GTK+ C++ wrappers
on
GTK-- vs. QT
·
· Score: 5, Informative
I worked on a rushed project earlier this year, and used the gtk-- C++ wrapper for GTK, as well as the gnome-- wrapper (then still very much under development) for the Gnome libraries, specifically the canvas.
Personally, I found it frustrating to use. As these wrappers are still being worked on, the documentation is sketchy. The object-owning semantics are confusing (at least to me) - I was forever leaking memory or prematurely destroying objects. Trying to destroy something from within a menu callback I recall was particularly noisome. The gnome-- canvas wrappers were a moving target, though they may have since stabilised, and didn't fully expose the canvas API. Writing one's own canvas items is done in C, and then wrapped.
Perhaps with more persistance I might have figured out how to set up keyboard acceleration, but it is at anyrate a real battle to find documentation that explains what is going on with it. AFAIK, there is no straightforward way of making a multiple file selection in GTK+ 1.2. In gnome canvas (not GTK+, but a close cousin) there is promised functionality that is simply not implemented - I'm thinking here of smoothed lines, for which the code reads:
/* FIXME */
I haven't used QT yet. It certainly looks pretty, and a quick glance at the example code and docs provided seems to indicate that it's not too complicated, and well documented. I'd certainly shy away from GTK+ if a C++ interface is required.
The new version of GTK under development should address many of the shortcomings of the current toolkit, and includes goodies such as Pango. It is not yet in a stable state, however, with the API still undergoing final tweaking I believe.
My home Linux system is based on the Linux From Scratch documents, with the added organisational principle of having every/usr destined package in its own directory, and links made into the/usr tree.
This has worked fine for me with ldconfig, which reports itself as being
version 1.9.9. My/etc/ld.so.conf doesn't require an entry per directory at all - provided there are symlinks from (say)/usr/lib/libblah.so to whereever the library really is.
There's certainly a lot to be said for the 'bottom-up' approach to nanotechnology. Cost for starters! One issue though is, how does one address these very tiny devices?
The problem with a whole bunch of identical tiny circuits is of course that they're all identical - there's no way to differentiate between them. There will have to be some way of distinguishing and interacting with these units.
A couple of ideas spring to mind though. One is to encode the position of one of these units in the unit itself as it is being assembled, by interacting with some sort of precisely engineered field. What would work (if anything) depends very much on the chemistry, but it could be something as simple as a gradient in an electrostatic field, to aligning with a very fine grid of polarized light. There are options, but it all sounds Hard. Schemes like this could attack the problem of differentiation, but there's still interaction and addressing.
One way to solve the addressing problem is to bypass it almost entirely. If these structures are sufficiently small, and can be engineered to act as a giant grid of finite-state automata with evolution rules based on neighbouring states, one can simulate a computational device with a version of Conway's Life on speed. Input and output can be done at the edges of the constructed array, which is probably going to be more simple than trying to address the middle of the structure. The problem lies in initialising the state of the array - clearing it is probably easy enough, depending on how state is stored, but priming it with a state that admits the computational task desired seems to be almost as hard as addressing the cells in the first place.
Another approach might be to give each cell some random state as it is constructed (and there should be plenty of sources of randomness at the molecular level to draw on.) Imagine that this state corresponds to an "activation key": when an appropriately modulated high frequency EM signal hits the cell, it pushes it over into an active state. Before this, it's effectively off (perhaps an off cell would simply propogate signals from its neighbours and do no computation). Give each cell some way of indicating that it has been activated (eg, it emits some light upon activation), and then fire random keys at the cells. This solves the addressing problem, and the interaction problem (one could use the same key for changing the cell's state) - but then one has no easy way of telling how the newly identified cell connects to the other addressable cells.
Do any slashdotters have any ideas? Or can point to literature where these problems are (ahem) addressed?
It does appear though, tbat this is a copyright issue rather than a patent one. Copyright has long been applicable to mathematical papers and diagrams (for good or ill), and has far less damaging potential consequences than patenting.
the European Patent Office just published a new examination directive which extends the realm of the European patent practice to software, business methods and mathematics.
Patenting mathematics is outright crazy. It's the same sort of crazy that allows the patenting of software, but in the past one could always say: patenting algorithms is like patenting mathematics, and thus clearly nonsense. reductio ad absurdum has come along and bit us all on the arse.
Trying to imagine a world where mathematics is patentable is both hard and disturbing. Can you imagine if only licensed physicists were allowed to use Hilbert space theory? If one needed to pay a levy every time one used Shannon's law to help design a product? Where would we be if the finite element methods could only be applied to engineering analyses with the blessings of its creators?
How much mathematical progress would be made, if every mathematician had to check whether the work they were building on was patent-encumbered? If every publication had to first get the approval of some patent holders, with the possibility of a required payment?
It quickly gets surreal. Many statements in mathematics are equivalent when viewed in the appropriate fashion. Many too are based on certain sets of axioms. What does patentable mean when viewed in this light?
This to me is a clear sign that extreme IP advocates have just completely lost the plot.
As the article you linked to states, it certainly is a high-risk field for new entrants. However it is hard to deny that those major labels that have survived, are profitting. The article itself states that the profit comes from their back catalogue - of the music which those companies own. I stand by my statement that the music itself is indeed a valuable asset.
I don't think I've been unfairly manipulative of the description of the current state of the music industry. When a few large companies hold the keys to commercial success, they get to make the rules. One of these rules is that artists no longer own their own creative works. Artists may have the right to be independent, but it's a very tough choice given the current situation. As you say, marketing and promotion are hard without a big label's backing.
Importantly though, I think you should be careful not to make a false dichotomy. The current system can be reprehensible while the alternative can be better than 'not getting paid at all for a recording'.
Many people engage in creative work, and many of these people create things that can be freely (if not necessarily legally) copied at little or no cost. Most do not rely upon huge popularity and royalties for recompense (for example, scientists, academics, visual artists,...). The arts are regarded as culturally valuable, and as such need to be supported. However the recording industry only rewards those whose work is wildly popular. The market is not the best solution to this particular problem, as it typically rewards the least offensive rather than the most valuable; relies on artificial protections to be viable; and when it does reward, does so out of all proportion to the energies that went into creation. Alternatives and their discussion are, I feel, too important and long to be discussed in a quick slashdot comment though.
The question is whether they're [peer-to-peer networks] going to be used -
whether they'll respect what artists create just like we in the recording
business respect what the business sponsors and software developers
in this audience create.
Note that she doesn't claim that they in the
recording business respect artists or their work themselves.
Courtney
Love's rant on the piracy of the recording industry makes for educational
reading. Later Rosen says,
Are the works of artists valuable? The answer,
in my view, is a resounding YES.
And of course they are. Look at the
profits of the major labels. The problem being of course, is that this
is monetary value, and further, they are much more valuable to the labels
than the artists once the rights have been signed away.
The language in the speech is emotive, as is to be expected.
But the kiddie porn quote is surely beyond the pale,
The fact that I was invited means that someone out there knows that peer-to-peer technology is getting bad rap.... The fact that it is also used as a transmitter of child pornography has not gone unnoticed by many federal and law enforcement agencies.
And the very companies that the RIAA represent publish and promote music with hate-lyrics.
We also have the old chestnut of referring to illegal copying as
theft. Repeatedly. This should be plain enough, but many people seem
to have bought the lie. Illegal copying is just that. It may well be
damaging to the creators of the material (which is probably wrong) as
well as to the distributors (which is not necessarily wrong - people
don't have a right to make a profit, remember!). What it is not
though, is theft. Let alone piracy. The debate on intellectual property is
muddied enough as it is, without resorting to misleading language.
I think the most poignant quote though is,
But as long as you're looking for whom piracy really hurts, ask the
guitarist in the coffee shop, or the group scratching out a living
touring in a beat-up van.
This is so true. Sadly, it's the piracy of
the recording industry - which has, among other things, managed to have
artists' work reclassified as work for hire (!) - that is responsible
for artists living in poverty while simultaneously having millions of CD
sales. The term piracy is much more applicable to this sort of action;
what these labels do is not illegal copying, but the wholesale transfer
of rights from the artist to themselves using the big stick of exclusive
access to mainstream distribution channels.
If you have an interest in the music industry and not yet read the Salon article linked above, you really ought. It's very educational.
PS: If you do want to support artists, there is always
Fairtunes.
I know it's not the same, but it does feel similar to what the MPAA tried to pull with Divx. Divx of course failed because customers didn't want to buy crippled equipment, and rightly so. Perhaps HP will face a similar response here.
Region coding is another example of crippling for profit, but unlike say Divx, it didn't affect the majority of customers. In the major markets of US and Japan, only a few would seek to play DVDs from outside their native region. Europe was more badly affected, and DVD still hasn't taken off in Australia really, due to the paucity of region 4 releases outside the big titles.
HP's crippling though would become apparent everytime one tried to record on it. What is Digital Audio Media, other than a disingenuous choice of name? I'm presuming it's the same as the (expensive) CD-Audio disks, which or course are just CD-R with magic mark on it for the benefit of (presumably) the RIAA.
Similar shenanigans killed DAT as a home medium, but maybe the other features of
the HP device will win out. Recording aside, it does look like a nice piece of kit.
An older board - the UP2000 - is a dual processor SDRAM (not DDR) based Alpha motherboard, which has 6 PCI slots, two of which are 64-bit.
This new board has DDR ram, but only 32-bit PCI, and then only three slots. While nice and all - DDR is good, and of course it's for the Alpha 21264B, not 21264A - this does seem a bit of a step backwards in the IO stakes. Especially when it's noted that the UP2000 has onboard Ultra-2 SCSI as well.
Perhaps this board was originally targetted at the 'lower-end' workstation segment? Does anyone know if a more server-oriented 21264B board is on the way? It seems sadly unlikely given the current circumstances.
If one wants to have 64-bit multiprocessing on a budget, what are the current alternatives?
From the discussion on the Apple discussion web site, the nature of the bug is as follows.
The original installer script has the lines
# if iTunes application currently exists, delete it
if [ -e $2Applications/iTunes.app ] ; then
rm -rf $2Applications/iTunes.app 2>/dev/null
fi
while the replacement (2.0.1) has
# if iTunes application currently exists, delete it
if [ -e "$2Applications/iTunes.app" ] ; then
rm -rf "$2Applications/iTunes.app" 2>/dev/null
fi
In these scripts, $2 corresponds to the volume on which iTunes is to be installed, and will be of the form/Volumes/name-of-volume.
For those unfamiliar with Bourne shell variable expansion, if $2 has spaces in it, the argument to the rm command in the first version of the script will expand to more than one word, and rm will try and delete both of these. The -rf tells rm to delete everything down recursively and not complain about it.
This is particularly a problem on the Mac, where filenames and volume names often have spaces in them., even at the beginning of the name. If one had multiple partitions mounted in/Volumes, and the one on which iTunes was to be installed was called, say, ' OS X', then the rm command would expand to
rm -rf/Volumes/ OS X/Applications/iTunes.app 2>/dev/null
and would then try and delete everything under/Volumes. This is clearly bad.
The second version, by including quotes around the argument, fixes the problem. The quotes force the argument to be treated as a single argument after variable expansion.
Traditionally, people have been super careful about destructive operations and shell expansions. I don't think I've ever seen something like this written in a 3rd party script before, in fact (let alone from the OS vendor!). This could well be an example of programmers new to a Unix-like platform still getting used to the Unix way of doing things, and getting bitten as a result.
Well, now that Fujitsu are concentrating on 2.5" format MO disks, this may not be too far fetched.
They already manufacture a 3.5" MO disk and drive which can store 2.3GB and is quite fast as well.
Minidisc is smaller (2.5") but can currently only hold about 150MB if I remember correctly. If Fujitsu can sell their upcoming technology to the audio markets as well as the computing market, then currently achieved densities would lead to a next generation minidisc storing ten times as much as they do currently. With ATRAC LP2, that'd be about 20 hours of music.
Not bad, eh?
Given that Sony and Fujitsu are apparently being very friendly on the MO front, this is not too far fetched. And given the ever onward march of progress in MO densities (it's quadrupuled in the last 2 years), it seems very likely that a new 2.5" MO disk from Fujitsu could well hold somewhat more than 1.5GB.
Perhaps it affects to small degree, but some people tend to believe it even defines the way people think. I think this is akin to superstition; the idea that words have magical powers.
'Magical' is probably not the best word for it, but words certainly do have spooky powers. A written or spoken word without understanding in the perciever is just scribble or noise. Language isn't just the physical manifestations, it's also the associations between these manifestations and mental states or symbols. These are by their very nature subjective, but one can't dismiss the subjective easily - each of us lives in a subjective world of our own, and we share enough of this world with our neighbours that we can communicate and interact.
Think of the words, "I promise", or in a wedding, "I do." These don't just describe a physical situation, they actually make a concrete change in a social environment.
Think also of the art of rhetoric, or the use of propoganda. There is a reason why we have very large and well funded institutes of marketing and advertising.
Words can be used to limit individuals (for example with self-supporting but very restrictive doctrine, as used by some fundamentalist religions) or to enable them (through educational texts or philosophical tracts.) They can have wide ranging effects through the influence they might have on a single powerful individual, or on a more subtle shaping of millions of minds.
In short, words have the potential to be immensely powerful.
How does this relate to the use of 'he' or 'she' or 'they' in English text? Proving that the indiscriminant use of 'he' affects the goals, aspirations and abilties of English speakers of either gender is probably hard - though no doubt it is something that has been investigated. But given the undeniable power of language, and the fact that a word's associations can come unbidden to the mind and flavour the ideas they describe (poetry anyone?)... it seems very easy to believe that gender-specific language could well have some effect, one which has the potential to be very harmful to a large proportion of the population, and in turn, to our society itself.
As we have perfectly good alternatives to gender-specific language, why not use them in order to minimise this risk? Especially when the risk seems to both eminently possible and very large in scope. Personally, I'm all in favour of "they" - it has precedence, it need not sound awkward when used intelligently, and is least disruptive to the reader. But alternating (consistent) use of he and she is another workable alternative.
The argument against such changes to our langauge use are I believe along the lines of: messing with the language inhibits clear communication; why should anyone be entrusted with the power to modify our language at all? (bringing in Orwellian connotations); and that we shouldn't fix what isn't broken. I think that it probably is broken, by the argument outlined above. Further, we're not being asked to change language by some central authority, but instead to change our language usage in an intelligent way so as to enable more members of our society to prosper. Lastly, with some care, it need not inhibit communication in the slightest.
Of course, as you say, ideas behind the words are more important than the words themselves. But how do those ideas get there? Typically through the use of language. Changing the way we speak and write won't in all likelihood change the way we ourselves think or the ideas we hold; it does seem very possible that it will affect those who are growing up in our society, and has the potential to have in the long term a great beneficial effect.
Microsoft's abuse of its monopoly position has hurt consumers both directly and indirectly.
Firstly, and most easily quantified, is the cost people pay for the operating system. Somewhere buried in the analysis that was floating about at the time of the trial, were the figures which showed that per-unit profits were not only high, but increasing over time. The argument is that this would not be sustainable in a competitive environment. In this situation, the only people benefitting were those with stock or some other interest in Microsoft.
Secondly, past introduced incompatibilities have inconvenienced or cost consumers. Such things as the DR-DOS debacle, or the incompatibility of 'standard' Microsoft file formats, or even the apropriation of file name extensions have put pressure on consumers to go the whole Microsoft way. This costs more money (or encourages copyright violation!), wastes time and is generally unhelpful.
Thirdly, the efforts which Microsoft have engaged in have slowed or stopped competition on a number of fronts. This has had an indirect effect on consumers through lack of options and alternatives. The situation with OS/2 springs to mind.
One can trawl the archives for more quantitative data, and other ways in which this situation has hurt consumers.
Another class of people hurt are developers, systems administrators and the like. Windows has never played nice in a mixed environment, and on occasion has been downright nasty. Mix that with the stability problems that have plagued many Windows versions, the lack of emphasis on security and so on, and it's a nightmare from a support point of view.
Oh! And then there are Macro viruses, Outlook-propogated viruses, and so on. A whole bunch of daft security decisions that have very much hurt consumers. Why would people stick with such virus-prone software? Monopoly perhaps?
I've been modded down before for being anti-Microsoft, but honestly, this is all based on personal experience and that information which has come to light through the anti-trust trials. This isn't malicious slander, it's simply true.
Part of the problem I think with the Ctrl, Alt, Esc thing in Emacs is that the PC keyboard has put two of these keys - Esc and Ctrl - in silly positions. Contrast it with (say) the 'Unix'-style Sun keyboards, where Esc is left of '1' and Ctrl is left of 'A'. Given the degree to which Ctrl is used compared to Caps Lock, their positions on the standard PC keyboards are just crazy.
Typically on any system which I am the sole or major user, I'll rebind Caps to Ctrl and have two Ctrl keys on the left hand side. This is easy in X11 and Windows 95/98 (with the right software), pretty straight forward for the Linux console, and really awkward in Windows 2k (obscure registry settings.)
It was interesting to read that Matt Dillon is supporting Rik van Riel's work on the VM in Linux.
There has been a lot of controversy ove the Linux VM, with a number saying that the latest one - based on Rik's work - had made a poor comprimise: more smooth operation for lower overall performance, making it better suited to interactive applications than server applications.
Not being a kernel-list follower, I don't know much about the details, but I'm sure there are other Slashdot readers who are much more familiar with them. I would have thought that lower expected latencies from the VM would improve most server-based tasks as well, at the possible cost of reducing the maximum amount of work the machine could do - but then, it seems counterproductive to talk about performance in such a heavily-loaded environment when the best solution would be typically to add more hardware (RAM, CPUs or split across multiple machines.) Do I have this all wrong?
Also, is anyone able to describe the degree to which the new VMs of both Linux and FreeBSD are similar? What are the concepts behind them that distinguish them from other and earlier VMs?
Some people have been asking, why Java? What's the point? The Dreamcast is dead!. In answer: lots of people still use their DC, so the more support, the better. More generally though, the DC is still a nice piece of kit.
As a gaming platform of course, the Dreamcast's days are numbered - it won't be long before the rate of release of new games declines to a trickle. Not that that should stop anyone from enjoying the many fine games already released of course!
Yet with the very low cost of the DC now, if you can find one, and the very wide distribution of information on the console, it has become a perfect, cheap platform for experimenting with embedded programming, or console game development, or even for trying out non-x86 based Unixes. Remeber that there is a version of BSD and Linux available.
The so-called 'coders cable' can provide connectivity for development, but for high speed access there is the official broadband adapter (hard to find and pricey). If you're feeling brave there has been some talk about the viability of a USB-Controller port adapter. The Maple Bus (used to connect the controllers) has been well documented.
An earlier slashdot story has already talked about a nice step-by-step for Linux on Dreamcast.
Minor nit: if we conflate the meanings of disinterested and uninterested, we lose a really quite useful distinction.
I'm sure you meant an uninterested populace: one which doesn't care and is indifferent. A disinterested populace would be unbiased by personal interest, which seems to be very much not the case. The only way to get people to vote at all seems to be to appeal to their lack of disinterest.
I never really understood the attraction of trolling. Why do it? Karma points to burn? Just like getting people worked up?
For the record, the use of such loaded terms as "stealing" and "piracy" to describe copyright violation doesn't help anyone. Let's move on from there, so that we can actually have rational, informed debate.
I really hope this doesn't get modded down to minusville, but I'm not confident.
I don't think the poster to whom you're referring was saying, "Sure it's okay to kill 5000 people and cause untold economic damage." That's not the point.
As the poster themselves say,
I think for most people in the U.S., bombing other countries is like an adult video game. They don't have any feeling for the pain that bombing causes.
These do not seem to be the words of someone who supports such actions.
The point as I see it, is that the US has caused a lot of grief to a lot of people, both directly as a supporter of regimes that are helping to support US interests (but not the interests of their own people) and indirectly through the huge arms industry exporting weapons around the world for profit.
This doesn't excuse the actions of terrorists who seek some way of striking back, but it does help to explain them. These are probably people who are not primarily concerned about US retaliation, as evidenced by the suicide pilots in their numbers. When their country gets bombed, are they going to be repentent? Seems unlikely. A violent response is only going to beget more violence, as it always has done in the past.
There are too many targets for terrorism; not everything can be adequately defended, and not every potential terrorist can be rounded up. So to do the most good, we (and I don't just mean the US) have to try and create an environment where people don't have this burning hatred, where they don't want to commit mass-murder. It's harder than bombing a country, but would do a lot of good.
This is not the same as giving in to the demands of terrorists (notice that in this instance, there seem to have been no demands to give in to, too.) Why are there terrorists to begin with? Chances are, they're not just random whackos.
PS: There is certainly historical comparison for the scope of damage done, especially in the midst of war. Dresden? Hiroshima? Certainly the latter killed nearly one hundred thousand people, as well as causing long-term posioning and the utter destruction of most of a city.
PPS:Given the evidence of past bombings/missle attacks by the US (Serbia, Afghanistan), the US does seem to take a rather cavalier attitude when it comes to the safety of other citizens. I'm sure the bombing of the pharmaceutical factory and the Chinese embassy were mistakes. But what mistakes! One could be forgiven for thinking that the US simply didn't care that much.
One could define metadata as any data associated with some other data, but isn't that a bit broad? It really does encompass everything from word count to plot description. Surely a more practical definition is data which is not just associated with, but additional to the original data in question.
That's a debatable point. In the broadest sense, then yes, size is metadata, but then so is the byte-frequency histogram of the data, or the number of odd-length words found in the data, and so on. The most broad definition of metadata is too broad to be useful. Perhaps there is a comprimise position, but I can't see anything wrong in working with that outlined above.
With the issue of size: sure, it has to be stored somewhere (or equivalently, with an EOF marker.) So do the bytes of the file. So do the error-correction bits stored on the hard disk (for example.) My point is that the singly-forked unadorned file - in the abstract - is a finite ordered set of bounded numbers (typically bytes). How that set is represented physically is an implementation concern, and will have to include an encoding method for the bytes themselves, probably some error correction at the lowest level, and of course the size, too. Finite sets by their very nature are delimitted somehow - if they weren't they wouldn't be finite!
The file-type of a file might be stored seperately (as metadata); it might be derivable consistently from the file contents (as metadata). It's not an intrinsic property of the ordered set of bytes the file represents - it's entirely up to interpretation within the computing context. This is unlike size. Size doesn't care what the content is. Size is a necessary and constant property of any finite set of data.
To take the example of the AFPDS data, the fact that you can tell it is an AFPDS file relies on information external to the simplest file abstraction. It requires an external context, that of the AFPDS file format, for the file type to be derived. There's nothing intrinsic to the ordered set of numbers that makes it so; it could well be weather-data to some other software.
I also need to address this statement:
SO, file size is most definitely metadata. There's no way of knowing what is in a set of data unless you have additional information which is not the data itself.
The size of the data is part of the data itself - well, more precisely - once you have described a set of data, you have also described its size. To be repetitive, storing the size, or an EOF marker, or somesuch is simply part of the representation of that set. Once could represent a file by a single huge integer, and then store that integer on disk, for example, impractical though it be. (Of course, the actual representation of the integer on disk would probably in the interests of space efficiency be delimited in some way.)
PS: I had never before encountered the acronym AFPDS. I looked it up in an acronym index, but it wasn't very enlightening:).
PPS: One can extend the abstract file notion to include streams, which are ordered possibly infinite sets of bytes (or whatever). Take for example,/dev/zero under Unix. In such a context file size doesn't make a lot of sense, but this is only tangentially related to the argument.
I was burnt so bad by DirectPlay in DirectX 3, 5, 6 and 6.1 that I just don't trust Microsoft to write an API for opening a paper bag, let alone implement it.
The API was badly organised (jumbled up layers of abstraction).
It was undocumented at the wire-level.
It tried to abstract over things for which there were already abstractions.
It promised features that weren't implemented.
The features that were implemented were big-time buggy.
It exercised windows threading bugs.
Add to this of course, that it was Windows-specific. Don't even think of trying to run a central server for a DX game on anything other than Windows. This one though was obvious from the start.
I inherited a DirectPlay code base for the networking code I wrote. Due to time constraints and the lack of documentation describing how bad DirectPlay was, we stuck with it and ultimately used it as no more than a UDP clone - none of the other functionality was reliable. We would have saved months, literally, just by coding our own over the standard Berkeley socket interface, perhaps using the async I/O extensions. This is exactly what we did in the next game.
I'm sure DirectPlay in DirectX 8 is much better. It would have to be. But I still don't trust it, and really fail to see the point of it. Why tie yourself to one platform unnecessarily?
There are two sorts of information at stake: one is data which is an adjunct to that contained in the file; the other is information that is a function of the data in a file. Size is a function of the file data, which is an ordered set of numbers. Sure, this number has to be stored somewhere, so that the files can be meaningfully read - but then so do the bytes of the files themselves! Storing the size is simply an implementation concern - the size is an intrinsic property of the data.
The more information you can extract from the file contents itself, the better in my opinion. Metadata - as the article pointed out - can easily be mangled or lost. In this regard, I'm a big fan of the default Unix scheme of/etc/magic, though I do strongly believe it could be improved. We can't throw out all our old file formats, but we can certainly agree on a standard form of file preamble, which say contained a mime type or other globally recognized unique type identifier, to be applied to new formats developed. Some mechanism for users to be able to extend the/etc/magic system would be nice, too.
Slashdot Mirror
I worked on a rushed project earlier this year, and used the gtk-- C++ wrapper for GTK, as well as the gnome-- wrapper (then still very much under development) for the Gnome libraries, specifically the canvas.
Personally, I found it frustrating to use. As these wrappers are still being worked on, the documentation is sketchy. The object-owning semantics are confusing (at least to me) - I was forever leaking memory or prematurely destroying objects. Trying to destroy something from within a menu callback I recall was particularly noisome. The gnome-- canvas wrappers were a moving target, though they may have since stabilised, and didn't fully expose the canvas API. Writing one's own canvas items is done in C, and then wrapped.
Perhaps with more persistance I might have figured out how to set up keyboard acceleration, but it is at anyrate a real battle to find documentation that explains what is going on with it. AFAIK, there is no straightforward way of making a multiple file selection in GTK+ 1.2. In gnome canvas (not GTK+, but a close cousin) there is promised functionality that is simply not implemented - I'm thinking here of smoothed lines, for which the code reads:
I haven't used QT yet. It certainly looks pretty, and a quick glance at the example code and docs provided seems to indicate that it's not too complicated, and well documented. I'd certainly shy away from GTK+ if a C++ interface is required.
The new version of GTK under development should address many of the shortcomings of the current toolkit, and includes goodies such as Pango. It is not yet in a stable state, however, with the API still undergoing final tweaking I believe.
My home Linux system is based on the Linux From Scratch documents, with the added organisational principle of having every /usr destined package in its own directory, and links made into the /usr tree.
This has worked fine for me with ldconfig, which reports itself as being version 1.9.9. My /etc/ld.so.conf doesn't require an entry per directory at all - provided there are symlinks from (say) /usr/lib/libblah.so to whereever the library really is.
There's certainly a lot to be said for the 'bottom-up' approach to nanotechnology. Cost for starters! One issue though is, how does one address these very tiny devices?
The problem with a whole bunch of identical tiny circuits is of course that they're all identical - there's no way to differentiate between them. There will have to be some way of distinguishing and interacting with these units.
A couple of ideas spring to mind though. One is to encode the position of one of these units in the unit itself as it is being assembled, by interacting with some sort of precisely engineered field. What would work (if anything) depends very much on the chemistry, but it could be something as simple as a gradient in an electrostatic field, to aligning with a very fine grid of polarized light. There are options, but it all sounds Hard. Schemes like this could attack the problem of differentiation, but there's still interaction and addressing.
One way to solve the addressing problem is to bypass it almost entirely. If these structures are sufficiently small, and can be engineered to act as a giant grid of finite-state automata with evolution rules based on neighbouring states, one can simulate a computational device with a version of Conway's Life on speed. Input and output can be done at the edges of the constructed array, which is probably going to be more simple than trying to address the middle of the structure. The problem lies in initialising the state of the array - clearing it is probably easy enough, depending on how state is stored, but priming it with a state that admits the computational task desired seems to be almost as hard as addressing the cells in the first place.
Another approach might be to give each cell some random state as it is constructed (and there should be plenty of sources of randomness at the molecular level to draw on.) Imagine that this state corresponds to an "activation key": when an appropriately modulated high frequency EM signal hits the cell, it pushes it over into an active state. Before this, it's effectively off (perhaps an off cell would simply propogate signals from its neighbours and do no computation). Give each cell some way of indicating that it has been activated (eg, it emits some light upon activation), and then fire random keys at the cells. This solves the addressing problem, and the interaction problem (one could use the same key for changing the cell's state) - but then one has no easy way of telling how the newly identified cell connects to the other addressable cells.
Do any slashdotters have any ideas? Or can point to literature where these problems are (ahem) addressed?
Thanks for the links!
It does appear though, tbat this is a copyright issue rather than a patent one. Copyright has long been applicable to mathematical papers and diagrams (for good or ill), and has far less damaging potential consequences than patenting.
According to the EuroLinux article,
Patenting mathematics is outright crazy. It's the same sort of crazy that allows the patenting of software, but in the past one could always say: patenting algorithms is like patenting mathematics, and thus clearly nonsense. reductio ad absurdum has come along and bit us all on the arse.Trying to imagine a world where mathematics is patentable is both hard and disturbing. Can you imagine if only licensed physicists were allowed to use Hilbert space theory? If one needed to pay a levy every time one used Shannon's law to help design a product? Where would we be if the finite element methods could only be applied to engineering analyses with the blessings of its creators?
How much mathematical progress would be made, if every mathematician had to check whether the work they were building on was patent-encumbered? If every publication had to first get the approval of some patent holders, with the possibility of a required payment?
It quickly gets surreal. Many statements in mathematics are equivalent when viewed in the appropriate fashion. Many too are based on certain sets of axioms. What does patentable mean when viewed in this light?
This to me is a clear sign that extreme IP advocates have just completely lost the plot.
Thanks for your comments!
...). The arts are regarded as culturally valuable, and as such need to be supported. However the recording industry only rewards those whose work is wildly popular. The market is not the best solution to this particular problem, as it typically rewards the least offensive rather than the most valuable; relies on artificial protections to be viable; and when it does reward, does so out of all proportion to the energies that went into creation. Alternatives and their discussion are, I feel, too important and long to be discussed in a quick slashdot comment though.
As the article you linked to states, it certainly is a high-risk field for new entrants. However it is hard to deny that those major labels that have survived, are profitting. The article itself states that the profit comes from their back catalogue - of the music which those companies own. I stand by my statement that the music itself is indeed a valuable asset.
I don't think I've been unfairly manipulative of the description of the current state of the music industry. When a few large companies hold the keys to commercial success, they get to make the rules. One of these rules is that artists no longer own their own creative works. Artists may have the right to be independent, but it's a very tough choice given the current situation. As you say, marketing and promotion are hard without a big label's backing.
Importantly though, I think you should be careful not to make a false dichotomy. The current system can be reprehensible while the alternative can be better than 'not getting paid at all for a recording'.
Many people engage in creative work, and many of these people create things that can be freely (if not necessarily legally) copied at little or no cost. Most do not rely upon huge popularity and royalties for recompense (for example, scientists, academics, visual artists,
Hillary Rosen says,
Note that she doesn't claim that they in the recording business respect artists or their work themselves. Courtney Love's rant on the piracy of the recording industry makes for educational reading. Later Rosen says, And of course they are. Look at the profits of the major labels. The problem being of course, is that this is monetary value, and further, they are much more valuable to the labels than the artists once the rights have been signed away.The language in the speech is emotive, as is to be expected. But the kiddie porn quote is surely beyond the pale,
And the very companies that the RIAA represent publish and promote music with hate-lyrics.We also have the old chestnut of referring to illegal copying as theft. Repeatedly. This should be plain enough, but many people seem to have bought the lie. Illegal copying is just that. It may well be damaging to the creators of the material (which is probably wrong) as well as to the distributors (which is not necessarily wrong - people don't have a right to make a profit, remember!). What it is not though, is theft. Let alone piracy. The debate on intellectual property is muddied enough as it is, without resorting to misleading language.
I think the most poignant quote though is,
This is so true. Sadly, it's the piracy of the recording industry - which has, among other things, managed to have artists' work reclassified as work for hire (!) - that is responsible for artists living in poverty while simultaneously having millions of CD sales. The term piracy is much more applicable to this sort of action; what these labels do is not illegal copying, but the wholesale transfer of rights from the artist to themselves using the big stick of exclusive access to mainstream distribution channels.If you have an interest in the music industry and not yet read the Salon article linked above, you really ought. It's very educational.
PS: If you do want to support artists, there is always Fairtunes.
I know it's not the same, but it does feel similar to what the MPAA tried to pull with Divx. Divx of course failed because customers didn't want to buy crippled equipment, and rightly so. Perhaps HP will face a similar response here.
Region coding is another example of crippling for profit, but unlike say Divx, it didn't affect the majority of customers. In the major markets of US and Japan, only a few would seek to play DVDs from outside their native region. Europe was more badly affected, and DVD still hasn't taken off in Australia really, due to the paucity of region 4 releases outside the big titles.
HP's crippling though would become apparent everytime one tried to record on it. What is Digital Audio Media, other than a disingenuous choice of name? I'm presuming it's the same as the (expensive) CD-Audio disks, which or course are just CD-R with magic mark on it for the benefit of (presumably) the RIAA.
Similar shenanigans killed DAT as a home medium, but maybe the other features of the HP device will win out. Recording aside, it does look like a nice piece of kit.
An older board - the UP2000 - is a dual processor SDRAM (not DDR) based Alpha motherboard, which has 6 PCI slots, two of which are 64-bit.
This new board has DDR ram, but only 32-bit PCI, and then only three slots. While nice and all - DDR is good, and of course it's for the Alpha 21264B, not 21264A - this does seem a bit of a step backwards in the IO stakes. Especially when it's noted that the UP2000 has onboard Ultra-2 SCSI as well.
Perhaps this board was originally targetted at the 'lower-end' workstation segment? Does anyone know if a more server-oriented 21264B board is on the way? It seems sadly unlikely given the current circumstances.
If one wants to have 64-bit multiprocessing on a budget, what are the current alternatives?
From the discussion on the Apple discussion web site, the nature of the bug is as follows.
The original installer script has the lines
while the replacement (2.0.1) has In these scripts, $2 corresponds to the volume on which iTunes is to be installed, and will be of the formFor those unfamiliar with Bourne shell variable expansion, if $2 has spaces in it, the argument to the rm command in the first version of the script will expand to more than one word, and rm will try and delete both of these. The -rf tells rm to delete everything down recursively and not complain about it.
This is particularly a problem on the Mac, where filenames and volume names often have spaces in them., even at the beginning of the name. If one had multiple partitions mounted in /Volumes, and the one on which iTunes was to be installed was called, say, ' OS X', then the rm command would expand to
and would then try and delete everything underThe second version, by including quotes around the argument, fixes the problem. The quotes force the argument to be treated as a single argument after variable expansion.
Traditionally, people have been super careful about destructive operations and shell expansions. I don't think I've ever seen something like this written in a 3rd party script before, in fact (let alone from the OS vendor!). This could well be an example of programmers new to a Unix-like platform still getting used to the Unix way of doing things, and getting bitten as a result.
Well, now that Fujitsu are concentrating on 2.5" format MO disks, this may not be too far fetched.
They already manufacture a 3.5" MO disk and drive which can store 2.3GB and is quite fast as well.
Minidisc is smaller (2.5") but can currently only hold about 150MB if I remember correctly. If Fujitsu can sell their upcoming technology to the audio markets as well as the computing market, then currently achieved densities would lead to a next generation minidisc storing ten times as much as they do currently. With ATRAC LP2, that'd be about 20 hours of music.
Not bad, eh?
Given that Sony and Fujitsu are apparently being very friendly on the MO front, this is not too far fetched. And given the ever onward march of progress in MO densities (it's quadrupuled in the last 2 years), it seems very likely that a new 2.5" MO disk from Fujitsu could well hold somewhat more than 1.5GB.
'Magical' is probably not the best word for it, but words certainly do have spooky powers. A written or spoken word without understanding in the perciever is just scribble or noise. Language isn't just the physical manifestations, it's also the associations between these manifestations and mental states or symbols. These are by their very nature subjective, but one can't dismiss the subjective easily - each of us lives in a subjective world of our own, and we share enough of this world with our neighbours that we can communicate and interact.
Think of the words, "I promise", or in a wedding, "I do." These don't just describe a physical situation, they actually make a concrete change in a social environment.
Think also of the art of rhetoric, or the use of propoganda. There is a reason why we have very large and well funded institutes of marketing and advertising.
Words can be used to limit individuals (for example with self-supporting but very restrictive doctrine, as used by some fundamentalist religions) or to enable them (through educational texts or philosophical tracts.) They can have wide ranging effects through the influence they might have on a single powerful individual, or on a more subtle shaping of millions of minds.
In short, words have the potential to be immensely powerful.
How does this relate to the use of 'he' or 'she' or 'they' in English text? Proving that the indiscriminant use of 'he' affects the goals, aspirations and abilties of English speakers of either gender is probably hard - though no doubt it is something that has been investigated. But given the undeniable power of language, and the fact that a word's associations can come unbidden to the mind and flavour the ideas they describe (poetry anyone?) ... it seems very easy to believe that gender-specific language could well have some effect, one which has the potential to be very harmful to a large proportion of the population, and in turn, to our society itself.
As we have perfectly good alternatives to gender-specific language, why not use them in order to minimise this risk? Especially when the risk seems to both eminently possible and very large in scope. Personally, I'm all in favour of "they" - it has precedence, it need not sound awkward when used intelligently, and is least disruptive to the reader. But alternating (consistent) use of he and she is another workable alternative.
The argument against such changes to our langauge use are I believe along the lines of: messing with the language inhibits clear communication; why should anyone be entrusted with the power to modify our language at all? (bringing in Orwellian connotations); and that we shouldn't fix what isn't broken. I think that it probably is broken, by the argument outlined above. Further, we're not being asked to change language by some central authority, but instead to change our language usage in an intelligent way so as to enable more members of our society to prosper. Lastly, with some care, it need not inhibit communication in the slightest.
Of course, as you say, ideas behind the words are more important than the words themselves. But how do those ideas get there? Typically through the use of language. Changing the way we speak and write won't in all likelihood change the way we ourselves think or the ideas we hold; it does seem very possible that it will affect those who are growing up in our society, and has the potential to have in the long term a great beneficial effect.
Microsoft's abuse of its monopoly position has hurt consumers both directly and indirectly.
Firstly, and most easily quantified, is the cost people pay for the operating system. Somewhere buried in the analysis that was floating about at the time of the trial, were the figures which showed that per-unit profits were not only high, but increasing over time. The argument is that this would not be sustainable in a competitive environment. In this situation, the only people benefitting were those with stock or some other interest in Microsoft.
Secondly, past introduced incompatibilities have inconvenienced or cost consumers. Such things as the DR-DOS debacle, or the incompatibility of 'standard' Microsoft file formats, or even the apropriation of file name extensions have put pressure on consumers to go the whole Microsoft way. This costs more money (or encourages copyright violation!), wastes time and is generally unhelpful.
Thirdly, the efforts which Microsoft have engaged in have slowed or stopped competition on a number of fronts. This has had an indirect effect on consumers through lack of options and alternatives. The situation with OS/2 springs to mind.
One can trawl the archives for more quantitative data, and other ways in which this situation has hurt consumers.
Another class of people hurt are developers, systems administrators and the like. Windows has never played nice in a mixed environment, and on occasion has been downright nasty. Mix that with the stability problems that have plagued many Windows versions, the lack of emphasis on security and so on, and it's a nightmare from a support point of view.
Oh! And then there are Macro viruses, Outlook-propogated viruses, and so on. A whole bunch of daft security decisions that have very much hurt consumers. Why would people stick with such virus-prone software? Monopoly perhaps?
I've been modded down before for being anti-Microsoft, but honestly, this is all based on personal experience and that information which has come to light through the anti-trust trials. This isn't malicious slander, it's simply true.
But thanks for taking the time to make the analogy!
Part of the problem I think with the Ctrl, Alt, Esc thing in Emacs is that the PC keyboard has put two of these keys - Esc and Ctrl - in silly positions. Contrast it with (say) the 'Unix'-style Sun keyboards, where Esc is left of '1' and Ctrl is left of 'A'. Given the degree to which Ctrl is used compared to Caps Lock, their positions on the standard PC keyboards are just crazy.
Typically on any system which I am the sole or major user, I'll rebind Caps to Ctrl and have two Ctrl keys on the left hand side. This is easy in X11 and Windows 95/98 (with the right software), pretty straight forward for the Linux console, and really awkward in Windows 2k (obscure registry settings.)
It was interesting to read that Matt Dillon is supporting Rik van Riel's work on the VM in Linux. There has been a lot of controversy ove the Linux VM, with a number saying that the latest one - based on Rik's work - had made a poor comprimise: more smooth operation for lower overall performance, making it better suited to interactive applications than server applications.
Not being a kernel-list follower, I don't know much about the details, but I'm sure there are other Slashdot readers who are much more familiar with them. I would have thought that lower expected latencies from the VM would improve most server-based tasks as well, at the possible cost of reducing the maximum amount of work the machine could do - but then, it seems counterproductive to talk about performance in such a heavily-loaded environment when the best solution would be typically to add more hardware (RAM, CPUs or split across multiple machines.) Do I have this all wrong?
Also, is anyone able to describe the degree to which the new VMs of both Linux and FreeBSD are similar? What are the concepts behind them that distinguish them from other and earlier VMs?
Some people have been asking, why Java? What's the point? The Dreamcast is dead!. In answer: lots of people still use their DC, so the more support, the better. More generally though, the DC is still a nice piece of kit.
As a gaming platform of course, the Dreamcast's days are numbered - it won't be long before the rate of release of new games declines to a trickle. Not that that should stop anyone from enjoying the many fine games already released of course!
Yet with the very low cost of the DC now, if you can find one, and the very wide distribution of information on the console, it has become a perfect, cheap platform for experimenting with embedded programming, or console game development, or even for trying out non-x86 based Unixes. Remeber that there is a version of BSD and Linux available.
The so-called 'coders cable' can provide connectivity for development, but for high speed access there is the official broadband adapter (hard to find and pricey). If you're feeling brave there has been some talk about the viability of a USB-Controller port adapter. The Maple Bus (used to connect the controllers) has been well documented.
An earlier slashdot story has already talked about a nice step-by-step for Linux on Dreamcast.
Minor nit: if we conflate the meanings of disinterested and uninterested, we lose a really quite useful distinction.
I'm sure you meant an uninterested populace: one which doesn't care and is indifferent. A disinterested populace would be unbiased by personal interest, which seems to be very much not the case. The only way to get people to vote at all seems to be to appeal to their lack of disinterest.
I never really understood the attraction of trolling. Why do it? Karma points to burn? Just like getting people worked up?
For the record, the use of such loaded terms as "stealing" and "piracy" to describe copyright violation doesn't help anyone. Let's move on from there, so that we can actually have rational, informed debate.
(I know I'm off topic but)
"A preposition is a terrible thing to end a sentence with."
"What did you bring that book I did not want to be read to out of up for?"
Can anyone provide attributions for these infamous quotes?
Oh, and there's always Churchill's "The rule which forbids ending a sentence with a preposition is the kind of nonsense up with which I will not put."
-- HF.
I really hope this doesn't get modded down to minusville, but I'm not confident.
I don't think the poster to whom you're referring was saying, "Sure it's okay to kill 5000 people and cause untold economic damage." That's not the point. As the poster themselves say,
These do not seem to be the words of someone who supports such actions.The point as I see it, is that the US has caused a lot of grief to a lot of people, both directly as a supporter of regimes that are helping to support US interests (but not the interests of their own people) and indirectly through the huge arms industry exporting weapons around the world for profit.
This doesn't excuse the actions of terrorists who seek some way of striking back, but it does help to explain them. These are probably people who are not primarily concerned about US retaliation, as evidenced by the suicide pilots in their numbers. When their country gets bombed, are they going to be repentent? Seems unlikely. A violent response is only going to beget more violence, as it always has done in the past.
There are too many targets for terrorism; not everything can be adequately defended, and not every potential terrorist can be rounded up. So to do the most good, we (and I don't just mean the US) have to try and create an environment where people don't have this burning hatred, where they don't want to commit mass-murder. It's harder than bombing a country, but would do a lot of good.
This is not the same as giving in to the demands of terrorists (notice that in this instance, there seem to have been no demands to give in to, too.) Why are there terrorists to begin with? Chances are, they're not just random whackos.
PS: There is certainly historical comparison for the scope of damage done, especially in the midst of war. Dresden? Hiroshima? Certainly the latter killed nearly one hundred thousand people, as well as causing long-term posioning and the utter destruction of most of a city.
PPS:Given the evidence of past bombings/missle attacks by the US (Serbia, Afghanistan), the US does seem to take a rather cavalier attitude when it comes to the safety of other citizens. I'm sure the bombing of the pharmaceutical factory and the Chinese embassy were mistakes. But what mistakes! One could be forgiven for thinking that the US simply didn't care that much.
Soon we'll have to take this to e-mail :)
One could define metadata as any data associated with some other data, but isn't that a bit broad? It really does encompass everything from word count to plot description. Surely a more practical definition is data which is not just associated with, but additional to the original data in question.
That's a debatable point. In the broadest sense, then yes, size is metadata, but then so is the byte-frequency histogram of the data, or the number of odd-length words found in the data, and so on. The most broad definition of metadata is too broad to be useful. Perhaps there is a comprimise position, but I can't see anything wrong in working with that outlined above.
With the issue of size: sure, it has to be stored somewhere (or equivalently, with an EOF marker.) So do the bytes of the file. So do the error-correction bits stored on the hard disk (for example.) My point is that the singly-forked unadorned file - in the abstract - is a finite ordered set of bounded numbers (typically bytes). How that set is represented physically is an implementation concern, and will have to include an encoding method for the bytes themselves, probably some error correction at the lowest level, and of course the size, too. Finite sets by their very nature are delimitted somehow - if they weren't they wouldn't be finite!
The file-type of a file might be stored seperately (as metadata); it might be derivable consistently from the file contents (as metadata). It's not an intrinsic property of the ordered set of bytes the file represents - it's entirely up to interpretation within the computing context. This is unlike size. Size doesn't care what the content is. Size is a necessary and constant property of any finite set of data.
To take the example of the AFPDS data, the fact that you can tell it is an AFPDS file relies on information external to the simplest file abstraction. It requires an external context, that of the AFPDS file format, for the file type to be derived. There's nothing intrinsic to the ordered set of numbers that makes it so; it could well be weather-data to some other software.
I also need to address this statement:
The size of the data is part of the data itself - well, more precisely - once you have described a set of data, you have also described its size. To be repetitive, storing the size, or an EOF marker, or somesuch is simply part of the representation of that set. Once could represent a file by a single huge integer, and then store that integer on disk, for example, impractical though it be. (Of course, the actual representation of the integer on disk would probably in the interests of space efficiency be delimited in some way.)PS: I had never before encountered the acronym AFPDS. I looked it up in an acronym index, but it wasn't very enlightening :).
PPS: One can extend the abstract file notion to include streams, which are ordered possibly infinite sets of bytes (or whatever). Take for example, /dev/zero under Unix. In such a context file size doesn't make a lot of sense, but this is only tangentially related to the argument.
I was burnt so bad by DirectPlay in DirectX 3, 5, 6 and 6.1 that I just don't trust Microsoft to write an API for opening a paper bag, let alone implement it.
Add to this of course, that it was Windows-specific. Don't even think of trying to run a central server for a DX game on anything other than Windows. This one though was obvious from the start.
I inherited a DirectPlay code base for the networking code I wrote. Due to time constraints and the lack of documentation describing how bad DirectPlay was, we stuck with it and ultimately used it as no more than a UDP clone - none of the other functionality was reliable. We would have saved months, literally, just by coding our own over the standard Berkeley socket interface, perhaps using the async I/O extensions. This is exactly what we did in the next game.
I'm sure DirectPlay in DirectX 8 is much better. It would have to be. But I still don't trust it, and really fail to see the point of it. Why tie yourself to one platform unnecessarily?
No the AC is not wrong on the issue of file size.
There are two sorts of information at stake: one is data which is an adjunct to that contained in the file; the other is information that is a function of the data in a file. Size is a function of the file data, which is an ordered set of numbers. Sure, this number has to be stored somewhere, so that the files can be meaningfully read - but then so do the bytes of the files themselves! Storing the size is simply an implementation concern - the size is an intrinsic property of the data.
The more information you can extract from the file contents itself, the better in my opinion. Metadata - as the article pointed out - can easily be mangled or lost. In this regard, I'm a big fan of the default Unix scheme of /etc/magic, though I do strongly believe it could be improved. We can't throw out all our old file formats, but we can certainly agree on a standard form of file preamble, which say contained a mime type or other globally recognized unique type identifier, to be applied to new formats developed. Some mechanism for users to be able to extend the /etc/magic system would be nice, too.