Your claim does not make sense to me. In fact, the idea that launchd CANNOT be used in Linux appears rather foolish and short-sighted.
You provided a reference to an FSF document to support your reasoning. The cited web page says, in part:
The FSF now considers the APSL to be a free software license with two major practical problems, reminiscent of the NPL:
It is not a true copyleft, because it allows linking with other files which may be entirely proprietary.
It is incompatible with the GPL.
For this reason, we recommend you do not release new software using this license; but it is ok to use and improve software which other people release under this license.
Granted, the APSL does not prohibit users of the software to link with proprietary libraries, thus is not a "copyleft" license. So, what? This is less restrictive than the GPL, not more. This, in and of itself does not preclude Linux users from using it on their systems.
The FSF concludes that it is ok to use and improve software which other people release under this license.
You would be allowed to compile the daemons using gcc and glibc libraries and use them with no problem from the APSL. You would also be allowed to link GPLd programs against the supporting APSL licensed frameworks.
The only limitation is that if you ship an improved version of this code that you make that code available to others under APSL terms. i.e. you provide source code so that Apple and and other users of the APSLed code benefit from the changes.
Insisting on re-inventing every wheel just so that everything is covered under GPL is a waste of effort. It steals time of those working on GPLed code from doing other work, and selfishly prevents others from benefit from you good ideas if you improve a fork of the work rather than the original work itself.
It strikes me as foolish that GNU/Linux people spend so much effort to mimic other people's work, re-implementing large subsystems just to get them under the GPL umbrella, rather than cooperating with others to re-use and improve the best code available.
You state that there are "practically zero ways to reduce CPU-heavy and memory heavy graphical effects."
This does not hold water considering that there are a number of actual and documented ways to reduce these effects such as turning off window miniaturize effects, magnification, etc. The support for intelligent reduction in animation overhead for dashboard which I described also contradicts your assertion.
I do not claim that old G3s are suitable for use in a rendering farm, realtime video encoding, gene sequencing or other graphically demanding or compute intensive tasks. However, I also do not consider that the majority of things that people use computers for to be trivial. Most people who use computers spend a great deal of time reading, writing, coding, and doing research. These rarely tax the hardware.
Web rendering on a G3 is fairly poky. However, doing so efficiently (and pleasantly) is still possibly. For instance, when on my Pismo, I use either OmniWeb or Safari via a tabbed interface. I read top level sites, and as I see stories that might interest me, I open them in another tab and continue looking for interesting items. By the time I close that window, I now have a number of articles in the queue ready to read. So, the way I use the web, even fairly demanding tasks like that do not slow me down and are enjoyable.
Do I prefer the greater responsiveness of a G4 or G5 to that of a lowly Pismo when browsing? Absolutely. However, I gain far more pleasure from performing the task in OmniWeb or Safari, than I do from any other browsers on whatever hardware.
Your professed failure to understand appears to be merely that: you seem not to understand that others have different priorities than you do. Your saying it many times does not make your personal opinion any more compelling than the first time.
Your work flow may very well demand the latest GPU or CPU performance. The tasks of the majority of users does not. This does not make them wrong, make you wrong, or vice versa.
As long as an older system has sufficient memory to make paging rare, they are quite suitable for many users. The bottom line is not whether they are the best choice, but whether the user is capable of performing their intended task and enjoys doing so. Will a user enjoy Tiger more than Panther on existing hardware? Most users probably would. Would they enjoy it more on newer hardware? Most users probably would. Despite this, many user will not derive sufficient utility or enjoyment from it to be worth the additional cost of new hardware. Who cares? Their preference is all that matters to them. Why does it seem to matter to you?
A slightly longer answer is No, but you can effectively disable it by simply excluding all or most of your system from the spotlight database.
The spotlight Preference Module contains a Privacy Tab. In this panel, you can add directories which are to be excluded from the index database. Presumably, adding / here would suffice to both save disk space and ongoing CPU costs. However, doing so broadly seems rather pointless. Certainly if you have confidential data on a network accessible volume you would be prudent to omit it. Likewise, if you have a subtree containing a large database, or collection of large files whose content is not usefully presented by spotlight, It might be worth excluding them.
On the whole, though, the incremental cost of maintaining the index is trivial and is correlated to the addition/modification of the files. This, in most environments, is both sporadic and requires negligible CPU and disk resources. If parts of your workflow have a file access pattern which makes spotlight less valuable to you, simply tailor spotlight to meet your needs.
Also, both system wide, and application specific spotlight queries are astonishingly efficient. Performing real time queries and displaying the results uses very little CPU and happens quite quickly. Even long queries (lasting seconds) do not appear onerous, since the result list is updated frequently as the search occurs, and incremental results are available.
The user decides which kinds of data are displayed for searches, and can tailor searches to a subset of volumes or systems when multiple disk (and remote volumes) are mounted.
Anyway, you can tailor the system to index less (or effectively nothing). Doing so, however, is unlikely to be of benefit. The system once primed appears so efficient that you would not save enough disk/time to make it worth your while. I suggest that rather than worrying about how to disable it to save processor cycles, you try it out for awhile and discover how it can save your brain cycles.
Spotlight is not a specific function or program. Rather it is a pervasive system. The base system provides a daemon which creates an initial index of all files, and subsequently handles requests for updating new or modified files. This process runs heavily niced in the background. While you can access a general Spotlight query tool using Command-Space, the real benefit of spotlight is its pervasiveness. Use the spotlight tool in the Preferences app to find out where a particular setting lives. (Note that Windows converts searching for a Windows-centric name will be presented with the Mac-centric counterpart.) Likewise in mail, the finder, and other programs, spotlight is available to help you find the context specific data you seek. Since developers can easily create spotlight plugins to parse data formats and export metadata, expect that most future applications will integrate well with the system.
It is important to note, that I found spotlight to be quite useful for a number of tasks, even though I only used it sporadically for testing purposes. Thus, I am inclined to be favorably biased towards it. On the other hand, I usually use a dual G5 Powermac and a fairly recent G4 17" powerbook. The fact that most of my use of spotlight was on a 400Mhz G3 powerbook suggests that my assessment of its efficiency is likely credible.
As always your time and your mood are the only true measures of a software tool, not my opinion.
You joke about how Tiger degrades gracefully on a PDP-11 but are not far from the truth.
For instance, I have one of the early firewire based PowerBooks. This system does not have a GPU powerful enough to offload 3D animation. It also has a 400Mhz G3. Thus, it does not have Altivec to offload graphics vector processing. As a result, instead of slowing down the system to perform 3D rotation of dashboard widgets when the user customizes them, it draws the obverse of the widget into offscreen memory then quickly replaces the live widget window with the widget customization window.
Thus, once the initial background indexing for spotlight has been completed, Tiger on this old PowerBook is fast enough for interactive use. In fact, I have been using it for a number of tasks in preference to my 17" G4 since I can get the job done faster in Tiger than on my newer system running Panther.
Actually Intel compatible clusters in the supercomputer rankings are not all that compelling. True, linux cluster did fare very well for several years as measured by price/performance. Also it is true that about 63% of the top 500 supercomputers are Intel or Intel compatible.
Despite this, the majority of systems at the top of supercomputer top 500 chart are based on the POWER architecture, not Intel chips.
The POWER based systems, including BlueGene and PowerPC systems, are all much better on both price/performance, and Watt/performance basis. Intel chips do have a per chip advantage over PowerPC on many work-flows. However, when scaled, they directly consume more power, and indirectly require even more power to run higher capacity air conditioning.
Looking at the top of chart reveals that Intel compatible systems are a small minority. In the top 10, as of November 2004, are one NEC, HP, SGI, Alpha, Xeon, and Itanium, and 5 members of the power family. IBM powers more than half of the top 100, with all other chip families dividing the remaining half.
It is cheaper to build and operate a supercomputer cluster using either Apple or IBM gear (running either Macos X or Linux) than to do so using Linux on Intel these days.
No. It's amazing considering Apple only has ~3 percent market share.
That number is incredibly misleading. It underestimates real users since, lumped into the Microsoft share, are the majority of the Intel PCs used as Point of Sale terminals, ATMs, informatioin kiosks, etc.
It also fails to account for the fact that Mac owners tend to keep their systems for several years longer than PC users. Thus the percentage of home users of Macs is probably somewhere between %8-%12, not the oft quoted 3%.
Wasn't NeXT the 1st computer to ship with a scroll wheel mouse?
No.
NeXT never shipped a scroll wheel mouse at all. Next used a 2 button mouse. By default the second button was identical to the first. The mouse preferences allowed the user to use the second button to display a copy of the application menu at the current cursor position. Which button (left or right) did this was user definable to allow lefties and righties to use the system comfortably.
I was a NeXT user for about 9 years before switching to Macos X. I don't really miss the second mouse button. The only aspect of the old NeXT/OpenStep experience that I prefer to Macos X is tear-off menus.
Honestly, they couldn't get much more unhelpful without actually breaking the LGPL.
Denial, "not just a river in Egypt". What a load of horse hockey.
You provide links to evidence which appears to justify the opposite conclusion.
Your first link, http://dot.kde.org/1097096753/1097113373/ is an email asking about what's going on with merging Apple changes back into KDE. The immediate reply states:
There is no shared resource for KHTML/KJS's source code, instead Apple decided to develop KHTML/KJS further in house and releases them as WebCore/JavaScriptCore. There never were and still aren't enough developers working on KHTML/KJS to really keep up with all the changes done by Apple which aren't available as patches nor are documented.
To me this suggests that apple added so much code to KHTML/KJS that it made more sense architecturally to split them into two frameworks "WebCore" and "JavaScriptCore", These are still released in their entirety under the GPL. There are so few developers on the KHTML side that they have been unable to keep up with the changes. Seriously, how is that Apple's fault?
Forking happens all the time. It would take far more effort and money for apple to continue to backport changes piecemeal to the KHTML/KJS trees than to fork the distribution. The sensible approach would be for the KHTML/KJS team to accept the architectural changes and simply start using the WebCore and JavaScriptCore, which is cleaner, better maintained, and of higher quality due to all the work Apple has done. Rather than grouse about how much effort it takes to backport changes piecemeal, why not adopt the superior libraries which Apple has produced and continue to improve them?
NIH syndrome works both ways.
The final link you give is entirely out of context. You state that Apple recommended doing a diff to see what has changed. Actually, the wording was "The best way to see every change line by line is to diff against the originals.". Immediately following, was about 14 pages (in my browser) of itemized descriptions of the changes performed, organized by functional description, subdirectory or even by function call.
Thus, your assertion is at least deliberately disingenuous, if not an outright lie.
I'm still waiting for them to submit Objective C++. As a matter of fact, one of the first successful FSF legal actions was forcing NeXT to release the source code to their GCC derivative.
This statement is misleading in several important respects.
First, NeXT's original plan to develop a variant preprocessor for the gcc toolchain and maintain it independently was indeed incompatible with the copyleft. However, it did not result in legal action. NeXT was informed about how and why it was incompatible and subsequently released the code. There was no law suit or formal legal action taken at all.
Also note that this was over 15 years ago, before even the BSD 4.4 source code was publicly freed in the law suit between the Regents of the University of California and AT&T. NeXT could not even share their changes to the BSD 4.3 kernel to anyone without a valid Unix Source License.
Regarding the subsequent divergence of the FSF and NeXT/Apple objective-c, it appears that the problem rests squarely with the FSF. The gcc maintainers have repeatedly rejected '#import' and other architectural features which are pretty major. They also deliberately allowed several aspect of the underlying runtime to diverge from the NeXT base because they either rejected or even backed out changes proposed by NeXT/Apple developers.
The code is there for all to see in the Apple branches published in Darwin. Apple developers propose changes to the primary gcc source tree, and argue for their adoption, but are often rebuffed.
I think that support for Frameworks, even for plain C, would benefit both Linux and BSD tremendously. However, unless gcc developers accept those changes they will remain unavailable.
Search the gcc developer mailing list over the past 5 years for discussion of Apple submitted changes, if you don't believe me.
The majority of state have similar laws - generally called "Uniform Trade Secrets Acts" or UTSA.
In each case the law states that it is illegal to publish, make use of, or otherwise benefit from information which the user has reason to know is protected as a trade secret.
The defendant knew for a fact that this information was protected, and that the information was gathered from those who were contractually obligated not to divulge it (it was covered under NDA).
Thus, it was a civil crime for his source to give the information. It was also a civil crime to publish those statements.
The remedies (under the California UTSA statute) are divided into both monetary damages, and court ordered innunctions.
In this particular case, fines, and injunctions to stop making use of this particular information are pointless. Instead Apple asked the court to order the defendant to divulge his sources, since that is the only remedy which can enable them to stop further damage.
Thus, this is not really an issue of first amendment freedom or freedom of the press. It is a request from Apple to have a judge provide legal remedy under the UTSA, in the only way which is relevant for them. Basically the author who was subpoenaed solicited information in breach of contract, knowingly published that information in breach of the UTSA, and then sought protection under the guise of free speech This seems specious to me. I side with Apple on this one.
I do believe that if used unwisely this law could have a chilling effect. However, the injunctive remedy is decided on a case by case basis at the judges discretion. Also the law punishes those who misuse the law by awarding damages and legal fees to the defendants if the plaintiff is in the wrong. This seems to provide a reasonable measure against abuse.
Either install a non-Windows OS on your existing hardware or buy a Mac. Linux, any BSD, or Macos X are simpler choices. BSD or Linux are harder in the short run but require less on-going maintenance once the user is settled in. Macos X requires changing both hardware and software, but is likely to be an easier transition for most users.
Whether you like it or not, the Wintel platform is no longer a very good choice for the average computer user, and has become a quite unpleasant environment for most people.
My friends is brand new, installed 2 weeks ago. His says he does not have the ability to record all episodes of a particular series without duplicates (like a tiov seasons pass does). As a result for one series (in rerun syndication) he set up M-F 1-2, M-F 2-3, and then every week or so, has to schedule additional recordings when the network moves the show to an early morning time slot (6 times over the next 2 weeks).
So it is appears to be unaware when particular episodes are shown in a different time slot.
Actually, drivers for some USB 2.0 ethernet dongles are now supported in the most recent release, 7.1a. My Belkin USB2.0 interfaces are about twice as fast as they were under version 4.x.
A one hour show in Basic mode takes about 8 minutes. In Best Quality, an hour of video transfers in under 30 minutes.
A friend of mine just got a Dish network DVR. I must say, that compared to Tivo, it truly sucks. It's not even like comparing Mac to windows for usability. it's more like comparing Mac to a graphical DOS app.
The remote control has far too many buttons, and some common functions on buttons that are poorly placed.
If you are watching a show, and either accidentally or deliberately go into the menus, the video buffer of what you were watching gets immediately flushed.
The layout of the menus, the UI, the inability to record shows based on name, etc. show a shoddy inattention to detail. Basically, you can search by program name, but only record by time of day.
I was shocked by how poor the interface was, and how unpleasant the system was to use.
There is a product called Xtrovert by Xtrememac which solves this problem. It attaches to an Xserve RAID unit to turn it into a stable, vertical tower. It is made from anodized aluminum with a matte silver powder coat finish that matches Apple's aluminum products. It costs around $280 bucks.
California, Massachessets (where the editor is in school), and most other states have each passed similar laws which specifically prohibit individuals from disseminating or unfairly using trade secrets. The california statute is titled the "Uniform Trade Secrets Act" (a.k.a UTSA). Under that statute the editor of Think secret appears guilty of the act of "misapropriation of a trade secret" under the definition (b)(2)(B)(ii).
(b) "Misappropriation" means: (2) Disclosure or use of a trade secret of another without express or implied consent by a person who: (B) At the time of disclosure or use, knew or had reason to know that his or her knowledge of the trade secret was: iii) Derived from or through a person who owed a duty to the person seeking relief to maintain its secrecy or limit its use;
The editor of Think Secret published confidential information, and clearly knew that the information was covered under NDA. His prior published statements reveal this general knowledge, and he had been contacted repeatedly by Apple lawyers asking him to remove infringing information and desist from future similar activities. He appears to have no viable defense against charges under the UTSA.
The remedies for the complainant for such misappropriation include, both damages and injunctions. Monetary damages may be levied based on losses (or potential losses) of the complainant, enrichment of the defendant (e.g. ad revenue or donations from the web site), exemplary punitive damages not to exceed twice the two kinds of damage mentioned above.
Injunctive relief takes 3 forms. The defendant may be enjoined from further use or dissemination of the information. The defendant may be forced to pay a royalty for use of the information. Finally, the plaintiff may request unspecified injunctive relief at the discretion of the court. (c) In appropriate circumstances, affirmative acts to protect a trade secret may be compelled by court order.
So, it appear that the defendant's alleged acts are clearly in violation of this statute, and he has misappropriated a trade secret as defined by the law. The defendant repeatedly ignored requests to desist from such activities. The monetary relief afforded by the law is inconsequential in this case. Thus Apple's goal is to seek injunctive relief. The first two forms of relief (enjoining them from further use of this specific information, or charging them royalties for its past use) are each irrelevant here: this specific cat is out of the bag, and no reasonable royalty will remedy this breach. Thus, they seek injunctive relief to compel the defendant to reveal his sources, so that they can sue the one who revealed the information to the editor and/or prevent further breach of contract.
Suing the Think Secret editor is clearly the only way to either sue the original offender for relief or ensure that they refrain from future illegal activities.
Claims that Apple is bullying the defendant or overstepping their legal rights appear poorly founded. They are defending their right under the law in the only way available to them. It is up to the judge to determine whether to compel the defendant's cooperation or not.
Is it clear that the public good is truly served by letting the defendant protect his source in this case? The states in which the corporation is based, the state in which the leak originated, and the state in which the defendant lives and works each have similar laws which specifically prohibit the defendants alleged activities. Each version of the law (UTSA) specifically allow the plaintiff to seek injunctive relief to both punish the offender and limit future damage of the plaintiff's interests.
The case is of interest because there is generally a clear public good in free speech and wide dissemination of information. At issue is whether judgments under the UTSA will erode journalistic integrity, or be used by corpo
Carl Jung wrote about a collective unconscious. Was he a crack pot?
It depends on how you wish to interpret his works.
Jung, unlike Freud and his followers, was concerned about scientific rigor, and his work appears to be valuable both pragmatically and theoretically. Many perform a naive unscientific reading, and attempt to use his works to support the rankest forms of BS. I find these interpretation both distasteful and unfounded.
When Jung began his study, the majority of clinical observation produced data on abnormal patients, not healthy individuals. The richest source of data on normal psychology came from literature and mythology. Clearly, familiar plots and psychological patterns appear as themes in the worlds fables, religions, and written works.
Jung fits right in, when read alongside the most recent works in evolutionary psychology and related disciplines. Those in the related disciplines of Neurology, Evolutionary psychology, etc. currently explore subsystems of our brain and how their structure and functions map to our behaviors. These modern scientists alternately describe observed behavior and attempt to discover their underlying causes in our physical and genetic systems. Jung attempted find common threads in myth, religion, and literature and discover their underlying ground in the human mind.
When Jung began his career the vast majority of psychological was directed at abnormal or industrial psychology. Most discussion of normal psychology was philosophical. As a result Jung looked at archetypal themes which recurred throughout myth and literature, and still resonated with individuals of his day. By his reasoning, these archetypal themes owe their power and endurance to the psychological makeup of individuals.
These themes are both popular and powerful because people are moved by them, relate to them, or understand them intuitively without requiring reflection or study. Since they are basic and independent of rational focus, he termed them unconscious. Since they appeared shared by all people he called it "collective unconscious".
People are inclined to behave similarly to one another in certain circumstances and interpret the world in similar ways. It is logical to conclude that one reason we tend to do so could because our brains are so similar to one another. Jung wrote of mind and psychological tendency, a higher level of abstraction. These are no incommensurable.
In my reading of Jung (direct English translations, since I don't know German) I find no mystical or unscientific connotations. Unfortunately, many readers misread the term collective unconscious. They jump to the conclusion that it has mystical connotations and dismiss the work as unscientific. Alternatively they are unscientifically minded and eagerly attempt to use his work to justify pseudo-scientific or mystical hogwash.
I am not saying his work is all correct, just that a number of his theories and observations are both consistent with the latest scientific theories and informative.
But while you're using the service the value for iTMS is 15$*number-of-months while Napster has a value of a few hundred thousand dollars if you calculate the value by utility
What a fanciful (and ludicrous) claim!
Since "a few" is imprecise, let's simplify that to "one". This smaller claim is that the "value by utility" of napster is $100,000 per month. There are 44,640 minutes in a 31 day month. Even if one could derive value from listening to songs 24 hours per day, this works out to $2.24 per minute for listening to music.
Also, I doubt that listening to music while sleeping is of any tangible subjective value. Here I'll be generous, and allow only 6 hours of sack time per day. This raises the cost to 2.98 per minute to listen to music. Since your numbers were presumably based on per/song pricing, If we choose the value of 3 minutes per song, your purported "value by utility" is $8.96 per song.
These numbers do not make any sense, thus your claim does not make any sense. I'll refrain from figuring out what you meant by "a few hundred thousand", since even 100,000 is so far off the mark.
In Oracle, a table or index may be subdivided into multiple sections called partitions.
Each partition is associated with a tablespace, which defines the underlying storage.
A range partition defines a range of keys which belong to a partition.
The general ability to store subsets of data or indices on separate table spaces provides an abstraction layer which may be used to optimize locality of reference, bandwidth, etc for subsets of data rather than for the set as a whole.
Other criteria can be used to define partition membership like, hashing or explicitly list of keys. Some examples of range partitioning are tuning the underlying storage characteristics base on regions, date range, product codes, etc.
Before you argue with me, remember the traditional way to set value is to sell it and see what the open market brings. EBay is great because it generally establishes the real market value.
Bull. This claim is naive and misleading.
Your concept of value is accurate only for fungible commodities which have no direct utility.
People purchase (or rent) music solely for its utility: i.e. in order to listen to it. Unless they are a collector of rare or old albums, they do not do so not because it has any intrinisic monetary value.
In the original example, it is also naive to claim that the value of the iTunes Music is "$360" or some other precise monetary value. However, the original proposition is substantially correct.
At the end of the time period, the rented music has neither any fungible monetary value nor any value derived through utility, since one can no longer listen to any of it. iTunes music, still has precisely the same utility as the day it was purchased. The owner may listen to it on a PC or Mac, play it on an iPod, and burn it to CDs which can be played on any CD playing device. With a small (but inexorable) loss of quality, one can rerip such a burned CD and encode via mp3, ogg, or whatever you wish, and listen to it on any device you want.
Durable utility is of direct value to the owner. Only that owner can accurately ascribe a monetary value to that utility. Thus claiming that it has a precise "value" of $360 is specious. Despite this, the rented music has precisely zero current or future value unless the subscription fee continues to be paid. iTunes Music, by virtue of retaining its utility, has a positive value. This utility, though not directly fungible, can be ascribed a monetary value by the individual owner. The fact that this monetary estimate of value will vary among consumers or by a consumer over time is irrelevant.
Certain people will prefer to pay a per song fee for such durable utility. Other people may prefer to pay a monthly service fee for listening to music. This is a matter of personal preference, thus not subject to rigorous argument.
BTW: Personally, I do not find renting music to be compelling as a long term proposition. I might however, consider subscribing for a very short period of time to augment iTunes offerings. I could rent music I am less familiar with to explore various artists or genres in detail in order to identify music I would like to own long term.
My first post was deliberately short, in order to prevent others from suffering a similar fate.
In it I stated that I had followed instructions and that the web site claimed my system was supported. So, yes, this means that I ran ioreg to verify before proceeding.
It's like you didn't even bother to read it. You need to give the kernel extension the appropriate permissions, otherwise it won't load and touching the touchpad will give you an instant kernel panic.
That is not the case.
Permission and ownership were correct. In both the initial try and subsequent testing failure occurred immediately after the driver load, and was not triggered by any trackpad input.
Please state if you installed this from the source distribution or the binary distribution. The binary XY driver forced a panic on my system, as I reported in an earlier post.
Slashdot Mirror
You provided a reference to an FSF document to support your reasoning. The cited web page says, in part:
Granted, the APSL does not prohibit users of the software to link with proprietary libraries, thus is not a "copyleft" license. So, what? This is less restrictive than the GPL, not more. This, in and of itself does not preclude Linux users from using it on their systems.
The FSF concludes that it is ok to use and improve software which other people release under this license.
You would be allowed to compile the daemons using gcc and glibc libraries and use them with no problem from the APSL. You would also be allowed to link GPLd programs against the supporting APSL licensed frameworks.
The only limitation is that if you ship an improved version of this code that you make that code available to others under APSL terms. i.e. you provide source code so that Apple and and other users of the APSLed code benefit from the changes.
Insisting on re-inventing every wheel just so that everything is covered under GPL is a waste of effort. It steals time of those working on GPLed code from doing other work, and selfishly prevents others from benefit from you good ideas if you improve a fork of the work rather than the original work itself.
It strikes me as foolish that GNU/Linux people spend so much effort to mimic other people's work, re-implementing large subsystems just to get them under the GPL umbrella, rather than cooperating with others to re-use and improve the best code available.
Warning!
Do not create new drum tracks in GarageBand!
That chip on your shoulder seems pretty big.
You state that there are "practically zero ways to reduce CPU-heavy and memory heavy graphical effects."
This does not hold water considering that there are a number of actual and documented ways to reduce these effects such as turning off window miniaturize effects, magnification, etc. The support for intelligent reduction in animation overhead for dashboard which I described also contradicts your assertion.
I do not claim that old G3s are suitable for use in a rendering farm, realtime video encoding, gene sequencing or other graphically demanding or compute intensive tasks. However, I also do not consider that the majority of things that people use computers for to be trivial. Most people who use computers spend a great deal of time reading, writing, coding, and doing research. These rarely tax the hardware.
Web rendering on a G3 is fairly poky. However, doing so efficiently (and pleasantly) is still possibly. For instance, when on my Pismo, I use either OmniWeb or Safari via a tabbed interface. I read top level sites, and as I see stories that might interest me, I open them in another tab and continue looking for interesting items. By the time I close that window, I now have a number of articles in the queue ready to read. So, the way I use the web, even fairly demanding tasks like that do not slow me down and are enjoyable.
Do I prefer the greater responsiveness of a G4 or G5 to that of a lowly Pismo when browsing? Absolutely. However, I gain far more pleasure from performing the task in OmniWeb or Safari, than I do from any other browsers on whatever hardware.
Your professed failure to understand appears to be merely that: you seem not to understand that others have different priorities than you do. Your saying it many times does not make your personal opinion any more compelling than the first time.
Your work flow may very well demand the latest GPU or CPU performance. The tasks of the majority of users does not. This does not make them wrong, make you wrong, or vice versa.
As long as an older system has sufficient memory to make paging rare, they are quite suitable for many users. The bottom line is not whether they are the best choice, but whether the user is capable of performing their intended task and enjoys doing so. Will a user enjoy Tiger more than Panther on existing hardware? Most users probably would. Would they enjoy it more on newer hardware? Most users probably would. Despite this, many user will not derive sufficient utility or enjoyment from it to be worth the additional cost of new hardware. Who cares? Their preference is all that matters to them. Why does it seem to matter to you?
The shortest answer is Yes.
A slightly longer answer is No, but you can effectively disable it by simply excluding all or most of your system from the spotlight database.
The spotlight Preference Module contains a Privacy Tab. In this panel, you can add directories which are to be excluded from the index database. Presumably, adding / here would suffice to both save disk space and ongoing CPU costs. However, doing so broadly seems rather pointless. Certainly if you have confidential data on a network accessible volume you would be prudent to omit it. Likewise, if you have a subtree containing a large database, or collection of large files whose content is not usefully presented by spotlight, It might be worth excluding them.
On the whole, though, the incremental cost of maintaining the index is trivial and is correlated to the addition/modification of the files. This, in most environments, is both sporadic and requires negligible CPU and disk resources. If parts of your workflow have a file access pattern which makes spotlight less valuable to you, simply tailor spotlight to meet your needs.
Also, both system wide, and application specific spotlight queries are astonishingly efficient. Performing real time queries and displaying the results uses very little CPU and happens quite quickly. Even long queries (lasting seconds) do not appear onerous, since the result list is updated frequently as the search occurs, and incremental results are available.
The user decides which kinds of data are displayed for searches, and can tailor searches to a subset of volumes or systems when multiple disk (and remote volumes) are mounted.
Anyway, you can tailor the system to index less (or effectively nothing). Doing so, however, is unlikely to be of benefit. The system once primed appears so efficient that you would not save enough disk/time to make it worth your while. I suggest that rather than worrying about how to disable it to save processor cycles, you try it out for awhile and discover how it can save your brain cycles.
Spotlight is not a specific function or program. Rather it is a pervasive system. The base system provides a daemon which creates an initial index of all files, and subsequently handles requests for updating new or modified files. This process runs heavily niced in the background. While you can access a general Spotlight query tool using Command-Space, the real benefit of spotlight is its pervasiveness. Use the spotlight tool in the Preferences app to find out where a particular setting lives. (Note that Windows converts searching for a Windows-centric name will be presented with the Mac-centric counterpart.) Likewise in mail, the finder, and other programs, spotlight is available to help you find the context specific data you seek. Since developers can easily create spotlight plugins to parse data formats and export metadata, expect that most future applications will integrate well with the system.
It is important to note, that I found spotlight to be quite useful for a number of tasks, even though I only used it sporadically for testing purposes. Thus, I am inclined to be favorably biased towards it. On the other hand, I usually use a dual G5 Powermac and a fairly recent G4 17" powerbook. The fact that most of my use of spotlight was on a 400Mhz G3 powerbook suggests that my assessment of its efficiency is likely credible.
As always your time and your mood are the only true measures of a software tool, not my opinion.
You joke about how Tiger degrades gracefully on a PDP-11 but are not far from the truth.
For instance, I have one of the early firewire based PowerBooks. This system does not have a GPU powerful enough to offload 3D animation. It also has a 400Mhz G3. Thus, it does not have Altivec to offload graphics vector processing. As a result, instead of slowing down the system to perform 3D rotation of dashboard widgets when the user customizes them, it draws the obverse of the widget into offscreen memory then quickly replaces the live widget window with the widget customization window.
Thus, once the initial background indexing for spotlight has been completed, Tiger on this old PowerBook is fast enough for interactive use. In fact, I have been using it for a number of tasks in preference to my 17" G4 since I can get the job done faster in Tiger than on my newer system running Panther.
Actually Intel compatible clusters in the supercomputer rankings are not all that compelling. True, linux cluster did fare very well for several years as measured by price/performance. Also it is true that about 63% of the top 500 supercomputers are Intel or Intel compatible.
Despite this, the majority of systems at the top of supercomputer top 500 chart are based on the POWER architecture, not Intel chips.
The POWER based systems, including BlueGene and PowerPC systems, are all much better on both price/performance, and Watt/performance basis. Intel chips do have a per chip advantage over PowerPC on many work-flows. However, when scaled, they directly consume more power, and indirectly require even more power to run higher capacity air conditioning.
Looking at the top of chart reveals that Intel compatible systems are a small minority. In the top 10, as of November 2004, are one NEC, HP, SGI, Alpha, Xeon, and Itanium, and 5 members of the power family. IBM powers more than half of the top 100, with all other chip families dividing the remaining half.
It is cheaper to build and operate a supercomputer cluster using either Apple or IBM gear (running either Macos X or Linux) than to do so using Linux on Intel these days.
I live outside of boston.
My friends are scientists, engineers, writers, etc.
Thus some (the non-techies) have been mac users for a long time.
The majority of the geeks have migrated from BSD and Linux to Macos X.
All told, my immediate environment is about 50% mac, and a few of us have been working with computers for 10-20 years but never owned a Microsoft box.
Perhaps 5% is a more realistic estimate.
No. It's amazing considering Apple only has ~3 percent market share.
That number is incredibly misleading. It underestimates real users since, lumped into the Microsoft share, are the majority of the Intel PCs used as Point of Sale terminals, ATMs, informatioin kiosks, etc.
It also fails to account for the fact that Mac owners tend to keep their systems for several years longer than PC users. Thus the percentage of home users of Macs is probably somewhere between %8-%12, not the oft quoted 3%.
Wasn't NeXT the 1st computer to ship with a scroll wheel mouse?
No.
NeXT never shipped a scroll wheel mouse at all. Next used a 2 button mouse. By default the second button was identical to the first. The mouse preferences allowed the user to use the second button to display a copy of the application menu at the current cursor position. Which button (left or right) did this was user definable to allow lefties and righties to use the system comfortably.
I was a NeXT user for about 9 years before switching to Macos X. I don't really miss the second mouse button. The only aspect of the old NeXT/OpenStep experience that I prefer to Macos X is tear-off menus.
Denial, "not just a river in Egypt". What a load of horse hockey.
You provide links to evidence which appears to justify the opposite conclusion.
Your first link, http://dot.kde.org/1097096753/1097113373/ is an email asking about what's going on with merging Apple changes back into KDE. The immediate reply states:
To me this suggests that apple added so much code to KHTML/KJS that it made more sense architecturally to split them into two frameworks "WebCore" and "JavaScriptCore", These are still released in their entirety under the GPL. There are so few developers on the KHTML side that they have been unable to keep up with the changes. Seriously, how is that Apple's fault?
Forking happens all the time. It would take far more effort and money for apple to continue to backport changes piecemeal to the KHTML/KJS trees than to fork the distribution. The sensible approach would be for the KHTML/KJS team to accept the architectural changes and simply start using the WebCore and JavaScriptCore, which is cleaner, better maintained, and of higher quality due to all the work Apple has done. Rather than grouse about how much effort it takes to backport changes piecemeal, why not adopt the superior libraries which Apple has produced and continue to improve them?
NIH syndrome works both ways.
The final link you give is entirely out of context. You state that Apple recommended doing a diff to see what has changed. Actually, the wording was "The best way to see every change line by line is to diff against the originals.". Immediately following, was about 14 pages (in my browser) of itemized descriptions of the changes performed, organized by functional description, subdirectory or even by function call.
Thus, your assertion is at least deliberately disingenuous, if not an outright lie.
I'm still waiting for them to submit Objective C++. As a matter of fact, one of the first successful FSF legal actions was forcing NeXT to release the source code to their GCC derivative.
This statement is misleading in several important respects.
First, NeXT's original plan to develop a variant preprocessor for the gcc toolchain and maintain it independently was indeed incompatible with the copyleft. However, it did not result in legal action. NeXT was informed about how and why it was incompatible and subsequently released the code. There was no law suit or formal legal action taken at all.
Also note that this was over 15 years ago, before even the BSD 4.4 source code was publicly freed in the law suit between the Regents of the University of California and AT&T. NeXT could not even share their changes to the BSD 4.3 kernel to anyone without a valid Unix Source License.
Regarding the subsequent divergence of the FSF and NeXT/Apple objective-c, it appears that the problem rests squarely with the FSF. The gcc maintainers have repeatedly rejected '#import' and other architectural features which are pretty major. They also deliberately allowed several aspect of the underlying runtime to diverge from the NeXT base because they either rejected or even backed out changes proposed by NeXT/Apple developers.
The code is there for all to see in the Apple branches published in Darwin. Apple developers propose changes to the primary gcc source tree, and argue for their adoption, but are often rebuffed.
I think that support for Frameworks, even for plain C, would benefit both Linux and BSD tremendously. However, unless gcc developers accept those changes they will remain unavailable.
Search the gcc developer mailing list over the past 5 years for discussion of Apple submitted changes, if you don't believe me.
The majority of state have similar laws - generally called "Uniform Trade Secrets Acts" or UTSA.
In each case the law states that it is illegal to publish, make use of, or otherwise benefit from information which the user has reason to know is protected as a trade secret.
The defendant knew for a fact that this information was protected, and that the information was gathered from those who were contractually obligated not to divulge it (it was covered under NDA).
Thus, it was a civil crime for his source to give the information. It was also a civil crime to publish those statements.
The remedies (under the California UTSA statute) are divided into both monetary damages, and court ordered innunctions.
In this particular case, fines, and injunctions to stop making use of this particular information are pointless. Instead Apple asked the court to order the defendant to divulge his sources, since that is the only remedy which can enable them to stop further damage.
Thus, this is not really an issue of first amendment freedom or freedom of the press. It is a request from Apple to have a judge provide legal remedy under the UTSA, in the only way which is relevant for them. Basically the author who was subpoenaed solicited information in breach of contract, knowingly published that information in breach of the UTSA, and then sought protection under the guise of free speech This seems specious to me. I side with Apple on this one.
I do believe that if used unwisely this law could have a chilling effect. However, the injunctive remedy is decided on a case by case basis at the judges discretion. Also the law punishes those who misuse the law by awarding damages and legal fees to the defendants if the plaintiff is in the wrong. This seems to provide a reasonable measure against abuse.
What could be simpler?
Either install a non-Windows OS on your existing hardware or buy a Mac. Linux, any BSD, or Macos X are simpler choices. BSD or Linux are harder in the short run but require less on-going maintenance once the user is settled in. Macos X requires changing both hardware and software, but is likely to be an easier transition for most users.
Whether you like it or not, the Wintel platform is no longer a very good choice for the average computer user, and has become a quite unpleasant environment for most people.
My friends is brand new, installed 2 weeks ago.
His says he does not have the ability to record all episodes of a particular series without duplicates (like a tiov seasons pass does). As a result for one series (in rerun syndication) he set up M-F 1-2, M-F 2-3, and then every week or so, has to schedule additional recordings when the network moves the show to an early morning time slot (6 times over the next 2 weeks).
So it is appears to be unaware when particular episodes are shown in a different time slot.
Actually, drivers for some USB 2.0 ethernet dongles are now supported in the most recent release, 7.1a. My Belkin USB2.0 interfaces are about twice as fast as they were under version 4.x.
A one hour show in Basic mode takes about 8 minutes.
In Best Quality, an hour of video transfers in under 30 minutes.
A friend of mine just got a Dish network DVR. I must say, that compared to Tivo, it truly sucks. It's not even like comparing Mac to windows for usability. it's more like comparing Mac to a graphical DOS app.
The remote control has far too many buttons, and some common functions on buttons that are poorly placed.
If you are watching a show, and either accidentally or deliberately go into the menus, the video buffer of what you were watching gets immediately flushed.
The layout of the menus, the UI, the inability to record shows based on name, etc. show a shoddy inattention to detail. Basically, you can search by program name, but only record by time of day.
I was shocked by how poor the interface was, and how unpleasant the system was to use.
There is a product called Xtrovert by Xtrememac which solves this problem. It attaches to an Xserve RAID unit to turn it into a stable, vertical tower. It is made from anodized aluminum with a matte silver powder coat finish that matches Apple's aluminum products. It costs around $280 bucks.
One source is here: http://www.smalldog.com/product/43611
California, Massachessets (where the editor is in school), and most other states have each passed similar laws which specifically prohibit individuals from disseminating or unfairly using trade secrets. The california statute is titled the "Uniform Trade Secrets Act" (a.k.a UTSA). Under that statute the editor of Think secret appears guilty of the act of "misapropriation of a trade secret" under the definition (b)(2)(B)(ii).
The editor of Think Secret published confidential information, and clearly knew that the information was covered under NDA. His prior published statements reveal this general knowledge, and he had been contacted repeatedly by Apple lawyers asking him to remove infringing information and desist from future similar activities. He appears to have no viable defense against charges under the UTSA.
The remedies for the complainant for such misappropriation include, both damages and injunctions. Monetary damages may be levied based on losses (or potential losses) of the complainant, enrichment of the defendant (e.g. ad revenue or donations from the web site), exemplary punitive damages not to exceed twice the two kinds of damage mentioned above.
Injunctive relief takes 3 forms. The defendant may be enjoined from further use or dissemination of the information. The defendant may be forced to pay a royalty for use of the information. Finally, the plaintiff may request unspecified injunctive relief at the discretion of the court.
(c) In appropriate circumstances, affirmative acts to protect a trade secret may be compelled by court order.
So, it appear that the defendant's alleged acts are clearly in violation of this statute, and he has misappropriated a trade secret as defined by the law. The defendant repeatedly ignored requests to desist from such activities. The monetary relief afforded by the law is inconsequential in this case. Thus Apple's goal is to seek injunctive relief. The first two forms of relief (enjoining them from further use of this specific information, or charging them royalties for its past use) are each irrelevant here: this specific cat is out of the bag, and no reasonable royalty will remedy this breach. Thus, they seek injunctive relief to compel the defendant to reveal his sources, so that they can sue the one who revealed the information to the editor and/or prevent further breach of contract.
Suing the Think Secret editor is clearly the only way to either sue the original offender for relief or ensure that they refrain from future illegal activities.
Claims that Apple is bullying the defendant or overstepping their legal rights appear poorly founded. They are defending their right under the law in the only way available to them. It is up to the judge to determine whether to compel the defendant's cooperation or not.
Is it clear that the public good is truly served by letting the defendant protect his source in this case? The states in which the corporation is based, the state in which the leak originated, and the state in which the defendant lives and works each have similar laws which specifically prohibit the defendants alleged activities. Each version of the law (UTSA) specifically allow the plaintiff to seek injunctive relief to both punish the offender and limit future damage of the plaintiff's interests.
The case is of interest because there is generally a clear public good in free speech and wide dissemination of information. At issue is whether judgments under the UTSA will erode journalistic integrity, or be used by corpo
Carl Jung wrote about a collective unconscious. Was he a crack pot?
It depends on how you wish to interpret his works.
Jung, unlike Freud and his followers, was concerned about scientific rigor, and his work appears to be valuable both pragmatically and theoretically. Many perform a naive unscientific reading, and attempt to use his works to support the rankest forms of BS. I find these interpretation both distasteful and unfounded.
When Jung began his study, the majority of clinical observation produced data on abnormal patients, not healthy individuals. The richest source of data on normal psychology came from literature and mythology. Clearly, familiar plots and psychological patterns appear as themes in the worlds fables, religions, and written works.
Jung fits right in, when read alongside the most recent works in evolutionary psychology and related disciplines. Those in the related disciplines of Neurology, Evolutionary psychology, etc. currently explore subsystems of our brain and how their structure and functions map to our behaviors. These modern scientists alternately describe observed behavior and attempt to discover their underlying causes in our physical and genetic systems. Jung attempted find common threads in myth, religion, and literature and discover their underlying ground in the human mind.
When Jung began his career the vast majority of psychological was directed at abnormal or industrial psychology. Most discussion of normal psychology was philosophical. As a result Jung looked at archetypal themes which recurred throughout myth and literature, and still resonated with individuals of his day. By his reasoning, these archetypal themes owe their power and endurance to the psychological makeup of individuals.
These themes are both popular and powerful because people are moved by them, relate to them, or understand them intuitively without requiring reflection or study. Since they are basic and independent of rational focus, he termed them unconscious. Since they appeared shared by all people he called it "collective unconscious".
People are inclined to behave similarly to one another in certain circumstances and interpret the world in similar ways. It is logical to conclude that one reason we tend to do so could because our brains are so similar to one another. Jung wrote of mind and psychological tendency, a higher level of abstraction. These are no incommensurable.
In my reading of Jung (direct English translations, since I don't know German) I find no mystical or unscientific connotations. Unfortunately, many readers misread the term collective unconscious. They jump to the conclusion that it has mystical connotations and dismiss the work as unscientific. Alternatively they are unscientifically minded and eagerly attempt to use his work to justify pseudo-scientific or mystical hogwash.
I am not saying his work is all correct, just that a number of his theories and observations are both consistent with the latest scientific theories and informative.
But while you're using the service the value for iTMS is 15$*number-of-months while Napster has a value of a few hundred thousand dollars if you calculate the value by utility
What a fanciful (and ludicrous) claim!
Since "a few" is imprecise, let's simplify that to "one". This smaller claim is that the "value by utility" of napster is $100,000 per month. There are 44,640 minutes in a 31 day month. Even if one could derive value from listening to songs 24 hours per day, this works out to $2.24 per minute for listening to music.
Also, I doubt that listening to music while sleeping is of any tangible subjective value. Here I'll be generous, and allow only 6 hours of sack time per day. This raises the cost to 2.98 per minute to listen to music. Since your numbers were presumably based on per/song pricing, If we choose the value of 3 minutes per song, your purported "value by utility" is $8.96 per song.
These numbers do not make any sense, thus your claim does not make any sense. I'll refrain from figuring out what you meant by "a few hundred thousand", since even 100,000 is so far off the mark.
In Oracle, a table or index may be subdivided into multiple sections called partitions.
Each partition is associated with a tablespace, which defines the underlying storage.
A range partition defines a range of keys which belong to a partition.
The general ability to store subsets of data or indices on separate table spaces provides an abstraction layer which may be used to optimize locality of reference, bandwidth, etc for subsets of data rather than for the set as a whole.
Other criteria can be used to define partition membership like, hashing or explicitly list of keys. Some examples of range partitioning are tuning the underlying storage characteristics base on regions, date range, product codes, etc.
Actually, the value is $0.
Before you argue with me, remember the traditional way to set value is to sell it and see what the open market brings. EBay is great because it generally establishes the real market value.
Bull. This claim is naive and misleading.
Your concept of value is accurate only for fungible commodities which have no direct utility.
People purchase (or rent) music solely for its utility: i.e. in order to listen to it. Unless they are a collector of rare or old albums, they do not do so not because it has any intrinisic monetary value.
In the original example, it is also naive to claim that the value of the iTunes Music is "$360" or some other precise monetary value. However, the original proposition is substantially correct.
At the end of the time period, the rented music has neither any fungible monetary value nor any value derived through utility, since one can no longer listen to any of it. iTunes music, still has precisely the same utility as the day it was purchased. The owner may listen to it on a PC or Mac, play it on an iPod, and burn it to CDs which can be played on any CD playing device. With a small (but inexorable) loss of quality, one can rerip such a burned CD and encode via mp3, ogg, or whatever you wish, and listen to it on any device you want.
Durable utility is of direct value to the owner. Only that owner can accurately ascribe a monetary value to that utility. Thus claiming that it has a precise "value" of $360 is specious. Despite this, the rented music has precisely zero current or future value unless the subscription fee continues to be paid. iTunes Music, by virtue of retaining its utility, has a positive value. This utility, though not directly fungible, can be ascribed a monetary value by the individual owner. The fact that this monetary estimate of value will vary among consumers or by a consumer over time is irrelevant.
Certain people will prefer to pay a per song fee for such durable utility. Other people may prefer to pay a monthly service fee for listening to music. This is a matter of personal preference, thus not subject to rigorous argument.
BTW: Personally, I do not find renting music to be compelling as a long term proposition. I might however, consider subscribing for a very short period of time to augment iTunes offerings. I could rent music I am less familiar with to explore various artists or genres in detail in order to identify music I would like to own long term.
My first post was deliberately short, in order to prevent others from suffering a similar fate.
In it I stated that I had followed instructions and that the web site claimed my system was supported. So, yes, this means that I ran ioreg to verify before proceeding.
It's like you didn't even bother to read it. You need to give the kernel extension the appropriate permissions, otherwise it won't load and touching the touchpad will give you an instant kernel panic.
That is not the case.
Permission and ownership were correct. In both the initial try and subsequent testing failure occurred immediately after the driver load, and was not triggered by any trackpad input.
Please state if you installed this from the source distribution or the binary distribution. The binary XY driver forced a panic on my system, as I reported in an earlier post.