Date: Tue, 21 Nov 2006 03:05:22 -0800 From: "Ted Mittelstaedt" <tedm@toybox.placo.com> To: "Chuck Swiger" <cswiger@mac.com> Cc: cremes.devlist@mac.com, Lonnie Cumberland <lonnie@outstep.com>, freebsd-questions@freebsd.org Subject: Re: MAC OS X connection to FreeBSD? Message-ID: <002e01c70d5c$f134cc30$3c01a8c0@coolf89ea26645> References: <454E9F7B.5010105@outstep.com> <454EB6D6.3030807@infowest.com><454EBEEC.1060002@u.washington.edu> <454F210C.9000602@outstep.com><004001c70706$0d571ec0$3c01a8c0@coolf89ea26645><62104DC1-8AD2-41E1-B469-51CAC91A8D8B@mac.com><002701c70c85$db821eb0$3c01a8c0@coolf89ea26645> <4561E19B.7050706@mac.com>
next in thread | previous in thread | raw e-mail | index | archive | help
----- Original Message ----- From: "Chuck Swiger" <cswiger@mac.com> To: "Ted Mittelstaedt" <tedm@toybox.placo.com> Cc: "Lonnie Cumberland" <lonnie@outstep.com>; <cremes.devlist@mac.com>; <freebsd-questions@freebsd.org> Sent: Monday, November 20, 2006 9:10 AM Subject: Re: MAC OS X connection to FreeBSD? > Ted Mittelstaedt wrote: > [ ... ] > >>> The biggest problem with MacOS X is that a lot of UNIX software that > >>> runs on FreeBSD and such, is not ported to MacOSX, and it's very > >>> difficult to compile on MacOSX. > >> This is completely wrong. Take a look at macports [1] (formerly > >> darwinports) for a large repository of UNIX software that compiles > >> very cleanly on OSX. It's nearly 7 years since OSX shipped to the > >> public. In that time, most opensource software was updated to compile > >> cleanly on OSX. The primary changes to allow this were to the > >> "configure" scripts so they recognize darwin as a base OS. If other > >> patches were necessary, most software maintainers accepted these > >> patches back into their trunk. > > > > I spent a lot of time trying to build KDE on Panther and end up at a > > dead end. There's many problems there. > > That's not a very specific bug report. It's at least possible that reading > this would help: > > http://developer.apple.com/opensource/tools/runningx11.html > I read that at the time plus a lot more. Yes the X11 system was installed. If you have never tried building KDE on Panther I suggest you try it. This isn't something that a single bug report can take care of. I assumed like your obviously doing that all the reports on the Internet that were several years old that said KDE was alpha on MacOS X were out of date and that by now the problems had been fixed. I discovered they wern't out of date. Like I said, I think that if I tried it on Tiger I would have had fewer problems and gotten a lot further. But even the people that have tried it are all saying it's alpha quality - what parts of it they can get up and going, that is. > > A lot I understand were fixed in Tiger and there's somewhat of alpha > > binaries out there - but they must use magic to build them. > > No magic. A guy called Torrey Lyons (sp?), who used to be a Darwin developer > and ran (runs?) the XonX project, IIRC, did most of the work to port and > integrate X11 with OS X, and got things like pasteboard integration and > rootless X11 support, way back in the Rhapsody and 10.1/10.2 days. > > 10.3 and 10.4 should have shipped with a reasonable X11R6 distribution > available as an optional installation item. > That's what I thought with 10.3 but I was wrong. I haven't tried 10.4 so far. > > In any case, Tiger won't run on anything that doesen't have firewire > > so you can just toss out your older iMacs right there. > > Apple recommends you run Tiger on G4 or later hardware, but I believe that > Tiger will run on G3 hardware if you can get a DVD drive attached. > Rubbish. Apple came out with option CD installation of Tiger you could send away for, I did and I have it. A DVD drive is not necessary. And many G3s came with a firewire port and can boot it. But not all of them did and the ones that didn't you cannot boot Tiger on. > Given that the G3 machines came out in what-- 1998?-- the situation is > probably more to Apple's credit than to its detriment. Frankly, if you need > to throw out 8-year-old machines in order to run the latest version of the OS, > and you've gotten 8 years of decent utilization out of the hardware during > that time, you're way ahead of the typical 3-year depreciation cycle for > business PCs. > Oh ho, so where did the much-vaunted "Mac's retain their resale value longer than PC's" go that I used to hear spewed so often? I have 8 year old PC's still in service. One is even running Windows 3.X I use it for firmware updating old HP JetDirect cards that run pre-A05.05 software and to run cvt100 to act as a dumb VT100 terminal for a stack of devices that have ASCII consoles. Another one runs the DOS version of ICVerify and processes several hundred dollars worth of transactions a day as part of an automated system. In the last 5 years it's never crashed. I have customers still running Windows 2K on desktops bought in 1999 during the great Y2K scare when everyone ashcanned their old hardware and bought new stuff. In the last 8 years Microsoft has revised upward it's "product lifecycle" schedules about 4 times, each time extending them. So much for the so-called "3 year depreciation cycle for business PC's" That's wishful thinking by Microsoft looking for increasing their revenue stream, and by "power users" in corporations that want to spend more time playing with the latest software doodad on company-supplied hardware than doing their jobs. > Ted, you do know that Win98 & ME simply won't run on modern PCIe motherboards, > right? You do know that XP and Vista don't run, or run with extreme > compromises to video support and performance, on Pentium-pro machines without > AGP or with maybe AGPx1 slots with say, a 4MB Riva128 card? > Do you know the vast majority of business hardware in service isn't this advanced? You need to read up on the following: http://www.microsoft.com/windows/lifecycle/default.mspx Win98SE - little under 5 years of support Windows 2K - 5 years of support Win XP - 7 years of support, and unlimited with Windows Product Activation, at least right now. Vista - TBA In case you didn't realize, thats a linear progression. Consider that automobiles have been manufactured for around a century. Today a typical lifespan of a passenger car, as judged by parts availability on the aftermarket for repair parts, is around 25 years. That is, from 25 years of date of manufacture of an automobile, there's still enough of them on the road to make money selling new repair parts for them. Why is this? It is because auto manufacturing reached a plateau years ago, there's been little advancement of any significance other than lower emissions, and even that has reached the area of diminishing returns. The IBM PC was introduced August 12 1981, this is the 25th year of the personal computer industry. We are a quarter of the way there. Eventually PC's will run out of innovations and the new models that come out will be merely styling changes, just like cars, just like a host of other products that I could name. We are already developing a good used market in personal computers. What proportion of people that run FreeBSD and Linux do so on brand new hardware they bought specifically for doing this, verses used hardware that they bought a few years ago? Granted, businesses don't buy cars and run them for 25 years. But someone is running those 25 year old cars. And the computer market is waking up to the fact that the entire market isn't going to be driven by what businesses spend on computers, forever. > > Command line programs do compile, but you have to make > > many changes in some of them. Macports is fine if the program you want is > > in there. If not and your rolling it yourself, then you better know > > what your doing. > > It helps to know what you are doing, regardless of the platform or the > software in use. > > >> OSX has excellent support for most UNIX software. > > > > Uh, huh. Yeah, right. > > OSX has excellent support for the Unix software I want to run. > OK, that's better. :-) Ted > > There's not a lot of UNIX software out there that supports all > > the major flavors of UNIX very well, which is why the FreeBSD ports > > system is so important for FreeBSD. > > Or pkgsrc for NetBSD, or Fink for MacOS X, or RPMs for Linux. > > > And unfortunately more and more > > UNIX software is being written and built on Linux and not tested > > on FreeBSD. Take mysql for example, it's heavily dependent on > > threads and while the FreeBSD threads implementation is superior > > to the Linux implementation, mysql didn't work at all with it until > > recently, and it doesen't work as well as under Linux. > > With due respect for the FreeBSD developers, it's not at all clear to me that > FreeBSD has a better threads implementation than Linux. I've seen lots of > results favoring one or the other, although it seems clear that mysql in > particular has been tuned closely with and for Linux. > > Of course, Solaris has been targetting highly parallel systems (with 64, 128 > or more CPUs or CPU threads of execution) since before Linux was first > released, and Sun's been optimizing Solaris for highly multithreaded Java apps > for quite some time. For that matter, multithreaded Java performance is a > fairly high priority at Apple, too, but Apple doesn't sell hardware with > dozens or hundreds of CPUs. > > There are plenty of other reasonable choices besides MySQL, however-- postgres > or OpenBase come to mind, or even the classic BerkeleyDB for some cases. I > think postgres and FreeBSD make a fine combination. > > -- > -Chuck > _______________________________________________ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscribe@freebsd.org" >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?002e01c70d5c$f134cc30$3c01a8c0>