If you don't know who Bruce Perens is you just haven't been paying attention. The list of big names generally runs from top down like Linus, Richard Stallman, Eric Raymond, Bruce Parens, Alan Cox, Mad Dog Hall, and so forth. Is he Linux top dog #1, no, but he's definately earned a table at the big kids table and is one of the most active names of the group, and he maintains a very high profile. But just for a refresher, so that we don't loose site of his contributions to the world, lets recap quickly: he's worked at Pixar and its predecessor NYIT, was "Senior strategist, Linux and Open Source" at HP from 2000 to 2002 untill the Compaq deal when he was terminated, he created BusyBox and wrote the aging but popular ElectricFence, he's been hacking UNIX kernels since 81 and Linux since 94, father of a beautiful son, founder of the LSB, Co-founder of the OSI, popular (and expensive) public speaker, HAM (K6BP) and co-founder of No-Code International, and most famously he wrote the original "Open Source Definition" which serves as the basis of much of the organized Linux and Open Source (notice the caps) movement. Depending on where you look, he says or other people suggest, that he actually coined or helped coin "Open Source" (see "Revolution OS" the movie) and trademarked it. Cut me alittle slack if I missed a little or mistated, but thats his basic quick bio.

So, from all this we basically get the idea, the d00d is important, been around awhile and isn't afraid to pick up a banner and go with it. He knows how business works, can interface and drive it, isn't afraid to trade mark something or start an organization or put his butt on the line (come on, putting your name on the OSD takes alittle guts, credit where its due). Bruce is a guy that deserves respect, admiration, and is someone that you can learn a lot from. So lets be clear about who we're dealing with before proceeding.

Because of his great works in days past and present one would think that he'd have a really good sense of historical significance, capability of the community, drive and determination, ability to take risks, and historical repetition. But recent words from Bruce don't tend to back that up. You'll find that he's an active /. commentor and contributer. Here's a nice little exchange with Bruce recently between Alex (no last name) and Bruce:

Thanks Bruce - maybe I haven't been clear enough, I don't understand what it matters if a person working on an open source project is paid or a volunteer.

Are you actually trying to suggest that the same code/communication written by a Sun staffer are somehow less worthy (and/or useful) than identical code/communcation written by a volunteer ?

If you are, I cannot believe I used to respect you. [Alex]

To which Bruce responds:

Don't say stuff like that before you even understand what I am talking about.

My point is that Sun has not been successful in building much of an outside developer community, for what should be the second most important software program in all of Open Source software, behind only the kernel. Of the people participating, a good many are Sun contractors, etc. And it's difficult to find very many volunteers. [Bruce]

Very interesting exchange indeed. Now, here they were talking about OpenOffice directly, but in reply to the "Sun Grants Access to 1,600+ Patents" story. This all translate in with Bruce's general comments on the suject of the CDDL, OpenSolaris, OpenOffice, and Sun's open sourcing efforts in general. This is raw and pure FUD of the worst kind. Based on his writtings you'd think that that he'd be thrilled by the patent allowance that Sun is providing and embrace Sun's move to provide indemdification in the CDDL. Is this a case of "works for me but not for you"? His position is certainly unclear at best.

I found an interesting artical that Bruce wrote for ZDNet back on June 13th, 2001 entitled Beware of wolves in agnostic's clothing. When LinuxDevices.com reported the article they, I think, more appropriately titled the mention "Bruce Perens: in defense of open source", which would have been a fitting title. So now we look back at this artical and wonder, was he defending open source (no caps) or defending Linux against Microsoft? Almost seems like the latter is true. My favorite paragraph is the following:

It's entertaining to watch the effect that open source seems to have on Microsoft executives. In February, Jim Allchin called us un-American. Recently, Steve Ballmer said Linux is a cancer! These are not the words of calm people. What is it about open source that makes Microsoft executives so uneasy? Could it be that after all of these years, competition has become so foreign an idea to them that they react with horror?[Bruce Parens]

Competition. Interesting... interesting... Because it now seems like those on the BSD front and those of us on the OpenSolaris front are now in those same crosshairs but Bruce and a minority of the Linux community are in the Microsoft seat and we're in his. Is it really fair that I'm perhaps bending his words against him? It might be. I'll admit that no one likes to write something and see it come back to bite them, even I. But the reactions to announcements even thus far by Bruce and other members of the Linux community are strikely similar and hail back to an earlier time. Isn't it irronic that Bruce himself taught us how to fight these battles. Perhaps, as Bruce wrote above, after all these years of success and going it alone competition has become foreign even to them.

I remember a time when those fighting the good fight were trying to win the legitamcy of open source software as a model of success and of a better and brighter future. They called for anyone that would come to rally to the innovation and community that open and free software can provide; come one, come all. And they did it, with hard work, and determination, they did it. What changed? What happened? Even if, and this isn't true, Sun only wanted to "cash in" on the open source idea, would that be a bad thing? By the old standards it wouldn't, because it would simply be validation that the model itself was the right one and it worked... but now, it seems nothing is quite good enough for them. There isn't anything you can say, nothing that you can do. Your with us, or your with them. Thats not what this is supposed to be all about... and I just don't understand it.

Since Bruce is a HAM I looked him up in the FCC database (one of the downsides to be a HAM, your record information is public) and it would seem that he lives here in the Bay Area. Perhaps one day I'll have the honor of meeting Bruce and discussing some of these things with him and trying to clarify the position he has. He seems like a really nice guy and in many ways he's a role model to many of us, because like I said before, he taught us how to fight this sort of fight. Maybe if he doesn't feel the need to put a pen through my temporal lobe we can have some coffee and see where this goes, but for now we'll just have to watch and wait and hope that we can all come together again as a single undivided open source community, whether you prefer Linux or BSD or OpenSolaris or HURD or whatever. I for one would like Bruce to be a friend to us all, bringing us together.

FUD, do you know what it stands for? Just saying it out loud gives you the idea. FUD is an acronym invented by Gene Amdahl after he left IBM to found his own company: "FUD is the fear, uncertainty, and doubt that IBM sales people instill in the minds of potential customers who might be considering [Amdahl] products." The acronym became particularly popular during the thriving years of Linux, when it was pointed at just about everyone. Linux verus the world.... and it certainly helped.

So this evening I find this, "Why I love the GPL", and its largely Linux FUD. Why this fella feels the need to defend either Linux or the GPL, I dunno, he seems to use the terms interchangably as though the two are joined at the hip, which really depends on who your talking to. Sure, Linux certainly is the undoubted poster child for the GPL, but very little of the GPL code out there is Linux specific (we've discussed this several times before) and the GPL was in use long before Linus choose it for his kernel.

I started to write a big long reply to this article but I'm not going to bother, it's fairly obvious whats wrong with this thing on it's own. I don't know if this artical was pointed in the direction of the CDDL or not, nothing indicates that it is but the timing is.... interesting. The GPL has its up's and down's, especially when compared to the BSD license (basically "live and let live, just keep my frickin' name on it"). Here are two choice lines that I personally love and sum up where this guy is coming from:

The GPL is often described as idealistic and altruistic. If the kernel developers were interested only in the code, wouldn't the BSDs be the ones with the huge development corps instead of Linux? That's what we're told all the time by the BSD-bigots, it's better technically.

BSD-bigots eh? Ok. Here's the next one:

Richard Stallman is a man. Brilliant, opinionated, and uncompromising. Many attacks on the GPL are made indirectly, by going after Richard Stallman, for no other reason than he is vulnerable to them, while the license itself is not.

Damn... I've done some butt kissing in my time, but wow. Are these statements untrue? Nah. In the "RMS vs ESR" flamewar I always came down on the side of RMS and free software rather than ESR and Open Source, but so be it. Is Richard really vulnerable to anything but a tailwind? I don't think so, go insult Richard and and find out what happens. If there is one guy in the whole open source world that doesn't need protecting its Richard.

Ok, so anyway, does the GPL need defending? I didn't really think so. Are we ramping up to a GPL vs the world fight? Sorta seems that way, but he's looping Linux and the GPL together alittle more closely than maybe he should. Is this guy afraid of the CDDL? Has he ever taken BSD seriously? Has he ever considered that their way isn't the only way? I don't know. But it definately meets the definition of FUD and leaves you wondering who they feel they need to fight. This is happening more and more. Are we looking at self implosion in the making? Lets just watch and see.

I'm an old NeXT fan, I even have a beautiful NeXTStation Color Turbo Slab (Pizza box, 33Mhz) out in my garage. So it was really kool to go to /. and find "Steve Jobs Demos NeXTSTEP 3.0", a decent Quicktime video transfer from an old NeXT VHS marketing tape. Pretty kool and brings back memories (all past, I was in the 6th grade I think when these things were hot) because I've had the fortune to watch alot of these old tapes over the last couple years.

The koolest of all the NeXT Marketing videos can be found in the Sun Microsystems Library on the Santa Clara Campus. When I was migrating iPlanet as a consultant into the new facility (the origonal Netscape/Sun merger, I came in to help move and migrate the systems, being that I did both Solaris, HP-UX, and AIX, all used for development and testing; Sun used Solaris and Netscape, at the time, mostly used HP-UX) I was restricted to access just in that building, however on the first floor was the Sun Microsystems Library. And, believe it or not, the let me check stuff out! I was in heaven!!! Even kooler was the fact that they had a video section with old press confrences, meetings, etc. Anyway, one day I was scanning the library of videos and found one with a NeXT logo, upon further inspection it was a coding contest between NeXT and Sun! Totally kool. Sure enough, I watched the tape that night and 2 coders, one from NeXT (black turtleneck) and one from Sun (blue shirt, jacket) sat down and hacked an app to the same basic specs, Sun using C with MOTIF and NeXT with ObjC and their tool set. I don't need to tell you the outcome, it's obvious, but you'd be shocked at how well the Sun guy did. It was a seriously kool video and I wish I'd had the ability to make a copy of the tape. I'd love to see something like that happen again... ObjC vs Java.

NeXT Rules! Well.... ruled. Sorta..... except that OSX is really next-gen NeXT which I blame for ruining my beloved BANG (Bay Area NeXT Users Group) which only talks about OSX and WebObjects anymore. Then again, I'd gladly trade my NeXT fetish for a Thinking Machines fetish instead. Or even Amdahl if I got to have coffee with Gene. :)

Since I'm posting pictures, I can't resist showing off this one that I took last night...

That pictures is completely and totally unposed! She'll walk up to the paino and give me the "lift me up there" sign and just hammer away. I can't really vouch for her playing well, but then she's only 14 months old... I'll cut her slack till she's at least 2. :) [If you haven't figure it out, that's my daughter Nova... yes, shocking that I have a child I know, even more shocking that she's so darn cute, but its true. You might even be more amazed that a weirdo like me is expecting again... Tamarah is 4 months pregnant with our second.]

About a year ago I was doing a late night shopping trip (we do all our shopping during the night, when the store is empty and freshly stocked) when I walked past these:

The things look nasty and they've been sitting ontop of my NeXT Turbo Slab (I love that thing) for awhile, so I wouldn't drink them if they were the last sources of moisture on planet earth, but even still I'm just geeky enough to think they're pretty kool. They also had an "Enlightenment" one, which I have around somewhere, but can't seem to find... I loved lining them all up and thinking that it was a divine message.

So Tamarah is at a meeting tonight and won't be back for about an hour and Nova crashed out early on me so I thought I'd run through some of the comments out there. The /. comments were pretty cut and dry, predominately positive, and because comments are threaded anything that was just plain wrong was quickly corrected. OSNews isn't quite so threaded, so I thought I'd reply to some of the interesting and/or negative comments to correct them or at least provide some feedback, at least from one community member. Remember, I don't work for Sun, I'm just a pilot member, which means I have some but not all of the inside scoop and all these opinions are my own. On with the comments:

solaris strength's are in the Kernel. ZFS, Dtrace, TCP/IP stack, etc.

The rest of the OS is playing catchup to linux. The tools? why the GNU tools for the most part are being ported to Solaris to improve those. Servers, interface? etc, are in general ports of standard F/OSS software. Maybe their version of X will be nice, but don't expect much else.

As for kernel toys, Dtrace is slowly being created by kprobes(which are now dynamic if you actually look),

The big thing is ZFS, which might get a GPL port.

As for developers, there are 100 of them now on 'Open Solaris' Geez how many developers did Sun have working on Solaris? It's not to hard to see where those first 100 came from. Also Sun said they would release the entire OS this month, now it's June. Let me Guess June will turn into August? [peragrin]

Okey, so thanks for pointing out the strengths of Solaris10. I'd add to that SMF (Service Management Facility, part of the Self Healing stuff) which is the first real major change to give us something better than the old RC Init system we've had, Janus for running Linux binaries (personally I don't think it's a big deal, almost all of these apps will recompile cleanly on Solaris anyway, so only closed source apps are really an issue), and my personal favorite: Zones (if you want to compare Zones, aka Grid Containers, with User-Mode Linux, Xen, or even Linux VServers, you just haven't seen Zones in action yet).

As for the tools, where talking about the base OS here, so the tools will be the standard *NIX toolset we all know and love. Has GNU really made superior tools to whats in Solaris? Not really. Is GNU Tar better than Solaris Tar? We'd all say yes, but only because some nice features like integrated decompression (-z or -j) are built in. So, the GNU replacement tools are nifty but really just add alot of neat features to the same tools we all know and love on just about any platform. Some things are speed up and stuff, but cat is cat. And, when it comes to OpenSolaris it frankly doesn't matter anyway. I can guarrentee you that one of the first things that'll happen post-release is that someone will build a GNU-ized distribution (this isn't a license issue because your not merging code), where you have GNU binutils, etc, ontop of a Solaris kernel. I imagine people will probly write in GNU like option handling to Solaris tools over time, they just won't look at GNU code when they do it.

KProbes? Please. Bryan Cantrill would like nothing more than to see every *NIX OS on the planet have a DTrace like functionality built in. And, I'd imagine he's hoping that it'll happen. He knows that people can and will try to do it, and he's commented on the KProbes progress for awhile. Will it be as solid as DTrace? So far it doesn't look like it, but it'd be kool if it was. I'm all behind KProbes really taking off and those guys are working hard.

ZFS won't port due to license restrictions, so thats out. GNU or someone else might try to build a filesystem to compete with ZFS, and if so, kool. Besides, I think the idea that kernel code can be ported back and forth has been overblown. It's just not that easy to yank out kernel bits and plug them in elsewhere. Can you port DTrace? Probly not, but even if you could, it would sure as hell be easier to take the idea behind DTrace and reimplement it from scratch, making the only place interesting to borrow code the D-language parsers.

But as for the number of developers and the release date. A lot of people are saying the same thing, that no one should trust that Sun will get OpenSolaris's buildable source out the door on time. But there is something to remember, there are (at last public say) 60+ pilot members external to Sun, which means if they are late on delivering the code one of the following is true: 1) 60+ developers are screaming bloody hell trying to figure out what the holdup is, or 2) There is something that needs serious fixing before it goes wide open that is a major show stopping issue. In either cases, you'll have a ton of us blogging about it. Solaris FCS releases are a diffrent beast than most because Sun realizes that it's not just geeks that run the OS or small companies, but banks, and governments, and hospitals. In a very real sense peoples lives depend on systems running Solaris and they take that dead (no pun) seriuosly, which is why Sun has absolutely no problem holding a release for acouple weeks to get it right.

Anyway, peragrin isn't a believer, but thats kool. I thought he made some good points. On to the next!

Anyone know what compilers are used to compile Solaris for x86? Intel compilers? SUN proprietary? [HIRE ME!]

Jonathan Adams pointed out that GCC 3.4.3 ships with Solaris10 and can build AMD64 today, but he didn't know about the SPARC side. Hireme continues...

Was just wondering since SUN sells a C compiler as a seperate product It seems to me that the SUN C compiler would be the prefered way to compile any OpenSolaris source code. Perhaps i am confusing the Sparc side of things with the x86 side but the source is supposed to be the same for both platforms. This opens up a whole new realm of porting. Imagine Solaris running on Power 5 or PA-RISC... [HIRE ME!]

I bring this up only because several people have really worried about it. Some people have even screamed "Well, it's not really free is it if you can't compile it!!!". I can not disclose plans, or direction, but I'll say this: Sun writes a lot of software, has hundreds of engineers and some of the smartest minds in the industry... do you really think they'd open the codebase, spend hundreds of millions of dollars, and more than 10 years on this effort and then just sorta forget that you can't compile it?!? Come on... there is a plan.

On a side note, Sun Studio 10 was just released which is an upgrade containing primarily X86 and AMD64 optimizations and performance improvements that (Sun says) produces much faster binaries than GCC. You can download a free trial for Solaris (X86, AMD64 and SPARC) and Linux (32 and 64bit).

Ok, last one.

Everytime there's a Sun thread, people are comparing Linux to Solaris and vice versa. Is there a particular reason why the BSDs are never compared to Solaris in these threads? [Salacious]

There were lots of other good comments to respond too, but this one really really interests me. He makes a really great point, why does everyone forget that the *BSD's exist? When I've been talking to the press (or even some people within Sun - you know who you are) they talk about how "Linux has more applications than Solaris!" or "Linux has more developers than Solaris!". Well, there are two major problems with this ancient bullshit old skool thinking, something that us little hacker people know but apparently industry analysts haven't figured out, how many "Linux developers" are working on the kernel? Not that many, I recently heard that by Linus' count it's about 200. Ok, so the rest of the "Linux developers" are working on libraries and applications, and all those goodies. Are the Xorg guys "Linux developers"? Are the Apache guys "Linux Developers"? No, they aren't writting Linux kernel code and their applications don't have to run on Linux. All these apps have been running on FreeBSD, OpenBSD, Solaris, HP-UX, AIX, IRIX, QNX and more for ages. In the open source community (not the "Linux community") developers on BSD and Linux often work in lock step. For instance, in the Enlightenment Project of which I am a part we have one developer who only works on PPC and very often OSX, another who only works on BSD, another (me) who does most work on Solaris, and a large number of people that run our applications on these platforms and more (IRIX, HP-UX, AIX, Tru64).

During the press interviews I've done in the last week each time a reporter has said "But what do you say to those who think that Solaris will never have the number of developers that Linux has?" I can't help but think to myself "Those poor BSD guys". Because the BSD guys work their frickin' asses off!!! I can't tell you how much respect I have for all the BSD guys (and gals) out there, because they are so massively commited and there are so many bad ass skilled people working on the BSDs that it's just unbelievable... but do they get any glory? Hell no. The question is an obvious, "WTF!?!?". And sadly I don't have an answer, but what I've found out during the last couple weeks is that almost all of the battles we're fighting with OpenSolaris is a battle that the BSD guys have been fighting for a long long time. And look at the legacy of FreeBSD, for instance; FreeBSD 1.0 released in December of 1993! In contrast Linux 1.0 released in March of 1994 and was pretty scary but moving along, BSD had stability galore. You call Solaris a toy and a joke, you forget that Linux was a disaster for a long time. But lets be honest, it was messy and nasty and bad by todays standards, but it was in it's infancy at the time and a remarkable achievement, and even with it's faults we loved it and held it and worked on it together to make it was it has become. For years and years Linux was just laughable when put side by side with BSD, but did that stop anyone or did that slow developement? Hell no, they worked harder (and ported the best bits) and puffed out there chests a bit and then worked some more. Arguably both FreeBSD and particularlly now OpenBSD have some things up their sleeves that really let them point and laugh at Linux, but for some reason it's just not that big a deal. Why the BSD guys don't throw a tantrum I dunno... I think they are just too kool and mature to scream like angry children about it (but I'm not). But everything aside it comes down to this, everything that the "Linux community" wants to take credit for and be proud about the BSD guys did first, and not just first but years ahead. When you look at the major movements lining up now, you can see that BSD and GNU both were light years ahead of the upstart Linux and Solaris has built heavily on BSD's lineage (which Scott points out frequently... maybe too frequently).

Okey, I don't really wanna start a flame war... I love Linux, I've defended Linux, and I run Linux as a desktop system because the driver support is just unbeatable right now and Solaris/X86 isn't a good Quake3 system for nVidia 6660GT AGP cards fully accellerated. I went to Linux10 and subsiquent events, I went to the first ever LinuxWorld, I have many friends who worked for VA, I pushed companies and all of my employeers to consider running Linux when Solaris/SPARC wasn't a good fit, etc, etc, etc. I've loved Linux as much as anyone else can and still do, but to make a bad analogy:

Oh man, I can see the hate mail now.... I better setup a new procmail filter tonight. But I hope maybe some of my point gets through... BSD, Linux, OpenSolaris, GNU, whatever, there are lots and lots of kernels and operating systems out there, especially in our space, the open source space, just because the snazzy catch phrase "open source" (personally I prefer free software) got tagged and slapped on Linux doesn't mean that everything else is inferior or somehow unblessed. A lot of people bust their asses on projects that don't have any direct correlation to Linux beyond the fact that the developer themselves and the bulk of the users probly are running Linux too. If someone called Apache a "Linux Application" and I were a FreeBSD user I'd be pissed off! If someone called KDE a "Linux Application" I'd be upset. We can all understand why people would say that, sure, because most of the users are gonna be on Linux, but we forget that to the larger old-skool world "Linux Application" means something completely diffrent! When the press hear "Linux App" they think of it like "Windows App". Remember when the reason people bought Windows over MacOS was the fact that "no software exists for Mac"? And so this war started between Microsoft and Apple to plead, beg, and bribe developers to come to their side of the table... well, sadly, in the modern world many people still think like that! They say "open source everything" out of one side and "Linux has more application than any other *NIX platform" on the other. Most open source apps can run on FreeBSD, OpenBSD, NetBSD, Solaris, HP-UX, Tru64, AIX, IRIX, or whatever with little to no porting needed, and most of that porting is actually only needed if your not willing to use GNU LibC on the other platform. Maybe I'm not the only person to point this out, but now trying to talk to people about OpenSolaris and where we're going I feel closer to the BSD guys than ever before.

Okey, so those are just my personal thoughts and no one else is responsable for them. Again, I don't work for Sun so don't flame mail them. Comments are welcome, if your really pissed and don't want to comment you can email me, but I think alot of us can admit that what we little people (coders, admins, g33ks) understand as "how things are" and what the tech press and businesses and managers and CIOs and folks understand are often not so similar as we want to think. In alot of ways they still jsut don't get it. It's like that famous line from the Simpson's when Bart goes to Australia to get booted and in the pub Marge says "COF-FEE" and the barkeep repeats "BE-ER", you know the one. In the end, at least at the bottom, we're all one big happy disfunctional family changing the world one subroutine at a time... we just stop occasionally to beat each other up. :)

So, it's been a wild ride the last acouple of weeks leading up to yesterdays preview. All of us in the pilot have been working together to provide feedback on what you saw yesterday and to shape things. It's been an amazing proccess, but a lot of work. Sleep deprivation really sucks. Learning the buildable codebase and tweeking and building have been really kool, and I wanted to get everything built, tested, installed, and figured out prior to the preview in order to be able to honestly say to anyone that I actually did have all the code and I was building, installing, and running the code. Because of that work I was able to provide the screenshots and version information that got passed around the net yesterday to prove it. Sadly, all that work ment having to stop working and playing with Enlightenment and the EFL for awhile, which sucks because there are some seriously kool things going on right now that I'm missing out on, and picking up on other projects. I've been working on a book for SAGE that's gotten put on the back burner due to OpenSolaris, and my editor is I think about ready to chop off my head, so I can now get back to that. Besides that, I actually do have a real job and I need to double down on some projects there. I'm also hoping to spend alittle time at the track too, but more than anything I miss hanging out with my wife and watching old movies all night, so that'll be fun.

But, even though this round is over, we're ramping up. You can see on opensolaris.org that a CAB is being assembled, and so we're working on that, as well as continueing work on the codebase and working on a wide variety of projects. One thing that I'm working on is to get back in touch with Solaris/X86 since I've been a SPARC man for so long, I'm seeing things that can be done to help new users and generally make X86 less mysterious. Plenty of us long time SPARC guys have never had to worry about where the kernel was stored, or what every single module did, as long as you knew the important ones (st, sg, ssd, etc) you were good to go; however now on X86 you can get confused as to which driver does what and how you forceload things, and why doesn't this match with that, etc, etc. These are issues that you run into on Linux too but genearlly know how to fix because it's documented somewhere and you find it via google; well, lets do that for Solaris/X86 too! I've resisted Solaris/X86 like alot of people for a long time, but with DTrace, Zones, SMF, and other features found in Solaris10 it's just really really really hard to not run it, especially when you consider that it still has more robust SMP and memory management than Linux, along with one of the best schedualers on the planet. I'm comin' around and I want to make it easier for other to do the same.

Lots to do, lots to work on, lots to gain. It's a fun and exciting time. But for now, I can look forward to sleeping alittle more than I've been able to for awhile in preperation for the big big push in Q2 when the whole project goes open.

Just a couple of things I should clarify to anyone reading this blog, since to some people it's not clear, I am not a Sun employee. I'm a member of the community. I wouldn't mind working for Sun, but I don't, and I'm not planning on it, so don't think that anything I say is some how just party line stuff. There are plenty of people inside of Sun that don't like me, and I've trashed certain things that Sun has done before and will continue to do so, whether people think it's wise or not. I'm behind Sun because thats where the future is. One of the kool things about not working for Sun is that when I say something sucks, people listen and something might happen, whereas if I was an employee I'd probly get repermanded or fear loosing my job. Believe it or not, I am objective. I'm interested in whats best for Sun, and thus myself, whether Sun or anyone else likes it or not.

A note to the press: I've been contacted by several diffrent press agencies already, and I think all of the community members are more that happy to talk to anyone about Sun, Solaris, and OpenSolaris to anyone. I could sit around and talk about Sun all day long (and sometimes I have). However, I've recently learned that the press can not (or will not) quote blogs for articles. I think thats kinda lame, but whatever, I don't make the rules. Anyway, if you contact individuals in the community please let us know what angle you want to hear from. Many of us are professional systems administrators, have worked for medium, large and Fortune500 corperations, but are also midnight hackers and open source developers in a variety of very kool projects, as well as the fact that most of us also have a long time relationship with Linux. Some members of the press are interested in the business side, some are interested in the community side, some are interested in the corperate side, and some are interested in all of them. Just let us know which hat you want us to put on and we are happy to discuss the impact OpenSolaris will have so we can help you get the story that your trying to report.

The announcements that occured today may seem simple and expected, but as a member of the pilot I've gotten to see the unbelievable commitment and hard work of many of the individuals working, often round the clock, on our behalf. Sun's new strategy is built around the community, and thus by definition can't succeed without the community. Jim Grisanzio was given the task of mobilizing and organizing parts of the community for the pilot phase, to manage and direct all of us. I've been a pain in his side more than once, but he's worked hard, long into many many nights, keeping us directed and pushing forward. There are hundreds of people in Sun to thank for the announcement today, Scott's team, Jonathan's team, Claire's team, Engineering (Mike, Keith, Brian), legal, just to name a few. But more than anyone I've gotten to see Jim really make Sun's vision a reality and to hold things together. Trust me, we're a surly bunch and all have our own ideas and opinions. Jim has had a hard couple weeks and so it seemed only fitting that I, and the community, make a gesture to show our thanks for all his hard work on all of our behalfs. This afternoon I asked Jennifer Bauer (best PM in all of Sun!!!!) to take these (see pic) up to Jim's office. Hopefully he shares alittle with everyone else, which is why we got 2 bottles.

Jim, on behalf of a greatful community, Thank You.

PS: Hey, IBM and HP, how many of your community members and customers would show you this kinda love? Maybe you should learn some leasons from Sun, eh?

choice  n.
   1. The act of choosing; selection.
   2. The power, right, or liberty to choose; option.
   3. One that is chosen.
   4. A number or variety from which to choose: a wide choice of styles and colors.
   5. The best or most preferable part.
   6. Care in choosing.
   7. An alternative.

Choice is freedom (2). Choice is empowering (2). Choice is liberating (7). Choice is difficult (6). Choice is rare (5). A lot of groups and vendors want to give you choice. HP will give you choice, IBM will give you choice, and Sun will give you choice. But at what cost? HP will give you choice, so long as your willing to choose something acouple of weeks or months later when Carli changes her mind. IBM will give you choice, so long as you can afford to ensure that choice with Global Services. But some how Sun's choice is diffrent... Sun provides choice that is true, that is reliable, and that is innovative. A year ago, if you approached HP you were given unparralled choice, you could choose Windows, or HP-UX, or Tru64, or OpenVMS, or MPE, or even Linux. But times changed, the business reorganized and customers that choose to leverage that choice were asked to choose again. At IBM we see a completely diffrent story, you could choose Windows, or choose z/OS, or choose AIX, or even Linux. But only Linux runs on multiple platforms, otherwise your running Windows on Intel, z/OS on z/Series (mainframes), and AIX on POWER, only Linux can run on all three, but the flexability and cost savings of Linux sort of fly out the window when you're bound to a service agreement. And even the IBM Linux offering is uninteresting when you look at the value and flexability provided by other vendors, unless of course you want to run Linux on z/Series mainframes, but who exactly is going to migrate their existing services to a mainframe running Linux? If you can afford to buy that machine you can't afford to run it on Linux; it makes as much sense as running Windows on Unisys. Sun stands alone. Even if you've never considered Sun you must admit that the choice is difficult to beat. Solaris is the most trusted UNIX platform in the world, on the widest range of systems from a single vendor, including Intel Xeon, AMD Opteron, Sun UltraSparc, and soon Niagra You get the choice of Solaris, or Linux, or now even Windows. You get the choice of small edge systems and massive high performance servers. You even have the choice to run Solaris on non-Sun boxes, or to buy Fujitsu PrimePower from Sun. You have a choice of partners, Oracle, BEA, SAP, and more. And shortly, for the first time, you even have a tested and tried enterprise operating system that you can rip open. And the choices from Sun, are always getting better through superior engineering and innovation in chip design, in system design, in operating system design, and even the design of the company itself. Sun's got all the keys..... like it or not, when you can buy a system running either Linux, Windows or Solaris on a system thats got more power and is cheaper, service included, than Dell? Thats just a no brainer.

But hey, people want to buy from HP over and over, or IBM over and over, not considering a possible move to Sun. Why? How can you not considering. You're only hurting yourself and your organization. Why not leave yourself open to choice, even if you don't want it? Because choice also means a choice to not choose to change, to stick with what you already like, but when you look at the direction of HP or IBM its hard to be certain you'll always have even that choice. Sun's management have held the wheel straight and true, evolving and growing with change, not just jumping on the bandwagon of the week, even though they get a lot of flack for it. I for one value my choice. I value my technology investments. I value the options I'm able to provide to my employeer and organization. Only Sun's got the value, the choice, and the proven track record to adapt to change in a safe, reliable, and controlled way.

As we shift into a new era at Sun, openning the doors to the Solaris operating system we continue down a long long road, started at Stanford Univerity Networks, based on BSD and commonidy hardware, bound together with what Sun does best: engineer a better solution for tomorow, today. OpenSolaris is just part of that journey, just as SPARC was, just as Java was. People were unsure then, even questioned the leadership of the company, but here we are today, on that same course, moving in the same direction, for the same purpose. And even a long haired guy at home waiting for his wife to come home from school can tell you that.

BUS	Clock	Width	Throughput
SBUS	25Mhz	64bit	200MB/s
PCI	33Mhz	64bit	264MB/s
PCI	66Mhz	64bit	528MB/s
cPCI	33Mhz	64bit	264MB/s
cPCI	66Mhz	64bit	528MB/s

PCI-X	66Mhz	64bit	528MB/s
PCI-X	133Mhz	64bit	1064MB/s
PCI-X	266Mhz	64bit	2128MB/s
PCI-X	533Mhz	64bit	4264MB/s

BUS	Lanes	Throughput
PCIe	x1	250MB/s
PCIe	x2	500MB/s
PCIe	x4	1GB/s
PCIe	x8	2GB/s
PCIe	x16	4GB/s
PCIe	x32	8GB/s

PCI certainly has a confusing array of options, being as the ones listed above are simply the popularly used varieties. A look at the PCI-Sig spec page reveals several PCI specs that you've never heard of and hopefully never will. Bottom line is that PCI-X is just conventional PCI on crack. PCI Express is the actual innovative new architecture (explore it more here) that will help push IO in the next generation as we approach 4GB and 10GB Fibre Channel. InifiniBand is still the dark horse though; when will we see it and how well will it catch on? No one seems to know at this point. The future seems pretty content on accepting PCIe 16x as the new replacement for the aging AGP (AGP 4x pushed data at 1GB/s and 8x at 2GB/s) providing double todays accepted throughputs with room to grow into 32lane PCIe when needed.

Being a storage guy, though, the best place to keep a handle on what is popular (read: useful) is to look at what the HBA makers are doing. Hence, Qlogic's HBA offerings by bus is very interesting. You'll notice the SANblade QLE2360 is a 2Gbps FC HBA and utilizes PCI Express x4, providing way more bandwidth than is actually needed - 5 times more. Odd bit is that most of the motherboards appearing provide one or more PCIe x16 slots for video (2 on the new SLI boards) and up to 2 PCIe x1 slots. AMD chipsets have been much slower to incorperate PCI Express, which is odd to me because thats the primary target for high performance gamers, but so be it. NVidia's nForce4 chipset has PCI Express finally and so we're seeing some progress. VIA supposedly already has a chipset with PCI Express capability but it's be slow to market. I have yet to see boards with PCIe x4 or x8, and I still haven't seen any non-HBA PCIe offerings besides video cards.

The future is always an uncertain thing. Will conventional PCI and PCI-X stick around for the next 3-5 years or is PCIe the way of the future on its own? Will see see a PCIe form factor similar to cPCI (AdvancedTCA) roll out? While PCI Express solves some bandwidth issues PCI still is a poor choice in the high end market because it's not inherently hot swapable, which is why I've always like Compact PCI options for servers like the SunFire 3800 and 4800. If we can migrate systems over the next 5 years to solely using AdvancedTCA with PCIe x16 capable slots we'd be in goooooooood shape.

I've been going through all sorts of diffrent articles on Sun; boning up for the coming storm and the next peice I'm doing on cuddletech, entitled "Sun's Strategy: WOW! or WTF?" that I'm hoping to complete soon.  It's digging me into Sun politics far more than I'd ever wanted to but perhaps thats unavoidable.  There's BusinessWeek's "Sun: A CEO's Last Stand" from July, Eric Raymond's recent Open Letter: "Java is not a bazaar", Eweek's "Critics Hit Sun Gear Giveaway" from the other day ago, Computerworld's "Sun's CEO pins future hopes on "old strategy"" from November, Bob Cringely's Predictions for 2005, and I've even been pouring over the latest Standard & Poors SUNW Report (thanks to my trading accounts research features) dated the 8th of January.  I now, more than ever, feel like I have a tough time trying to tell the execs of a $14billion corp how to run shop. :)