that falls under the "your quarrel is with IBM, not with individual Linux users" part of the document.
And I can tell you that if they're polling at 1 a second of *anything*, they don't "know what they're doing". That is complete overkill, there's no way the amount of bandwidth being used for testing is worth the 59-second jump on knowing what went wrong. Humans generally have to react to it, that kind of resolution is just crazy.