Slashdot Mirror
Workingmac.com Interview With Jordan Hubbard
LiquidPC writes: "workingmac.com has an interview
with Jordan Hubbard (one of the founders of the FreeBSD project, and
currently works for Apple on development of OS X). Questions range from
'How do open-source operating systems compare to closed-source operating systems?' to
'What does the future hold for FreeBSD?'" It's a quick interview, but a good read. Interesting that to talk about the Mac OS now is to talk about UNIX.
I doubt it's much of a problem. Almost all of Darwin is Open Source. Those parts that are encumbered are only done so, because of restrictive third-party licenses (such as Lucent's code that Apple uses in it's Airport 802.11 driver). Most of this encumbered code is being moved out the kernel proper and into loadable kernel modules (KEXT's), so that none of Darwin itself will be encumbered. And some of this encumbered code can be replaced by Open Source code. (Going back to the Airport driver, see Rob McKeever's Wavelan driver for an Open Source replacement. )
http://homepage.mac.com/robm/WirelessDriver.tgz
As to parts of Darwin (kernel or userland) being optimized for OSX and not Darwin proper, that is unlikely. It's not like NT 4 and up where the GUI is running in the kernel and they can do all kinds of funny tricks.
As far as I know, Jordan Hubbard does not actually write FreeBSD code. He is a doc writer, FreeBSD advocate, and release manager (Linux users might not be familiar with this concept).
cpeterso
Damn, you would figure you guys would read a bit more before you post ( but heck, this IS slashdot ).
Part of his duties, besides being involved in dariwn, are to keep working on FreeBSD.
Why you ask?
Because apple is source syncing many parts of Darwin with FreeBSD and there's soon to be a move to start syncing it with the 4.x branch.
Besides, that was the numero uno premise jkh had before he went to work at apple. According to his emails to the freebsd mailing lists, we wanted to assure everyone that his role in freebsd would not be compromised by his work at apple. If anything, freebsd is gonna get more benefits out ot it.
PowerPC port anyone?
"FreeBSD... because a pc is a terrible thing to waste."
he writes quite a bit of code still..
... hehe
One of the things he's currently working on is the rewrite of sys install into something that is humanly understandable
Plus he's very involved in packaging it all together.. after all, he is the "release engineer" also..
It is really quite simple. You wonder: is Mac OS X the largest unix distro? Go count the figures. Microsoft has a 90 percent marketshare, Apple has 5 percent and all the Unix versions (AIX, HP/UX, Solaris, ***BSD and the many linuxes) combined have another 5 percent marketshare. Since Apple is now shipping all macs with os x installed (wether people actually start using it instead of keeping hanging around in mac os 9 is a different issue), it must be the largest unix distro by now because it has about as much marketshare as all other unices combined. Further more: Apple sold loads of cd's too. Considering the fact that the company has sold more than 10 million Macs that are able to run os x (G3 and G4) in the last couple of years, and quite a number of those users (that is the impression one gets in the newsgroups and all) has been running os x (either a licensed copy or a illegal one), I think we are talking quite a big distro here... This is no strange thing. Their unix has been tweeked and interfaced to be understandable for mere mortals (the 95 percent of the population that doesn't understand and doesnt' want to understand Unix CLI).
and that's a pretty big difference when it comes to how one copes with problem analysis and recovery when you hit something the operating system does not handle well or at all.
However, I have yet to see a single sysadmin tweak the source of a Linux or BSD kernel because they found a bug or performance bottleneck. I mean really, who expects the average sysadmin to go in and fix a kernel if something breaks. No, they submit bug reports and hope someone else fixes it soon. Just like in closed source.
Not to disparage open source software, but to think the average maintainer is going to dive in and fix things when he notices a problem is stretching things a bit. The people who actually can fix things in the code generally are not sysadmins.
No, Thursday's out. How about never - is never good for you?
the Feds should fund the development of an operating system and office suite.
One of your points is that it wouldn't take very much money, by government standards. And that is true.
But just because government is already spending piles of money it shouldn't spend, doesn't make it right to have government spend even more money it shouldn't.
You know, safe cars are important. Instead of letting the car makers come up with safe cars, maybe the federal government should spend a few bombers worth of money on safe cars? And I hate it when little children fall into swimming pools. Should the federal government spend money on developing a really good fence to put around swimming pools? Just because something is good and worthwhile, doesn't mean that it is appropriate to have the federal government spend money on it. There is no end to cool projects they could fund, and if I approve of them doing the one I like, I have no moral grounds to disapprove of them doing the ones I don't like.
Besides, I would hate to see a free software project made into a pork-barrel project. It would not be pretty.
One of the problems: government loves attaching little strings to the money. Federal money for software development! But 60% of it must be spent in Senator Kennedy's district. And the new products have to be approved by a comittee that will check it to make sure none of the messages offend anyone. (And I mean anyone.) And the armed forces will make up a 10,000 page document of requirements that the program must meet, and it won't be shipped until it meets them all, and the requirements will be continually updated... the whole thing would be a nightmare.
It is possible for government to do things quickly and efficiently, but that's not the way to bet!
Now, an idea I would approve of: require all federal offices to gradually phase out proprietary software and phase in free software, over (say) the next 10 years. If a department finds that the free software doesn't meet its needs, it can fund developers to add the missing features. 10 years is very generous; they could do it much sooner than that if motivated.
steveha
lf(1): it's like ls(1) but sorts filenames by extension, tersely
care to back that up with some fact? there are ALOT of Solaris boxes out there.
Apple did $19 million in sales of Mac OS X the first weekend it was out. Assuming everyone in that number paid the full $129 price, you get about 150,000 right out of the gate. This doesn't take into account all the beta testers (100,000 in all) that got a $30 discount. Then you tally the developers, educators, and Apple specialists that get it for substanially less or free outright (this revenue might not even been counted in the $19m since it's through special channels).
Then add in that every machine Apple ships now comes with Mac OS X, and that they shipped 827,000 machines last quarter (which ended June 30). And they've certainly sold some copies of OS X off the shelf in the last 5 months. And none of that counts burned copies of Mac OS X or Darwin users.
I suspect it's a relatively big number, probably over 1 million at this point. I'm guessing that it will be at least 2 million year's end. And note that all of this happens despite the fact Apple has run zero TV or magazine advertisements for Mac OS X at this point. They are certainly holding off on marketing campaign until at least until 10.1, and very likely until Office 10 comes out (sometime this fall). I'm not clear on whether the measurement for largest Unix vendor is the number of units sold in a particular period of time (which I suspect Apple is kicking ass on), or the total installed base.
- Scott
Scott Stevenson
Tree House Ideas
As to your various other claims there is no "BSD Mach Microkernel" though MacOS X is based on a derivative of the Mach microkernel originally developed at CMU (I know - those three letter school acronyms all sound alike..)
Mach's " Principal Investigator " was Rick Rashid, with Avadis "Avie" Tevanian who was " principal designer and engineer of the Mach operating system. BTW Avie Tevanian left CMU to continue the development of Mach at Next and is now Sr. VP of SW Engineering at Apple.
First of all I'm neither clueless nor a Mac evangelist, second off... Just where is your "second of all?I don't read ACs: If a post isn't worth so much as a nom de plume to its author then I wont bother either.
Just a correction: that's A/UX, not AIX. AIX is IBM's commercial UNIX.
The kernel for OS X isn't a UNIX(TM) kernel, it's Mach, developed at Carnegie Mellon. BSD is implemented as a userland server atop the microkernel (much like OSF/1). Theoretically, if anyone deigned to write Win32 as a subsystem atop Mach, you could have Windows applications running on top of the server.
I'm not sure -- does the Classic environment run on top of BSD on top of Mach, or does it run directly on top of Mach?
Three Step Plan:
1. Take over the world.
2. Get a lot of cookies.
3. Eat the cookies.