October 22, 2004
Okay, so there are times I see things on the Mac web, and I wonder, I really, truly wonder...
Do people even make an attempt at logic before they post "news"?
I speak of the article on silicon.com entitled "Apple's Tiger delayed at Amazon"
Since when has Amazon become the controller of Apple's release schedule?
Now, Amazon is a really cool site. If they don't have it, it only exists on eBay, or that's how it seems sometimes. But why would anyone use a date posted on Amazon and think that date was when Apple is shipping Tiger, when they haven't announced it yet on their own site. Steve Jobs has issued no press releases. No, Apple is now going to use Amazon to announce ship dates.
Am I the only one who looked at those dates, and went, "oh yeah, right"? I know I'm not, Shawn King at Your Mac Life did the same. but evidently, Jo Best over at silicon.com decided that like the final word to Virginia about Kris Kringle, if Amazon says it's so, it's so:
Apple is cutting it fine to make good on its boast that the latest version of its Mac OS, Tiger, will ship in the first half of next year.
Amazon.com has now changed the date from which it expects to ship Tiger: initially when taking pre-orders, the e-tailer gave a release date of 31 March 2005. That's now slipped to 30 June.
Apple isn't cutting anything fine. They haven't announced a ship date. Heck, you could say Tiger will ship on St. Swiven's day and be as accurate as Amazon.
I know it's real hard to dial a ten digit number, or email someone at Apple, but come on, let's at least make a vague attempt to double-check a dubious claim before stating that Tiger is "delayed".| Comments ()
October 21, 2004
Time for some heresy
Not that me being heretical surprises anyone who knows me, but since he's getting air time on this again:
Jef Raskin did not create the Mac, Steve Jobs did.
Well, more technically, Burrell Smith built the first Mac, but the Mac as we know it was Steve Job's baby.
Jef Raskin created a project called the Macintosh. This is true, it is undeniable. It was a low cost appliance, designed to be simple to use, and affordable. Sounds like the Mac doesn't it?
Well, not when you start looking at it.
The project Raskin was working on was not the Mac at all. It was originally based on a Motorola 6809, a 16 - bit processor. Raskin hated the mouse, and still does. It was called the Macintosh, and he worked for Apple. That's about where the similarities end.
But that's kind of immaterial really. Jef Raskin did not drive the decision to use an improved version of the Lisa's UI. That was Jobs, and a couple others. He didn't push the M68000, that was Burrell Smith and Bill Atkinson. He didn't create the basic icons that still define a huge part of the UI even today, that was Susan Kare, and others.
When you look at the Mac, even the 128K Mac, it's not Jef's machine. It's Steve's. It's Bill's, Bruce's, Andy's, Susan's, Bud's, Chris's, etc. (It's not Tog's either). Jef may have started the concept of the Macintosh, the idea of a simple, easy to use, inexpensive computer. But that's like saying that Da Vinci gets credit for the F-22.
He had an idea. A good one. A damned fine idea. But not for what he wants credit for. He also managed to put together a great team, one that can legitimately be called "Legendary". Hell, just for hiring those folks, he deserves a good spot in the Book of Computer History. But could he have created the Mac? No way. He still opposes most of the concepts in the Mac. Some of his points are valid, but if you read enough of his work, and listen to him enough, you'd swear that he's the only person to realize that for many things, a keyboard is faster, and we need consistency in the UI. Well, he's not. Keyboard control of the UI is increasing with each new version of Mac OS X and Windows, and Linux, not decreasing. Other folks have good ideas too, but instead of insisting we chuck all the other good parts, and Be Like Jef, these features are being integrated into the UIs on every platform. So in a sense, he's being proven right, just not with the fanfare he seems to want.
Now, being an egotistical person, if I were to see that the things I had started were kicking ass and taking names, I'd be ecstatic. I mean, talk about validation, his core ideas are at the heart of personal and business computing. Wow. But every time you see him, he has to remind you that he, Jef Raskin, created the Apple Macintosh. Go to his home page. In the first sentence, he created the Apple Macintosh. But it's not really true is it?
No, and in a sense that's sad. Jef started the concept, but in the end, what came to be was not that product. He didn't do the work, the redesigns. He didn't decide that the Swedish Campground symbol for "Something interesting here" should be the meta key. He was no where to be found for the concept of the resource manager. When you look at all the work that was done in the Mac, at the actual work, it's not Jef's spirit. It's Steve's. The hundreds and thousands of ideas that created something truly new, (which his Macintosh would not have been), were not his. They weren't even inspired by him. Steve Jobs was at the heart of it.
He did a really cool thing, but he didn't do what he really, desperately wants people to think he did. He got kicked out of his own project so early on, (and while that may have been a despicable act on the part of Jobs, it was over twenty years ago, get over it) that it became something completely different. Not even the form factor was what he wanted. But, like Richard Stallman whining about "It's GNU/Linux!!!", (which would be like Snap-On wanting Boeing to rename all their planes to have the Snap-On name), knowing that he started the ball rolling, even if the ball became a tesseract that flies, isn't enough. it will never be enough until everyone is chanting his mantra, that he, Jef Raskin, invented the Apple Macintosh.
I feel bad for him, he got screwed. But it was over twenty years ago. Do something new. Really.| Comments ()
October 12, 2004
Conversation with EMC and Dantz about EMC's new Acquisition
Some backstory on my relations with EMC in the SMB, (Small to Medium Business) market first:
"Not a pleasant experience" would be the way to describe it. Even as recently as three weeks or so ago, trying to talk to EMC when your budget was well under 100K elicited a reaction that came across as "Oh god, more poor people. Here, something you can afford, can we go now?" (No, they didn't say that. But once our budget was laid on the table, the tone of the meeting took a decidedly worse tone.) There was a complete disconnect between the capabilities of the SMB market and what EMC, as a company had to offer. Even the Dell/EMC partnership didn't seem to help much on the EMC side. Dell obviously understands this market very well, but it seemed that EMC was still living in a world of million - dollar budgets and vendor lock in. As well, EMC has historically viewed the Mac market as something other people dealt with, when they acknowledged Apple at all.
So when I saw the press release on this, and read the FAQ, I was not feeling terribly confident. In addition to writing hither and yon, I am also an IT manager. I have been for well over a decade, and have been using Retrospect for most of that time. There's a professional concern about this acquisition both journalistically, and in my "real" job. Since I figured that real information, (as much as possible) is better than guessing, I gave EMC and Dantz a call.
My conversation with them shows that they get that this deal is being viewed warily by Dantz's Mac user base. The EMC folks all said "We just don't have the DNA to move in the SMB market." Allowing for the overloading of DNA as a term, that's a pretty big admission. "We don't know what we're doing here." For a company like EMC, that has made its money from knowing precisely what it was doing, that's a large bite of crow to get down.
EMC's position on this is that the Mac market was one reason they bought Dantz. They realize that in the SMB world, there is a strong Apple presence, and that they have even less experience with the Mac market than they do with the SMB market. So they view Dantz as their best path to success in the SMB market. EMC is seeing Dantz's history in the Mac market, the SMB market, their heterogeneous product line, and the infrastructure that Dantz has in place for doing business in the SMB market as their best chance to make up for past missteps in that arena. As Larry Zulch said, "If EMC was not interested in the Mac market, they could have easily bought another company that has no Mac presence".
Dantz's point of view on this is that they're getting access to more resources for Retrospect, which will allow them to concentrate those resources on their products, in particular, the Mac server product, which Larry Zulch admitted was as patched as it could be, and that they were working on, (no dates given, sorry) the replacement for the Mac server which will be, as he puts it, "a product done the right way". He admits that it's hurting them short term, but that it would be, in the long term, worse to continue patching a code base that has been wrung out as far as it can go. As he put it, Retrospect 6 on the Mac is their first "real" Mac OS X product. I can't say that I disagree with him.
Both Larry Zulch and the EMC representatives admit that they have a lot of work to do. Dantz has made missteps that are well-known in the Mac market, and EMC's bungled handling of the SMB market is legendary among IT managers in that space. They all said that they are committed to making this work, not just for Windows users, but for Mac users too. Larry Zulch also said that there would be announcements at Macworld in January. He didn't say what they were, so I don't have any details for you on that. (The cynic in me says that every vendor at Macworld has an announcement, the realist says "wait and see".)
I realize there's not a lot of hard information. Well, to be fair, the press release is hours old as of this writing, so there's not a lot of hard information to be had. "Keep an open mind" is what Dantz and EMC are asking for, and that makes sense. It's going to take a month just to sort out email routing and HR issues, so expecting much of anything prior to Macworld is probably unrealistic. But they were very explicit as to continuing Mac support, and even increasing support for Dantz's Mac products. As to what that translates to, we'll have to wait and see.
Now, as to what I would like to see at Macworld, as an IT manager running Mac networks:
- A more definite announcement for the Mac Server product. The standard line of "It'll be better, really" and excuses about old code bases have lost all but a thin, frayed thread of tolerance. We've been waiting on this "Great Update" since Mac OS X came out. Version 6 is what Version 5 should have been, so in a very real sense, we're way behind on this.
- Real Mac OS X support in EMC's product line. Notes about NFS support aren't going to cut it. For a lower end system, and considering that all the parts of the OS EMC needs to deal with is open source, there has to be real support for Mac OS X for things like the Clarion AX100 level stuff. No one realistically expects EMC to support OS X on the high end, but if they're serious about the SMB market, then that market definition has to include Mac OS X in more than an incidental manner. As well, EMC showing real interest in the Mac market will go a long way towards allaying fears about Retrospect Mac.
- Some realism when talking about SANs and other products in the SMB market. I understand that for their high-end products, EMC wants tight control over the configuration, but in the SMB market, a company can't afford to trash all their existing hardware, not even for EMC. Even a modest, (to EMC) cost increase can kill a project or a deal. That means that if a customer has Linux on multiple platforms, a *BSD box or two, some Macs as servers and clients, Windows, etc, that all of those need to be treated with legitmacy. Telling a 100 - person company "You'll have to replace your non-Windows and non-x86 Linux servers with Windows and Red Hat on Intel, and you'll have to buy all new fiber - attached RAID" will kill any chance of that deal happening.
- Finally, just listen to what a customer wants and needs and sell them something for that. Not even EMC knows my business better than I do, and I have these odd ideas that I, and not EMC, am better suited to integrating new storage and back up products and choosing directions for my business. I know that FC drives have better specs than SATA, but if I only have a SATA budget, then play nice with me anyway.
So while I'm not skipping through the daisies after this conversation, I'm also not ready to dump Retrospect either. I guess my attitude at this point, is open - minded, but cautious, or as the late President Reagan said:
"Trust, but verify".| Comments ()
October 7, 2004
Installing and Setting up Nagios 1.2 on Mac OS X Server 10.3.5
So these are my notes on setting up Nagios 1.2 on Mac OS X Server 10.3.5. If nothing else, it'll add to the Google Database. Nagios, for those unfamiliar with it, is an open source network monitoring application, ala Intermapper, or Lansurveyor. (Although if you don't know what Nagios is to begin with, you're going to want to check out the Nagios site before you do all the work I talk about here.) Note, this is not with MySQL support compiled in, since that's going away in favor of an n-tier structure. There are a few requirements for Nagios, namely zlib, libpng, libjpeg, and GD lib. if you have current versions of them already, you're golden. otherwise, follow the appropriate steps (If you're installing this on a stock Server setup, you don't have them). You also need Apple's Developer Tools, otherwise you have nothing to compile with. All of this is from my experience installing this on a G4 Tower running Server 10.3.5 and now into the breach...
Following the rough order I used.. (Credit where credit is due section. This is based on an earlier article from the most excellent AFP548.com site. If you're doing this kind of stuff a lot, look there first, you'll be glad you did.)
download the latest zlib from http://sourceforge.net/projects/libpng/
unpack the tar.gz file with gnutar -xzvf path/to/tarball
cd to the zlib directory
sudo make install
sudo ranlib /usr/lib/libz.a
download the most current version of LibPNG, at LEAST 1.2.7 to avoid nasty security holes and bugs from: http://libpng.sourceforge.net/, get the tar.gz version
unpack the tarball with gnutar: gnutar -xzvf path/to/tarball
cd into the libpng-1.2.7 directory
cd into the scripts subdirectory
cp ./makefile.darwin ../makefile
sudo make install
sudo ranlib /usr/local/lib/libpng.a
get the latest libjpeg source from ftp://ftp.uu.net/graphics/jpeg/, currently jpegsrc.v6b.tar.gz
unpack the tar.gz file with gnutar -xzvf path/to/tarball
cd to the jpeg-6b directory
sudo makedir /usr/local/man/man1 (if you don't already have one)
sudo ranlib /usr/local/lib/libjpeg
GD lib setup
download the latest from http://www.boutell.com/gd/ , currently 18.104.22.168
unpack the tar.gz file with gnutar -xzvf path/to/tarball
sudo make install (Note that 22.214.171.124 may not work right in Panther, so if you get a lot of errors on the make, then download 1.8.4 if 2.X makes and installs correctly for you, ignore the rest of this section)
unpack the 1.8.4 tar.gz file with gnutar -xzvf
cd to the 1.8.4 directory
no ./configure, just make
sudo make install : TAKE NOTE OF THE INSTALL LOCATIONS, YOU'LL NEED THEM FOR NAGIOS!
PLEASE, read through all the steps FIRST so you understand what's going to happen and there's no surprises. Surprises in this app are a right proper pain.
if it doesn't already exist, sudo mkdir /usr/local/nagios
In System Preferences, create a new user called nagios. This will, under Panther, also create the nagios group.
In NetInfo manager:
make sure that the users nagios and www are in the nagios group
delete the home directory entry for user nagios
change the passwd for user nagios to *
make sure that user nagios has no shell (changing it to /usr/bin/false works)
Because of the tendency for Nagios to not find the GD libraries when you do a 'normal' make, we'll have to specify their locations for the configure step. Those options are:
-with-gd-lib=/usr/local/lib for the gdlibs
--with-gd-inc=/usr/local/include for the gd includes
Standard configuration options are:
--prefix=/usr/local/nagios the root nagios directory (you did create this already, right?)
--with-cgirul=/cgi-bin/ the path for the CGIs
NOTE: There is a tendency to try and specify the physical CGI path that exists in the Mac OS X finder. DON'T DO THIS!!! By default, Apache on Mac OS X Server has this aliased correctly so that cgi-bin can and SHOULD be used. If you try to use the /Library/WebServer/CGI-Executables path, you're GOING to cause yourself a LOT of problems! You have been warned!
-with-htmlurl=/ the root directory in nagios for the web display. It's actually going to be /usr/local/nagios/share
so our ./configure command is going to be:
./configure --with-gd-lib=/usr/local/lib --with-gd-inc=/usr/local/include --prefix=/usr/local/nagios --with-cgirul=/cgi-bin/ -with-htmlurl=/ --with-nagios-user=nagios --with-nagios-group=nagios
Go ahead and run that. If you get any errors, make sure you pathed everything correctly, run make clean, then rerun the ./configure
sudo make install
sudo make install-commandmode
sudo make install-config
Congratulations! If you didn't get any errors, then you now have Nagios installed. Now, you're still not ready to go yet. you have to install the plugins, and configure some things so it will work right.
To build the plugins, you can download them from the Nagios Plugin site, and just build them yourself. It's not too hard to do, I'll leave that as a challenge for you if you feel like doing it :-)
Nagios config files
Nagios comes with a complete set of example config file. Now, you won't be able to use them like that, (you just have to rename them to make them “real”...it's pretty obvious how, they just have to end in .cfg), and you'll have to plan your setup a bit, but the Nagios documentation is really excellent, and will, if you read it, give you everything you need to configure Nagios correctly. (The configs are so variable based on your needs that there's no point in me doing samples, you'd have to do almost as much editing on them as you'll have to on the sample configs anyway.)
Nagios HTML pages
Out of the gate, you're going to have to do some retouching of /usr/local/nagios/share/side.html. Basically, Nagios assumes a Linux-like environment, and doesn't grok how Apple changed the Apache directories. So, it sets the path to /nagios/cgi-bin/ and that won't work, so you need to make sure that all the cgi-bin references are /cgi-bin/nameofcgi.cgi, otherwise Nagios can't find the CGI. If you put your CGIs somewhere else, well, you have to change side.html to reflect your setup.
Nagios CGI file locations
Since by default, Mac OS X Server's Apache uses /Library/WebServer/CGI-Executables/ for /cgi-bin, you have to copy the CGI's from /usr/local/nagios/sbin to that directory. Once you've done that, make sure they're owned by nagios, group is www, and the perms are set to 755. Setting up CGI permissions. You don't want just everyone to have unlimited access to the details of Nagios, so you want to lock down the CGI access.
To do that, we're going to make a couple of changes to httpd.conf, and set up a .htaccess file for locking down access. httpd.conf changes: Note: This is the simplest way to set this up. There are other changes you can make to httpd.conf to do the same thing, and they will work just as well. this one works in my situation.
Open /etc/httpd/httpd.conf in your text editor of choice (I use BBEdit) and look for the following directive:
Allow from all
and change it to read:
# Options None
Allow from all
This will allow you to run the CGI's after you have authenticated in your browser. To do that, we need to set up a .htaccess file in the cgi-bin directory, and then set up the passwords. Create file named .htaccess in /Library/WebServer/CGI-Executables/ and set it up thusly:
AuthName “Nagios Access”
Once you've done that save it and close it. Now, before we create the password file for this, you want to know all the users that will be authenticating. At the very least some sort of catchall Nagios admin user. For our example, we used nagiosadmin, but you can add others. Run the following:
sudo htpasswd -c /usr/local/nagios/etc/htpasswd.users nagiosadmin
This will create the user file that our .htaccess file is looking for, and set the first user to nagiosadmin. you'll be asked for the password for this user, set it to what you feel is correct. To add other users, you run sudo htpasswd /usr/local/nagios/etc/htpasswd.users
If you read that file, you'll also note this is where you set which nagios user can do what. The docs that are a part of Nagios explain this thoroughly, so I don't have to.
Server Admin Web Server Setup
Finally, you have to set up OS X Server's Web Server so that it knows about Nagios. You do this in Server Admin. In my case, since this machine is the only thing running as a Web server on this box, I just pointed the primary site at /usr/local/nagios/share/ and that was it. You MUST enable CGI Execution here, (otherwise this won't really work) and you'll really want to kill the performance cache for this site too. (It's a nice idea, but a pain in the tushie the way it works.) Once thats done, you should be set. Just point your browser at the machine running Nagios and enjoy! If you are going to be running Nagios on your network, I would highly recommend joining the nagios-users mailing list. It has a good s/n ratio, and the responses on it are quite helpful in my experience.| Comments () | TrackBacks (1)