Bzzt! From what I've read, HT can and does slow down some applications
For a good analysis, read this article over at Ars. In particular, it does point out that the likely cause of slowdowns in some apps is down to cache contention. Near the end, it also says:
With the wrong mix of code, hyper-threading decreases performance, just like it can increase performance with the right mix of code
In short, sometimes it helps, sometimes it hinders.
Finally, you don't need multithreaded apps to take advantage of SMP/HT; if you're running a cpu intensive application on one CPU, the other is free for interactive stuff. You do, however, get much more benefit in a multi-threaded application.
I guess the real question is how much research money goes into chip development? IBM sell many different products (including intel servers, Unix servers, software products, consultancy, mainframes etc, etc) while Intel sell (mainly) Pentium CPUs with some sidelines in graphics etc. So while IBM is 4 times the size of Intel, I'd imagine Intel probably spends more on CPU development.
Unfortunately, they have to do this to maintain their trademark. Trademark law is fairly clear on this; "protect it or lose it". If they let this web site use the trademark, another company or individual could come along and abuse it.
What they should do is negotiate something which (a) maintains their trademark and (b) allows the website to continue.
The EU recently (circa 95/96) extended copyrights to 70 years. When this extension is about to run out, I expect frantic lobbying by corps in the US to extend it again and the US will threaten (sorry, lobby) other countries to extend as well.
Like it or not, we need international copyright laws to make this stick, but unfortunately we're following America's lead because that's where the big bucks (Hollywood/Disney) are.
Not trusting user input is one of the major things in Web apps. A few years back, while I was working at a University, a lecturer said that Javascript form checking was a great idea as it meant that you didn't have to check at the server side.
That theory didn't last long with her as I pointed out the flaws.
That said, the client side form checking is a useful tool, as it saves the incorrect data being sent and thus (a) saves the user time and (b) saves server-side CPU. Client side checking is optional, but desirable; server-side checking is mandatory.
Of course, it use to be "Eight Megs And Constantly Swapping", but given that most people have considerably more than that in their graphics cards, it's a little outdated.
.NET is their buzzword compliant strategy including SOAP, XML, Web services and their latest plan to crush competitors. Somewhere in there is the ditched Hailstorm/Passport plan for world domination.
Palladium is the DRM, sorry, secure platform where the idea is that a Palladium enabled OS will only run signed apps, presumably adding security by not running any viruses, worms and any haxxor tools. Of course, this means any open source will not work in a Palladium OS because of the difficulty of getting an open source app signed.
That's my understanding of the two, but I'm not 100% sure; it's been difficult trying to work out exactly what.NET really means...
Like I said, you can get similar things and Sun sell some resource management software which is used to guarantee that users/groups get a specified amount of CPU time. In short, there can be a need in Unix, although the schedulers do tend to work a little better.
The whole personification of corporations is something which is required for a number of things in business like contracts as only a "person" can make a contract. When you agree to sell 5,000 widgets to company X's procurement guy, he hasn't agreed to buy them, he is acting as an agent for his company in order that the company can buy them from your company. After all, he probably doesn't have the money himself and you probably don't have the 5,000 widgets. By making the company a "person" which can "agree" to contracts (actually the employees agree as agents of the company), you enable business to take place.
Yes, there are ways to abuse this notion of a corporate "person", especially in limited companies, but that's the price paid for allowing business to take place. Should something be done about it? *shrug* I'm not up to date on the legal stuff enough to make comments.
Ulimit isn't the same thing; ulimit limits on total CPU time, not % CPU at any given time.
Thus, if they run an app for some time (e.g. over a few days) they might run up 30 minutes total CPU time with no real problems. On the other hand, someone could start an app, peg it at 99% for 10 minutes and completely hammer the server.
Operating systems like Solaris have add-ons to achieve system limits like CPU allocations to individual users/groups, but that is NOT ulimit.
A Sunblade 100 with minimum spec is $995 list. You can buy a PC for half that with similar or better performance. To get performance better than a P3/4, you're really looking at something like a blade 1000 which starts off at around $5,000 (off the top of my head).
Yup, the blade 100 is nice, but it doesn't have price/performance of Intel yet. What it does have is binary compatibility with an F15K, so you can run the same 64-bit app on your cheap-ass desktop as you do on the big iron in the data centre.
The poster is right; most hackers don't have a Sun on their desk as they are comparatively expensive. Given the choice between a $1000 Sun which plays very few games and has OK performance or a $1000 PC which will play 99% of games, what do you think the average hacker is going to buy?
Yup, IRIX was good, but maintaining a full OS takes a lot of money. This way, they can piggy-back on investments made by other people & companies while still having a modern OS. They've already integrated XFS into linux, and it wouldn't surprise me to see other SGI/IRIX technologies coming into linux in the same way. Similarly, IBM have migrated JFS into linux.
Well, as SpamAssassin is open source, you should be able to add new rules. In any case, you can easily modify the scores attached to certain messages in your ~/.spamassassin/user_prefs file
Not quite what you mean, perhaps, but one of the great things about smit/smitty in AIX is that the GUI (or curses based version of it) is simply a frontend to the OS level commands. For example, "smit mkuser" brings up a dialog to set all the parameters for creating a user. Once you've run the command, you can use the "View command" option to see what was actually run.
In a lot of cases, I'll use smit to do something, then repeat from the command line to run over multiple instances (e.g. to mirror multiple logical volumes).
Better yet, get some preview licenses from the companies for evaluation, do a quick run of some sample sets you want to test and then spend the money on the one which works best for you. At the very least, buy a single license for each and test them, but you should try to get eval licenses if you can squeeze the vendors properly.
Moved? AFAIK, Solaris 9 is still only uses NFSv3 (or optionally v2) so we haven't "moved" to NFSv4 yet. Dunno about any GSSAPI stuff in there, but here is Secure NFS (generally considered an oxymoron) which can use Diffie-Helman or Kerberos.
Slashdot Mirror
For a good analysis, read this article over at Ars. In particular, it does point out that the likely cause of slowdowns in some apps is down to cache contention. Near the end, it also says:
In short, sometimes it helps, sometimes it hinders.Finally, you don't need multithreaded apps to take advantage of SMP/HT; if you're running a cpu intensive application on one CPU, the other is free for interactive stuff. You do, however, get much more benefit in a multi-threaded application.
I guess the real question is how much research money goes into chip development? IBM sell many different products (including intel servers, Unix servers, software products, consultancy, mainframes etc, etc) while Intel sell (mainly) Pentium CPUs with some sidelines in graphics etc. So while IBM is 4 times the size of Intel, I'd imagine Intel probably spends more on CPU development.
What they should do is negotiate something which (a) maintains their trademark and (b) allows the website to continue.
Like it or not, we need international copyright laws to make this stick, but unfortunately we're following America's lead because that's where the big bucks (Hollywood/Disney) are.
That theory didn't last long with her as I pointed out the flaws.
That said, the client side form checking is a useful tool, as it saves the incorrect data being sent and thus (a) saves the user time and (b) saves server-side CPU. Client side checking is optional, but desirable; server-side checking is mandatory.
Of course, it use to be "Eight Megs And Constantly Swapping", but given that most people have considerably more than that in their graphics cards, it's a little outdated.
Seems like a lot of crap just to give client-server with XML to me...
Palladium is the DRM, sorry, secure platform where the idea is that a Palladium enabled OS will only run signed apps, presumably adding security by not running any viruses, worms and any haxxor tools. Of course, this means any open source will not work in a Palladium OS because of the difficulty of getting an open source app signed.
That's my understanding of the two, but I'm not 100% sure; it's been difficult trying to work out exactly what .NET really means...
D'oh! You're right, of course...
Like I said, you can get similar things and Sun sell some resource management software which is used to guarantee that users/groups get a specified amount of CPU time. In short, there can be a need in Unix, although the schedulers do tend to work a little better.
Won't someone pleeeeeaaaase think of the children
Yes, there are ways to abuse this notion of a corporate "person", especially in limited companies, but that's the price paid for allowing business to take place. Should something be done about it? *shrug* I'm not up to date on the legal stuff enough to make comments.
Thus, if they run an app for some time (e.g. over a few days) they might run up 30 minutes total CPU time with no real problems. On the other hand, someone could start an app, peg it at 99% for 10 minutes and completely hammer the server.
Operating systems like Solaris have add-ons to achieve system limits like CPU allocations to individual users/groups, but that is NOT ulimit.
There's also the issue about licenses; while there may be licences for the server, buying licenses for all the clients might be expensive.
Yes, I may well be unhip and "not with it"; heck it took me about a fortnight to find something which told where the "all your base" came from...
Yup, the blade 100 is nice, but it doesn't have price/performance of Intel yet. What it does have is binary compatibility with an F15K, so you can run the same 64-bit app on your cheap-ass desktop as you do on the big iron in the data centre.
The poster is right; most hackers don't have a Sun on their desk as they are comparatively expensive. Given the choice between a $1000 Sun which plays very few games and has OK performance or a $1000 PC which will play 99% of games, what do you think the average hacker is going to buy?
Yup, IRIX was good, but maintaining a full OS takes a lot of money. This way, they can piggy-back on investments made by other people & companies while still having a modern OS. They've already integrated XFS into linux, and it wouldn't surprise me to see other SGI/IRIX technologies coming into linux in the same way. Similarly, IBM have migrated JFS into linux.
"Imagine a Beowulf cluster of those..."
Well, as SpamAssassin is open source, you should be able to add new rules. In any case, you can easily modify the scores attached to certain messages in your ~/.spamassassin/user_prefs file
- Being the target of rants/raves
- posting duplicate articles
*sigh* I thought everyone knew that...Been there a while. My guess is that it's there to try and push you towards a subscription and/or to give /. a bit more revenue.
In a lot of cases, I'll use smit to do something, then repeat from the command line to run over multiple instances (e.g. to mirror multiple logical volumes).
Better yet, get some preview licenses from the companies for evaluation, do a quick run of some sample sets you want to test and then spend the money on the one which works best for you. At the very least, buy a single license for each and test them, but you should try to get eval licenses if you can squeeze the vendors properly.
Moved? AFAIK, Solaris 9 is still only uses NFSv3 (or optionally v2) so we haven't "moved" to NFSv4 yet. Dunno about any GSSAPI stuff in there, but here is Secure NFS (generally considered an oxymoron) which can use Diffie-Helman or Kerberos.