A Look at IRIX 6.5.17

XFS writes "OSNews got their hands on the latest version of IRIX, 6.5.17 (released in August), and they have published an interesting article about it and they explain why IRIX was and still is, one of the best workstation Unices out there. Especially when it comes to multimedia/GL performance. I hope SGI will do something with IRIX though, as they seem to have let it fall behind and be one of these great technologies that get lost through various corporate focus shifts..."

Comments

What about...

[Sc00ter]

What happened to SGI's big push for Linux?? looking back at past SGI stories there's a lot there about Linux work.

Re:What about...

[SanLouBlues]

Thus the poster's comment about hoping sgi doesn't let Irix die. The truth is even Solaris has caught up to Irix in terms of usability, 3d speed, and multimedia-ness unless you get a stack of Onyx machines. And the expense of an Onyx is what is driving Irix towards AIX in terms of market size.
And we all know IBM is even trying to migrate to Linux over AIX, so why shouldn't SGI with their fewer resources than IBM do the same. And they are.
But there are enough Irix diehards (Woot!) that SGI can't just drop it in one fell swoop. Plus SGI hardware does not go obsolete even a third as fast as PCs do.

Re:What about...

[duffbeer703]

That's funny.

AIX has more (and growing) marketshare than Irix has had in years. The Power4 and Power5 chip is attracting alot of business away from Sun.

SGI has been obsolete since 1996. A $2,500 Dell Workstation can do as much as a $25,000 SGI workstation.

Wake up McFly! It's 2002 calling!

Re:What about...

[Refried Beans]

The truth is even Solaris has caught up to Irix in terms of usability, 3d speed, and multimedia-ness unless you get a stack of Onyx machines.

What kind of crack are you smoking? Sun released a new graphics workstation last year that they said could compete with an Octane 2. At the same time SGI released Fuel which roughly doubles performance over Octane 2. At SigGraph Sun was showing an in-development high end graphics targetted at InfiniteReality3. SGI released at the show InfiniteReality4 and also backed it up with 128p (on the show floor). Sun has not caught up.

Oh my god

[Anonymous Coward]

And now we get all these posts from guys who know a bit of Linux who think they know *NIX because they post on slashdot.

Fallacy of benchmarks

[Anonymous Coward]

I think it is pretty interesting that the benchmark that they used measured memory throughput of the graphics CPU, as opposed to, say, an actual workload-handling of the OS. In other words, this is a synthetic benchmark, versus a real-world benchmark. They say, "Look! We can do memory transfers really really fast!"

Unfortunately, memory transfers are not the world when it comes to multiprocessor multimedia boxes. The overhead comes in when you're trying to synchronize a large number of threads/CPUs to do a large task. For example, an Oracle database.

Sun has proven that it scales up the tree very well with large numbers of processors. But from my understanding, Linux is more efficient with a low processor count, and less and less efficient with more processors.

I question its ability to do anything with a real workload. And I've even more suspicious because they use a benchmark I've never heard of to push its superiority on a single-aspect synthetic benchmark.

Re:Fallacy of benchmarks

[Mac Degger]

Uh, I do a bit of 3d, but only just enought to be intelligently stupid about it...but I don't use Oracle databases for my asset management...I couldn't care less what a database did on my CG box...memory transfers however is exactly what I would want to be fast.

And syncing large numbers of threads/cpu's, well, that is handy when doing a distributed render, but remember here: Irix is NOT Linux...just as linux is not unix :)

Anyway, I'd say they used the correct benchmark...especially for a workstation running a 3d app.

But as I said, IANA3DExpert...

Re:Fallacy of benchmarks

[fgodfrey]

Uh, SGI sells more than multimedia workstations. They sell 1024 processor supercomputers too and memory bandwidth *IS* the world in high performance computing. That's why supercomputers like high end SGI's and Cray's can still beat the crap out of a cluster on a number of applications. That said, Irix is also quite good at scaling up to large processor counts. If you can find me another example of a single kernel OS that scales to 1024 processors, I'd be quite shocked. The only SSI OS on that scale that I can think of is Unicos/mk (the Cray T3E operating system).

I do, however, agree that benchmarks are often quite useless. The way it any machine performs is highly dependant on what mix of jobs/applications you plan to run on it.

Hmmmmm

[IamTheRealMike]

One of the most innovating things about the IRIX in the '90s were the vector icons it uses for its desktop and file/icon managers. IRIX had vector support by default in its desktop long before MacOSX ever existed.

Not sure what this is implying, but it seems to be a surprisingly common misconception that MacOSX has vector based artwork. Not so. GNOME can do, and I think KDE3.1 can as well, via SVG. MacOS icons though are just bitmaps in a variety of sizes, with some scaling/blending algorithms applied.

The SGI desktop is of course based on a heavily modified commercial X Server. And here I will stop for a second, get a big breath and say: 'wow'. I have never seen an X server being so fast, on a 5-year old machine (no matter if this is an SGI machine or not).

I'd kind of expect this given that IRIX comes as a bundle with the hardware. When you choose the hardware as well as the software you can of course optimize the drivers a lot, so you will get good speeds out of it. XFree has to deal with a lot of different hardware, and the driver manufacturers are sometimes less than helpful. Probably worth remember that IRIX won't have some of the newer X extensions like XRender.

Re:Hmmmmm

[spitzak]

NeXT (I programmed these things in the 1980's) certainly supported drawing using PostScript paths, but all "icons" were tiff files. In fact they added a special tag to tiff to represent their 2-bit gray screen.

Until recently vector-based icons were way too slow. Except for the Irix ones, which were neceessarily quite simple. This simplicity did add to their appeal I think, though they never really put some good graphics designers on it.

KDE and Gnome and OS/X all render the vectors into pixmaps and then blast the pixmaps on the screen. OS/X certainly supports pixmaps and all the icons that appear to be airbrushed are bitmaps. They scale quickly and nicely because they have them carefully rendered at several resolutions and use mipmapping (the same technique your fancy graphics card uses for textures when it is in it's highest-quality mode) to scale.

Ignoring the drawing speed vector icons are much more efficient and take far less memory. In fact the earliest icons could be considered vector-based, they were drawn on vector screens by machines where 8K of memory was expensive.

Dual 195MHz machine?

[qurob]

SGI was very kind to send us in this dual Octane 2x195 Mhz MIPS machine accompanied with a 24" SGI-branded Trinitron monitor.

Otherwise, they would have had to shell out a whole $799.00 on eBay for one.

What did SGI do, pull one from the junk bin?

They should have sent some relatively modern hardware....

Re:Dual 195MHz machine?

[commodoresloat]

I think it's worth remembering that Mhz means different things over in SGI land.

Yeah - it means "megaherz" as opposed to "inches of penis length" as it means here in slashdot-land.

Do something with IRIX?

[Highly Motivated Ano]

What exactly do you want SGI to do with IRIX? Put it in a box, shrinkwrap it, and make it run your overclocked AMD chip-of-the-week? Probably won't happen.

Since IRIX 6.5, SGI has continued its promise to release quarterly updates. Each release introduces changes to the feature and maintenance stream.

I guess I'm confused as to what your hopes for IRIX are.

I haven't read the OSNEWS.com article yet, but I hope it isn't one of those "OS review" articles where they look at the installer and give it a rating.

-David

I'm sure they've heard this before, but...

[sam_handelman]

IRIX Machines are huge in scientific computing.

However, since SGI announced that they wouldn't support IRIX anymore, everyone has concluded that they need to shift over to Linux machines.

Most people I know buy Dell machines. The cost savings is actually less of a concern for scientists (although it is an issue,) than keeping up with the state of the art.

If SGI released their IRIX source code, that would do a lot to help them recover their scientific market share; scientists would pay the extra money for SGI hardware if they aren't worried that support for the OS is going to evaporate entirely, and a Linux distro with lots of SGI-specific code imported from IRIX ought to fit that bill nicely. I'm a biologist, though, so maybe I'm missing something.

Re:I'm sure they've heard this before, but...

[lovebyte]

True, SGI is huge in scientific computing. But about 3 years ago, many scientific computing groups decided to stop using SGI (because SGI MIPS where not so fast, were too expensive and SGI itself seemed to be about to die) and move to Compaq alpha chips (for many applications they were the fastest chips, and by far). Ah! The same people are now puzzled about what to do or have gone to using Linux clusters.

Re:I'm sure they've heard this before, but...

[MarkTina]

SGI have announced that they're not supporting IRIX ??!?!? Since when ? Please provide a link to your source of information. And why should any company release for free the source code of their product ? They've put many years of effort into IRIX why should they release it to the great unwashed masses ? All (!) SGI have to do is get a good entry level system back into their range of workstations and price it to attract. Linux is OK but it always lives in BETA or ALPHA versions of everything, people start projects then give up when it gets to the boring bits like documentation or bug fixing, this is where "proper" commercial software wins, it's supported by people whose only job is to support that bit of software and things are documented so that if the lead programmer gets run over by a bus the project can continue. Regards Mark

Re:I'm sure they've heard this before, but...

[mysticbob]

However, since SGI announced that they wouldn't support IRIX anymore, everyone has concluded that they need to shift over to Linux machines.

false. sgi does and will continue to support irix, virtually forever. period. ask them if you don't believe me. or, even better, back up your claim with a press release, web page on sgi's site, etc. you will not find either, anywhere.

Re:I'm sure they've heard this before, but...

[Anonymous Coward]

"IRIX Machines are huge in scientific computing."

Well, they used to be, but are fading fast. At least in my area (Space systems operations, simulation & analysis).

Why?
- Their floating point performance sucks.
- PC graphics performance is now good enough for most applications.
- They are hugely expensive

2 years ago I ported a large astrodynamics simulation from Irix to Linux. Had a brand new Octane and a brand new dual PIII 700MHz running Redhat. I don't remember what chips the Octane had, but they were the best available at the time. And the Octane had twice the RAM.

This very floating point intensive simulation ran about 3x faster on the PC which cost about 10x less ($3K vs. ~$30K).

My boss just asked me if we should buy a service agreement for one of our older Octanes. $5K/year. I told him no. If it dies, we'll spend half that to upgrade to a faster Linux box.

Don't get me wrong. If you need the kind of huge bandwidth and massive multi-processing that higher end SGI's can give you, there is no question that SGI rocks. But at the workstation level, forget it.

And, FWIW, IRIX is absolutely the least stable Unix I've ever worked with. And I've worked with a few.

If there only was...

[Antity]

If there only was some kind of free (as in beer, please don't stop reading, this is not supposed to start a flamewar :-)) IRIX distribution, maybe slimmed-down, so at least more people could get hands on it and actually try it out.

I mean: Many of us have read lots about IRIX, how it works neatly for graphics workstations etcpp., but how many of you actually were able to try it out?

IRIX could gain a huge boost in popularity if people could "try it at home" on cheap x86 hardware and then - maybe - convince people at work to buy it if it is ok for the job. Even a 30-day evaluation copy would be great.

IMHO, it was a great idea of Sun to give away SunOS/x86 for free for personal use. So I had the possibility of fiddling around with it at home and improve my work with Solaris at work.

Anyone out there providing ssh'd remote X access to an IRIX box so one could have a look?

Re:If there only was...

[Xpilot]

There'd need to an x86 IRIX first, and that would take some time and money to develop. And giving it away free simply to promote it doesn't really sound like a good strategy for SGI. If I remember correctly, SGI is facing hard times these days during the slump of the tech sector. Hollywood's growing usage of Linux clusters for rendering isn't helping either.

SGI machines are simply too expensive to be a commodity machine, and anyone who buys one already has some application or requirement that made the purchase necessary (not simply because he thought it was sort of fun to play with the free x86 version).

Re:If there only was...

[gmack]

"Anyone out there providing ssh'd remote X access to an IRIX box so one could have a look?"

Yeah any scientific intitution still running last year's IRIX.

Seriously though IRIX is very tied to the hardware and by the time they ported it to x86 they would have a mostly new OS.

SGI wants to go x86 but the best way for them seems to be just having their own engineers port the best IRIX features to Linux. SGI is still active on linux-kernel so they havn't changed their minds yet.

"The OS looks dated" - WTF?

[Gothmolly]

Recipe for technology article:
1 paragraph fluff
1 paragraph spin
1/2 cup FUD

Mix in HTML editor.
Publish.

Seriously, this article is light on details, and filled with inane comments like "the OS looks dated". While there were some good comments, half of the time it was gushing over the X server, or cheering over the fact that the author can run XMMS. What about performance? Applications? Hardware compatibility/expandibility? Talk to us about the box - does SGI/IRIX know about USB, for instance? FireWire?

Details please..

In other SGI news:

[Sn4xx0r]

They just hired Jon "maddog" Hall to 'help SGI to sponsor and encourage a community-based "Extreme Linux" movement around SGI's NUMAFlex architecture.'

X11 credit's where credit is due

[Diabolical]

This is a fine example that X11 is a good graphics-server.

Although alot of X11 bashing has been going on IRIX shows us that X11 is actually a very viable and capable graphics-server and certainly gives the finger to all those X11 implementations which have been done BAD.

So please, next time you go and blame your sucky graphics on X11, take a good look at the implementation of it in your system.

Irix: What Linux wants to be when it grows up...

[Anonymous Coward]

Simple fact: Irix is elegant. The 4DWM is far superior to almost anything that you can put on linux ( poor 5DWM project went away.. it was the salvation) For the end user 4DWM was simple, uncomplicated, and most of all fast.. I personally have several IRIX boxen that I wouldn't trade in for anything. The installation of free software http://freeware.sgi.com couldn't be made easier. recompiling the kernel ( one command) and the fact that when I stick in a second display sub-system it automatically sets it up for me. True the machines are starting to show wear and age, and for the most current hardware you are paing a fortune, but if you NEED a ferrari you pay for a ferrari, you don't take a Yugo ( PC) and stick a ferrari logo on it... There is the whole fact that IRIX is expensive, but you do pay for what you get. I can effectively use a R4000 or R4400 with the latest version of IRIX, and it runs just fine. Try that with a stock install of Linux on a machine that is 6 or 7 years old, you won't be happy.

Re:Mmmmkay...

[uncleFester]

Seriously, IRIX is about as stable as Windows 95. Have you ever used it?

I think it depends on your point of reference. When I worked as a student Unix lab luser, we just acquired a bunch of Indigos with 4.0.5f, and it seems fairly stable compared to what we were running in the lab before (HP diskless workstations (not even SWAP local.. don't ask)). If anything would happen, it typically would be a graphics hang curable with the Vulcan Death Grip.

The last job had a handful of systems (an Origin and a few Octanes): one Octane had semi-frequent graphics hangs but I think that was hardware on its way out. The other Octanes never had problems... and this was running typically pissey software (HKS Abaqus, PATRAN, ANSYS, ProE), either locally or over X/network (again, don't ask). These were systems running 6.5.x (all were at .15 when I was laid off).

My biggest beef with IRIX was the piss-poor default security. 4.0.x releases were HORRIBLE: things have become better in time but SGI's security for these boxes is still kinda shoddy at best. And with cheaper PC graphics hardware (or cheaper workstation-level graphics like HP's fx cards) SGI can no longer rest on their 'visualization' laurels.

.. but I do love the fact I own an Indigo2 now. :)

-fester

IRIX isn't what killed SGI for us.

[nutznboltz]

What killed SGI for us was their hideous treatment of customers. We had some SGI boxes with 10-Base T as web servers in the past. When we went to look at a 100-Base T card we discovered that all SGI wanted to do was to sell us new boxes. They priced then network card around $6000.

Re:IRIX isn't what killed SGI for us.

[bmajik]

Well, i dont blame them. The 100mbit cards for the machines that didn't ship with 10/100 were expensive and SHITTY. For instance, for the indigo2, you could get an EISA phobos ethernet card which is basically a 3c597 with a different rom. This was spendy and you would never saturate the 100mbit card and it would generate a crapload of interrupts (slam cpu utilization).

Similar story for SGI Indy/Challenge S. The 100mbit workstation cards were 3rd party, spendy, and not so hot.

you could get SGI labeled 100mbit cards for the VME boxes (Challenge DM, for isntance) that worked pretty well.

I Actually Found the Article Quite Bad

[RAMMS EIN]

I read the article, but didn't bother to submit a story to /. because I found the article quite bad. IRIX may be quite good, but the author is all enthusiastic about features that I wouldn't think great wins myself. OK, you might dismiss this as a difference in taste, but I still think the author could look around a little more.

``The Guest account has quite some privillages by default, I was even able to install software, for example some KDE libraries and applications, so it was good enough to keep me going.''
It sounds like the author applauds this. Think about it, though. Would you really have Natalie Netuser log in to your box and have her install her own software? Apart from the security issues (which might not be there...I don't know exactly what kind of software you can or can't install), I think you'd better order that new hard drive already.

``The great thing about IRIX is that a lot of open source applications have been ported over to the proprierty X11 of IRIX''
Right. So IRIX is great because it can run all those open source apps that were developed with Linux and BSD in mind? OK, this might make IRIX better than some other proprietary OSen, but that doesn't necessarily make it great.

``X just works''
Yes, and so it would on Linux if the OS came preinstalled and tailored to the machine you ordered. That's not a feature of IRIX, it's the logical result of writing software for specific hardware (which, IMHO, shouldn't be necessary - standards should take care of that).

``Because there is one IRIX, one company behind it, and very specific versions, there are virtually no dependancy problems. Installations just work.''
Because there is one RedHat Linux | Windows | Mac OS, one monopoly behind it, ...
I mean, this sort of Just Works (WOW) goes against flexibility and freedom of choice. I don't know about IRIX, but I know that RedHat's packaging system gets confused when you install software via other means. Windows is a disaster (install from _what_ source?), and Apple is getting it right with OS X. ports rules!

``The window manager included on IRIX is the 4Dwm, while the toolkit used is the king of the Unix toolkits, Motif.''
I don't like 4Dwm, but I can see why others would. But Motif the king of Unix toolkits? Come on, speak for yourself, man. I don't even have Motif installed. All apps I use are either console or GTK, and there are a number of apps that would be cool to have, but not really worth installing Qt for. Motif _was_ king, yes, but it's reign is over.

---
Timeout error: Operator fell asleep while waiting for NT to complete boot sequence

Irix is painful and unpleasant

[Montressor]

IRIX might have enterprise-level performance, but it suffers enormously in the usability department. I work in a lab where IRIX is standard, because 64-bit memory addressing and extreme graphics performance using ImageVision is a must. However, I keep running into issues with the development tools. Most impotantly, SGI's cc (c compiler) is slow and hard to customize flags on, especially for debugging. Furthermore, frequently, if my program commits a memory fault, it receives a SIGKILL rather than a segfault which makes it very difficult to debug (this usually happens if the malloc pool gets corrupted or while using ImageVision).

The ImageVision library (an OpenGL-based image processing system) hsa great performance and features. However, it refuses to link with programs not built with cc (thus, no gcc!). Furthermore, programs that seem to follow spec mysteriously die with a SIGKILL during deallocation. I certainly realise that I might be doing something wrong in the way I call the library, but it does not provide any error
message, exception, or fault.

Finally, IRIX standard header files are a colossal mess and almost impossible to use. Standard C and C++ objects are casually redefined throughout the header structure.

No more Eugenia, please

[BurritoWarrior]

All of her (his? its?) articles are completely myopic, focusing on "how nice/useable" is this as a desktop.

Evaluating Irix on its UI alone is completely useless.

Re:No more Eugenia, please

[BurritoWarrior]

I see. So, if I have an interest in farming, it would be OK (and somehow useful) for me to do a review on whether or not a Corvette makes a good tool for plowing.

What is the logic for evaluating the desktop/gui of something that isn't designed for such, and therefore, logically, does not have much emphasis or time dedicated to it?

In order for a review of something to be useful, it should focus on what the tool is designed to do, not what your "interest" is.

usb and firewire

[green pizza]

The newest SGIs support USB and a few (maybe just O3K?) have firewire. On the big iron, USB has been a *really* nice upgrade from PS/2, especially on machines configured with multiple graphics pipes and multiple users. Rather than installing extra BaseIO modules for additional PS/2 hookups for additional users, you can now just plug in as many keys/mice as you'd like, bind each set to a certain number of graphics pipes. Helps us keep our Onyx 3800 flexible... most of the time it's running each of its three graphics pipes seperately... we have a config that'll drive three sets of keys/mice for three users, one graphics pipe driving two monitors per user. But when we need the power, we have one user driving all three pipes on a single multi-projector panoram screen. It's not totally plug and play, but it's a lot easier than it used to be.

SGI still rules on installations

[zrk]

I preface this with the fact that I haven't worked with Solaris 9 yet.

If you want to do a custom installation of the OS, Sun's OS installer for version 8 IS HORRIBLE. I changed jobs three years ago, and I moved from a predominantly SGI environment to one that's predominantly Solaris.

The OS installation tools for IRIX three years ago still are BETTER than what Sun currently offers up in Solaris 8 TODAY. What's there is putrid and annoying. The IRIX installation allowed you to select packages based on package names and wildcarding, whereas Sun uses a number-based scheme, and the numbers change from monthly release to monthly release. IT SUCKS!!!!!!

We do a lot of customized OS installations (the most minimal core OS, plus some additional packages) for security reasons, and we don't have enough common system types to make the Solaris automated installation worthwhile.

I wish Solaris would join the 21st century in this regard.

Re:Why is SGI not switching to FreeBSD ?

[fgodfrey]

How on earth did this comment get moderated "insightful"?!?! I'm not really sure where to start here:

crash-resistant, high-performance file system. Ever heard of "XFS"??? It's journaled and has been around almost longer than the FreeBSD project.

First multithreaded kernel: Um.... Right... Multithreaded kernels have been around for probably a decade if not more. FreeBSD is hardly the first. Irix has had kernel threads for ages. The first reference I can find to them is in '95 (and I suspect they have been around longer than that) when FreeBSD didn't even run on multiprocessor systems.

First "compact" kernel: What is a "compact" kernel? The FreeBSD kernel is a monolithic BSD kernel. Irix is a monolithic System V kernel. Even Linux is a monolithic kernel (of Linus + other's design). Microkernels haven't lived up to their initial hype (though MacOS X uses one), but neither they nor monolithic kernels are "obsolete".

Now don't get me wrong, FreeBSD is a great OS. I have run it in the past and regularly use it. But it doesn't run on 1024 processors, have multiple tens of terabytes of storage in a single filesystem, and manage a terabyte of RAM. It's not designed for that. Irix is.

IRIX is SVR4, not BSD based.

[cant_get_a_good_nick]

Minor quibble, but it mentions it's different than Linux because it's UCB BSD based. It's not (well, it is up to teh point that SVR4 took a lot of BSDisms) but it's SVR4 machine. Linux distros generally take a bit from classic BSD, a bit from SVR4, and a bit of whatevehell else they want, so they're all a bit different.

sinking ship

[JDizzy]

SGI has lost all their jedi developers, and as Eugenia mentions how Dominic Giampaolo went from SGI, to Be. Dominic now works for Google, as of 2 years ago. But that isn't the only person who has jumped ship at SGI, countless other faces, and names no longer reside at the old 'cool place to work'. The stock price was a penny-stock for awhile. Their hardware bussiness is lagging behind, and basically not moving forward. There has not been any new inovations from SGI in the past 3 or 4 years. In short, the industry finally caught up with SGI, and they have lost their graphics nitche. What Eugina doesnt' know, among many many many things, is that Sun is filling the void that was left open by SGI.

Anyways, his review of Irix is a version that is now old. It would be one thing if this was a hot new release of Irix 6.6 (non-existent), or some major point release. This version is a maintance release, and is old now. We have version 6.5.18, as of a week ago or so.... SGI sent me the new disc's. Also, booting the system up, and playing with the OpenGL crap is lame. Sorry, but it might impress an idiot, or people who have never played quakeII. Mozilla is not that impressive, and neither is the port of KDE or Gnome for Irix. These are things that, well, are not very interesting considering these tools run on just about all *nix variants now. Only a noobie would think otherwise. His bench mark analysis is to simply say "x seems to run faster on this old box", with no numbers or anything. Basically there is no relevance to his claims.

What I see is a guy who got a new toy to write about, and is all wet behind the ears. I use SGI computers evry day, and they are not all that! I have everything from O2, indiego, to bing honking 12-way Onyx clusters.

So let me explain what is nice about IRIX, for somebody that really does use it, and isn't still inthe first day experience level. Think about it, when you first tried Linux, or FreeBSd for the first time, as in never touched *nix before. If red hat was your first distro, say aorund version 7.*... your review might look something like Eugina's: noobie'ish! Sure, you can click'ity'clickty around the menus, launch softwareyou have never seen beofre.... ohh... ahhh...ooo.... wow! Whatever! The good thing about Irix is the fact that evrything is doable with a pretty gui tool. It was apparent from the early stages in Irix that people at that company were tired of the command line. For example, their package manager (aka software install tool) swmgr is fully graphical, and probably the best software installer for Unix there is, hands down. The upside is they also have another just-as-good version of the tool for the commandline. Sun could take some hints from what SGI has done inteh swmgr tool. For example, it has pie charts of filesystem utilization, with colors that represent what the other softwre packs take up, and what it would take to isntall this new peice. Everything look perfect. On the other hand, the X window system in general is lacking, the toolchest is gay. Lets face it, the SGI default desktop is kinda bleak, and empty. Maybe I'm a bit too used to CDE, KDE, or whatever.... but the first thing everyone I know does is install KDE to get some real work done. The day to day work of a developer wis what makesIrix nice to work on. The diff tool highlites the changes in files in an inteligent way, the ps program is graphical, or not, and is easy to spot problems with. The NIS, NFS, AFS work with gui tools to make things easy, yet all these tools could be used in a command line only mode. For those subterainian-commandlien dwealers, your still taken care of, and nicely too.

XSF is not like BFS, no matter how much Eugina want to think they are the same. They are not! It is true that XFS is more unixlike where bfs was more Be like. Both are 64 bit namespace safe, both have extensible attributes, but on XFS you have to really work hard to mess with these features. This is one area SGI needs to improve. The tools that ship with IBM's JFS are the best, but the features of XFS are probably better than JFS. Basically put, SGI XFS just works, without much tweaking. Or if you want, you can mess with the XFS. These days SGI is getting out of the graphics biz, and moving into the storage server biz. So maybe they will improve the XFS options/administration.