MPEG audio decoding requires a good deal of CPU perfor-
mance, especially layer-3. To decode it in realtime, you
should have at least a Pentium, Alpha, SuperSparc or
equivalent processor. You can also use the -singlemix
option to decode mono only, which reduces the CPU load
somewhat for layer-3 streams. See also the -2 and -4
options.
Isn't everything in OS X late-binding?
on
Is Mac OS X Slow?
·
· Score: 3, Interesting
Doesn't Objective-C suffer from the same performance problem as Java in that there is no early-binding by a linker of the explicit functions/methods that will be called in an application?
Is late-binding the largest cause of poor performance in OS X? And, if so, does this mean that GNUStep is a bad idea?
I would say that the first place to start is Ad Aware to remove the spyware, but www.lavasoft.nu seems to be down. I haven't been able to get a refupdate in quite awhile.
How about a compromise? Any researcher finding a mechanism of disease inherent within a genetic sequence can patent the sequence and receive no more than $100 per test for the diseased genes.
Or, if that amount is too low for some of the more esoteric diseases (which will not be often tested), how about a sliding scale?
We should have some legislative mechanism in place to reduce the maximum payout per test as the number of tests performed rises.
It is absolutely unreasonable to grant an exclusive patent on my genetic function (and I assume that men carry this gene sequence as well, even if it is inactive) without my personal consent. If the drug companies refuse to compromise on this issue, then they should expect wholesale disregard for their patents, as is proving to be the case.
Imagine that you have a library, and a librarian is filing away new books. When she is done filing them, she puts entries into the card catalog downstairs for the new books. The card catalog represents a filesystem's metadata.
Now imagine that the librarian falls out of a 2nd story window into a dumpster and is carted away before she finishes filing the books and updating the catalog. You have no idea what books were filed; you have to perform an exhaustive search of the library to ensure that the card catalog is correct, which takes a long time. This was fsck before journaling.
Servers with large amounts of disk space cannot afford extensive fsck times after a crash. It can take hours.
Now imagine that the librarian keeps a small notepad of the books that she is filing, and when she meets her sticky end, the new librarian can read the notepad, check and verify the new entries, then update the card catalog to a consistent state. We assume that the notepad is updated before the book is filed, so if we have an incomplete notepad entry, the librarian died and the entry can be disregarded. The notepad corresponds to the journal in a journaling file system.
It takes time to write a journal, so journaling filesystems will always be at least a little slower than non-journaling equivalents, design improvements aside.
Most journaling filesystems will only guard the card catalog (metadata). Some, such as VxFS and ext3, can also be made to journal the books (data), but performance goes down because so much more goes through the log.
Another feature to look for in journaling filesystems is dynamic inode creation. ext3 does not have this feature - you can only have so many card catalog entries, and when you exceed them, you can't add any more new books. XFS, for example, can create new inodes on the fly as long as you have disk space.
For Sun people, it is always a surprise to find that Sun's UFS does journaling (you don't have to buy Veritas VxFS), but you have to turn it on with an option in/etc/vfstab.
Wild idea: what would happen if the FSF formed a not-for-profit music label? One that passed on as much profit as possible to the artists?
The RIAA is the greatest threat to free software today, without question. Microsoft has certainly not resorted to the legal system to destroy the movement; the RIAA has and is.
It's time that we took the threat for its real value and struck back, hard.
As a citizen, your duty is not to rewrite the law in the courtroom should you ever be called to serve on a jury. Nor is it your duty to encourage others to break laws you think are 'bad'.
So tell me, if congress makes jaywalking a federal offense punishable by 10 years in prison and a $100k fine, it's not the responsibility a jury to refuse to enforce it? You're patently wrong.
If I could believe that I was guilty, then perhaps.
If I believed that I was wrongly convicted, that I was the victim of a DMCA witch hunt or other unreasonable persecution of dubuious constitutional footing, then no, it would be time to leave, and never look back, save to pull all the assets I could out of the system.
There really should be a FAQ somewhere for people who need to leave in a hurry. I'm surprised that it isn't done more often.
Under his situation, where he is currently at liberty but will begin a prison sentence soon, how difficult would it be for him to put himself beyond the reach of the criminal justice system?
For myself, if I were on the jury, I would have used some of the powers described at FIJA to ensure that he received a maximum of 6 months in prison (assuming that I believed in his guilt and I agreed with the law, which I probably wouldn't).
However, assuming that members of the jury were duped or otherwise misled into this unreasonable sentence, how easy is it to leave and where should he go? Is Brazil the best destination (since they lack an extradition treaty)?
I would like to know this, for the day when these sentences are doled out for Kazaa users.
The suggestion I would have is put the GNU stuff in/usr/local/bin for now - and this is exactly what Sun is doing.
So you're telling me that Sun's current attitude towards GNU is right? You don't deal with perl much, do you?
The situation is technically similar to RedHat 7, where two versions of gcc were included - it's not such a drastic piece of brain damage, but the effect is much longer. The system is fragmented and is not true to itself.
Here is what I really suggest that Sun do: take Red Hat Linux - remove the Linux, insert the Solaris kernel, and the minimum of supporting utilities to make it run. Maintain LSB status if at all possible.
Then offer Solaris/Classic and Solaris/GNU. Certify each for the high-end. Announce that they will merge into sunos 3 within 5 years.
If Sun really wants to set the world on fire, GPL the kernel, then integrate SGI's XFS into Solaris as a native filesystem.
And the day that HP-UX/Itanium was available for purchase, Solaris/Itanium should have been available for free - the port is finished and is sitting on a shelf at Sun - but I digress.
Really, Sun can end the Linux question anytime they want - just by opening their kernel. Who else can claim this? Who else has come so close?
First and foremost, Solaris patchchk. This utility, written in perl, generates an HTML page which must be read by a browser. Now, RedHat up2date, which I like a lot, is python-based, and RedHat goes a little overboard with python support in the OS. patchchk is awful compared to up2date, but I must confess that I haven't used it in some time.
Solaris perl is built with Sun's C compiler, and this perl version cannot be extended by gcc. If you lack Sun cc and you need to extend perl, then you must reinstall in a separate location and then manage two installations. You cannot uninstall Sun's perl without breaking lots and lots of things.
The Solaris package system is an abomination. Please, PKZIP is something we should have left behind in our DOS days. Let's integrate bzip2 into something with the speed of RPM.
If you want an LVM, you have to load DiskSuite, and the documentation leaves a great deal to be desired.
UFS, Sun's native file system, supports journaling, but is loaded by default without it and very little mention is made of the importance of turning it on.
Solaris is certainly better than HP-UX with/etc/system (versus rebuilding the HP-UX kernel every time you change SHMMAX). However, AFAIK, Solaris must be rebooted for changes to/etc/system to take effect. I know that HP-UX is very recently getting dynamic kernel tunables, I hope Solaris is as well. I certainly enjoy them in Linux.
Some of this stuff is really old. Seriously, do we really need both awk and nawk? HP-UX standardized on nawk, but really we should all just switch to gawk.
Along these lines, it's time to remove every SysV utility that can be replaced by a GNU equivalent. Every commercial UNIX should be doing this.
Both Sun and HP are still trying to get out of Motif/CDE. What's the holdup? CDE on my workstation makes other people ask me if I drive an Edsel.
Solaris suffers from the same problem as all commercial UNIX: the question of GNU integration. They now rely upon GPL utilities in a BIG way, but they are hesitant to integrate them properly and make them work well. In the meantime, there is enough SysV cruft that hasn't been touched in years that you could realistically call this OS "Solaris the Living Dead."
It's time for Sun to concentrate on the OS components that it does well, and throw everything else to GNU.
AFAIK, Ciprofloxacin is one of the first totally synthetic antibiotics - it acts by preventing DNA replication and hence cell division. Most antibiotics previously came from fungus and mold.
Isn't Ciprofloxacin patented by a major drug company?
And you [i]certainly[/i] never know when a tweak could corrupt and endanger your machine.
While I've never used it, Mac OS X is based on BSD UNIX, and enjoys protected memory and filesystem permissions when configured properly. Any GUI tweaks that do not involve root authority should not impact other accounts or system hardware. It is quite common for UNIX users to create separate accounts to run untrusted apps, and this can be done with moderate to high confidence on patched systems. Unless a root exploit is involved, the worst a rogue app can do is trash your account (fork bombs excluded).
I realize the fear that many Mac users have of applications that crash the system. Under UNIX, this propensity is greatly reduced if not eliminated.
Slashdot Mirror
You never made a decent BIOS anyway.
It can take a lifetime to gain a customer and only a moment to lose one. Adobe should be good company for you. Keep up the good work.
Why isn't Sistina's LVM making it into the kernel? SUSE has been including it as standard in their distribution for some time.
I don't track the LKML at all. I'm curious why XFS made it in, but LVM did not.
I really didn't care for some of the later stuff on earth, but I really liked the earlier shows.
Everybody have a look at http://www.mercedesproblems.com/ before you even think of buying one of these clunkers.
I'm giving Volvo a try now.
Sure, Comcast, go ahead and get upset when I download an ISO image of Red Hat at peak hours. But give me a way to get the ISO during non-peak times.
This needs to be implemented by a "download agent" installed on my system that can consult yours and operate only when traffic is not saturated.
If you don't have this, then don't complain.
The original Romanavoratnalunda had a certain Je ne said quois.
I admit, I'm somewhat attached.
From man mpg123:
MPEG audio decoding requires a good deal of CPU perfor- mance, especially layer-3. To decode it in realtime, you should have at least a Pentium, Alpha, SuperSparc or equivalent processor. You can also use the -singlemix option to decode mono only, which reduces the CPU load somewhat for layer-3 streams. See also the -2 and -4 options.Doesn't Objective-C suffer from the same performance problem as Java in that there is no early-binding by a linker of the explicit functions/methods that will be called in an application?
Is late-binding the largest cause of poor performance in OS X? And, if so, does this mean that GNUStep is a bad idea?
I would say that the first place to start is Ad Aware to remove the spyware, but www.lavasoft.nu seems to be down. I haven't been able to get a refupdate in quite awhile.
As in Doric, Ionic, and Corinthian columns.
I wonder what the Master was doing in Greece? Hmm.
Dumping the panic to swap seems more sensible than allocating another partition. This is how other OSes do it.
How about a compromise? Any researcher finding a mechanism of disease inherent within a genetic sequence can patent the sequence and receive no more than $100 per test for the diseased genes.
Or, if that amount is too low for some of the more esoteric diseases (which will not be often tested), how about a sliding scale?
We should have some legislative mechanism in place to reduce the maximum payout per test as the number of tests performed rises.
It is absolutely unreasonable to grant an exclusive patent on my genetic function (and I assume that men carry this gene sequence as well, even if it is inactive) without my personal consent. If the drug companies refuse to compromise on this issue, then they should expect wholesale disregard for their patents, as is proving to be the case.
Imagine that you have a library, and a librarian is filing away new books. When she is done filing them, she puts entries into the card catalog downstairs for the new books. The card catalog represents a filesystem's metadata.
Now imagine that the librarian falls out of a 2nd story window into a dumpster and is carted away before she finishes filing the books and updating the catalog. You have no idea what books were filed; you have to perform an exhaustive search of the library to ensure that the card catalog is correct, which takes a long time. This was fsck before journaling.
Servers with large amounts of disk space cannot afford extensive fsck times after a crash. It can take hours.
Now imagine that the librarian keeps a small notepad of the books that she is filing, and when she meets her sticky end, the new librarian can read the notepad, check and verify the new entries, then update the card catalog to a consistent state. We assume that the notepad is updated before the book is filed, so if we have an incomplete notepad entry, the librarian died and the entry can be disregarded. The notepad corresponds to the journal in a journaling file system.
It takes time to write a journal, so journaling filesystems will always be at least a little slower than non-journaling equivalents, design improvements aside.
Most journaling filesystems will only guard the card catalog (metadata). Some, such as VxFS and ext3, can also be made to journal the books (data), but performance goes down because so much more goes through the log.
Another feature to look for in journaling filesystems is dynamic inode creation. ext3 does not have this feature - you can only have so many card catalog entries, and when you exceed them, you can't add any more new books. XFS, for example, can create new inodes on the fly as long as you have disk space.
For Sun people, it is always a surprise to find that Sun's UFS does journaling (you don't have to buy Veritas VxFS), but you have to turn it on with an option in /etc/vfstab.
...to form our own music label?
Wild idea: what would happen if the FSF formed a not-for-profit music label? One that passed on as much profit as possible to the artists?
The RIAA is the greatest threat to free software today, without question. Microsoft has certainly not resorted to the legal system to destroy the movement; the RIAA has and is.
It's time that we took the threat for its real value and struck back, hard.
So tell me, if congress makes jaywalking a federal offense punishable by 10 years in prison and a $100k fine, it's not the responsibility a jury to refuse to enforce it? You're patently wrong.
If I could believe that I was guilty, then perhaps.
If I believed that I was wrongly convicted, that I was the victim of a DMCA witch hunt or other unreasonable persecution of dubuious constitutional footing, then no, it would be time to leave, and never look back, save to pull all the assets I could out of the system.
There really should be a FAQ somewhere for people who need to leave in a hurry. I'm surprised that it isn't done more often.
Umm no, Sun's Linux is a repackaged Red Hat Advanced Server. I believe that Red Hat has some preferential access to OpenOffice in exchange.
There is definitely a higher-end crowd that wants a stable platform. There is also a higher-end crowd that likes the latest tools.
And if Sun takes backward-compatibility so seriously, then I'll just mail you my IPX and you can load Solaris 9 on it.
And gawk is several steps backward from awk/nawk? Really?
Under his situation, where he is currently at liberty but will begin a prison sentence soon, how difficult would it be for him to put himself beyond the reach of the criminal justice system?
For myself, if I were on the jury, I would have used some of the powers described at FIJA to ensure that he received a maximum of 6 months in prison (assuming that I believed in his guilt and I agreed with the law, which I probably wouldn't).
However, assuming that members of the jury were duped or otherwise misled into this unreasonable sentence, how easy is it to leave and where should he go? Is Brazil the best destination (since they lack an extradition treaty)?
I would like to know this, for the day when these sentences are doled out for Kazaa users.
So you're telling me that Sun's current attitude towards GNU is right? You don't deal with perl much, do you?
The situation is technically similar to RedHat 7, where two versions of gcc were included - it's not such a drastic piece of brain damage, but the effect is much longer. The system is fragmented and is not true to itself.
Here is what I really suggest that Sun do: take Red Hat Linux - remove the Linux, insert the Solaris kernel, and the minimum of supporting utilities to make it run. Maintain LSB status if at all possible.
Then offer Solaris/Classic and Solaris/GNU. Certify each for the high-end. Announce that they will merge into sunos 3 within 5 years.
If Sun really wants to set the world on fire, GPL the kernel, then integrate SGI's XFS into Solaris as a native filesystem.
And the day that HP-UX/Itanium was available for purchase, Solaris/Itanium should have been available for free - the port is finished and is sitting on a shelf at Sun - but I digress.
Really, Sun can end the Linux question anytime they want - just by opening their kernel. Who else can claim this? Who else has come so close?
Easy, we call it SunOS 3.
What, you had scripts that broke between SunOS 4.1.3 and Solaris 2?
Sun wants to be on the desktop. Solairis is a dinosaur. Linux isn't going to work out anytime soon. Something has to change.
Sun submits patches to the relevant projects that guarantee behavioral compatibility.
The "we can't upgrade because stuff will break" crowd really gets on my nerves sometimes.
Solaris suffers from the same problem as all commercial UNIX: the question of GNU integration. They now rely upon GPL utilities in a BIG way, but they are hesitant to integrate them properly and make them work well. In the meantime, there is enough SysV cruft that hasn't been touched in years that you could realistically call this OS "Solaris the Living Dead."
It's time for Sun to concentrate on the OS components that it does well, and throw everything else to GNU.
AFAIK, Ciprofloxacin is one of the first totally synthetic antibiotics - it acts by preventing DNA replication and hence cell division. Most antibiotics previously came from fungus and mold.
Isn't Ciprofloxacin patented by a major drug company?
While I've never used it, Mac OS X is based on BSD UNIX, and enjoys protected memory and filesystem permissions when configured properly. Any GUI tweaks that do not involve root authority should not impact other accounts or system hardware. It is quite common for UNIX users to create separate accounts to run untrusted apps, and this can be done with moderate to high confidence on patched systems. Unless a root exploit is involved, the worst a rogue app can do is trash your account (fork bombs excluded).
I realize the fear that many Mac users have of applications that crash the system. Under UNIX, this propensity is greatly reduced if not eliminated.