Personally, I avoid anime dubs, and for lots of reasons, but I don't mind other people seeing them. Besides, I know enough Japanese that I can nearly completely understand most anime raw - I saw Mononoke Hime 'raw' before I saw a sub, and 'got' most of it - so I don't need dubs or subs much. Might be interesting to see a good dub for once though...
Some little points I thought I'd share with the rest: *) The movie came out in Japan 2 years ago. *) Ashitaka comes from a race known as the 'ainu' (I think it was) who (in real history) died out somewhere between the 5th and 10th centuary or something. Originally, "shogun" used to be short for something that meant something like "army leader and suppresor of the Ainu". (I don't have some of my history books around, so I'm just doing it from memory) *) "mononoke" means something like a "vengeful ghost/spirit" - doesn't translate very well. "hime" is a name suffix which normally translates as 'princess', though in some periods it can be a bit more like 'lady'. *) The "wolves" are "mountain dogs" in the original *) In Japan, firearms were removed from the whole country by the government in the early 17th centuary. They were originally introduced by Portugease missionaries, who offered them to local lords in exchange for the lord and people going christian. *) I like this film a lot, though that doesn't mean it'll be like that for everyone.
IDC doesn't make their surveys available for free, so you have to rely on the reporting of the people who do buy them. Bit of a shame. Still, I thought some people might be interested in a Sun press-release which quotes figures from a Q2 1999 server survey.
btw, these IDC general servey surveys divide into 3 groups - entry level ($1,000,000). It also splits into Unix and general server. For Sun, "entry level" equates to its E450 and below, mid-range = Ex500 range, and high-end = Starfire. Over the last year Sun has more than doubled it's shipments of Starfires! (They recently finished building a new factory for them) For almost all of Sun's products, demand is exceeding supply - Sun has nearly $1billion in unfulfilled orders at the moment! It's also continuing to grow pretty steadily at 20-25% per year, and is still getting that despite it being some time since it introduced new hardware - their nice UltraSparc-III is rather late.
The only big computer hardware company that is growing faster than Sun, is Dell. However, Sun is pure-Unix and Dell is pure-PC, while the other biggers (HP, IBM and Compaq) are all mixed...
I haven't seen this article posted on Slashdot yet, so I thought I'd mention it. However, I've been off for a few days, so might have missed it.
Anyway, according to this InfoWorld article, "Sun will release" a Java 1.2 port for Linux in "early 2000" with the Blackdown porting group. Not big on details, and Sun haven't done a press release. It's not clear if Sun will be making it available to download from it's site, but that's the implication. However, the article also states that this release will include Sun's nice HotSpot compiler.
While it's not going to win any awards for bangs-for-bucks, Pixar are using a bunch of nice Sun hardware for their render-farm - 120 E4500s and 4.5 terabytes of disc storage, totaling 1680 CPUs. List price has lots of zeros.... Sun press release.
Just breaking up the company doesn't do a great deal by itself, I'd say, though that does depend on how and where you break them up. I think they should also be forced to openly and publically publish all their APIs, specs, development information, etc well in advance of any product release. MS currently use them to undermine their competition - force others to have over IP (interlectual property) or similar in return, threaten to or delay giving out information etc etc. Also, the price they charge OEMs for Windows should be based only on volume and other 'normal' things, rather than doing it all secretly. For instance, MS threatened to make OEMs pay $5-10 more unless they include IE (pre Windows 98 days), or if they do include any 'competitors' (ie Netscape). They also really screwed IBM over the price for Windows 95 because IBM wanted to continue selling OS/2 and Windows 3.1
As a little aside, Scott McNealy (Sun CEO) said that he didn't think MS should be broken up - he said it could be like those horror movies where you cut the monster into bits each of which turn into a new monster ^-^. Also, he said that breaking up should only really apply when a company has a total monopoly - there is still competition remaining. Last I heard, Larry Ellison (Oracle CEO) was in favour of breaking MS up...
One problem facing the DoJ is that most of their real options involves something that would cause MS's stock price to collapse. Since the US's stock prices have built up to a massive bubble, this could trigger the bubble to burst. However, that bubble is going to burst sooner or later, and the sooner the better really, though it'd be better if it was gentler...
And finally, I can't remember the words exactly, but in the DoJ's proposed Findings of Fact, they said (with regards to MS's attack on Netscape) "First they said they did not shoot the victim, then that everybody shot the victim, then that the victim wasn't harmed" ^-^
Btw, for running Windows and other 'PC' stuff, you get Sun's PCi-card, which has an AMD K6-2 on it.
The article were simply pointing out that Sun is offering this low-end unix workstation at, wait for it, the market for low-end workstations. Sun never target any of their computers at the home market!
Also, why assume Sun are targeting Linux with this? Is that the only possible reason Sun could have for reducing the price!? Come on! PC prices, for the same specs, drop all the time, and Sun just move a bit less frequently.
Interesting times up ahead for CPUs... Sun's UltraSparc-III should be selling by December, and looks pretty damn speedy. More and faster Alpha's coming. Merced is just a test/development platform btw, and won't be that great anyway - the IA-64 design itself has some designed-in limitations, and the Merced design is already a bit of a hack. (anyone want details?) btw, I was reading up some interesting info about Sun's MAJC chip, which is aimed at embedded designs with high-speed data processing, is in a couple of major ways it's actually quite like the IA-64 design, except it has a bunch of other extra spiffy things to make it faster. (want info...?)
The developers and managers (and a 3rd party name consultancy firm Sun hired) came up with a top 10 list... Then Sun's lawyers picked the one that'd be easiest to trademark from that list.
No, completely wrong, Sun is definately not a services company.
85% of Sun's revenue comes from hardware, the rest from software and services. Sun's gross margins on hardware are also about double that on software/services. ie they make about 90-95% of their profit from hardware.
(however, their fastest growing devision is the Enterprise Services division - ie supporting Starfire systems... But that's from a small start)
I'd like to make some points: *) Sun's SCSL license is not open source, nor trying to be.
*) Not all the source code will be made available straight away. Like with their other stuff released under the SCSL, you'll only have to pay Sun money if you plan to make a profit on it yourself. Sun also require you to fully publish the specs for any new APIs you add, and are very hard on compatability - ie they don't want it corrupted/code forked. btw, you will need Sun's compilers to compile it, and they haven't (yet) made the compilers SCSL.
*) I don't think this is really aimed at the general public (of coders) - it's mostly aimed at commercial compains who currently want to liscence Solaris to make their own products, and there are a few, and also at developers who already use Solaris.
*) I think the two main advantages they hope to gain from this is more developer interest, and better quality software - by getting better feedback. The reliability of Solaris is very important to Sun - there are managers whose salary and bonus are tied to it's reliability.
*) This is part of Sun's relatively new, general attitude towards development - first with Java last year, some of their microprocessor designs, some high end software (Sun Cluster Tools 3), Star Office, new software (Jini and Jiro) and now Solaris. I've heard it said that Sun plan to make all their software available under the community source unless there is a good reason not to.
*) Sun have very good reasons to worry about protecting their software - Microsoft would love to damage Sun, like it did Netscape.
*) The descision to go with their "community source" lisence would not be new. Sun have very long lead times on development for Solaris, and because Solaris contains quite a lot of other people's IP (which they'd have to get a new lisence for, or do their own version) as well as tidy up the source for public release, they would have to make the descision very early in the product cycle. Solaris 8 (which will be the first to have some source code freed) went into alpha about 1 year ago, and has probably been in code freeze for about 3-6 months, and it'll be released in about 4-6 months. And Sun aren't even adding that many new features with Solaris 8. Co-incidentally (probably not), when Solaris first went into alpha, was about the time I first head Sun execs talking about making Solaris open source.
*) Some people have said that Sun might be worried about Linux, or Project Montetery wiping them out or something. Currently, as with the past few years, Sun has been having very stable and reliable growth (20-25% per year) and I haven't seen the slightest indication that they're "hurting" from Linux (the opposite in fact) and Monterey isn't even finished yet, and even at best won't take off for another 2 years.
Btw, what's new in Solaris 8 you might ask? Well, they're putting in their cluster tools software as standard (currently a seperate product), doing IPv6, including perl (perl 5.0005_03 to be precise) as well as some modules to access parts of Solaris, some bits for Java, new diagnostics tools and such. btw, Sun will support everything that comes with Solaris for 5 years after they stop selling it.
*) Just because you found someone several years ago, who thinks that Sun would be worse than MS, doesn't mean it's true. Even if Sun gained the same position MS is in, I doubt they'd act the same - MS has been acting like they are now for pretty much their entire history. For example, I've never heard of Sun trying to co-opt other people's standards, but MS does it all the time. Sun does good backwards compatability on all it's products - compile something for Solaris 2.5, and it'll almost certainly work fine on Solaris 2.7, or 2.8 when it comes out. MS mostly uses the cheapest programmers it can get, unlike Sun - and Solaris is far far more stable than any MS product. Sun's standard policy for anything that comes with Solaris is that they will continue to support for 5 years after they stop selling it - they're still supporting Solaris 2.3, unlike MS which tends to drop support for older versions the minute something newer is available. If you've followed the details of the DoJ vs MS trial in detail, like I have, you'll get a pretty clear idea of what lengths MS will go to force people to do what they want. The biggest reason my MS execs are so agressive is because of the share options - by working themselves to the bone to get raises and by improving MS's outlook, they can make loads of money. While Sun do offer share options, they mostly offer a good salary and a good working environment. btw, some Sun managers pay is tied to the reliability of Sun hardware/software solutions.
*) Sun aren't stopping development for StarOffice - they are keeping all 200 developers. If they wanted to kill off StarOffice at a later date, then why the heck would they release the source code and keep the developers?
*) Unless MS just dies very quickly, Sun would have very little to gain by suddenly dropping StarOffice - that would be a gift to MS. The 'StarPortal' thing with the Java client needs a server - so it'll be very hard to make a home user to switch. Corportations would be a bit different, but they're hardly likely to drop their general purpose PCs for a pure Sun solution just because Sun drop StarOffice. The StarPortal thing is more new markets, not current ones.
*) The SunRay 1 is a very focus solution, aimed at things like call centers, or where you would currently have terminals. It is absolutely not in any way at all a general solution or intended to be a replacement for all PCs. (just the ones that are doing simple basic things) Sun currently only supply Java for Solaris and Windows - if they drop StarOffice for the Java-only version (which might well require Java 1.3 which has just gone into beta), that helps MS, unless Sun massively increase the support they directly provide for Java on other OSs.
*) Wabi itself was dropped by Sun 2 years ago. It seems Caldera were doing 'Wabi for Linux' and it was them that dropped it in March this year. (I'm not sure of some of the details)
*) Sun is a hardware company - 85% of their revenue comes from hardware, and they have very good, and high end server solutions. It's hardly surprising that they'll play to their strengths. It would also be unreasonable to expect Sun to give away all their technology and software to Linux, as they only make a profit on their software stuff because it sells more hardware.
*) I've seen lots of people say that Sun is hurting because of Linux. Actually it's the other way around - Linux is growing and helping the general unix market, which helps Sun. Finantially, Sun is doing just as well as they've always been doing (20-25% growth per year, 2nd only to Dell for a large computer manufacturer) and I haven't seen any indications to suggest that they are being hurt by Linux. Scott McNealy actually said recently that he doesn't want Sun to grow more than 25% per year - can't hire good people fast enough to grow faster reliable. Besides, though Sun is somewhat vulerable in terms of hardware sales for boxes for software development, or web servers and the like, there is little
*) As for the liscense - I'm pretty damn sure it prevents you from say giving a copy of the binary to a friend, or using one download to install it on a load of machines in an office. Also, (though it's not officially confirmed yet I think), AOL will be distrubuting StarOffice on their standard CD-ROM, and Compaq and other PC manufacturers will be pre-installing it on some of their computers.
They could have mentioned that an ex Sun guy also provided funding for the company. (ex-Microsoft guy, ex-Sun guy, and Linus... getting along together? We could learn something from these guys...)
Anyway, with regards to the speculated ability to execute binaries for x86, PowerPC etc, that's sort of similar to some of the things Sun's MAJC chip will do, which has also been in development for a long time. The MAJC chip doesn't literally execute Java byte-code but it's easier to translate to than other architectures - you still need a JVM. You can just run C/C++ binaries normally too - so yes, you can port Linux to it. However, what this means is that you don't have to support old versions and you can optimise best for each generation.
Here is a transcript of a live chat with the MAJC architecture lead designer, and they discuss some interesting things. (yes, someone does ask how easy it would be to do a Linux port ^-^)
Actually, the next generation Starfire will have a COMA architecture. (might have ccNUMA too though... not sure yet). See Project Serengheti, though products won't be out for at least another year.
"Project Sun Screen" is IBMs name for their plan. For commercial unix, Sun have been pretty much stealing the show for quite a while now. I think HP are the only other large company to have better than a small increase in revenue recently in commercial Unix. HP have also been targeting Sun more recently, as are others - if you are the market leader, then you get targeted the most.
I don't know how IBM are comparing their performance to Sun's Starfire. Pretty meaningless without giving any details. btw the Starfire is over 2 years old now. I don't think Sun are yet officially supporting their 450Mhz UltraSparc-II in volume on Starfire yet, even though it's been out for a while. (btw, you can get US-II 450's with 8MByte of 2nd level cache - clocked at 450Mhz!) Also, Sun's UltraSparc-III will be shipping in volume this December, starting at 600Mhz, and from early SPEC 95 benchmarks I've heard of it's about 10% faster (in fp) than an 600Mhz EV67 (Alpha 21264A) and they haven't even finished optimising it yet. SPEC int should be very good too.
Latest SPEC results here - 600Mhz Athlon has SPEC int/fp of 27.2/21.6. 667Mhz EV67 (Alpha 21264A) has SPEC int/fp of 37.5/65.5. The more competition, the better! (that includes competition between OS's) .
See info on Intel's web site.. About time too - the current StrongARM hasn't been updated much in 3 years. The new one (to be produced on a 0.18micron fab) can run up to 600Mhz (while consuming just 0.45W), while in low power mode it consumes just 0.04W while running at 150Mhz.
For all users, not just business users, AOL (who are kinda close to Sun these days) plan to distribute it as standard on all their CDs. Compaq are apparantly going to pre-installed it on some of their PCs, and others might follow.
Also, Sun will be launching their next-gen "JavaStation" in about 5-6 days, I believe.
You can download StarOffice 5.1 for free here - support for Solaris SPARC/Intel, Linux, OS/2, Windows 95/98/NT, and several languages. Unfortunately, you do have to register, but if you've already got a username/password for say the Java (or Solaris) Developers Connection then you can use that.
And finally, Sun will be making available Solaris 8 early access - ie open beta to the public, for the first time for a new version of Solaris. You'll probably see the source code available later...
Yes, I know Sun isn't the only one with this approach - I said that in my post.
I have no idea if Sun plan to do redundancy checking with multiple pipelines with the UltraSparc-V. (some IBM, and other, chips do this...) They might do it as an option, but I would currently guess they're doing it mostly for performance.
EBay's reliability problems are mostly related to poor management decisions (it seems) rather than EBay's (or Sun's or Oracle's) techs. Doing the above kind of checking wouldn't have helped EBay either. It doesn't matter what OS you use, if you have a screwed up setup, you'll get problems. And you'll be surprised/horrified at just how long it can take screwed setups to be fixed if the site's already gone live. (I know from experience. and no, it wasn't my screwed up setup.)
They haven't really released enough details (on their website) just yet, but it does look interesting. One of the more obviously different attitudes the specification takes is highly customisable implimentations - you design a variation targeted at a particular application, whatever that might be - graphics accelerator, MP3 player/decoder, MPEG2/DVD decoder, or a more general purpose chip. Since it is mostly being targeted at embedded applications this is not surprising though.
Some other interesting aspects include:
'Support' for JIT/access-time compilers - not only does this help Java, but it is to make backwards compatability with older versions quite simple. This seems a bit like what Transmeta are doing, which was co-founded by an ex Sun guy btw.
Hardware support for ultra-fast thread switching - so fast that if one thread stalls waiting for DRAM access (which can take up to 100 clock cycles), you can switch to another thread rather than go idle. On many current OSs threads will be switched if the current one has to do some slow I/O say (ie read from disc) - so this is quite an improvement.
A more general approach to improving parallelism - you can have more than one CPU core in a single physical chip, which might or might not share their 1st level caches. (read this Microprocessor Report article for some background on this.) IBM are apparantly going to do a version of the PowerPC G4 which has 2 CPUs on one core, and I kinda suspect Sun might be planning something similar for their UltraSparc-V.
I'm not sure how Sun plan to make money of the design. It seems pretty likely they might do something like their "community source" model - you can get the design for free, but if you want to use it commercially you pay a license. ARM is doing well just licensing their CPU designs. I'd image Sun using to 'assist' their servers as add-on boards for doing heavy multi-media/3D graphics stuff - can you say "render farm"? Also, since Sun like selling their servers, they'd be happy for people to make lots of little, cheap devices that connect to nice big Sun servers.
Like the original poster said, IEEE Micro will probably have some interesting stuff, but it seems Sun aren't releasing all the details yet - looks like we'll have to wait until the Microprocessor Forum in October. I liked the article (written by the Sun engineers) about the UltraSparc-III - not only was it interesting (and I like Sun's approach) , it helped me figure out the inherant problem with the IA-64 architecture...
This is a LONG article from SunWorld mag, and you have to go down a fair bit to get to the info about file journaling, and such, but it's a good read: Getting to know the Solaris filesystem, Part 1
Selling boxes for 'portals' and internet companies in general is seen a a 'growth' market - and some (like HP) are only now just starting to target this area. However, Sun really do have both the mindshare and marketshare for high-end internet-related stuff. SGI, which isn't doing very well at the moment (it made a 'surprise' profit in the last quarter, when it has been making losses recently) will have to work really hard to be able to make a good grab for this market - on the high end they'll be mostly up against Sun (on their home turf too) while on the low end they'll be against a pretty crowded market, and jumping on the Linux bandwagon won't help a great deal here.
Shame really - SGI have some really sweet technology. Been rather letdown by management lately it seems. I know some people who are kinda traditional buyers of high-end (ie Origin 2000s) SGI kit and they aren't too impressed with SGI's moves recently.
If SGI are going to drop/lose/sell some of their high-end stuff, I wonder who's going to buy... Sun bought the physical design for the Starfire about 3 years ago for $100M. Now they're making about as much revenue off the Starfire (including attached storage and services) as SGI are making in total...
Some little points I thought I'd share with the rest:
*) The movie came out in Japan 2 years ago.
*) Ashitaka comes from a race known as the 'ainu' (I think it was) who (in real history) died out somewhere between the 5th and 10th centuary or something. Originally, "shogun" used to be short for something that meant something like "army leader and suppresor of the Ainu". (I don't have some of my history books around, so I'm just doing it from memory)
*) "mononoke" means something like a "vengeful ghost/spirit" - doesn't translate very well. "hime" is a name suffix which normally translates as 'princess', though in some periods it can be a bit more like 'lady'.
*) The "wolves" are "mountain dogs" in the original
*) In Japan, firearms were removed from the whole country by the government in the early 17th centuary. They were originally introduced by Portugease missionaries, who offered them to local lords in exchange for the lord and people going christian.
*) I like this film a lot, though that doesn't mean it'll be like that for everyone.
btw, these IDC general servey surveys divide into 3 groups - entry level ($1,000,000). It also splits into Unix and general server. For Sun, "entry level" equates to its E450 and below, mid-range = Ex500 range, and high-end = Starfire. Over the last year Sun has more than doubled it's shipments of Starfires! (They recently finished building a new factory for them) For almost all of Sun's products, demand is exceeding supply - Sun has nearly $1billion in unfulfilled orders at the moment! It's also continuing to grow pretty steadily at 20-25% per year, and is still getting that despite it being some time since it introduced new hardware - their nice UltraSparc-III is rather late.
The only big computer hardware company that is growing faster than Sun, is Dell. However, Sun is pure-Unix and Dell is pure-PC, while the other biggers (HP, IBM and Compaq) are all mixed...
Anyway, according to this InfoWorld article, "Sun will release" a Java 1.2 port for Linux in "early 2000" with the Blackdown porting group. Not big on details, and Sun haven't done a press release. It's not clear if Sun will be making it available to download from it's site, but that's the implication. However, the article also states that this release will include Sun's nice HotSpot compiler.
While it's not going to win any awards for bangs-for-bucks, Pixar are using a bunch of nice Sun hardware for their render-farm - 120 E4500s and 4.5 terabytes of disc storage, totaling 1680 CPUs. List price has lots of zeros.... Sun press release.
As a little aside, Scott McNealy (Sun CEO) said that he didn't think MS should be broken up - he said it could be like those horror movies where you cut the monster into bits each of which turn into a new monster ^-^. Also, he said that breaking up should only really apply when a company has a total monopoly - there is still competition remaining. Last I heard, Larry Ellison (Oracle CEO) was in favour of breaking MS up...
One problem facing the DoJ is that most of their real options involves something that would cause MS's stock price to collapse. Since the US's stock prices have built up to a massive bubble, this could trigger the bubble to burst. However, that bubble is going to burst sooner or later, and the sooner the better really, though it'd be better if it was gentler...
And finally, I can't remember the words exactly, but in the DoJ's proposed Findings of Fact, they said (with regards to MS's attack on Netscape) "First they said they did not shoot the victim, then that everybody shot the victim, then that the victim wasn't harmed" ^-^
Btw, for running Windows and other 'PC' stuff, you get Sun's PCi-card, which has an AMD K6-2 on it.
The article were simply pointing out that Sun is offering this low-end unix workstation at, wait for it, the market for low-end workstations. Sun never target any of their computers at the home market!
Also, why assume Sun are targeting Linux with this? Is that the only possible reason Sun could have for reducing the price!? Come on! PC prices, for the same specs, drop all the time, and Sun just move a bit less frequently.
Interesting times up ahead for CPUs... Sun's UltraSparc-III should be selling by December, and looks pretty damn speedy. More and faster Alpha's coming. Merced is just a test/development platform btw, and won't be that great anyway - the IA-64 design itself has some designed-in limitations, and the Merced design is already a bit of a hack. (anyone want details?) btw, I was reading up some interesting info about Sun's MAJC chip, which is aimed at embedded designs with high-speed data processing, is in a couple of major ways it's actually quite like the IA-64 design, except it has a bunch of other extra spiffy things to make it faster. (want info...?)
Scary... but sensible too.
85% of Sun's revenue comes from hardware, the rest from software and services. Sun's gross margins on hardware are also about double that on software/services. ie they make about 90-95% of their profit from hardware.
(however, their fastest growing devision is the Enterprise Services division - ie supporting Starfire systems... But that's from a small start)
I'd like to make some points:
*) Sun's SCSL license is not open source, nor trying to be.
*) Not all the source code will be made available straight away. Like with their other stuff released under the SCSL, you'll only have to pay Sun money if you plan to make a profit on it yourself. Sun also require you to fully publish the specs for any new APIs you add, and are very hard on compatability - ie they don't want it corrupted/code forked. btw, you will need Sun's compilers to compile it, and they haven't (yet) made the compilers SCSL.
*) I don't think this is really aimed at the general public (of coders) - it's mostly aimed at commercial compains who currently want to liscence Solaris to make their own products, and there are a few, and also at developers who already use Solaris.
*) I think the two main advantages they hope to gain from this is more developer interest, and better quality software - by getting better feedback. The reliability of Solaris is very important to Sun - there are managers whose salary and bonus are tied to it's reliability.
*) This is part of Sun's relatively new, general attitude towards development - first with Java last year, some of their microprocessor designs, some high end software (Sun Cluster Tools 3), Star Office, new software (Jini and Jiro) and now Solaris. I've heard it said that Sun plan to make all their software available under the community source unless there is a good reason not to.
*) Sun have very good reasons to worry about protecting their software - Microsoft would love to damage Sun, like it did Netscape.
*) The descision to go with their "community source" lisence would not be new. Sun have very long lead times on development for Solaris, and because Solaris contains quite a lot of other people's IP (which they'd have to get a new lisence for, or do their own version) as well as tidy up the source for public release, they would have to make the descision very early in the product cycle. Solaris 8 (which will be the first to have some source code freed) went into alpha about 1 year ago, and has probably been in code freeze for about 3-6 months, and it'll be released in about 4-6 months. And Sun aren't even adding that many new features with Solaris 8. Co-incidentally (probably not), when Solaris first went into alpha, was about the time I first head Sun execs talking about making Solaris open source.
*) Some people have said that Sun might be worried about Linux, or Project Montetery wiping them out or something. Currently, as with the past few years, Sun has been having very stable and reliable growth (20-25% per year) and I haven't seen the slightest indication that they're "hurting" from Linux (the opposite in fact) and Monterey isn't even finished yet, and even at best won't take off for another 2 years.
Btw, what's new in Solaris 8 you might ask? Well, they're putting in their cluster tools software as standard (currently a seperate product), doing IPv6, including perl (perl 5.0005_03 to be precise) as well as some modules to access parts of Solaris, some bits for Java, new diagnostics tools and such. btw, Sun will support everything that comes with Solaris for 5 years after they stop selling it.
On a somewhat different subject there's also this article: Nasdaq hacked through MS security hole
*) Sun aren't stopping development for StarOffice - they are keeping all 200 developers. If they wanted to kill off StarOffice at a later date, then why the heck would they release the source code and keep the developers?
*) Unless MS just dies very quickly, Sun would have very little to gain by suddenly dropping StarOffice - that would be a gift to MS. The 'StarPortal' thing with the Java client needs a server - so it'll be very hard to make a home user to switch. Corportations would be a bit different, but they're hardly likely to drop their general purpose PCs for a pure Sun solution just because Sun drop StarOffice. The StarPortal thing is more new markets, not current ones.
*) The SunRay 1 is a very focus solution, aimed at things like call centers, or where you would currently have terminals. It is absolutely not in any way at all a general solution or intended to be a replacement for all PCs. (just the ones that are doing simple basic things) Sun currently only supply Java for Solaris and Windows - if they drop StarOffice for the Java-only version (which might well require Java 1.3 which has just gone into beta), that helps MS, unless Sun massively increase the support they directly provide for Java on other OSs.
*) Wabi itself was dropped by Sun 2 years ago. It seems Caldera were doing 'Wabi for Linux' and it was them that dropped it in March this year. (I'm not sure of some of the details)
*) Sun is a hardware company - 85% of their revenue comes from hardware, and they have very good, and high end server solutions. It's hardly surprising that they'll play to their strengths. It would also be unreasonable to expect Sun to give away all their technology and software to Linux, as they only make a profit on their software stuff because it sells more hardware.
*) I've seen lots of people say that Sun is hurting because of Linux. Actually it's the other way around - Linux is growing and helping the general unix market, which helps Sun. Finantially, Sun is doing just as well as they've always been doing (20-25% growth per year, 2nd only to Dell for a large computer manufacturer) and I haven't seen any indications to suggest that they are being hurt by Linux. Scott McNealy actually said recently that he doesn't want Sun to grow more than 25% per year - can't hire good people fast enough to grow faster reliable. Besides, though Sun is somewhat vulerable in terms of hardware sales for boxes for software development, or web servers and the like, there is little
*) As for the liscense - I'm pretty damn sure it prevents you from say giving a copy of the binary to a friend, or using one download to install it on a load of machines in an office. Also, (though it's not officially confirmed yet I think), AOL will be distrubuting StarOffice on their standard CD-ROM, and Compaq and other PC manufacturers will be pre-installing it on some of their computers.
Well, that's enough for now...
Anyway, with regards to the speculated ability to execute binaries for x86, PowerPC etc, that's sort of similar to some of the things Sun's MAJC chip will do, which has also been in development for a long time. The MAJC chip doesn't literally execute Java byte-code but it's easier to translate to than other architectures - you still need a JVM. You can just run C/C++ binaries normally too - so yes, you can port Linux to it. However, what this means is that you don't have to support old versions and you can optimise best for each generation.
Here is a transcript of a live chat with the MAJC architecture lead designer, and they discuss some interesting things. (yes, someone does ask how easy it would be to do a Linux port ^-^)
Actually, the next generation Starfire will have a COMA architecture. (might have ccNUMA too though... not sure yet). See Project Serengheti, though products won't be out for at least another year.
I don't know how IBM are comparing their performance to Sun's Starfire. Pretty meaningless without giving any details. btw the Starfire is over 2 years old now. I don't think Sun are yet officially supporting their 450Mhz UltraSparc-II in volume on Starfire yet, even though it's been out for a while. (btw, you can get US-II 450's with 8MByte of 2nd level cache - clocked at 450Mhz!) Also, Sun's UltraSparc-III will be shipping in volume this December, starting at 600Mhz, and from early SPEC 95 benchmarks I've heard of it's about 10% faster (in fp) than an 600Mhz EV67 (Alpha 21264A) and they haven't even finished optimising it yet. SPEC int should be very good too.
Latest SPEC results here - 600Mhz Athlon has SPEC int/fp of 27.2/21.6. 667Mhz EV67 (Alpha 21264A) has SPEC int/fp of 37.5/65.5. The more competition, the better! (that includes competition between OS's) .
See info on Intel's web site.. About time too - the current StrongARM hasn't been updated much in 3 years. The new one (to be produced on a 0.18micron fab) can run up to 600Mhz (while consuming just 0.45W), while in low power mode it consumes just 0.04W while running at 150Mhz.
Also, Sun will be launching their next-gen "JavaStation" in about 5-6 days, I believe.
You can download StarOffice 5.1 for free here - support for Solaris SPARC/Intel, Linux, OS/2, Windows 95/98/NT, and several languages. Unfortunately, you do have to register, but if you've already got a username/password for say the Java (or Solaris) Developers Connection then you can use that.
And finally, Sun will be making available Solaris 8 early access - ie open beta to the public, for the first time for a new version of Solaris. You'll probably see the source code available later...
I have no idea if Sun plan to do redundancy checking with multiple pipelines with the UltraSparc-V. (some IBM, and other, chips do this...) They might do it as an option, but I would currently guess they're doing it mostly for performance.
EBay's reliability problems are mostly related to poor management decisions (it seems) rather than EBay's (or Sun's or Oracle's) techs. Doing the above kind of checking wouldn't have helped EBay either. It doesn't matter what OS you use, if you have a screwed up setup, you'll get problems. And you'll be surprised/horrified at just how long it can take screwed setups to be fixed if the site's already gone live. (I know from experience. and no, it wasn't my screwed up setup.)
MAJC home page . See the docs home page - introduction, and a "community" page .
They haven't really released enough details (on their website) just yet, but it does look interesting. One of the more obviously different attitudes the specification takes is highly customisable implimentations - you design a variation targeted at a particular application, whatever that might be - graphics accelerator, MP3 player/decoder, MPEG2/DVD decoder, or a more general purpose chip. Since it is mostly being targeted at embedded applications this is not surprising though.
Some other interesting aspects include:
'Support' for JIT/access-time compilers - not only does this help Java, but it is to make backwards compatability with older versions quite simple. This seems a bit like what Transmeta are doing, which was co-founded by an ex Sun guy btw.
Hardware support for ultra-fast thread switching - so fast that if one thread stalls waiting for DRAM access (which can take up to 100 clock cycles), you can switch to another thread rather than go idle. On many current OSs threads will be switched if the current one has to do some slow I/O say (ie read from disc) - so this is quite an improvement.
A more general approach to improving parallelism - you can have more than one CPU core in a single physical chip, which might or might not share their 1st level caches. (read this Microprocessor Report article for some background on this.) IBM are apparantly going to do a version of the PowerPC G4 which has 2 CPUs on one core, and I kinda suspect Sun might be planning something similar for their UltraSparc-V.
I'm not sure how Sun plan to make money of the design. It seems pretty likely they might do something like their "community source" model - you can get the design for free, but if you want to use it commercially you pay a license. ARM is doing well just licensing their CPU designs. I'd image Sun using to 'assist' their servers as add-on boards for doing heavy multi-media/3D graphics stuff - can you say "render farm"? Also, since Sun like selling their servers, they'd be happy for people to make lots of little, cheap devices that connect to nice big Sun servers.
Like the original poster said, IEEE Micro will probably have some interesting stuff, but it seems Sun aren't releasing all the details yet - looks like we'll have to wait until the Microprocessor Forum in October. I liked the article (written by the Sun engineers) about the UltraSparc-III - not only was it interesting (and I like Sun's approach) , it helped me figure out the inherant problem with the IA-64 architecture...
This is a LONG article from SunWorld mag, and you have to go down a fair bit to get to the info about file journaling, and such, but it's a good read:
Getting to know the Solaris filesystem, Part 1
Eclipse update: Apple PR stunts shocks World
Register article
news.com article
AP Business article on Yahoo
Shame really - SGI have some really sweet technology. Been rather letdown by management lately it seems. I know some people who are kinda traditional buyers of high-end (ie Origin 2000s) SGI kit and they aren't too impressed with SGI's moves recently.
If SGI are going to drop/lose/sell some of their high-end stuff, I wonder who's going to buy... Sun bought the physical design for the Starfire about 3 years ago for $100M. Now they're making about as much revenue off the Starfire (including attached storage and services) as SGI are making in total...
It's more like a general purpose CPU with some bits to help make it easier to speed up Java, and other languages.
See this JavaWorld article from a former Sun engineer. This is about Java from the embedded point of view.