Yes most of the 2000 packages that come preinstalled with your Linux distro can be obtained or source compiled for Solaris, including the GUI stuff you prefer. But it's a lot of work. Why do the work when PC hardware is better for a desktop in the first place?
It's a fine idea to let users type keywords or names in order to go to the site they want.
The stupid part is the assumption that someone else can manually make the appropriate mapping between keywords and sites.
Google is so good because it fundamentally doesn't rely on this (even when the "submit my site" feature and meta tags trick webmasters into believing that it does).
You're wrong. First, Google is a service to users: in order to be maximally effective, it should rank those results as highest that users will search for most often, and those are the most popular results, no matter what you and I, or any other minority, may think of their quality. Second, by rating based on explicit links to a page, Google does use a form of quality assessment, namely, that of the author of the linking page's.
The author seems so misinformed about Google and search engines in general that he must be suspected to be working for Altavista.
Google supports all the "advanced" queries mentioned, including negation, and the site:foo -improbableword trick really isn't that hard to find.
Google's database is actually surprisingly up to date, probably more than Altavista's.
Explicit submissions of sites to a search engine shouldn't have much impact on that search engine's overall quality. Search engines find pages by roaming the web. If explicit submission is going to make your page appear a few days or weeks sooner, that doesn't mean it will pop up in search results. It probably shouldn't. Search results must be relevant to the user; this is what the previous Slashdot thread on Google was about.
Explicit submission is a toy to make webmasters feel good.
This idea just doesn't work, except when you translate it into terms that the search engine and the user actually agree on. Two methods are known to reach that agreement:
Create a vocabulary, expose it to the user, expect the user to understand it before using it. This is what library cataloguing systems (Dewey decimal etc.) and web-based subject indexes like DMOZ do. You can't use them effectively without being familiar with them. There are better and worse ways of creating these indexes, in the sense that some will be easier to learn than others, but creating a subject tree that is immediately obvious to any new user is simply impossible. In such a catalog, you can play your dialog, to some extent, by navigating the subject tree, but you are likely to run into trouble.
Use terms whose meaning are already obvious to the user. This means: full text indexing. Actual occurrences of actual strings on the actual page are the only kinds of search terms that a user can understand without having to be trained in advance. Not only is it less work to create than any kind of indexing based on abstract search terms, it is also going to be more useable, because users actually understand what the search engine is doing in response to their search terms. SMART, Altavista and Google have proved the superiority of this method over the use of abstract subject indexes. Additional technology, such as word stemming and the kind of ranking that Google uses, can seriously improve the results, but 'fuzzy logic' and any other idea based on abstract concepts or search terms is fundamentally inferior - they are still important when the full text content isn't available, of course.
With Google you can play your dialog by pphrasing it in terms of keywords to be found on the pages, and this is exactly what I do when I use Google:
[user] "Mary Jane Carpenter"
[google] (the first 1-10 links return bio, official website, etc. etc.)
[user] (doesn't find the MJC wanted; determines a distinctive attribute that is likely to appear on somebody's webpage) "Lewis"
[google] returns more specific results
There are plenty of keywords to try - if I have trouble finding something, it probably means that it isn't on the Web in the first place, not that Google doesn't give me a way to locate it.
(But that claim can be researched, of course.)
Of course you have to learn how to spell. Google can employ soundex to return "Louis" matches in response to "Lewis" queries. But if it did this kind of thing automatically, its results would deteriorate, because false positives with unknown causes are much more disruptive to the user than an excess of genuine hits; the latter can simply be filtered away by adding more search terms.
the only problem was that sometimes a file would be edited using Notepad or something, that shouldn't have been, which introduced ^M characters that confused CVS.
side note: you can have the CVS server convert such things automatically
The British pound has been doing very well in recent years, but it hasn't always, and I doubt that the strong pound is a major factor in Britain's decision to pull out of Europe.
One major reason for the Euro is political. Most countries are prepared to sink billions of Euros into compromise if this is what it takes to avoid another "World" War, or Cold War. We've had enough of that already.
Another is that with the infrastructure improving all the time, those petty national borders and differences are becoming more and more serious obstacles to trade, in particular for the big and the powerful. Every step towards the freedom of trade and unity of language enjoyed in the USA is welcome.
From the article "By the completion of the ENIAC project in 1946, work that once took 30 hours to compute instead took 15 seconds."
Since most of us were born after the advent of computers we take for granted that mundane computation tasks can be automated for fairly low cost and at great time savings. However, for all that technological progress has been hailed in the last 20 years, is there any task that we have received this kind of improvement in efficiency on?
Yes, the WWW and Google achieve a similar speedup for looking up information. Only for more advanced topics in non-computer related areas we have to resort to libraries or phone calls these days.
This position of yours is very much against my common sense, but also contrary to copyright law, which explicitly forbids you to publish derivative works, and that's precisely what you're advocating here. in the case of editing it could be arguedthat the original hasn't been published yet - I don't know if that plays a reole, I am not a lawyer.
Of course you can make something new that is merely inspired on the original work, and the line between that and derivative work can be hard to draw - but what you're suggesting is certainly derivative.
The answer is that nobody knows.
Hasbro certainly doesn't seem to care.
Freeciv doesn't actually compete with any of the commercial Civs, which have thousands of times as many players.
You can represent the structure, but you can't manipulate it using standard relational logic.
For example, take a table representing a parent-child relationship. Now try to sort the persons in the table by their number of descendants. SQL has only recently been extended to allow this query to be posed. Perhaps your relational database can handle this kind of query, where you have arbitrary-depth path walking, ybut ou can't expect it to handle them efficiently.
No you ***, (s)he's not telling you that engineers are good, or MBAs are bad. The point is to say that if you lack certain expertise and hire an expert to make up for it, it's not going to do much good if you subsequently ignore the expert's advice.
You're right, the article itself makes the same mistake of linking Microsoft to what is a Unix->Unix switch.
Switches from a commercial Unix to Linux take little re-training, have no drawbacks except to suits or in exceptional situations, and many real benefits to those who have to work with the systems or pay for them. Many companies have made this switch in the last 10 years and it's hardly worth reporting.
Switching from MS products to Linux on the other hand is more difficult (except for those who know Unix already) and therefore, much more noteworthy. It is also very rare.
Therefore, it is very misleading to suggest that the switch at Amazon has anything to do with Microsoft software.
The real concern is the inherent dangers of central databases. That way, you create a single point of failure, attack, and control.
This is independent of who is in charge of the thing: Microsoft, the central goverment, whoever else. It is an argument about the inherent dangers of a star-like infrastructure. A technical issue rather than a political one.
The whole point of public key cryptography is to allow secure transactions and verified identities without the need for a central controlling database like Passport. The alternative you're asking for is a distributed net of parties able to check each other's credentials, just like PGP and SSL work today.
To stress the inherent vulnerability of this approach, the article makes the point that Microsoft's software cannot be expected to be without flaws. This is where you see Sun speaking. We all know that Sun software isn't really any better, and it would be foolish to suggest that this problem can be solved by using different software, or improving the implementation. Note that the article doesn't do this.
If by 'technical' arguments you mean arguments concerning how.NET or Passport is implemented, you've missed the point - they are not the issue here.
Copyright resides - as expected - on computer programs; my country's copyright law explicitly includes "computer programs and preparatory materials".
Copyright exists on a "work", the visible/tangible results of someone's expression. It does not apply to techniques or methods, such as a programming interface to a piece of hardware.
The problem with your message is the phrase "this is where Linux messes up". On the contrary, Linux software can do this at $0 cost, with next to no admin costs, in other words, you mostly pay for the hardware.
I don't think it is fair to say that America acts in the world only to serve its own ends. When viewing the matter on a national scale it may be true that America always keeps its own interests in mind, but this does in no way diminish he selfless sacrifices and devotion that Americans have been displaying - we Europeans only have to think of WWII.
And in many instances, we can only be thankful for American aid and intervention. But this is not always so. Furthermore, the outright arrogance and narrowmindedness that we see so often from Americans - Republicans, especially - are hard to stomach.
Terrorism can never be excused, and we don't know what drove the attackers to their deeds, but the present government would do well to fix its attitude problem, just in case.
Slashdot Mirror
play Freeciv. you'll last longer
Q/A number 2 in the Freeciv FAQ.
don't underestimate the importance of gcc
For your information, freaking Germany has one third the population of freaking USA, and many of them read Slashdot.
Yes most of the 2000 packages that come preinstalled with your Linux distro can be obtained or source compiled for Solaris, including the GUI stuff you prefer. But it's a lot of work. Why do the work when PC hardware is better for a desktop in the first place?
The problem is not the business model.
It's a fine idea to let users type keywords or names in order to go to the site they want.
The stupid part is the assumption that someone else can manually make the appropriate mapping between keywords and sites.
Google is so good because it fundamentally doesn't rely on this (even when the "submit my site" feature and meta tags trick webmasters into believing that it does).
On Unix: troff + eqn.
On IBM mainframes: Script (which spawned GML, which spawned SGML, which spawned XML), but I don't know how well it did mathematical formulas.
You're wrong. First, Google is a service to users: in order to be maximally effective, it should rank those results as highest that users will search for most often, and those are the most popular results, no matter what you and I, or any other minority, may think of their quality. Second, by rating based on explicit links to a page, Google does use a form of quality assessment, namely, that of the author of the linking page's.
Explicit submission is a toy to make webmasters feel good.
With Google you can play your dialog by pphrasing it in terms of keywords to be found on the pages, and this is exactly what I do when I use Google:
[user] "Mary Jane Carpenter"
[google] (the first 1-10 links return bio, official website, etc. etc.)
[user] (doesn't find the MJC wanted; determines a distinctive attribute that is likely to appear on somebody's webpage) "Lewis" [google] returns more specific results There are plenty of keywords to try - if I have trouble finding something, it probably means that it isn't on the Web in the first place, not that Google doesn't give me a way to locate it.
(But that claim can be researched, of course.)
Of course you have to learn how to spell. Google can employ soundex to return "Louis" matches in response to "Lewis" queries. But if it did this kind of thing automatically, its results would deteriorate, because false positives with unknown causes are much more disruptive to the user than an excess of genuine hits; the latter can simply be filtered away by adding more search terms.
side note: you can have the CVS server convert such things automatically
The Labour party is just as divided on Europe.
The British pound has been doing very well in recent years, but it hasn't always, and I doubt that the strong pound is a major factor in Britain's decision to pull out of Europe.
One major reason for the Euro is political. Most countries are prepared to sink billions of Euros into compromise if this is what it takes to avoid another "World" War, or Cold War. We've had enough of that already.
Another is that with the infrastructure improving all the time, those petty national borders and differences are becoming more and more serious obstacles to trade, in particular for the big and the powerful. Every step towards the freedom of trade and unity of language enjoyed in the USA is welcome.
Since most of us were born after the advent of computers we take for granted that mundane computation tasks can be automated for fairly low cost and at great time savings. However, for all that technological progress has been hailed in the last 20 years, is there any task that we have received this kind of improvement in efficiency on?
Yes, the WWW and Google achieve a similar speedup for looking up information. Only for more advanced topics in non-computer related areas we have to resort to libraries or phone calls these days.
This position of yours is very much against my common sense, but also contrary to copyright law, which explicitly forbids you to publish derivative works, and that's precisely what you're advocating here. in the case of editing it could be arguedthat the original hasn't been published yet - I don't know if that plays a reole, I am not a lawyer.
Of course you can make something new that is merely inspired on the original work, and the line between that and derivative work can be hard to draw - but what you're suggesting is certainly derivative.
The answer is that nobody knows.
Hasbro certainly doesn't seem to care.
Freeciv doesn't actually compete with any of the commercial Civs, which have thousands of times as many players.
You can represent the structure, but you can't manipulate it using standard relational logic.
For example, take a table representing a parent-child relationship. Now try to sort the persons in the table by their number of descendants. SQL has only recently been extended to allow this query to be posed. Perhaps your relational database can handle this kind of query, where you have arbitrary-depth path walking, ybut ou can't expect it to handle them efficiently.
No you ***, (s)he's not telling you that engineers are good, or MBAs are bad. The point is to say that if you lack certain expertise and hire an expert to make up for it, it's not going to do much good if you subsequently ignore the expert's advice.
Switches from a commercial Unix to Linux take little re-training, have no drawbacks except to suits or in exceptional situations, and many real benefits to those who have to work with the systems or pay for them. Many companies have made this switch in the last 10 years and it's hardly worth reporting.
Switching from MS products to Linux on the other hand is more difficult (except for those who know Unix already) and therefore, much more noteworthy. It is also very rare.
Therefore, it is very misleading to suggest that the switch at Amazon has anything to do with Microsoft software.
As (nearly) always, this is about a shift from commercial Unix to Linux. It has nothing to do with Microsoft.
Please read before you post.
The whole point of public key cryptography is to allow secure transactions and verified identities without the need for a central controlling database like Passport. The alternative you're asking for is a distributed net of parties able to check each other's credentials, just like PGP and SSL work today.
To stress the inherent vulnerability of this approach, the article makes the point that Microsoft's software cannot be expected to be without flaws. This is where you see Sun speaking. We all know that Sun software isn't really any better, and it would be foolish to suggest that this problem can be solved by using different software, or improving the implementation. Note that the article doesn't do this.
If by 'technical' arguments you mean arguments concerning how .NET or Passport is implemented, you've missed the point - they are not the issue here.
Copyright exists whether it is claimed or not.
Copyright resides - as expected - on computer programs; my country's copyright law explicitly includes "computer programs and preparatory materials".
Copyright exists on a "work", the visible/tangible results of someone's expression. It does not apply to techniques or methods, such as a programming interface to a piece of hardware.
See a posting by an Afghan in America that seems to sum up the situation in Afghanistan quite well.
The problem with your message is the phrase "this is where Linux messes up". On the contrary, Linux software can do this at $0 cost, with next to no admin costs, in other words, you mostly pay for the hardware.
I've never seen a Linux crash in 5 years of desktop use, but it did hang often (say, 3 times a year) when memory gets overcommitted.
Newer kernels may address this problem; I haven't been running xanim on Linux in a while.
I don't think it is fair to say that America acts in the world only to serve its own ends. When viewing the matter on a national scale it may be true that America always keeps its own interests in mind, but this does in no way diminish he selfless sacrifices and devotion that Americans have been displaying - we Europeans only have to think of WWII.
And in many instances, we can only be thankful for American aid and intervention. But this is not always so. Furthermore, the outright arrogance and narrowmindedness that we see so often from Americans - Republicans, especially - are hard to stomach.
Terrorism can never be excused, and we don't know what drove the attackers to their deeds, but the present government would do well to fix its attitude problem, just in case.