|
Technology Computing, programming, science, electronics, telecommunications, etc. |
View Poll Results: What is the worst Operating System ever? | |||
HP-UX | 0 | 0% | |
HP-UX | 1 | 16.67% | |
HP-UX | 4 | 66.67% | |
PHUX (er, HP-UX) | 1 | 16.67% | |
Voters: 6. You may not vote on this poll |
|
Thread Tools | Display Modes |
02-08-2002, 02:26 PM | #16 |
Radical Centrist
Join Date: Jan 2001
Location: Cottage of Prussia
Posts: 31,423
|
IBM says it has invested $1B in putting Linux on its big iron.
It also says it has already recouped that investment. Everything's changing. |
02-08-2002, 07:39 PM | #17 | |
Punisher of Good Deeds
Join Date: Apr 2001
Posts: 183
|
Quote:
"And with the capacity for four 64-bit Linux processors, zSeries for Linux has the flexibility to grow with your business." If you re-read my post, I referred to systems with up to 4 processors. This is one of those small servers. Thanks for backing me up. |
|
02-08-2002, 07:48 PM | #18 | |
Punisher of Good Deeds
Join Date: Apr 2001
Posts: 183
|
Quote:
Thus, the 'one big server running really big processes doing really really big stuff' (somewhat simplified post still stands - I do not see Linux in five years being a true mainframe-level system. It may successfully scale to 8 processors, maybe 16. Since those will be 64bit processors, a lot of the work will have to be rethought; although much of Linux has been ported to 64bit architectures, the big iron Unices have a lot more experience under their belt on 64bit. I truly don't see Linux catching up to them in five years' time. It would be nice, of course, but it's not going to happen anytime soon. We need to abandon the techie mentality and look at it from the accountant's viewpoint. Reliability is what matters, and the track record of AIX/Solaris is considered to be too good. Whether that is true or not is another matter altogether. HP-UX is considered very highly as well, by the way - for systems that need to stay up and process a *LOT* of data. X. |
|
02-08-2002, 11:26 PM | #19 | |
Read? I only know how to write.
Join Date: Jan 2001
Posts: 11,933
|
Quote:
Basically, an IBM mainframe is really nothing more that an odd number of single chip computers assembled on a hybrid wafer sharing other support functions such as a dedicated encryption processor. Even more interesting, should any one processor fail, spare processors on the hybrid pick up the load without any external indication or failure. This is but one of many hardware features that makes IBM mainframes now so competitive to Sun's 'too many servers' solution. From the other side, Sun is being eaten alive by the PC world. Their position (based upon proprietary MIPS technology) would have been worse if Itanitum had not stumbled on a major architectural defect. Sun are in the same position of Germany in WWII Africa - with Montgomery on one side and Patton on the other. Which competitor is more dangerous? Previously, minicomputers would eat the low end market from the mainframe business. IBM will successfully reversed that trend at Sun's expense. Selling AMD based PCs is only a rear guard defense against the PC industry. Sun must confront IBM to survive - or discover new markets. Sun's position in not enviable. Currently they must do too much in house. Furthermore, Sun has yet to really decide - Solaris or Linux. That indecision only worsens Sun's current position. They will have to make a decsion soon or suffer. One great advantage for IBM is that their worse problem - incompatible hardware - has been solved by Linux. Going 100% to Linux was a no-brainer for IBM because of that hardware legacy created by MBA educated management John Akers and his predecessor. Linux solves a company wide problem created over 20 years ago by those MBA managers. |
|
02-11-2002, 02:23 PM | #20 |
Professor
Join Date: Jan 2001
Posts: 1,788
|
Sun is based on proprietary MIPS technology? That would be news to Sun Microelectronics, that's for sure.
Note that unlike Intel, Sun has been shipping a 64-bit processor for some time now.... As for Itanium, I remember when it was Merced. Untenable vaporware then, untenable vaporware now. |
02-11-2002, 09:57 PM | #21 |
Lecturer
Join Date: Jan 2001
Location: Carmel, Indiana
Posts: 761
|
just curious, what itanium defect now?
And by the way:
1. The Linux mainframes still run Z/OS and Z/VM for the virtual machine capabilities. That VM software is the best out there. 2. Those mainframe processors are totally unlike anything else. IBM fabs them and designs them on their own. They're just incredibly different, and 4 of them with Z/OS and Z/VM can easily outdo 100 PC servers, which leads me to: 3. IBM mainframes also have I/O better than any Intel or Sun box due to ESCON and FICON, as well as other wicked bus technologies they own the patents to . They're designed to run several Linux instances getting Slashdotted while there's several apps and DB2/390 running in other partitions. 4. IBM boxes are completely hot-swappable. Even the processors and RAM. Tandem and a very few others can claim that as well. However, they run Linux and a bunch of standard software now, under Z/OS and Z/VM, which abstract the nasty details of how to make an entire machine hot-swappable from you the user. 5. Sun got into the Intel box and AMD box business from buying Cobalt (maker of small Rackmount boxes and the Qube), and from offering SunPCI cards so you can run Windows software on your Sun Ultra workstations. They shipped those with Celeron and AMD K6 CPU's. They've been offering Intel and AMD boxes through Cobalt. So they signed with AMD to provide them . 6. Sun's actually marketing the Cobalt boxes now. They didn't for a while due to Sun's ignorance of what they actually sell. Look at NetDynamics, which they let die partially because they didn't know how to market it correctly, and partially because they didn't develop it . Sun has so many products that they don't sell because their salespeople do not even know what they are. They finally came to the realization that they do sell AMD boxes. 7. Cobalt RAQ servers kick butt. They beat Windows 2000 for an easy install. Now if they were only developed more. What's wrong with Itanium now besides the fact it emulates a 32-bit PC at the speed of a Pentium 100, and it's got the most complex compiler ever developed? |
02-11-2002, 10:26 PM | #22 | |
Read? I only know how to write.
Join Date: Jan 2001
Posts: 11,933
|
Re: just curious, what itanium defect now?
Quote:
Itanium was originally to be introduced in the Merced version. But a major architectural flaw (little discussed) delayed Merced by two(?) years. By that time it was too slow to be successful and sold for development and prototype systems. I believe a WSJ article that laid out that history. Intel has literally consumed the entire processor development groups of both Hewlett Packard and Compaq - intact - for Itanium development projects. Each group was assigned different versons of Itanium. Itanium chips are coming. It was seriously delayed by an architectural problem that occured, ironically, about the same time as an architectural problem happened during P4 development. |
|
02-11-2002, 10:58 PM | #23 |
Lecturer
Join Date: Jan 2001
Location: Carmel, Indiana
Posts: 761
|
you forgot the compiler divisions too :)
The Itanium only has a Pentium 100 in it, apparently. Good for Word 97 and not much else.
Then again, Intel has had 64-bit SQL Server and Windows 2000-based apps for quite a while due to them prototyping them on the AlphaServers. That's why they have Windows out along with the chips already. I saw a quad Alpha running 64-bit Windows 2000 with SQL Server 7.0 almost 3 years ago. It smoked. No, let's put it bluntly. It was 10x the speed of SQL Server 7 on 32-bit for certain queries. It's too bad that: A. Oracle is not completely optimized for 64-bit yet. I want to play with it on SPARC. They do have a 64-bit version, however it's not as complete. B. Compaq dropped the ball and didn't want to finish the Win2K port. From what I saw, it was complete to a beta 3 level. Darnit. C. I want a 64-bit database server that runs true 64-bit software so I can really crunch numbers. The compiler group has swallowed the old DEC group at Compaq, Intel's own (which is not a bad one! Their compiler is great for x86), and HP's. Insane. And let's not mention the Microsoft people trying to work with Intel C++ and Visual C++ on IA64. At least the Itanium II, McKinley, is being done mainly by HP. Say what you want about HP-(S)UX, but the PA-RISC chips are really good. HP's own design teams can pull this off. |
02-12-2002, 02:22 AM | #24 |
Guest
Posts: n/a
|
HP's hardware is awesome. Don't get me wrong.
However, it's only as good as the software it runs, and the operating system is <b>shit</b>. |
02-12-2002, 09:21 AM | #25 |
Lecturer
Join Date: Jan 2001
Location: Carmel, Indiana
Posts: 761
|
Have you tried Linux or NetBSD for it yet
Dhamsaic,
Have you tried either of those with it? I know how bad their OS is, however it's coupled with awesome hardware and a good compiler. |
02-12-2002, 09:33 AM | #26 |
Guest
Posts: n/a
|
I haven't tried any of the BSD's yet, unfortunately, because the machines are at work and I have to actually Get Stuff Done.
Linux on the PA-RISC processor may be swell, but I've yet to see an actual full distribution. I know that Debian was working that way, but I haven't checked up on it in nearly a year, simply because they were so far behind that I didn't think I'd ever get a chance to actually do it. Ideally, Slowaris would be ported to HP hardware and all would be happy Unfortunately, that's not going to happen. I've heard good things about NetBSD on PA-RISC hardware, however. Like I said though, we only have some 6 usable HP boxes and I don't have the liberty to blow one away and try Linux or NetBSD on it "just for kicks". It's in that interest that I've been looking at picking up a cheap B-180L off eBay or something, but I've yet to come across any decently recent hardware at decent prices. The government may have $5,000 to spend on a HP-UX box for us to develop on, but I don't have that kind of money to (potentially) waste after I get a box and realize I can do nothing useful on it. |
03-13-2002, 03:57 PM | #27 |
highly unstable
Join Date: Mar 2002
Location: Minneapolis
Posts: 11
|
worst OS ever?
Not too sure on any of the *nixes, I've run a couple, but never done any hardcore fooling around with them (i.e., any sort of network admin). But, by far, the worst operating "system" (if it can even be called that) was Microsoft Bob. Not only was it the fore-runner of the evil Clippy, but it also had security holes large enough to drive trucks through (like the "feature" that allowed a user to recreate their password if they typed it incorrectly three times).
__________________
"Rock music must give birth to ORGASM AND REVOLUTION! -High Noon (Devil's Advocate), DJ Shadow |
03-13-2002, 04:20 PM | #28 |
Guest
Posts: n/a
|
Bob was actually, I think, way ahead of its time. It was just implemented in the most awful manner possible.
Let's not forget that Melinda French was one of the head designers on Bob, which is probably why it ever saw the light of day - she was knocking up the boss. BTW, it's not really an operating system - more like an operating shell. |
03-13-2002, 05:40 PM | #29 |
retired
Join Date: Dec 2001
Location: Toronto, Canada
Posts: 1,930
|
knock up Verb. To make pregnant. E.g." I knocked her up and now her brother's after my hide."
I get irritable when words are misused, too. :whofarted dham doesn't like to say "fucking" Last edited by Nic Name; 03-13-2002 at 05:52 PM. |
03-13-2002, 07:15 PM | #30 |
still says videotape
Join Date: Feb 2001
Posts: 26,813
|
I believe he got mixed up with the ruralism "knocking boots". As in making "love" in the back of a pickup with your shit kickers on.
Last edited by Griff; 03-13-2002 at 07:17 PM. |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Display Modes | |
|
|