This is performance as in speed you're talking about, and not some other kind of performance I haven't met before?
If you've got Linux, Solaris, or Windows handy, try grabbing Sun's JDK1.3. It's very fast, especially if you use the -server option when you start it. I use it for heavy-duty number crunching on a daily basis. I guess I might be able to make some of the code go a little faster if I rewrote in C, but the months it would take to do this easily outweight any the (small) performance gain.
I seem to remember that there's an information theory proof that wavelets are, the most efficient way of encoding the information in any given image (although I think there was a caveat that the `ideal' set of wavelets is likely to be very hard to find. It's a while since I've looked at the wavelet literature, and I don't have the reference to hand, but it can probably be turned up somewhere...
Anyway, I guess this is why most interest is in wavelets now -- and I saw a JPEG2000 implementation listed on freshmeat the other day!
Inbreeding certainly is a bad thing. But the reason behind this is that most individuals carry some `recessive lethal' genes. That is, genes which cause no trouble if you only have one copy, but which cause problems (or even death) if you have two bad copies. Inbreeding increases the chance of combining a bad copy from each parent.
In principle (given sufficiently advanced genetic technology) it might be possible to identify and eliminate the recessive lethal alleles from the tiger gene pool. It would then be safe for cloned animals to interbreed. Of course, this is still a long way off. But then again, I'd guess there's some way to go before technology is advanced enough to do this cloning, anyway.
There's a separate issue, given a population founded from just a few individuals, genetic diversity will be extremely low, which could make the population rather vulnerable to environmental changes and disease. Again, this might not be an insurmountable problem -- it might even be possible to add some genes from closely related living species to some individuals, in order to boost diversity in critical areas (like the immune system). But once again, we're still a long way off.
My two nucleotides, and probably completely wrong...
I'd love to agree with this ('cos I want to go to mars, too;), but really a few more unmanned probes are probably in order before we try sending people there. If anyone is planning to stay there for more than a few days, we'll want to know a lot more about the chemistry of the martian surface. And a lot more about the water which seems to be available there.
That said, I hope they start work on the manned trip soon...
According to the release notes, Mutt 1.2 supports this (along with many other improvements to the IMAP support). I'm not using IMAP on my current machines so I haven't tested this, but it's certainly worth a try if you need it.
Yes, mutt does support IMAP these days -- has done for a couple of years now. It used to be a bit flakey, but I think it's starting to shake down okay in the 1.2.x releases.
IMAP is a very nice technology, especially since you can put clever backends behind it (mail in a proper database has always struck me as a Good Thing), but it's generally received quite a cool reception in the UNIX world. That's changing now, but a lot of admins still aren't quite happy with it. *sigh*.
This isn't necessarily true at least from a legal point of view, although it fits in with a lot of current thinking, especially in the computer and entertainment industries.
Copyright used to be a limited right. It expired, and was limited by things like the principle of first sale. Software licences have always taken the view that they can tell you exactly what you can and can't do with a piece of software (often, these days, to the extent that licences are made non-transferrable). The exact legality of this is a mess (the UCITA maybe clears this up a bit, but we know that has problems of its own...).
I'm not in a position to say whether GPL is enforcable or not, but I think it might be premature to think it's the author's choice.
Hmmm, I take it you haven't tried IBM's JDK for Linux -- mixed mode virtual machine with an in-house JIT, and it performs great for me. The Blackdown.org people are working on linux builds of the Sun HotSpot system, which also goes at a pretty respectable speed. Those enough for you?
Of course, none of this is much reassurance if you're running Linux on Alphas or PowerPCs, but nobody would think of running Linux on non-Intel hardware, would they?;)
Right now, the Internet could really do with some tools which empower the user. This looks like another way for big content providers to herd users where they want them. Traditional search engines have always been a bit of a battle, with content provides trying to find new ways to `stuff' the search results, and make sure that their pages came out on top. Now you don't need to do that any more -- just pay these guys whatever they're asking, and you can display what you want. Including images, by the look of things, which sounds cool, but really just lets provides grab your attention.
Of course, if this were heavily user-moderated, I guess it might just work. But don't hold your breath. I'll be sticking to Google...
I'd guess that the main problem would be in terms of sleep patterns of certain animals. The albedo of the moon's natural surface is really pretty low (somewhere around 0.15, from memory). If this were increased to near 1.0, the full moon would be very bright. We already know that some species are sensitive to the moon's cycle (look at seahorses, for instance). I'd not like to predict how a brighter moon would affect these.
This got posted while I was in the middle of typing up a related question, which addressed the other side of the coin. The dynamic linking/RPC/whatever issue certainly is something that needs some consideration, especially with (for example) Java applications potentially being made up from hundreds of.class files, linked at run time.
But I'm rather worried about a reaction to this being taken to its logical extreme, and giving a very broad definition of `derived work'. But let's try taking this to its logical extreme:
[thomasd@unixbox]$ mycommercialapp | grep "hello"
It's possible to argue that I'm setting up a `derived work' of the commercial application and grep (we'll assume that this is a GPL'ed version of grep). There's a defined protocol being used between the two components (UNIX stream I/O). Now, I fully accept that this is a contrived example, but I think it does show that this is an area where ther is a rather fine line between freedom (of users to do what they want with software) and protecting the Free nature of the software components themselves.
Of course, some of this can be dealt with by individual developers of componentized applications making a clear statement about their interpretation of the `plugin issue' in the documentation. But it would be nice to have some more debate in the community to see what the general view is.
I'm a great believer in free/open software development. But to me, it's even more important that, wherever possible, software is developed in the form of open components with defined APIs (the exact means of communication don't really matter -- could be dynamic linking, CORBA, or something else). I see this as a potentially positive trend for both programmers and end-users. The GIMP is a good example of a Free product which already seems to be benefitting from good plugin support.
When the {L,}GPL first appeared, dynamic linking was quite rare, and CORBA still over the horizon. Now these technologies are ubiquitous (and Java, in particular, doesn't have a concenpt of static linkage at all). Are we going to see updated licences which are more aware of componentized software? And do you believe that Free programs should only be used with Free plugins (and vice versa), or might it be reasonable to allow the end user to mix Free and proprietary components, so long as they are communicating via a well defined, and completely open, interface?
Hell, there's a major rural/urban divide in telecomunications, even in the UK (this is even true in the south-east, although obviously more pronounced in places like scotland/wales/dartmoor). Okay, we can blame some of this on the telecos, but realistically geography is always going to affect telecoms provision.
To what extent is DSL rolling out in.au? This won't reach into remote areas, but it does have a range of several miles, which could make it viable in at least some rural areas, so long as demand is high. I believe IDSL (which uses an ISDN-like encoding) has a longer viable range than the more familiar DSL technologies, and this might help matters.
For remoter areas, I'd suspect that some kind of wireless system is the only approach that's likely to be viable in the forseeable future. Aren't the remoter areas of Australia already quite reliant on radio communications? This might set a precendent for wireless IP.
Anyway, I hope things do move along. Like most UK-based geeks, I'm entirely sympathetic to the cause of getting telecos to roll out decent internet access technologies...
Nice to see a mention of microtransactions...
on
RMS On eBooks
·
· Score: 3
RMS is just right about the `click here to send $1 to the author'. That sounds a fair price for a book (and, incidentally, substantially more than authors get for paperback fiction right now). The real requirement is some infrastrucure for making small payments, and also for allowing anyone who wants it to easily (and cheaply) set up the necessary infrastructure for recieving these payments.
Sadly, I'm not sure that anyone with the clout to create a new type of financial service is really interested in doing this. Certainly, while microtransactions have been talked about for a while now, they seem more distant today than they did a couple of years back. Until they happen, e-commerce is going to be pretty much the exclusive preserve of the big guys.
There's something beautiful about a book
on
RMS On eBooks
·
· Score: 1
Purely by coincidence, I was re-reading Farenheit 451 last night. There are quite a few references to the texture and smell of Real Books. Yeah, I'm a technophile by almost anyone's standards, but there's no way I'm trading my favourite paper volumes for anything.
Anyway, I strain my eyes enough while coding (and reading slashdot...)
You know, I was thinking much the same thing all the way through the article. The `laptops' in Ender's Game are interesting on the basis that they are (probably intentionally) hackable. Sadly, I'm not sure any current educational authority would follow that kind of logic...
They've worked very hard on standards-compiliance. The cause of confusion is that they've implemented the W3C's DOM level 1 (and most of DOM level 2 now, I believe). Netscape 4 used an old, proprietary DHTML system. Up till now, DHTML authors have generally used `browser-sniffing' code, and write one version of the page for Netscape and another for IE (which already uses DOM level 1). Right now, a lot of developers haven't updated things to use the W3C DOM with Netscape 6. Don't worry, it will come.
So long as you're careful, it should be possible to write standards-compliant, cross browser DHTML which will work (without browser-sniffing) in at least IE5+ and Netscape 6+. Problems in the short term, but really this is great news.
I can't comment on IE5 (hardly ever used it), but Netscaoe Classic version 4.5+ does something similar -- they call it Internet Keywords. But it's easy to disable in the browser preferences. Same goes for the `Smart Browsing': this tells the Netcentre which pages you're looking at in exchange for a list of `What's related' links, but again it can be turned off quite easily.
After a rather slow start, open source seems to have worked out pretty well for Netscape. There are some quite big community projects still going on (XSLT support, for instance).
I'd actually go as far as to say that the release of a commercial Netscape 6 will be the point when Mozilla.ORG really starts coming into its own. Just imagine: a commercial browser with millions of users, but still (almost) all the code out there for anyone to grab and hack on.
Methinks the future looks very bright. Now we just need a working OpenJava plugin for Linux...
Slashdot Mirror
If you've got Linux, Solaris, or Windows handy, try grabbing Sun's JDK1.3. It's very fast, especially if you use the -server option when you start it. I use it for heavy-duty number crunching on a daily basis. I guess I might be able to make some of the code go a little faster if I rewrote in C, but the months it would take to do this easily outweight any the (small) performance gain.
Anyway, I guess this is why most interest is in wavelets now -- and I saw a JPEG2000 implementation listed on freshmeat the other day!
...which I've always taken to mean HTTP. If you need to have multiple FTP (or whatever) servers, I don't see that this is going to be a problem.
In principle (given sufficiently advanced genetic technology) it might be possible to identify and eliminate the recessive lethal alleles from the tiger gene pool. It would then be safe for cloned animals to interbreed. Of course, this is still a long way off. But then again, I'd guess there's some way to go before technology is advanced enough to do this cloning, anyway.
There's a separate issue, given a population founded from just a few individuals, genetic diversity will be extremely low, which could make the population rather vulnerable to environmental changes and disease. Again, this might not be an insurmountable problem -- it might even be possible to add some genes from closely related living species to some individuals, in order to boost diversity in critical areas (like the immune system). But once again, we're still a long way off.
My two nucleotides, and probably completely wrong...
*sigh*
Pity my laptop only does 1024x768... but I did need a machine I could carry around.
That said, I hope they start work on the manned trip soon...
But I believe the company which is using those banners is none other than the speakeasy.net mentioned by the original poster...
According to the release notes, Mutt 1.2 supports this (along with many other improvements to the IMAP support). I'm not using IMAP on my current machines so I haven't tested this, but it's certainly worth a try if you need it.
IMAP is a very nice technology, especially since you can put clever backends behind it (mail in a proper database has always struck me as a Good Thing), but it's generally received quite a cool reception in the UNIX world. That's changing now, but a lot of admins still aren't quite happy with it. *sigh*.
Copyright used to be a limited right. It expired, and was limited by things like the principle of first sale. Software licences have always taken the view that they can tell you exactly what you can and can't do with a piece of software (often, these days, to the extent that licences are made non-transferrable). The exact legality of this is a mess (the UCITA maybe clears this up a bit, but we know that has problems of its own...).
I'm not in a position to say whether GPL is enforcable or not, but I think it might be premature to think it's the author's choice.
Verbal Diarrhoea :(
I cannot count syllables
Six on the first line
Haiku appropriate
From one who will emulate
David Brin's dolphins
Of course, none of this is much reassurance if you're running Linux on Alphas or PowerPCs, but nobody would think of running Linux on non-Intel hardware, would they? ;)
Of course, if this were heavily user-moderated, I guess it might just work. But don't hold your breath. I'll be sticking to Google...
I'd guess that the main problem would be in terms of sleep patterns of certain animals. The albedo of the moon's natural surface is really pretty low (somewhere around 0.15, from memory). If this were increased to near 1.0, the full moon would be very bright. We already know that some species are sensitive to the moon's cycle (look at seahorses, for instance). I'd not like to predict how a brighter moon would affect these.
But I'm rather worried about a reaction to this being taken to its logical extreme, and giving a very broad definition of `derived work'. But let's try taking this to its logical extreme:
It's possible to argue that I'm setting up a `derived work' of the commercial application and grep (we'll assume that this is a GPL'ed version of grep). There's a defined protocol being used between the two components (UNIX stream I/O). Now, I fully accept that this is a contrived example, but I think it does show that this is an area where ther is a rather fine line between freedom (of users to do what they want with software) and protecting the Free nature of the software components themselves.
Of course, some of this can be dealt with by individual developers of componentized applications making a clear statement about their interpretation of the `plugin issue' in the documentation. But it would be nice to have some more debate in the community to see what the general view is.
When the {L,}GPL first appeared, dynamic linking was quite rare, and CORBA still over the horizon. Now these technologies are ubiquitous (and Java, in particular, doesn't have a concenpt of static linkage at all). Are we going to see updated licences which are more aware of componentized software? And do you believe that Free programs should only be used with Free plugins (and vice versa), or might it be reasonable to allow the end user to mix Free and proprietary components, so long as they are communicating via a well defined, and completely open, interface?
To what extent is DSL rolling out in .au? This won't reach into remote areas, but it does have a range of several miles, which could make it viable in at least some rural areas, so long as demand is high. I believe IDSL (which uses an ISDN-like encoding) has a longer viable range than the more familiar DSL technologies, and this might help matters.
For remoter areas, I'd suspect that some kind of wireless system is the only approach that's likely to be viable in the forseeable future. Aren't the remoter areas of Australia already quite reliant on radio communications? This might set a precendent for wireless IP.
Anyway, I hope things do move along. Like most UK-based geeks, I'm entirely sympathetic to the cause of getting telecos to roll out decent internet access technologies...
Sadly, I'm not sure that anyone with the clout to create a new type of financial service is really interested in doing this. Certainly, while microtransactions have been talked about for a while now, they seem more distant today than they did a couple of years back. Until they happen, e-commerce is going to be pretty much the exclusive preserve of the big guys.
Anyway, I strain my eyes enough while coding (and reading slashdot...)
You know, I was thinking much the same thing all the way through the article. The `laptops' in Ender's Game are interesting on the basis that they are (probably intentionally) hackable. Sadly, I'm not sure any current educational authority would follow that kind of logic...
So long as you're careful, it should be possible to write standards-compliant, cross browser DHTML which will work (without browser-sniffing) in at least IE5+ and Netscape 6+. Problems in the short term, but really this is great news.
Bottom line: configurability is the key.
I'd actually go as far as to say that the release of a commercial Netscape 6 will be the point when Mozilla.ORG really starts coming into its own. Just imagine: a commercial browser with millions of users, but still (almost) all the code out there for anyone to grab and hack on.
Methinks the future looks very bright. Now we just need a working OpenJava plugin for Linux...