I wouldn't call anything VA did "overpriced". Their servers are worth every penny. The fact that they have a high margin is great for me, because it gives them more money to test equiptment before they ship it to me. They even use Teflon cables for internal SCSI cabling. That's an extra $150 just to be extra sure that you don't have SCSI problems. VA machines are wonderful.
I'm guessing they probably also have the vulnerability to change the item description, which might be even worse. Think about it. If someone allows you to change the description on the web site, you could link to their page, with a specially coded URL that adds "Bag of Crap" to their cart for $5. Or you could add a "B1 Bomber" for $10. This would be even better than setting prices, and there's probably some vulnerable software packages. Noone bothers to check input these days. Where do these programmers come from?
a) they get 30 days to tune your server
b) after those 30 days, you have to pay for the time they took to tune your server. At Oracle consulting rates, that's a _lot_ of money
Even more than just creating new TLDs, why not replace the root name servers. It would probably be pretty easy to have a new set of root name servers that simply queried the old ones for names that are not found. And it wouldn't even require a plugin, just change the entry int/etc/resolv.conf, or wherever windows people change it.
I think you're misunderstanding what they are calling "embedded". It's not for embedded devices. Embedded Mozilla are for projects like Galeon/skipstone/nautilus/whatever that use mozilla as a component inside their application, "embedded" in it.
OSX only works on new-world Macintosh machines. Are you _sure_ the machine you're trying Nautilus on compares equally? If its an x86 box it needs (a)double the megahertz, (b)128MB RAM. In addition, as has been said before, I'm guessing that OSX takes better advantage of hardware accelleration than X does.
Minimum system requirements for Mozilla are a Pentium 233, so making 386 would be pointless because its a totally unsupported architecture. You might as well optimize for people who actually _can_ use your product.
Quite true. I don't know what the graphics capabilities are of this machine, but I am certain that Solaris can take better advantage of them than Linux.
Actually, I believe UltraLinux has shown to be faster on Sparcs, at least for the single processor case. One of the reasons is that the Linux system call is much lighter weight, because traditional UNIX uses the stack to pass system call parameters, while Linux just puts them in registers. Now, there are several things Linux can't do, like utilize special sun firmware for kernel/hardware debugging. However, for most servers, Linux is faster for single processor machines.
Of course not, unless you're getting the gas from the tanks under their gas station, without their permission. That's the closest parallel I can come up with - not sure how good it is...
No, its not. The "idea" that's there, the location, cost Texaco millions in research. Someone opening a competing gas station next door is no less stealing ideas than copying a CD is.
As far as the math thing goes, the question isn't "can they be paid", but instead, "should it monopolized by a single entity". Math shouldn't be patented, even if there were noone to pay the mathematicians. The fact is, mathematicians realized this, and found ways of making money that don't require monopolizing ideas.
If musicians wanted to, they too can make money without monopolizing their song. One method is the patronage model, which in fact is how Larry Wall gets paid. Race-car drivers, too. Basically, a large organization pays money to be identified with a popular person. Another way musicians could get paid is for writing songs to be used in commercials and such. Also, don't forget concerts. However, prices for these will have to go up if free music lowers CD sales. However, there is no evidence yet that totally free music will lower artist's CD sales.
Of course, no matter what happens, nothing can screw over music artists any worse that current record companies.
Re:I have had a fearful thought....
on
KDE 2.1 Is Out
·
· Score: 2
Actually, Mozilla is usably fast, if you embed it in something else. What makes Mozilla slow is the fact that much of the web-browsing logic is written in JavaScript. So, if you use Mozilla embedded in Galeon, it is much, much faster.
Re:I have had a fearful thought....
on
KDE 2.1 Is Out
·
· Score: 2
I think you're missing the point. Mozilla is not _just_ a browser. It's a development platform. And thus, its taken a long time to build. Were it just a browser, you could say that it took a long time to build. In addition, Mozilla is _not_ a GNOME project. It doesn't even really use GTK.
Instead of thinking of Mozilla as a browser, think of it as a replacement for GNOME/KDE rather than a component of them. Look at what it has:
it's own Object Model, with its own Object Protocol
It's own widget set
It's own theming engine
It's own language
The ability to house many applications that aren't just web-browser specific
Its a whole system to itself. Its _not_ just a browser. Now, whether or not it was a good idea to be a whole system to itself is up for debate. But don't compare Mozilla to Konqueror. If anything, compare Mozilla to GNOME or KDE.
I'm very aware of how much money it takes. Do you think it should be illegal for me to open up a gas station next to a popular Texaco because Texaco spent millions researching the location? Do you think that mathematical formulas should be patented simply because the researcher spent decades coming up with it? The idea that ideas are separate entity from things (like, that music is separable from the CD) is a really bad concept. Probably one of Plato's worst (don't get me wrong, I like Plato and Socrates, just not the whole "world of ideas" thing). And, in addition, you never answered the question of whether I should be able to duplicate a hammer that I own. Doubtless the company that made the hammer spent R&D money building it, so should I be able to copy the hammer and give it away?
You shouldn't be expect to be paid anything doing anything unless there is a contract saying so. Also, what right does someone else have to say what I can do with my own CDs? Do I not have a right to share? What is immoral about copying what I have to give it to someone else? If I owned a hammer, could I not build an identical one to give to my neighbor?
Actually, I think Mozilla's performance is more related to the fact that the browser is actually written in JavaScript. Also, the scope of Mozilla is much, much bigger than Konquerer. Mozilla is its own XML-based development environment. It has its own object model, it's own widget set, etc. Whether this is a good idea or not, I don't know, but that is more of the cause of these problems than language choice. Also, the fact that Mozilla is extremely cross-platform, including its own portability library makes the fact that they've done so much in so little time, absolutely amazing.
Did you read _why_ he said that? It's because "intellecutal property" covers too many issues to be dealt with under a single term. HE DID NOT SAY THAT YOU SHOULD AGREE WITH HIM ABOUT HIS OPINIONS, he just said that there are too many different issues to comment on them sanely as a group. Copyright, Trademark, and Patent issues are so disparate, that trying to talk about them together under the term "intellectual property" is wreckless and almost absurd.
I think you're forgetting something. Who are the testers of open source projects? The _users_. That's why they let untested software out to the users, because they _are_ the testers. If you don't want untested software, either (a) use well-commended versions of software, or (b), test it yourself. You might say "Oh, a commercial organization shouldn't have to test software itself!" But then you would be living in a dream world. Every medium-to-large organization has to test its software. If you pay millions of dollars to Oracle to get the latest version of Oracle Applications, then you, yes you, will still have to put in a LARGE amount of work testing. In fact, some releases of Oracle Applications are completely non-functional. AND THIS IS WHAT THEY ACTUALLY SELL TO END-USERS! So, as the user, you have to test free software. And you have to test proprietary software, too. Which would you choose?
I wasn't saying that. I was just saying that it was a better choice. Engineering decisions aren't about leaping into the unknown, it's about making wise choices that last for the long-term. And their use of OO C was along those same lines, too.
I wasn't saying _anyone_ was breaking the spec. I'm just saying that Enlightenment is pretty-much a full-fledged desktop environment (like WindowMaker). Thus, if you have Enlightenment + GNOME, you have a lot of duplication of code which leads to slowness and memory hogging. I have nothing against Enlightenment, I just think if you use it with GNOME it takes more space and processing power than it needs to. Enlightenment by itself works great, as well as GNOME with SawFish or other lightweight WMs.
Encapsulation is very good. However, even if something is encapsulated well, does not mean that you don't need to look under the hood. If a certain function is becoming a bottleneck, you need to be able to see what that function is doing, and how you can help it out. For example, you might thing that a sort() algorithm would be a fairly closed-box solution. However, some sorts perform dramatically better with differently-ordered sets. If you can tweak previous code to get it to a form that works better with that type of sort, then you have gained an advantage. Encapsulation helps the learning curve, but in order to masterfully use a tool, you also must know how it works underneath. Trust me, there is NO SUCH THING as a completely black box. The actual implementation always has an effect. A good encapsulation will make the interface easy to learn and predictable, but it is no substitue for masterful knowledge.
Yes, the strength of GNOME lies in the choice of application languages - because they used C as their base language. That was my point.
As far as dealing with technical issues first, they are actually using CORBA, as opposed to KParts. With CORBA, they didn't come up with something that just works for that purpose, they chose something with more long-term applicability.
Slashdot Mirror
I wouldn't call anything VA did "overpriced". Their servers are worth every penny. The fact that they have a high margin is great for me, because it gives them more money to test equiptment before they ship it to me. They even use Teflon cables for internal SCSI cabling. That's an extra $150 just to be extra sure that you don't have SCSI problems. VA machines are wonderful.
I'm guessing they probably also have the vulnerability to change the item description, which might be even worse. Think about it. If someone allows you to change the description on the web site, you could link to their page, with a specially coded URL that adds "Bag of Crap" to their cart for $5. Or you could add a "B1 Bomber" for $10. This would be even better than setting prices, and there's probably some vulnerable software packages. Noone bothers to check input these days. Where do these programmers come from?
Interestingly, SAP is releasing their DB backend under the GPL.
If you read the fine print you'll notice
a) they get 30 days to tune your server
b) after those 30 days, you have to pay for the time they took to tune your server. At Oracle consulting rates, that's a _lot_ of money
So, they're actually trying to jip you.
Even more than just creating new TLDs, why not replace the root name servers. It would probably be pretty easy to have a new set of root name servers that simply queried the old ones for names that are not found. And it wouldn't even require a plugin, just change the entry int /etc/resolv.conf, or wherever windows people change it.
If it locked up your box, its not Mozilla's problem.
I think you're misunderstanding what they are calling "embedded". It's not for embedded devices. Embedded Mozilla are for projects like Galeon/skipstone/nautilus/whatever that use mozilla as a component inside their application, "embedded" in it.
There's a few things to think about.
OSX only works on new-world Macintosh machines. Are you _sure_ the machine you're trying Nautilus on compares equally? If its an x86 box it needs (a)double the megahertz, (b)128MB RAM. In addition, as has been said before, I'm guessing that OSX takes better advantage of hardware accelleration than X does.
Minimum system requirements for Mozilla are a Pentium 233, so making 386 would be pointless because its a totally unsupported architecture. You might as well optimize for people who actually _can_ use your product.
Quite true. I don't know what the graphics capabilities are of this machine, but I am certain that Solaris can take better advantage of them than Linux.
Actually, I believe UltraLinux has shown to be faster on Sparcs, at least for the single processor case. One of the reasons is that the Linux system call is much lighter weight, because traditional UNIX uses the stack to pass system call parameters, while Linux just puts them in registers. Now, there are several things Linux can't do, like utilize special sun firmware for kernel/hardware debugging. However, for most servers, Linux is faster for single processor machines.
Of course not, unless you're getting the gas from the tanks under their gas station, without their permission. That's the closest parallel I can come up with - not sure how good it is...
No, its not. The "idea" that's there, the location, cost Texaco millions in research. Someone opening a competing gas station next door is no less stealing ideas than copying a CD is.
As far as the math thing goes, the question isn't "can they be paid", but instead, "should it monopolized by a single entity". Math shouldn't be patented, even if there were noone to pay the mathematicians. The fact is, mathematicians realized this, and found ways of making money that don't require monopolizing ideas.
If musicians wanted to, they too can make money without monopolizing their song. One method is the patronage model, which in fact is how Larry Wall gets paid. Race-car drivers, too. Basically, a large organization pays money to be identified with a popular person. Another way musicians could get paid is for writing songs to be used in commercials and such. Also, don't forget concerts. However, prices for these will have to go up if free music lowers CD sales. However, there is no evidence yet that totally free music will lower artist's CD sales.
Of course, no matter what happens, nothing can screw over music artists any worse that current record companies.
Actually, Mozilla is usably fast, if you embed it in something else. What makes Mozilla slow is the fact that much of the web-browsing logic is written in JavaScript. So, if you use Mozilla embedded in Galeon, it is much, much faster.
I think you're missing the point. Mozilla is not _just_ a browser. It's a development platform. And thus, its taken a long time to build. Were it just a browser, you could say that it took a long time to build. In addition, Mozilla is _not_ a GNOME project. It doesn't even really use GTK.
Instead of thinking of Mozilla as a browser, think of it as a replacement for GNOME/KDE rather than a component of them. Look at what it has:
it's own Object Model, with its own Object Protocol
It's own widget set
It's own theming engine
It's own language
The ability to house many applications that aren't just web-browser specific
Its a whole system to itself. Its _not_ just a browser. Now, whether or not it was a good idea to be a whole system to itself is up for debate. But don't compare Mozilla to Konqueror. If anything, compare Mozilla to GNOME or KDE.
I'm very aware of how much money it takes. Do you think it should be illegal for me to open up a gas station next to a popular Texaco because Texaco spent millions researching the location? Do you think that mathematical formulas should be patented simply because the researcher spent decades coming up with it? The idea that ideas are separate entity from things (like, that music is separable from the CD) is a really bad concept. Probably one of Plato's worst (don't get me wrong, I like Plato and Socrates, just not the whole "world of ideas" thing). And, in addition, you never answered the question of whether I should be able to duplicate a hammer that I own. Doubtless the company that made the hammer spent R&D money building it, so should I be able to copy the hammer and give it away?
You shouldn't be expect to be paid anything doing anything unless there is a contract saying so. Also, what right does someone else have to say what I can do with my own CDs? Do I not have a right to share? What is immoral about copying what I have to give it to someone else? If I owned a hammer, could I not build an identical one to give to my neighbor?
Actually, I think Mozilla's performance is more related to the fact that the browser is actually written in JavaScript. Also, the scope of Mozilla is much, much bigger than Konquerer. Mozilla is its own XML-based development environment. It has its own object model, it's own widget set, etc. Whether this is a good idea or not, I don't know, but that is more of the cause of these problems than language choice. Also, the fact that Mozilla is extremely cross-platform, including its own portability library makes the fact that they've done so much in so little time, absolutely amazing.
I'd also like a moderation -1, poster either misread or totall misunderstood the article
Did you read _why_ he said that? It's because "intellecutal property" covers too many issues to be dealt with under a single term. HE DID NOT SAY THAT YOU SHOULD AGREE WITH HIM ABOUT HIS OPINIONS, he just said that there are too many different issues to comment on them sanely as a group. Copyright, Trademark, and Patent issues are so disparate, that trying to talk about them together under the term "intellectual property" is wreckless and almost absurd.
I think you're forgetting something. Who are the testers of open source projects? The _users_. That's why they let untested software out to the users, because they _are_ the testers. If you don't want untested software, either (a) use well-commended versions of software, or (b), test it yourself. You might say "Oh, a commercial organization shouldn't have to test software itself!" But then you would be living in a dream world. Every medium-to-large organization has to test its software. If you pay millions of dollars to Oracle to get the latest version of Oracle Applications, then you, yes you, will still have to put in a LARGE amount of work testing. In fact, some releases of Oracle Applications are completely non-functional. AND THIS IS WHAT THEY ACTUALLY SELL TO END-USERS! So, as the user, you have to test free software. And you have to test proprietary software, too. Which would you choose?
I wasn't saying that. I was just saying that it was a better choice. Engineering decisions aren't about leaping into the unknown, it's about making wise choices that last for the long-term. And their use of OO C was along those same lines, too.
I wasn't saying _anyone_ was breaking the spec. I'm just saying that Enlightenment is pretty-much a full-fledged desktop environment (like WindowMaker). Thus, if you have Enlightenment + GNOME, you have a lot of duplication of code which leads to slowness and memory hogging. I have nothing against Enlightenment, I just think if you use it with GNOME it takes more space and processing power than it needs to. Enlightenment by itself works great, as well as GNOME with SawFish or other lightweight WMs.
I run LinuxPPC as my primary desktop platform - have been for over a year. What exactly are you talking about?
Encapsulation is very good. However, even if something is encapsulated well, does not mean that you don't need to look under the hood. If a certain function is becoming a bottleneck, you need to be able to see what that function is doing, and how you can help it out. For example, you might thing that a sort() algorithm would be a fairly closed-box solution. However, some sorts perform dramatically better with differently-ordered sets. If you can tweak previous code to get it to a form that works better with that type of sort, then you have gained an advantage. Encapsulation helps the learning curve, but in order to masterfully use a tool, you also must know how it works underneath. Trust me, there is NO SUCH THING as a completely black box. The actual implementation always has an effect. A good encapsulation will make the interface easy to learn and predictable, but it is no substitue for masterful knowledge.
Yes, the strength of GNOME lies in the choice of application languages - because they used C as their base language. That was my point.
As far as dealing with technical issues first, they are actually using CORBA, as opposed to KParts. With CORBA, they didn't come up with something that just works for that purpose, they chose something with more long-term applicability.