You are using a subtly different definition of "popular" than Ian Clarke is.
Your definition of popular seems to be "stuff that people like and agree with".
Ian Clarke's is "stuff that people download".
Take an example: "Mein Kampf". That's an "unpopular" work, in the sense that few people agree with it. But it might be "popular" in that many people will download it.
In fact, many of the people who download it probably disagree with it. Me, for instance. I know I disagree with Hitler's view on Jews, but I'm still interested in finding out exactly what he wrote, so I can decide why, exactly, I think he was wrong.
With Microsoft, it's not that they think they are going to make any money off the old software... they just want to force people to buy the new software.
Microsoft depends on people always, always having to upgrade.
Imagine if Windows 95 and Office 95 were good enough that companies could still use then today. They would just buy nice new computers from Dell with no operating system, and install their old, perfectly legal copies of Windows 95 on them.
This would be fatal for Microsoft. This is why they have always forced old software to become unusable. They refuse to sell it, refuse to support it, and make their new software incompatible with it.
Microsoft is getting caught in a tight spot. If they don't make their software better, Free Software will take over. But if they make it good enough, people will buy it once and never upgrade. Either way, Microsoft is screwed.
This is why Microsoft is making moves to a subscription model of software, where you cannot buy the software. You will only be able to lease it on a yearly basis.
If Microsoft succeeds in doing this, they have it made. They won't even need to put out new software anymore - just minimal support, and let the money roll in.
Actually, I think the author is correct. Here's an example. Remember a thread is an object in Java.
Imagine that your program "FooServer" creates a thread "SubThreadStarter".
SubThreadStarter runs for a while and creates a whole bunch of "ListenerThreads", which attach to a port and wait for input, kind of like a web server.
Then, SubThreadStarter stops, and FooServer destroys it.
But now the garbage collector will notice that all those ListenerServers's are unreachable from FooServer, and they all get destroyed, although they may be running and doing useful stuff.
This might not be what you wanted to happen!
This may sound a little contrived, but stuff like that happens. Fortunately it is easy to work around, just by keeping a master list of threads in the FooServer program (which you would probably want to do anyway, actually).
Disclaimer: I am not a Java expert. Gentle corrections are welcome.
I bought a handspring, and love it. Two months after getting it, I tried to put it into my shirt pocket while walking up some stairs. I missed the pocket, it fell, hit the stairs, bounced over the edge...
my heart stopped...
It fell three floors and landed flat on concrete. If it wasn't in it's padded case it would have flew apart into a thousand pieces. As it was, the screen and case were totally fractured.
Since I bought it on VISA I had a replacement warranty through that. But when I called up Handspring to get the info I needed for VISA, and explained that it was destroyed through my own stupid mistake, they said "Oh, don't bother, we will just send you a free replacement."
Wow.
It arrived three days later by FedEx, and had nice clear instructions and a FedEx shipping label to send the dead one back. All I had to do was HotSync and I was up and running.
Put the power supply off in the corner somewhere, and just run a wire bundle up to the computer.
Cut the power cables, one at a time so you don't confuse them and splice in a nice 8-foot set of wires of at least the same gauge (thickness). As an added advantage, you get the noise of the cooling fan further away.
Whatever you do, never go with an gutless power supply. That will doom you to a flakey computer, especially with today's power-hungry chips. Use at least a 250 watt supply.
My machine is 256 MB, two 7200 rpm hard drives, and CD ROM + CDR. With a 250 watt power supply I had occasional spontaneous reboots, especially if both the CDR and the CD were spinning up at the same time. Upgrading to a 300 watt power supply fixed the problem.
You're welcome! I damn near wore my fingers off typing it, trying to keep up with the audio, and haven't been rewarded with much moderation. (yet?)
I'm surprised that this story isn't getting more attention, not just from moderators, but from people posting.
There's some really interesting stuff here - IBM obviously supports Linux in a huge way now - they've ported all their key software, they support the community, they are contributing to the kernel... and the guy who made the presentation obviously gets it. He had great responses for the standard questions about "Linux Fragmentation", how to make money with Linux, and some minor FUD about "the roadmap" for the next kernel. IBM is even thinking that in a few years, they may not need AIX anymore!
IBM may end up being the first, or even only profitable Linux based company! This is significant!
But everyone is off posting and moderating in some soon-to-be-forgotten story about some guy trying to get arrested at the Republican convention... sheesh.
Actually, there is a parallel between biological viruses and the internet, and I think that's what the original article was getting at.
The parallel is the spread of ideas, not files. Memes, if you like that word. The particular idea in this case is "I can get music for free... I should get music for free!".
That is the idea that Napster "infects" people with. I'm not saying if it's good or bad, but it's sure an infectious idea!
There's some really good stuff in here, especially in the Q and A section.
Perhaps most important part of Linux strategy:
- IBM believes Linux will do for applications what the Internet did for networking.
- providing a common set of interfaces that all application developoers can adapt that will facilitate the development of applications which can be deployed across a variety of platforms
- just like internet standards brought the internet together, Linux and linux standards are going to help develop lots of applications.
- Some of IBM's efforts around Linux apps, working with the community, using tools from IBM and the communiuty to allow a common Linux development platform to emerge.
- target is a world class application development environment.
- Since Linux interfaces can be provided on other platforms like AIX, it is a good development target.
- IBM has put together a Linux developers kit, thousands have been distributed.
- No charge for development use, includes full versions of WebSphere, DB2, Lotus Domino, Visual Age for Java, IBM Java tools, documentation, etc.
- Also a version for Japan.
More efforts around the world.
- A few weeks ago in Europe, announced a major ($200M) effort to assist application developers to port their apps to Linux.
- Development centers around europe with technical help, support, for application dev on IBM servers.
- Software and server specialists
- Once application is ported to Linux, it can be deployed on many platforms.
Community efforts
- established open source community efforts, the Linux Technology Center.
- Open source is better for everyone - customer, developer, etc.
- They have open sourced a number of things, and are planning to open source more.
- As many companies (besides IBM) open source their technology, Linux will become more popular and better.
===============
Q and A session
Q: OS priorities: AIX vs Linux, no mention of Monterey? and, if clustering becomes more sophisticated, will that eliminate the need for AIX on the high end? If Linux accelerates migration to industry standard platforms, what are the implications for IBM's business?
A: Monterey was the effort to convert AIX to 64bit for Power architechture and also IA64. That has been done.
- Responding to market demand for Linux, IBM has integrated Linux into strategy.
- regarding clustering: Clustering on Linux is very good today, good for internet applications, good for supercomputers. This gives "horizontal scalability", using many single, dual, or quad processor machines.
- Where Linux is NOT so good today, and where AIX has the advantage, is to run transaction processing applications - where data must be shared at very high speed, mission critical transactions, and high-level SMP where data must be shared. Linux doesn't do it yet, but the 2.4 kernel will be much better.
- Over time it is possible that Linux will approach the capabilities of enterprise. If so, the enterprise stuff from AIX might be migrated to Linux.
- There is a lot more to an enterprise solution than just the kernel.
- This may take several years, and that's why they are continuing their work with AIX.
- This strategy will adjust in the future as Linux progresses.
- Yes, Linux is an excellent example of the industry moving to standards.
- This started with the Internet, and now others like XML, SOAP, are other examples. Linux is part of this.
- This is important becayse it is only by embracing standards that we will be able to build the applications required by their customers. (ebusiness, etc.)
- Market is embracing standards,
- IBM is happy with this because they make their money selling hardware, middleware, and services.
- As more of that hardware, middleware, services is supporting standards, you can build bigger and bigger solutions, and it's good for IBM's business.
Question 2:
Q: re incremental opportunity for Linux. To what extent is Linux demand incremental, vs. displacing other operating systems?
A: It is early in the process, so hard to tell. However, he believes that Linux will cause a big expansion in the market. Includes an expansion in the number of products, including consumer products that will integrate IT technology. The TIVO is a good example. - computer tech is going further into consumer electronics, telecommunications, and others. All those consumer appliances can become network applicances. Believe the IT industry is going to grow - the more devices you connect, the more services you need, bigger services.
Clearly the growth of Linux will also come by displacing existing legacy systems, but this will not be as significant as the incremental growth.
Question 3:
Q: (1) re Linux kernel: existing kernel is "long in the tooth", how to we get the predicted commitment to Linux without more visibility on when the next kernel will arrive and what it's capabilities will be?
(2) re revenue growth rate: in Linux server market - the real opportunity in the market may be embedded systems, with money to be made on management.
A: (1) regarding the kernel being "long in the tooth". Operating systems are like fine wine - they need time to mature. The more mature they are, the better they get in the sense of quality, scalability. THe best operating systems in the world, they are not that young. Compared to them, Linux is actually quite young! (comments on early development of kernel by Linus, etc.) In 1998, 1999, it hit the big time. Kind of like TCP/IP. It had been around for a long time, used by the research community, until it hit the big time in 94/95.
The next version of Linux, 2.4, is scheduled to come out this fall, deploying next year, higher levels on SMP, concurrency, etc. The community is working on this, IBM's tech center is trying to help the community. Other vendors like Silicon Graphics, are also contributing, this makes Linux better and better for applicances, desktops, as well as larger and larger servers.
2. Regardiung the explosion of linux appliances. IBM views this as a wonderful business opportunity. They need technology and they need services. Technology like microprocessors, very small storage devices, analog stuff like integrated circuits for communications. This is driving IBM's tech business. Then they need integration services, this feeds the software business. A lot of work on WebSphere is dedicated to support of pervasive devices. The more of them you have out there, the more you need stuff like WebSpere on systems like the 390.
Services is the biggest part of the e-business opportunity. 60% of the revenue. 5% is hardware, the rest is software, more or less.
Big opportunity. If you have 10 billions of chatty little devices each doing lots of transactions, that's the kind of load that will keep the system 390 mainframes busy. Theres a lot of demand for server capacity.
Middlewhere is also a big part of the equation. A big part of WebSphere is targeted to this.
Question 4.
Q: IBM is working with many distributors. Do you see a possibility of Linux fragmenting, and if so, do you see a possibility of IBM providing their own version of Linux?
A: We are working primarily with the 4 major worldwide distributers: Red Hat, TurboLinux, SUSE, and Caldera.
And we do different things with each. Suse does the system 390, for example. Suse is big in europe, while Turbo is big in Asia. Red Flag Linux is the biggest in China. Mandrake is big in France, and on the desktop. Hard Hat Linux is big on embedded devices. So, over 150 distributors. They don't need IBM to be a distributor. They are better off working with distributors to integrate the whole system for their customers.
If the problem is in the Linux kernel, IBM can get together with the appropriate distributors.
There is an organization called the Free Standard Group, and just about all the major distributors are part of it and are committed to avoiding fragmentation. Look at other efforts like XML - the industry is doing well for keeping things together.
Here's my notes on the first half hour of the one hour presentation. Sorry these aren't really well formatted. I may have misheard a few things, but I think this is pretty accurate summary:
===========================
- They (IBM) are aggressively embracing Linux, and have been for about 12 months now.
- Linux is very good technology:
very modular
very flexible
- Linux is becoming very popular all over the world, including Asia, Europe
- Linux is great for business solutions - delivering solutions on the appropriate platform.
- Linux is second most popular server operating system by volume, second only to Microsoft, and has achived this very rapidly.
- Linux is expected to become most popular server operating system
- IDC expects Linux to continue to be the fastest growing server operating system.
- IBM expects Linux to be a huge force in the market, and this will attract developers and application development.
- Some discussion of the mllions of servers expected.
- Note that it's hard to predict the growth, just like trying to predict the growth of the internet in 1994.
- IBM expects both volumes and revenues to go up.
Where is Linux most popular?
- Internet types of apps: web servers, ISP's, application service providers
- Very popular with internet companies, and also with large enterprises for internal intranets
Linux activities within IBM:
In the 12 months or so since IBM started seriously working on Linux, they have:
- Linux-enabled just about all IBM platforms,
- ported much of their software to Linux.
- using Linux widely within IBM
Linux Servers
- Majority of Linux use is on servers, specifically PC servers.
- Netfinity servers running Linux are being deployed many places - some examples:
Example 1:
- 250 branches of some insurance broker in the UK, didn't catch the name
Example 2: Millenium Partners
- a NY based security firm
- have consolidated trading floor apps on Linux
- intelligent support for trading on a server. Desks are running IBM "screens" User interface?
Example 3: weather.com
- one of the most heavily trafficed web sites on the net, especially during weather disturbances.
- Needed a web site that could handle very high peak loads, but at a low cost.
- Using an IBM Netfinity cluster.
- They ported their Solaris based system.
- Much more capacity at much lower cost.
Linux on Supercomputers
Example: 256-way cluster in New Mexico
- Small room on 24 racks.
- Peak 275 Gigaflops (375?)
- IBM expects many more research groups to be installing clusters like this
Linux on System 390
- Linux ports to many architechtures
- runs very well on 390 mainframes, called "Linux390".
- Is becoming very popular with system 390 customers
- already 4000 downloads of the code
- Last week, IBM had an InstallFest (virtual community meeting) where 60 customers around the world got the system up and running.
- Customers can designate specific "engines" on a 390 G5 or G6 for linux applications
- This affects pricing for the other applications running on the 390, since pricing is often done by "engine", and the Linux designated engines don't count toward the pricing of the other apps. So no software price penalty for expanding a mainframe, adding engines if they are used for Linux.
- Announced prices for selected software on Linux 390, much closer to prices to other distributed platforms, to encourage customers to use 390s instead of distributed systems.
Some customers planning (are?) using Linux390
- Many customers run mission critical databases on 390s, and web servers on different platforms that talk to the 390.
- This gives network performance problems, and raises security concerns.
- Instead, they put Linux390 on the mainframe and port the web server based apps to the 390.
- Thus the communication is entirely within the mainframe, solving performance and some of the security issues
- Multi-Image facility: special version of VM for Linux gives the ability to run hundreds or thousands of Linux images on a 390. Very simple and effective way to manage all the Linux installations.
- IBM expects Linux390 to be a big part of the future of the 390 mainframes.
IBM's Unix Strategy
- Integrated Linux into Unix strategy, which is basically to have two flavors of Unix:
- AIX for enterprise-class applications for high-availability, scalability, on the RS6000's and other.
- AIX has been ported to IA64, in beta.
- Linux complements AIX for simpler / high volume / distributed / small server applications.
- Linux at the lower end, AIX at the higher end, gives customers the widest range, flexibility, etc.
- To make the strategy coherent, they are working to facilitate porting between the platforms.
- So, they are adding a Linux compatibility layer to AIX to make it as Linux compatible as possible.
- Customer can begin on Linux, and if they later find they need more scalability or reliability, they can recompile to get it onto AIX, for the largest Unix systems in the world.
Linux on Clients, in particular embedded systems.
- TiVo is a good example. Intelligent VCR. Based on IBM powerpc processor, running Linux.
- IBM is working with other companies working on the embedded market (Hard Hat Linux)
- is being tested on the next generation powerpc processors.
- IBM is part of the embedded linux consortium.
- expect that over next few years, Linux will be taking a stronger and stronger role in this space.
Linux Software
- IBM has taken their critical distributed software offerings and made sure they work on Linux, period.
- WebSphere, including the advanced parts for ecommerce3
- Domino runs on Linux
- DB2 runs on Linux
- background messaging stuff
- Tivoli running on Linux
Also Services
- Big opportunity for services.
- Providing technical support, one-stop shopping for services
- cooperating with Linux distributors
- IBM can manage the overall contract, bringing in a particular Linux distributor for their expertise
- As linux becomes more of part of ecommerce, ebusiness, lots of work in linking Linux to other applications.
------------------
I'll post another summary of the next half hour after I've listened to it...
I think DOOM came with that capability from the beginning, but I'm not sure. There must be some huge DOOM FAQ's out there that will tell you one way or another.
Don't you think this is a little too close to cheating? You mentioned setting the second PC to a 3rd-person view of yourself.
Wouldn't it have been much more useful to have it chasecam on your opponents? I think that would definitely count as cheating - at least for a head-to-head game.
I know that when I would play my roommate, we would yell at each other "Don't look at my screen!" when we were head-to-head in the same room. We often did sneak peeks, though. There's a huge advantage in knowing where your opponent is, what weapon they have, and where they are going.
So... is it cheating?
Regardless, it's a neat trick getting it running.
I might stick that old Voodoo 1 card back in my spare machine and try it myself. Torrey Hoffman (Azog)
Yes, you read it wrong. Microsoft does not actually disagree with the facts of the Gartner analyst who started this whole story, they just disagree with the "tone". (Last paragraph of the article.) In other words, they want to put a spin on it.
The undisputed fact is (4th paragraph of article), if you buy a bunch of PCs with Windows 2000 on them, and you are in the Microsoft Select program, and you use Ghost or something to wipe the hard drives and install a complete new image of the same OS, the apps, the utilities, and everything else, you are screwed.
All big companies do this. But Microsoft says, "No, the Windows 2000 that you bought with the computer is not the same as the Windows 2000 you are putting onto it with Ghost."
So you have to pay for it again. And you can't get free support anymore from the computer manufacturer, you have to pay Microsoft $375 per incident.
The only way to avoid this would be to purchase the computers without anything on the hard drives. Good luck! Linux, BSD, and BeOS users already know how hard that is!
I predict Microsoft will change this soon, otherwise they will take a big hit next time somebody does a comprehensive study of the TCO (Total Cost of Ownership) of Windows 2000 vs. Linux. I would expect that Linux will become competitive in such studies sometime next year.
Imagine a company with 5000 PC-using employees, and a few dozen support staff. (Same as Gartner's scenario in the article.) The money Microsoft wants out of them for software they already bought is about $600,000 dollars.
Now, that much money could hire a couple of Linux experts for a year, plus pay for several days of concentrated classroom instruction for the rest of the tech staff to get up to speed on Linux.
Actually rolling out Linux would still cost quite a bit, but once all the staff is trained and you have a couple of experts on board, you can probably make the jump without too much pain and disruption. I would bet that the transition to Linux would easily pay for itself in a couple of years, with lower software costs, lower hardware costs, better software reliability, and better security.
I predict this will begin to happen more and more next year.
Previous articles from Moody that show his total bias and ignorance, for your reference and reading pleasure:
A story on how biased the judge was"Virtually since the beginning of the trial, though, Jackson has brushed aside every request and argument made by Microsoft, and sided enthusiastically with the U.S. Department of Justice"
How nice those poor Microsofties are"I keep hearing Microsoft-employee-misfortune stories so powerful that they tug not only at the human heart, but at the journalist's, too."
Just below the note that Mr. Fred "Hi, I'm a Microsoft Tool!" Moody wrote a book about Microsoft on the multimedia frontier (*cough*), the ABC site has a search button... which is pre-filled in with...
Search for more on... Microsoft.
So, ABC has classified this as "Microsoft" news, despite the headline being "Linux". Things that make you go "hmmm".
I hope Slashdot gets around to a follow up on this one. I really, really want to see if Moody does a correction article. If he has a trace of honesty, he will do another article which admits:
1. He is wrong, because he added Red Hat's numbers to the numbers for all of Linux. If he had done it correctly, he would have reached the opposite conclusion.
2. His so-called methodology for determining the "Suckiness" of an OS is ridiculous anyway, so even ignoring pt. 1, his article is balony.
The Difference Engine was William Gibson and Bruce Sterling.
I was disappointed by the book when I read it, but I don't remember much else of the plot, characters, or settings. I guess I'll have to dig it out and re-read it - maybe my tastes have changed and I will appreciate it better the second time around.
ObOnTopic:
Whoo! That Photon Mapping is seriously cool! It looks better than ray-tracing, and it can be done in milliseconds! Does anyone know of a downloadable demo of it, or more information than was given in the article?
for (1), I think that it won't so much be Caldera soliciting their customers to move to Linux, it will just try to capture all the customers who have already decided to go to linux and are just beginning to implement it.
For (2), I agree with the earlier poster who stated that SCO Unix can never be opened completely because SCO doesn't even own all of it. Likely they will open parts of it, and port parts of it to Linux.
My expectation is that they will create a so-called "high-end" Linux distribution, with some SCO software (mostly userland) in it, specifically targeted to people migrating from SCO Unix.
Not a distribution for you and me, but very attractive for some people.
I agree - very few games from the 90's will still be played in 2010.
Probably the original Quake will be one of them. It spawned a whole community, hundreds of web sites, and the concept and practice of professional video game players actually making money to frag.
Now that it's been GPL'ed, it can never die, despite the inevitable changes that will come to hardware and operating systems.
I just had a chilly moment of foreboding... In 10 years, some kid is going to see me playing Quake and say: "Hey! that's one of those old games, like... uh... Brother Mario, right? That was for a... um... Nuntindo Main Frame?"
(sigh.) I should be too young to anticipate feeling that old.
This is not an attack on anonymity. Go read the actual IETF draft. You will see that the only thing it helps with is tracing back packets with SPOOFED originating IP's.
This will help prevent things similar to the attack on kuro5hin. Unfortunately, if attackers are using compromised machines, all it will (or can) do is help to quickly find the real IP addresses of the machines that have been compromised. You see, someone doing a denial of service attack right now can cause the servers they are using to output IP packets that look like they are from somewhere else. When those packets arrive at the target, 10 hops later, it is nearly impossible to find the real machines that is causing the attack. That's what this proposal solves.
This has nothing to do with eliminating privacy or anonymity. Every time you connect to a web site now, they can find out the IP address you are coming from. Duh! How else can they send the web page back to you??? If you spoof your originating address, you cannot have a two way conversation.
IP source spoofing is ONLY useful for denial of service attacks, and that is the ONLY thing this proposal addresses.
The so called solutions you are advocating, like restricting access to the net would be far, far worse for invading privacy. Think about it... how are you going to make sure that only "authorized people" use the internet? Well, you will have to identify all of them. With examinations, meeting criteria, getting what is equivalent to an "internet license"... well damn, there goes privacy! Just like anyone who sees your license plate on your car can find out who the car owner is. No privacy there either. Did you think about this?
The IETF proposal is not a perfect solution. You are correct that there probably isn't one. However, it is a good one and 100% better than your suggestion.
It would be trivial for the attackers to change just a few letters at the end of the post, completely changing the MD5 hash.
What would probably work better is some sort of "diff" approach. If a post is "too similar" to a lousy one it would be rejected. It isn't actually that hard to do similarity checking, and the load is not that high. After all, even on a busy day Slashdot only gets about 3000 posts or so.
This would get rid of the Jenna Elfman, Penisbird, Latin Lessons, and other fools. (I read one of the stories at -1 yesterday. Unbelivable!) I wonder if this would count as censorship? Perhaps instead of deleting the post, the system should just instantly moderate it down to -1.
I think that instead of a moderation system that just lets people bump a post +1 or -1, moderators should be able to pick a level for the post.
Then the computer would average all those moderations together. For example: Somebody makes a good post, and it starts out at 1. Moderator A rates it "Insightful" and "5". Moderator B rates it "Interesting" and "3". Average them all together, (1 + 5 + 3)/3 and it would show up as "3".
You might need to simplify a little. Perhaps instead of the numbers, have a scale like "Crap, Useless, Ordinary, Good, Very Good, Excellent". Maybe just use that and drop the Insightful, Funny, Interesting part of moderation.
I think that would be simpler for people, even Arts students. (heh. You said it, not me! *ducks*)
However, I think Slashdot works amazingly well, considering how many trollers and losers are attacking the system. So if you think your site is going to be big, or will have a lot of losers on it, you might want to stick with what is known to work.
Well, if you browse at +2, the signal to noise isn't too bad. A lot of "me too" type posts that don't add too much, a few morons, and a few really good posts.
If Slashdot didn't have the discussion area, nobody would be here. Instead, everyone would just check Wired, CNet, Blues News, and about 4 other sites. Between them, they get at least 90% of the stories that come up here, plus a lot more, and are more accurate.
But the level of trolling sure has increased. I try to read every response to any of my posts, and if I ever lower the threshold to -1 I'm amazed at the amount of junk down there.
By the way, this post is borderline between a "me too!" and interesting...:-) Torrey Hoffman (Azog)
Slashdot Mirror
You are using a subtly different definition of "popular" than Ian Clarke is.
Your definition of popular seems to be "stuff that people like and agree with".
Ian Clarke's is "stuff that people download".
Take an example: "Mein Kampf". That's an "unpopular" work, in the sense that few people agree with it. But it might be "popular" in that many people will download it.
In fact, many of the people who download it probably disagree with it. Me, for instance. I know I disagree with Hitler's view on Jews, but I'm still interested in finding out exactly what he wrote, so I can decide why, exactly, I think he was wrong.
Torrey Hoffman (Azog)
Ahhh.... Ok. I didn't realize the garbage collector started from more places than just the original program for determining reachability.
I suppose I should have checked more carefully - I have a couple of Java references here - but I rarely do any significant Java hacking.
Thanks for the info.
Torrey Hoffman (Azog)
With Microsoft, it's not that they think they are going to make any money off the old software... they just want to force people to buy the new software.
Microsoft depends on people always, always having to upgrade.
Imagine if Windows 95 and Office 95 were good enough that companies could still use then today. They would just buy nice new computers from Dell with no operating system, and install their old, perfectly legal copies of Windows 95 on them.
This would be fatal for Microsoft. This is why they have always forced old software to become unusable. They refuse to sell it, refuse to support it, and make their new software incompatible with it.
Microsoft is getting caught in a tight spot. If they don't make their software better, Free Software will take over. But if they make it good enough, people will buy it once and never upgrade. Either way, Microsoft is screwed.
This is why Microsoft is making moves to a subscription model of software, where you cannot buy the software. You will only be able to lease it on a yearly basis.
If Microsoft succeeds in doing this, they have it made. They won't even need to put out new software anymore - just minimal support, and let the money roll in.
Torrey Hoffman (Azog)
Actually, I think the author is correct. Here's an example. Remember a thread is an object in Java.
Imagine that your program "FooServer" creates a thread "SubThreadStarter".
SubThreadStarter runs for a while and creates a whole bunch of "ListenerThreads", which attach to a port and wait for input, kind of like a web server.
Then, SubThreadStarter stops, and FooServer destroys it.
But now the garbage collector will notice that all those ListenerServers's are unreachable from FooServer, and they all get destroyed, although they may be running and doing useful stuff.
This might not be what you wanted to happen!
This may sound a little contrived, but stuff like that happens. Fortunately it is easy to work around, just by keeping a master list of threads in the FooServer program (which you would probably want to do anyway, actually).
Disclaimer: I am not a Java expert. Gentle corrections are welcome.
Torrey Hoffman (Azog)
You want awesome service? Try Handspring.
I bought a handspring, and love it. Two months after getting it, I tried to put it into my shirt pocket while walking up some stairs. I missed the pocket, it fell, hit the stairs, bounced over the edge...
my heart stopped...
It fell three floors and landed flat on concrete. If it wasn't in it's padded case it would have flew apart into a thousand pieces. As it was, the screen and case were totally fractured.
Since I bought it on VISA I had a replacement warranty through that. But when I called up Handspring to get the info I needed for VISA, and explained that it was destroyed through my own stupid mistake, they said "Oh, don't bother, we will just send you a free replacement."
Wow.
It arrived three days later by FedEx, and had nice clear instructions and a FedEx shipping label to send the dead one back. All I had to do was HotSync and I was up and running.
Excellent! HandSpring Rules!
Torrey Hoffman (Azog)
Put the power supply off in the corner somewhere, and just run a wire bundle up to the computer.
Cut the power cables, one at a time so you don't confuse them and splice in a nice 8-foot set of wires of at least the same gauge (thickness). As an added advantage, you get the noise of the cooling fan further away.
Whatever you do, never go with an gutless power supply. That will doom you to a flakey computer, especially with today's power-hungry chips. Use at least a 250 watt supply.
My machine is 256 MB, two 7200 rpm hard drives, and CD ROM + CDR. With a 250 watt power supply I had occasional spontaneous reboots, especially if both the CDR and the CD were spinning up at the same time. Upgrading to a 300 watt power supply fixed the problem.
Torrey Hoffman (Azog)
You're welcome! I damn near wore my fingers off typing it, trying to keep up with the audio, and haven't been rewarded with much moderation. (yet?)
I'm surprised that this story isn't getting more attention, not just from moderators, but from people posting.
There's some really interesting stuff here - IBM obviously supports Linux in a huge way now - they've ported all their key software, they support the community, they are contributing to the kernel... and the guy who made the presentation obviously gets it. He had great responses for the standard questions about "Linux Fragmentation", how to make money with Linux, and some minor FUD about "the roadmap" for the next kernel. IBM is even thinking that in a few years, they may not need AIX anymore!
IBM may end up being the first, or even only profitable Linux based company! This is significant!
But everyone is off posting and moderating in some soon-to-be-forgotten story about some guy trying to get arrested at the Republican convention... sheesh.
Torrey Hoffman (Azog)
Actually, there is a parallel between biological viruses and the internet, and I think that's what the original article was getting at.
The parallel is the spread of ideas, not files. Memes, if you like that word. The particular idea in this case is "I can get music for free... I should get music for free!".
That is the idea that Napster "infects" people with. I'm not saying if it's good or bad, but it's sure an infectious idea!
Torrey Hoffman (Azog)
There's some really good stuff in here, especially in the Q and A section.
Perhaps most important part of Linux strategy:
- IBM believes Linux will do for applications what the Internet did for networking.
- providing a common set of interfaces that all application developoers can adapt that will facilitate the development of applications which can be deployed across a variety of platforms
- just like internet standards brought the internet together, Linux and linux standards are going to help develop lots of applications.
- Some of IBM's efforts around Linux apps, working with the community, using tools from IBM and the communiuty to allow a common Linux development platform to emerge.
- target is a world class application development environment.
- Since Linux interfaces can be provided on other platforms like AIX, it is a good development target.
- IBM has put together a Linux developers kit, thousands have been distributed.
- No charge for development use, includes full versions of WebSphere, DB2, Lotus Domino, Visual Age for Java, IBM Java tools, documentation, etc.
- Also a version for Japan.
More efforts around the world.
- A few weeks ago in Europe, announced a major ($200M) effort to assist application developers to port their apps to Linux.
- Development centers around europe with technical help, support, for application dev on IBM servers.
- Software and server specialists
- Once application is ported to Linux, it can be deployed on many platforms.
Community efforts
- established open source community efforts, the Linux Technology Center.
- Open source is better for everyone - customer, developer, etc.
- They have open sourced a number of things, and are planning to open source more.
- As many companies (besides IBM) open source their technology, Linux will become more popular and better.
===============
Q and A session
Q: OS priorities: AIX vs Linux, no mention of Monterey? and, if clustering becomes more sophisticated, will that eliminate the need for AIX on the high end? If Linux accelerates migration to industry standard platforms, what are the implications for IBM's business?
A: Monterey was the effort to convert AIX to 64bit for Power architechture and also IA64. That has been done.
- Responding to market demand for Linux, IBM has integrated Linux into strategy.
- regarding clustering: Clustering on Linux is very good today, good for internet applications, good for supercomputers. This gives "horizontal scalability", using many single, dual, or quad processor machines.
- Where Linux is NOT so good today, and where AIX has the advantage, is to run transaction processing applications - where data must be shared at very high speed, mission critical transactions, and high-level SMP where data must be shared. Linux doesn't do it yet, but the 2.4 kernel will be much better.
- Over time it is possible that Linux will approach the capabilities of enterprise. If so, the enterprise stuff from AIX might be migrated to Linux.
- There is a lot more to an enterprise solution than just the kernel.
- This may take several years, and that's why they are continuing their work with AIX.
- This strategy will adjust in the future as Linux progresses.
- Yes, Linux is an excellent example of the industry moving to standards.
- This started with the Internet, and now others like XML, SOAP, are other examples. Linux is part of this.
- This is important becayse it is only by embracing standards that we will be able to build the applications required by their customers. (ebusiness, etc.)
- Market is embracing standards,
- IBM is happy with this because they make their money selling hardware, middleware, and services.
- As more of that hardware, middleware, services is supporting standards, you can build bigger and bigger solutions, and it's good for IBM's business.
Question 2:
Q: re incremental opportunity for Linux. To what extent is Linux demand incremental, vs. displacing other operating systems?
A: It is early in the process, so hard to tell. However, he believes that Linux will cause a big expansion in the market. Includes an expansion in the number of products, including consumer products that will integrate IT technology. The TIVO is a good example. - computer tech is going further into consumer electronics, telecommunications, and others. All those consumer appliances can become network applicances. Believe the IT industry is going to grow - the more devices you connect, the more services you need, bigger services.
Clearly the growth of Linux will also come by displacing existing legacy systems, but this will not be as significant as the incremental growth.
Question 3:
Q: (1) re Linux kernel: existing kernel is "long in the tooth", how to we get the predicted commitment to Linux without more visibility on when the next kernel will arrive and what it's capabilities will be?
(2) re revenue growth rate: in Linux server market - the real opportunity in the market may be embedded systems, with money to be made on management.
A: (1) regarding the kernel being "long in the tooth". Operating systems are like fine wine - they need time to mature. The more mature they are, the better they get in the sense of quality, scalability. THe best operating systems in the world, they are not that young. Compared to them, Linux is actually quite young! (comments on early development of kernel by Linus, etc.) In 1998, 1999, it hit the big time. Kind of like TCP/IP. It had been around for a long time, used by the research community, until it hit the big time in 94/95.
The next version of Linux, 2.4, is scheduled to come out this fall, deploying next year, higher levels on SMP, concurrency, etc. The community is working on this, IBM's tech center is trying to help the community. Other vendors like Silicon Graphics, are also contributing, this makes Linux better and better for applicances, desktops, as well as larger and larger servers.
2. Regardiung the explosion of linux appliances. IBM views this as a wonderful business opportunity. They need technology and they need services. Technology like microprocessors, very small storage devices, analog stuff like integrated circuits for communications. This is driving IBM's tech business. Then they need integration services, this feeds the software business. A lot of work on WebSphere is dedicated to support of pervasive devices. The more of them you have out there, the more you need stuff like WebSpere on systems like the 390.
Services is the biggest part of the e-business opportunity. 60% of the revenue. 5% is hardware, the rest is software, more or less.
Big opportunity. If you have 10 billions of chatty little devices each doing lots of transactions, that's the kind of load that will keep the system 390 mainframes busy. Theres a lot of demand for server capacity.
Middlewhere is also a big part of the equation. A big part of WebSphere is targeted to this.
Question 4.
Q: IBM is working with many distributors. Do you see a possibility of Linux fragmenting, and if so, do you see a possibility of IBM providing their own version of Linux?
A: We are working primarily with the 4 major worldwide distributers: Red Hat, TurboLinux, SUSE, and Caldera.
And we do different things with each. Suse does the system 390, for example. Suse is big in europe, while Turbo is big in Asia. Red Flag Linux is the biggest in China. Mandrake is big in France, and on the desktop. Hard Hat Linux is big on embedded devices. So, over 150 distributors. They don't need IBM to be a distributor. They are better off working with distributors to integrate the whole system for their customers.
If the problem is in the Linux kernel, IBM can get together with the appropriate distributors.
There is an organization called the Free Standard Group, and just about all the major distributors are part of it and are committed to avoiding fragmentation. Look at other efforts like XML - the industry is doing well for keeping things together.
Torrey Hoffman (Azog)
Here's my notes on the first half hour of the one hour presentation. Sorry these aren't really well formatted. I may have misheard a few things, but I think this is pretty accurate summary:
===========================
- They (IBM) are aggressively embracing Linux, and have been for about 12 months now.
- Linux is very good technology:
very modular
very flexible
- Linux is becoming very popular all over the world, including Asia, Europe
- Linux is great for business solutions - delivering solutions on the appropriate platform.
- Linux is second most popular server operating system by volume, second only to Microsoft, and has achived this very rapidly.
- Linux is expected to become most popular server operating system
- IDC expects Linux to continue to be the fastest growing server operating system.
- IBM expects Linux to be a huge force in the market, and this will attract developers and application development.
- Some discussion of the mllions of servers expected.
- Note that it's hard to predict the growth, just like trying to predict the growth of the internet in 1994.
- IBM expects both volumes and revenues to go up.
Where is Linux most popular?
- Internet types of apps: web servers, ISP's, application service providers
- Very popular with internet companies, and also with large enterprises for internal intranets
Linux activities within IBM:
In the 12 months or so since IBM started seriously working on Linux, they have:
- Linux-enabled just about all IBM platforms,
- ported much of their software to Linux.
- using Linux widely within IBM
Linux Servers
- Majority of Linux use is on servers, specifically PC servers.
- Netfinity servers running Linux are being deployed many places - some examples:
Example 1:
- 250 branches of some insurance broker in the UK, didn't catch the name
Example 2: Millenium Partners
- a NY based security firm
- have consolidated trading floor apps on Linux
- intelligent support for trading on a server. Desks are running IBM "screens" User interface?
Example 3: weather.com
- one of the most heavily trafficed web sites on the net, especially during weather disturbances.
- Needed a web site that could handle very high peak loads, but at a low cost.
- Using an IBM Netfinity cluster.
- They ported their Solaris based system.
- Much more capacity at much lower cost.
Linux on Supercomputers
Example: 256-way cluster in New Mexico
- Small room on 24 racks.
- Peak 275 Gigaflops (375?)
- IBM expects many more research groups to be installing clusters like this
Linux on System 390
- Linux ports to many architechtures
- runs very well on 390 mainframes, called "Linux390".
- Is becoming very popular with system 390 customers
- already 4000 downloads of the code
- Last week, IBM had an InstallFest (virtual community meeting) where 60 customers around the world got the system up and running.
- Customers can designate specific "engines" on a 390 G5 or G6 for linux applications
- This affects pricing for the other applications running on the 390, since pricing is often done by "engine", and the Linux designated engines don't count toward the pricing of the other apps. So no software price penalty for expanding a mainframe, adding engines if they are used for Linux.
- Announced prices for selected software on Linux 390, much closer to prices to other distributed platforms, to encourage customers to use 390s instead of distributed systems.
Some customers planning (are?) using Linux390
- Many customers run mission critical databases on 390s, and web servers on different platforms that talk to the 390.
- This gives network performance problems, and raises security concerns.
- Instead, they put Linux390 on the mainframe and port the web server based apps to the 390.
- Thus the communication is entirely within the mainframe, solving performance and some of the security issues
- Multi-Image facility: special version of VM for Linux gives the ability to run hundreds or thousands of Linux images on a 390. Very simple and effective way to manage all the Linux installations.
- IBM expects Linux390 to be a big part of the future of the 390 mainframes.
IBM's Unix Strategy
- Integrated Linux into Unix strategy, which is basically to have two flavors of Unix:
- AIX for enterprise-class applications for high-availability, scalability, on the RS6000's and other.
- AIX has been ported to IA64, in beta.
- Linux complements AIX for simpler / high volume / distributed / small server applications.
- Linux at the lower end, AIX at the higher end, gives customers the widest range, flexibility, etc.
- To make the strategy coherent, they are working to facilitate porting between the platforms.
- So, they are adding a Linux compatibility layer to AIX to make it as Linux compatible as possible.
- Customer can begin on Linux, and if they later find they need more scalability or reliability, they can recompile to get it onto AIX, for the largest Unix systems in the world.
Linux on Clients, in particular embedded systems.
- TiVo is a good example. Intelligent VCR. Based on IBM powerpc processor, running Linux.
- IBM is working with other companies working on the embedded market (Hard Hat Linux)
- is being tested on the next generation powerpc processors.
- IBM is part of the embedded linux consortium.
- expect that over next few years, Linux will be taking a stronger and stronger role in this space.
Linux Software
- IBM has taken their critical distributed software offerings and made sure they work on Linux, period.
- WebSphere, including the advanced parts for ecommerce3
- Domino runs on Linux
- DB2 runs on Linux
- background messaging stuff
- Tivoli running on Linux
Also Services
- Big opportunity for services.
- Providing technical support, one-stop shopping for services
- cooperating with Linux distributors
- IBM can manage the overall contract, bringing in a particular Linux distributor for their expertise
- As linux becomes more of part of ecommerce, ebusiness, lots of work in linking Linux to other applications.
------------------
I'll post another summary of the next half hour after I've listened to it...
Torrey Hoffman (Azog)
I use it all the time. It comes up like this:
Torrey Hoffman (Azog)
I think DOOM came with that capability from the beginning, but I'm not sure. There must be some huge DOOM FAQ's out there that will tell you one way or another.
Torrey Hoffman (Azog)
Don't you think this is a little too close to cheating? You mentioned setting the second PC to a 3rd-person view of yourself.
Wouldn't it have been much more useful to have it chasecam on your opponents? I think that would definitely count as cheating - at least for a head-to-head game.
I know that when I would play my roommate, we would yell at each other "Don't look at my screen!" when we were head-to-head in the same room. We often did sneak peeks, though. There's a huge advantage in knowing where your opponent is, what weapon they have, and where they are going.
So... is it cheating?
Regardless, it's a neat trick getting it running.
I might stick that old Voodoo 1 card back in my spare machine and try it myself.
Torrey Hoffman (Azog)
Yes, you read it wrong. Microsoft does not actually disagree with the facts of the Gartner analyst who started this whole story, they just disagree with the "tone". (Last paragraph of the article.) In other words, they want to put a spin on it.
The undisputed fact is (4th paragraph of article), if you buy a bunch of PCs with Windows 2000 on them, and you are in the Microsoft Select program, and you use Ghost or something to wipe the hard drives and install a complete new image of the same OS, the apps, the utilities, and everything else, you are screwed.
All big companies do this. But Microsoft says, "No, the Windows 2000 that you bought with the computer is not the same as the Windows 2000 you are putting onto it with Ghost."
So you have to pay for it again. And you can't get free support anymore from the computer manufacturer, you have to pay Microsoft $375 per incident.
The only way to avoid this would be to purchase the computers without anything on the hard drives. Good luck! Linux, BSD, and BeOS users already know how hard that is!
I predict Microsoft will change this soon, otherwise they will take a big hit next time somebody does a comprehensive study of the TCO (Total Cost of Ownership) of Windows 2000 vs. Linux. I would expect that Linux will become competitive in such studies sometime next year.
Imagine a company with 5000 PC-using employees, and a few dozen support staff. (Same as Gartner's scenario in the article.) The money Microsoft wants out of them for software they already bought is about $600,000 dollars.
Now, that much money could hire a couple of Linux experts for a year, plus pay for several days of concentrated classroom instruction for the rest of the tech staff to get up to speed on Linux.
Actually rolling out Linux would still cost quite a bit, but once all the staff is trained and you have a couple of experts on board, you can probably make the jump without too much pain and disruption. I would bet that the transition to Linux would easily pay for itself in a couple of years, with lower software costs, lower hardware costs, better software reliability, and better security.
I predict this will begin to happen more and more next year.
Torrey Hoffman (Azog)
Previous articles from Moody that show his total bias and ignorance, for your reference and reading pleasure:
A story on how biased the judge was "Virtually since the beginning of the trial, though, Jackson has brushed aside every request and argument made by Microsoft, and sided enthusiastically with the U.S. Department of Justice"
How nice those poor Microsofties are "I keep hearing Microsoft-employee-misfortune stories so powerful that they tug not only at the human heart, but at the journalist's, too."
Gates is nicer than Allen "...Gates is emerging as a man with his heart in the right place"
How well Gates handled stepping down as CEO "... Gates neatly avoids a massive pitfall..."
Can anyone out there take this guy seriously?
Torrey Hoffman (Azog)
This is amusing.
Just below the note that Mr. Fred "Hi, I'm a Microsoft Tool!" Moody wrote a book about Microsoft on the multimedia frontier (*cough*), the ABC site has a search button... which is pre-filled in with...
Search for more on... Microsoft .
So, ABC has classified this as "Microsoft" news, despite the headline being "Linux". Things that make you go "hmmm".
I hope Slashdot gets around to a follow up on this one. I really, really want to see if Moody does a correction article. If he has a trace of honesty, he will do another article which admits:
1. He is wrong, because he added Red Hat's numbers to the numbers for all of Linux. If he had done it correctly, he would have reached the opposite conclusion.
2. His so-called methodology for determining the "Suckiness" of an OS is ridiculous anyway, so even ignoring pt. 1, his article is balony.
Torrey Hoffman (Azog)
Minor correction:
The Difference Engine was William Gibson and Bruce Sterling.
I was disappointed by the book when I read it, but I don't remember much else of the plot, characters, or settings. I guess I'll have to dig it out and re-read it - maybe my tastes have changed and I will appreciate it better the second time around.
ObOnTopic:
Whoo! That Photon Mapping is seriously cool! It looks better than ray-tracing, and it can be done in milliseconds! Does anyone know of a downloadable demo of it, or more information than was given in the article?
Torrey Hoffman (Azog)
I sort of agree with you...
for (1), I think that it won't so much be Caldera soliciting their customers to move to Linux, it will just try to capture all the customers who have already decided to go to linux and are just beginning to implement it.
For (2), I agree with the earlier poster who stated that SCO Unix can never be opened completely because SCO doesn't even own all of it. Likely they will open parts of it, and port parts of it to Linux.
My expectation is that they will create a so-called "high-end" Linux distribution, with some SCO software (mostly userland) in it, specifically targeted to people migrating from SCO Unix.
Not a distribution for you and me, but very attractive for some people.
Torrey Hoffman (Azog)
I agree - very few games from the 90's will still be played in 2010.
Probably the original Quake will be one of them. It spawned a whole community, hundreds of web sites, and the concept and practice of professional video game players actually making money to frag.
Now that it's been GPL'ed, it can never die, despite the inevitable changes that will come to hardware and operating systems.
I just had a chilly moment of foreboding... In 10 years, some kid is going to see me playing Quake and say: "Hey! that's one of those old games, like... uh... Brother Mario, right? That was for a... um... Nuntindo Main Frame?"
(sigh.) I should be too young to anticipate feeling that old.
Torrey Hoffman (Azog)
Excellent post. A small nit to pick:
The expression is, "setting the fox to watch the henhouse". Not the chicken.
Best wishes, and may your karma be ever increasing.
Torrey Hoffman (Azog)
The itrace packets will have an authentication section. Read the ietf draft, it explains some of the possibilities.
At any rate, spoofed itrace packets will be detectable.
Torrey Hoffman (Azog)
This is not an attack on anonymity. Go read the actual IETF draft. You will see that the only thing it helps with is tracing back packets with SPOOFED originating IP's.
This will help prevent things similar to the attack on kuro5hin. Unfortunately, if attackers are using compromised machines, all it will (or can) do is help to quickly find the real IP addresses of the machines that have been compromised. You see, someone doing a denial of service attack right now can cause the servers they are using to output IP packets that look like they are from somewhere else. When those packets arrive at the target, 10 hops later, it is nearly impossible to find the real machines that is causing the attack. That's what this proposal solves.
This has nothing to do with eliminating privacy or anonymity. Every time you connect to a web site now, they can find out the IP address you are coming from. Duh! How else can they send the web page back to you??? If you spoof your originating address, you cannot have a two way conversation.
IP source spoofing is ONLY useful for denial of service attacks, and that is the ONLY thing this proposal addresses.
The so called solutions you are advocating, like restricting access to the net would be far, far worse for invading privacy. Think about it... how are you going to make sure that only "authorized people" use the internet? Well, you will have to identify all of them. With examinations, meeting criteria, getting what is equivalent to an "internet license"... well damn, there goes privacy! Just like anyone who sees your license plate on your car can find out who the car owner is. No privacy there either. Did you think about this?
The IETF proposal is not a perfect solution. You are correct that there probably isn't one. However, it is a good one and 100% better than your suggestion.
Torrey Hoffman (Azog)
Good idea, except for one thing...
It would be trivial for the attackers to change just a few letters at the end of the post, completely changing the MD5 hash.
What would probably work better is some sort of "diff" approach. If a post is "too similar" to a lousy one it would be rejected. It isn't actually that hard to do similarity checking, and the load is not that high. After all, even on a busy day Slashdot only gets about 3000 posts or so.
This would get rid of the Jenna Elfman, Penisbird, Latin Lessons, and other fools. (I read one of the stories at -1 yesterday. Unbelivable!) I wonder if this would count as censorship? Perhaps instead of deleting the post, the system should just instantly moderate it down to -1.
Torrey Hoffman (Azog)
I think that instead of a moderation system that just lets people bump a post +1 or -1, moderators should be able to pick a level for the post.
Then the computer would average all those moderations together. For example: Somebody makes a good post, and it starts out at 1. Moderator A rates it "Insightful" and "5". Moderator B rates it "Interesting" and "3". Average them all together, (1 + 5 + 3)/3 and it would show up as "3".
You might need to simplify a little. Perhaps instead of the numbers, have a scale like "Crap, Useless, Ordinary, Good, Very Good, Excellent". Maybe just use that and drop the Insightful, Funny, Interesting part of moderation.
I think that would be simpler for people, even Arts students. (heh. You said it, not me! *ducks*)
However, I think Slashdot works amazingly well, considering how many trollers and losers are attacking the system. So if you think your site is going to be big, or will have a lot of losers on it, you might want to stick with what is known to work.
Torrey Hoffman (Azog)
Well, if you browse at +2, the signal to noise isn't too bad. A lot of "me too" type posts that don't add too much, a few morons, and a few really good posts.
:-)
If Slashdot didn't have the discussion area, nobody would be here. Instead, everyone would just check Wired, CNet, Blues News, and about 4 other sites. Between them, they get at least 90% of the stories that come up here, plus a lot more, and are more accurate.
But the level of trolling sure has increased. I try to read every response to any of my posts, and if I ever lower the threshold to -1 I'm amazed at the amount of junk down there.
By the way, this post is borderline between a "me too!" and interesting...
Torrey Hoffman (Azog)