Basically, all other things being equal, the games are big collections of statistics, and a clear statistical leader probably has an at least marginally better chance of victory on game day.
No, that's the point, they don't. The Patriots do not have the best stats, especially individually, and yet they are the best team in the NFL to this point, having won more games in a row than any team in the history of the league, except for the 1972 Miami Dolphins, If statistics mattered that much, the Patriots would not have won 14 games in a row (including two wins each against the Colts and Titans).
And how about two years ago, when the Rams were the clear statistical leader over the Pats? Oops!
Or if you just ment any 14 games (combinations as opposed to permutations) well Sony's odds just shot way up.
No, I mean the Patriots won 14 regular season games, and have won 14 games in a row (including playoffs), something no one predicted would happen, because if you look at their statistics, there's no reasonable way you would guess that. The point is that stats do not matter sometimes, except for the on in the W column. The Pats have shown, two of the last three years, that they don't need good stats to win, they only need to win.
I am not saying the Pats will necessarily win; of course, there's only one way to know that, and that's playing the game. I think they will win, although I always think they will win.:-) Still, the Patriots have the clear edge. They have a at least as good (if not better) offense, better special teams, a more balanced offense, and the proof of winning the second most consecutive games in history. And perhaps most importantly, they have a coaching staff that has been the best the league's seen in many years, that prepares the team better than any other.
Only one thing is for sure: we will find out which team is better in 2.5 days. But even if Carolina wins, this simulation, which doesn't take the whole season and intangibles into account, is still useless.
How many of these games would have predicted the Patriots would win 14 games in the regular season? Not a one. And if they can't do that, why trust them now?
This is not a serving issue, it is a client issue. iCal does not allow multiple authors to a calendar. You either subscribe to a calendar, or you publish it: not both. If iCal allowed it, Server would easily handle it, the same way it handles it now, using DAV in Apache. It's pretty easy to set up. You set a directory to work with DAV, you set users with write permissions, you create a URL for that calendar, etc.
There's nothing wrong with your review for what it is, and it's cool that you've spent the time to share your experiences, but I don't think most people dropping $500 or $1000 on server are going to be using it for that purpose.
As noted in the opening paragraph to Part 1: that was the point! To find out if it was worth it. And then in Part 2, I said, no it isn't. I didn't think it was hard to understand.:-)
the comments have been much more informative than the actual review
You must be new to Slashdot. That is kinda the point of Slashdot: the readers are a part of the experience. I didn't have time to do a full review that would satisfy everyone's needs, so I did one that satisfied mine, confident the readers would add their comments. Which they did. You're welcome!
The only thing I need to run with a GUI on my server is iTunes. However, moving files around is sometimes easier with direct access too, as AFP sometimes doesn't deal with files in the same way as direct Finder access, and sometimes the shell is just a pain, depending on what I am doing.
Instead of VNC, I use ARD. It's got more Mac-specific features, and the "server" (what Apple calls "client") is already built-in.
The fax sharing issue is even harder to figure out. Why on earth would you disable this? Does apple have a fax sharing product?
I am pretty sure it is because it uses the same sharing mechanism that is missing for the printer sharing. It's a side effect.
There is no justification for removing functionality present in the client which deals with serving resources over the network when you release your "server" product. "New Mac OS X Server, with ten percent less serving!" It just doesn't make sense.
It is lame -- don't get me wrong, after all, I am the one who bitched about it -- but I can't say there is no justification, as I am not on the team behind it. My guess is they had technical problems merging the systems.
I think your information is out of date. Perhaps I wasn't clear: the client mechanism is not available in Panther Server, it is only available in Jaguar Server. And I could not find any documentation on the matter.
I've never set up DNS before, I had to look up what a PTR record is (though I basically understand what a reverse lookup is), and I basically knew what A, CNAME, and MX records are. I've maintained/etc/hosts files. I've seen zone files before, and sorta understood the gist. So, that gives you more of an indication of what "not much" is. I still don't understand everything in the UI, but it was enough to get me going, and to set up all the records for my own domain, in a short amount of time.
What *I* wanted to know about Server is what it offered in terms of tools, so that is what I researched and did a review on. I didn't think I was alone in wanting to understand this better, and from the comments, I am sure I was right. Yes, a comparison would be nice, but I don't really have the time to do it, and the readers have offered their own opinions on that anyway.:-)
Heh, I did have NTP set up on the Server, but some of my computers are laptops that I take outside the house every once in awhile, and it was just easier to select the Apple NTP server than set up my own and make sure I can hit it from outside the LAN.
It used to be. Apple has since changed it to "Apple Filing Protocol". Shades of Dave Winer and "RSS," except that Apple had a good reason behind its change. I spelled out the acronym in Part 1, should've re-spelled-it-out for Part 2. Maybe I will do that now!
You misunderstand what I meant. What I meant is that anyone on the network can use this feature, whether they are on the machine or not, as long as they have an admin password, whereas with su - username, the person needs to be actually on the machine as a user. It's not a huge difference -- if you have an admin password, chances are you CAN get on the computer as a user -- but it is a difference.
Yes and no - AppleScript is no longer so limited. It's not as nice as a fully scriptable application's interface is, but you can now use AppleScript to send GUI Events to any applications, allowing you to script previously unscriptable applications.
Yes, if you follow all the caveats (have the right thingys turned on in Universal Access), and it is still a PITA to do.
BTW, to help you see this, here are two examples, one from AppleScript, one from Perl:
tell application "iTunes" of machine "eppc://Sweeney.local" to play
use Mac::Glue; my $itunes = new Mac::Glue 'iTunes', (eppc => iTunes => 'Sweeney.local'); $itunes->play;
When you run either (the latter assumes Mac::Glue is installed and a "glue" for iTunes has been created with the command line tool), you are asked your username/password for Sweeney. That user must match the current user session on Sweeney. You can optionally store the username/password in the keychain so it doesn't keep asking you. And then it executes the operation.
Slashdot Mirror
Basically, all other things being equal, the games are big collections of statistics, and a clear statistical leader probably has an at least marginally better chance of victory on game day.
:-) Still, the Patriots have the clear edge. They have a at least as good (if not better) offense, better special teams, a more balanced offense, and the proof of winning the second most consecutive games in history. And perhaps most importantly, they have a coaching staff that has been the best the league's seen in many years, that prepares the team better than any other.
No, that's the point, they don't. The Patriots do not have the best stats, especially individually, and yet they are the best team in the NFL to this point, having won more games in a row than any team in the history of the league, except for the 1972 Miami Dolphins, If statistics mattered that much, the Patriots would not have won 14 games in a row (including two wins each against the Colts and Titans).
And how about two years ago, when the Rams were the clear statistical leader over the Pats? Oops!
Or if you just ment any 14 games (combinations as opposed to permutations) well Sony's odds just shot way up.
No, I mean the Patriots won 14 regular season games, and have won 14 games in a row (including playoffs), something no one predicted would happen, because if you look at their statistics, there's no reasonable way you would guess that. The point is that stats do not matter sometimes, except for the on in the W column. The Pats have shown, two of the last three years, that they don't need good stats to win, they only need to win.
I am not saying the Pats will necessarily win; of course, there's only one way to know that, and that's playing the game. I think they will win, although I always think they will win.
Only one thing is for sure: we will find out which team is better in 2.5 days. But even if Carolina wins, this simulation, which doesn't take the whole season and intangibles into account, is still useless.
How many of these games would have predicted the Patriots would win 14 games in the regular season? Not a one. And if they can't do that, why trust them now?
This is not a serving issue, it is a client issue. iCal does not allow multiple authors to a calendar. You either subscribe to a calendar, or you publish it: not both. If iCal allowed it, Server would easily handle it, the same way it handles it now, using DAV in Apache. It's pretty easy to set up. You set a directory to work with DAV, you set users with write permissions, you create a URL for that calendar, etc.
There's nothing wrong with your review for what it is, and it's cool that you've spent the time to share your experiences, but I don't think most people dropping $500 or $1000 on server are going to be using it for that purpose.
:-)
As noted in the opening paragraph to Part 1: that was the point! To find out if it was worth it. And then in Part 2, I said, no it isn't. I didn't think it was hard to understand.
Yes ... which is what, I said, that I used.
the comments have been much more informative than the actual review
You must be new to Slashdot. That is kinda the point of Slashdot: the readers are a part of the experience. I didn't have time to do a full review that would satisfy everyone's needs, so I did one that satisfied mine, confident the readers would add their comments. Which they did. You're welcome!
Yes, I've used daapd. I even got it working with mDNS. It's not as good as iTunes though, so I went back to iTunes.
Ever heard of Quicktime Streaming Server 5?
:-) Or if it does, I couldn't see how.
Yes. Ever hear of daap? It's the protocol iTunes uses, that I need, that QTSS doesn't do.
I meant network access, not physical access.
The only thing I need to run with a GUI on my server is iTunes. However, moving files around is sometimes easier with direct access too, as AFP sometimes doesn't deal with files in the same way as direct Finder access, and sometimes the shell is just a pain, depending on what I am doing.
Instead of VNC, I use ARD. It's got more Mac-specific features, and the "server" (what Apple calls "client") is already built-in.
The fax sharing issue is even harder to figure out. Why on earth would you disable this? Does apple have a fax sharing product?
I am pretty sure it is because it uses the same sharing mechanism that is missing for the printer sharing. It's a side effect.
There is no justification for removing functionality present in the client which deals with serving resources over the network when you release your "server" product. "New Mac OS X Server, with ten percent less serving!" It just doesn't make sense.
It is lame -- don't get me wrong, after all, I am the one who bitched about it -- but I can't say there is no justification, as I am not on the team behind it. My guess is they had technical problems merging the systems.
I think your information is out of date. Perhaps I wasn't clear: the client mechanism is not available in Panther Server, it is only available in Jaguar Server. And I could not find any documentation on the matter.
I've never set up DNS before, I had to look up what a PTR record is (though I basically understand what a reverse lookup is), and I basically knew what A, CNAME, and MX records are. I've maintained /etc/hosts files. I've seen zone files before, and sorta understood the gist. So, that gives you more of an indication of what "not much" is. I still don't understand everything in the UI, but it was enough to get me going, and to set up all the records for my own domain, in a short amount of time.
To the first, you can run the built-in NTPd. No build-in GUI to get it configured and running though.
Yes there is. Server Admin -> Server Name -> Settings -> Advanced -> Enable NTP.
But yeah, if I have a group of users, esp. in a business setting, this is a feature I'd turn off.
What *I* wanted to know about Server is what it offered in terms of tools, so that is what I researched and did a review on. I didn't think I was alone in wanting to understand this better, and from the comments, I am sure I was right. Yes, a comparison would be nice, but I don't really have the time to do it, and the readers have offered their own opinions on that anyway. :-)
Heh, I did have NTP set up on the Server, but some of my computers are laptops that I take outside the house every once in awhile, and it was just easier to select the Apple NTP server than set up my own and make sure I can hit it from outside the LAN.
It used to be. Apple has since changed it to "Apple Filing Protocol". Shades of Dave Winer and "RSS," except that Apple had a good reason behind its change. I spelled out the acronym in Part 1, should've re-spelled-it-out for Part 2. Maybe I will do that now!
You misunderstand what I meant. What I meant is that anyone on the network can use this feature, whether they are on the machine or not, as long as they have an admin password, whereas with su - username, the person needs to be actually on the machine as a user. It's not a huge difference -- if you have an admin password, chances are you CAN get on the computer as a user -- but it is a difference.
Yes, check out the Xserve, mentioned in Part 1.
Why would this be any more of a security hole than someone being logged in as root and then doing "su - " ?
Because you don't need to be logged in as anyone to do this. Any user who has access to the machine can do it.
Isn't this new feature "masquerading" known traditionally as 'su'?
No. su is on the command line, this is via AFP. It is similar, though, yes.
Oh! I'm sorry. My apologies. That's in Part 2. I shall slink away now!
Yes and no - AppleScript is no longer so limited. It's not as nice as a fully scriptable application's interface is, but you can now use AppleScript to send GUI Events to any applications, allowing you to script previously unscriptable applications.
Yes, if you follow all the caveats (have the right thingys turned on in Universal Access), and it is still a PITA to do.