For example, a good word processor provides tools to help with writing the document, such as outline modes or views. All the great typesetting in the world is useless if the words you are typesetting are poor.
And a good LaTeX editor (and there are more than one) provides a lot of the same sorts of thing, including structure views (an outline view), spell checking, section folding and so on. A good typesetting program takes care of all the finer points of layout and formatting (in prepared documentclasses) elegantly so you can get on with worrying about what you are writing rather than how you're presenting it. All the lovely WYSIWYG options are useless if they just keep you distracted from the actual content of what you are writing.
And word processors are pretty fucking useless when you can't even expect to be able to give a saved file to another user and have them open it without problem. At least LaTeX gets around this problem by using a plain text format.
How is opening a raw LaTeX file really any different from opening the XML zipped inside an OpenDocument file? In practice both use a program to render the raw file into a more presentable format. if you don't have the program then you can't get the presentable format. Both are readable in the raw if that's required though.
Just because MS made a particularly fucked up closed binary format for their word processor doesn't mean that word processors are evil, it simply means MS isn't particularly good at making portable formats.
Why even bother with word processors these days when LaTeX is more than capable of the vast majority of document typesetting needs? It does take a bit longer to learn that Word, but everyone I know who has learned it has become far more efficient and can produce documents that are far more professional.
This is, quite simply, a remarkably stupid comment. I use LaTeX. For pretty much all my documents and presentations. I write my own document classes. Previously I have written LaTeX document classes reproducing the format of company Word and Powerpoint templates so I could produce my documents and presentations in LaTeX instead of MS Office - and yes, I did get that cleared with marketing. I am quite intimately familiar with all the power, flexibility and benefits that LaTeX has to offer. The fact remains that word processors are remarkably fast efficient and easy to use and entirely suitable for the majority of users. Most of the real benefits of LaTeX simply aren't of sufficient importance for most casual and business needs to bother - and it's not like word processors these days don't have their on benefits (usually relating to integration with the rest of an "Office Suite" package.
LaTeX is truly wonderful, and if you know how then by all means use it. But don't pretend that it's a replacement for a word processor - they are really filling different niches, and have quite different areas at which they excel, and at which they are weak. The right tool for the job and all that.
"no cheap and lazy 'holodeck' plot contrivances" Well you had space ships with wings that banked like airplanes, I do not remember how they solved that whole FTL thing. And if I remember right they often had the fighter jocks land there space planes and become ground pounders. It may have been enjoyable but no cheap and lazy contrivances?
Less than a lot of shows. It compares pretty well to the new BSG actually. The ships actually had control jets front and back (similar to BSG) and while they did have a tendency to do too much airplane style maneuvering, they did at least have some acknowledgement of the sorts of control systems required. They solved the whole FTL thing the same way everyone else does - they cheated and had some sort of hyperspace "jump". Find me an SF TV show that has some careful considered and well explained FTL system. And as for the last point - they were supposed to be "marines" thus mixing space and land combat - it's not an entirely unreasonable suggestion and, for instance, BSG does much the same thing: the pilots often find themselves acting and land troops.
Which is not to say it wasn't without plot contrivances, but they weren't of the "cheap lazy" form that all the holodeck and alternate dimension adn so on episodes of Trek.
Indeed, how that went missing is a bit of a puzzle. Strangely they managed to pick up Nowhere Man which was an interesting series with huge debts to The Prisoner, yet completely missed the superior source material.
I have to agree. As someone else said of Farscape, there are better SF shows, but there certainly aren't 50 of them! How it could be left off (along with Farscape and Red Dwarf apparently) is a mystery to me. Sure, it started poorly, but the last 2/3rds of the season were great. It had a lot more darkness and grit than many SF shows.
Also a little cnfusing is that while they managed to pick up Nowhere Man (which was a surprisingly good show all things considered) they somehow neglected The Prisoner (to which Nowhere Man owes a great deal).
And how exactly does Firefly finish behind Voyager and Xena?
Thirdly, a key is a REALLY good idea - something analogous to the Rosetta Stone.
Not exactly replying to your post as simply having my memory spurred with regard to something relevant: if you're really interested in storing information for future generations then The Rosetta Project is an interesting on. They seek to have as many distinct languages as possible printed on a small disk, beginning in large print but decreasing in size as it spirals inwards to the point where it is micro-etched. It's easy enough to figure out how to read it, and as long as you cna build tools to magnify it you can read everything on it.
And everybody gets cheaper goods in return. What's the difference to the American economy if money goes to some rich guy in China as opposed to some rich guy/company in the US?
Because the US pays for those goods with US dollars. A rich guy in China can't spend US dollars on goods where he lives, so those dollars need to get sold. In the end those dollars need to make their way back to US because they can only really buy goods in the US. Normally the money gets back to the US when it is used to buy US goods, and that's how the global economy works. The problem is that if the US is spending more US dollars on imports than it is buying back via exports then there are a lot of excess US dollars floating around out there. What good are those extra US dollars exactly? Well they are good for buying US goods, but there are more dollars piling up than there is value of US goods to buy. This tends to cause depreciation of the US dollar on global markets: people are less keen to buy US dollars because, having already bought all the US goods they want, they really aren't worth anything. As the US dollar depreciates it costs more and more in US dollars to buy foreign goods. That, in turn, promotes inflation in the US. If it gets bad enough total chaos ensues and you end up like Argentina with rampant run away inflation and total economic meltdown.
There are of course, many other factors involved, and many reasons why such economic chaos is quite avoidable, but fundamentally you simply can't keep sending all your money to China via a trade deficit, it just won't work.
What an excellent effort at misrepresenting what he was trying to say. The web is indeed document-centric. That doesn't mean you are restricted to reading, you can have natural extensions to interactive documents such as basic forms, which covers forums, online banking, interactive maps,and whatever else. Trying to cram entire programs with complicated interfaces down a document-centric pipe is not just a simple extension however. His point was that for those purposes there are other systems and other ports specifically designed for transferring complex GUIs and other things that may require significant client side computation. Trying to shoehorn everything into port 80 makes as much sense trying to do all your cooking in a microwave. Microwaves are great, but sometimes frying, roasting, baking, or grilling produces better results for some dishes.
Is it just me, or is slashdot actually using CSS now? It seems a little spotty (I get unformatted pages being rendered occasionally) but, um, otherwise it looks good... well, the same, but using style sheets. Way to go slashdot crew - I hope the kinks get worked out shortly.
It will be interesting to see how this pans out because it will have a lot to say about how copyright and intellectual property are being interpreted in the courts.
On the one had the authors do have a point: regardless of how little of the copyrighted works Google exposes to people searching, the fact is that Google itself is copying and making use of the whole work. Google is a for profit enterprise, and making books available for searching is part of that endeavour, so having a copy of the text is worth something to Google, yet they haven't sought any agreement with the authors to do so.
On the other hand, this is just stupid! What the fuck are they thinking? Google is effectively providing free advertising for them. Moreover such a service is obviously invaluable to the wider public, making it much easier for them to find (and then buy) the information they want.
Well I just had a look at AucTex and it has improved since I last used it, but things that would be nice are: automatic refernce and citation completion, folding that doesn't suck, a more integrated dvi/ps/pdf/html viewer, document structure, project and file, and reference/figure/table tag trees, a symbol picker (for those more obscure symbols), automated project archiving, context sensitive integrated LaTeX help, and a clean looking environemnt with antialised fonts.
Kile offers spellchecking and dynamic abbreviations. You don't have an entire lisp environment and email, webbrowser, newsreader and kitchensink, but then I'm more interested in editing LaTeX.
Look, I love Emacs. You wouldn't believe how much I used to use (pretty much everything), but in the end different tools were all slightly better at different specific tasks and Emacs attempt to e everything to everybody left it looking like an ugly stepchild. If you honestly think the speedbar is an elegant and beautiful solution, you need help.
I like Emacs and its brethren, however, I don't do software development in Emacs anymore. Full blown IDEs like Eclipse are really worth the power, especially since they are able to semantically understand source code.
I used to be a huge Emacs user and had all manner of interesting custom macros and functions written specialised for some of the things I did. In practice though the full scale IDEs have plowed ahead while Emacs is still wed to its way of doing things. Eclipse simply offers more features and a much more pleasant interface for editing code. Kile is so much better for editing LaTeX documents than the Emacs modes that it isn't funny. If I just want to edit config files then vi or nano are so much faster to load that it isn't worth breaking out Emacs.
Emacs works great though when I have to interact in a complex way with the shell. For instance, I find it very useful when used in conjunction with command line SQL clients for Postgres, Mysql, or Oracle.
This is the one area Emacs still stands out as being a better option than most. It works nicely for R and a number of other similar applications as well. I think, however, that such a lead is going to diminish. Already things like JEdit offer similar sorts of tet editor extensibility features that make customised functions and macros easily available in a more modern editing environment.
I like LaTeX, but you know the default presentation in the standard document classes was only meant to be a quick demo, right? It was assumed that serious writers/publications would all create their own classes using sensible typesetting preferences.
I do tend to write my own documentclasses, though often for reasons other than just to fiddle with margins. And as much of a quick demo as it may have been, it remains a remarkably good default all things considered, and is certainly better than what Word has to offer as its defaults.
That's what I meant. The more text you cram on the page, the less readable, so you're trying to maximise the function of "readable + amount of text on page".
Personally I prefer LaTeX and send pdf files. That works ok till I am working alone. But if we have to work and interact, keeping track of changes is not the easiest thing to do in LaTeX.
Keeping track of changes is as easy as RCS/Subversion/version control system of choice (I've even used Visual SourceSafe when I was in an MS shop). Sharing changes can be done easily enough via PDF annotations, or LaTeXdiff depending on what tools you have available.
LaTeX also offers possiblities that simply aren't available in word processors like MS Word and OO.o Writer. Using packages like xcomment it is possible to write a single document that is both a paper report and slide presentation - just change the document class and recompile. I've written document classes such that I have a couple of extra environments available: \begin{summary} and \begin{shared}. Anything in a summary environment is included in the presentation, but not in the report, and anything in shared is in both report and presentation. Anything not in either environment is left out of the presentation. With that done it is easy enough to start writing your report, adding a little set of bullet points summarising each paragraph in a summary environment as you go (and sharing any equations and diagrams as needed) and once you're done you've got your presentation complete as well as your report. You've also go the whole package encapsulated in a single file: any changes are easy to propogate from report to presentation of vice-versa, and maintenance is far easier. Try that with your standard office suite.
So, were your margins a bit too big? Or do you mean you sent him the latex source and he printed that out? Just because it's Latex doesn't mean you can't make shitty looking documents with it.
I suspect the OP was referring to the fact that MS Word has what are, by typesetting standards, very narrow margins that make for long lines of text. In practice narrow text actually proves to easier to read, requiring less left-right scanning with the eyes and making the end-of-line to start-of-next-line shift much easier and less prone to error. Professional typesetters are not idiots and have been studying and refining such things for a very long time. LaTeX defaults to the same margins you'll find in professionally typeset books and other publications - the same margins professional typesetters have come to use after years and years of experience and refinement. They look large if you're used to MS Word documents, but are by most other measures, the margin size that maximise readability and amount of text on the page.
In the past, I've made the argument that forcing people to declare types lowers productivity and decreases legibility.
The one good rebuttal I got was that it really helps to declare types on function parameters; it serves as a kind of documentation of the intended use and operation of a function. Plus, of course, it allows the compiler to catch errors where the function you write doesn't have the type you say it does.
I really wish using contracts (preconditions and postconditions) for this purpose would catch on. It provides better more exacting documentation of the function (not just the type on inputs and outputs, but specifics of what they must be, and what sorts of results you can expect in return), with an OO language that supports it (like Eiffel or D) it fits very naturally with inheritance (you cna weaken inherited preconditions, but not strengthen and vice versa for post-conditions), and it automatically provides an extremely powerful test harness for all your code by simply setting a switch to check all contracts at runtime.
If this sounds interesting and you program in Java, please take a look at JML and related tools. It lets you specify contracts via a JavaDoc style notation that
(1) Gets added to your javadoc documentation. (2) Can be used to automatically generate a unit testing framework. (3) Allows you to use extra static checking tools (like ESC/Java2) that can catch far more errors than static types alone - all prior to runtime.
If you're bothering to document your functions why not formalise the documentation a little further to allow generation of unit tests and extra static checks? It just makes sense.
So because the function says it takes an integer, and I fed it some integer, everything is going to be fine by definition? I think not. You are still going to need to exercise every line of code in runtime tests.
Static types are just a very weak form of formal specification. A stronger form would be something like contracts (preconditions, postconditions and invariants) as used in things like Eiffel and D and SPARK, which allow you to specify not just the type of the inputs and outputs, but specific properties that are required of them. Another step up the formality chain again requires you to be very explicit in your specification allowing the contracts to be statically checked prior to compilation - that covers all the code paths and doesn't require runtime checking - and properties of the software to be formally proved.
Static typing finds maybe 10% of the possible potential errors in your program at the expense of doubling the effort to write and understand it. In some cases static typing may be a necessary evil to attain the required execution speed, but as a "safety" feature any benefits it provides are outweighed by the complexity it adds to your code.
Whether the benefits you gain from any level of formal specification, be it static typing or full formal methods, are worth it depends on how badly you need to be sure the system works and how costly any errors are. For some systems it really isn't that important - having a prototype or working code is more important that catching every last glitch. For other systems (security, avionics, banking and finance software etc.) any sort of glitch, bug or exploit is sufficiently costly that the extra effort to catch as many errors statically as possible is going to be worth it. There is also a spectrum of different needs in between.
As another note, static types needn't complicate your code as much as you might imagine. Try looking at a language with good static types (SML or Haskell come to mind) and you'll see that static types can be quite easy and require very little extra work (in comaprison to static types in other languages).
In my experience static vs. dynamic typing is really a matter of how well I've managed to nail down exactly what sorts of data structures I'm going to use. If I'm still doing exploratory/prototype/research code then dynamic types win out every time because they provide incredible flexibility. If I have nailed down what data structures I want to use then static types aren't any extra work and provide and extra buffer of safety.
C# is strongly typed, Ruby and Python (not sure about Delphi) are dynamically typed.
First off it's static typing that C# has and Python and Ruby don't. Python and Ruby have strong typing, while Perl and C have weak typing. Delphi has strong static typing.
As a side note, it looks as if Python is going to get optional static typing with type inference in future versions, which could make for an interesting and flexible language.
I think a certain amount of his argument comes via mischaracterising how most popular and prevalent open source software is actually developed. He ants to envision it as the collection of 1000 lonely teenage hackers all chipping away at the code in their parents basement. Admittedly OSS has promoted itself that way in the past, but that doesn't seem to be how things really work.
Conceptual Integrity
Marshall tries to make the argument that because OSS projects are a pure design by committee hodge podge effort from hundreds of developers they have no overriding conceptual design to create a solid architecture. While that is true, to some extent, of a whole OSS OS, utlities and applications (such as a Linux distribution) it isn't really true of most of the major OSS projects. Sure there are plenty of little projects that may work that way, but most of the big ones like Linux, OpenBSD, Firefox, Samba, GIMP, most anything that's actually achieved significant mindshare, tends to have an iconic leader, or a small team of gatekeepers that head up the project and concern themselves with conceptual design and conceptual integrity. It is very rare indeed to find a major open source project that is truly an open free-for-all with no small group dictating the overall design. Hell, there seem to be a lot of people bitching abotu the fact that GNOME is run by a small group of people who are hewing tightly to their particular conceptual design. This doesn't seem to be anything like the issue Marshall makes it out to be.
Professionalism
Marshall tries to argue that OSS is like the games industry of the early 80's with a plethora of basement hackers turning out a fine array of crap. Again, in some sense this is true: trawl through Freshmeat or Sourceforge and you will find no end of half-assed poorly written barely functional open source projects. That's mostly because anyone can write something and call it open source. Take a look at the world of shareware Windows applications and you'll see the same thing.
If you take a look at the major software and applications in the OSS world, from the Linux kernel to OpenOffice, a large amount of the work done is done by professionals working for major companies. You see, major companies are interested in having a good kernel, or office suite, or desktop environment, or whatever - and they are often willing to pay people to work on those things. Think of it as an opportunity for IBM, Sun and Novell to work together on an Office Suite that they can all get to use. Increasingly OSS development is professional development paid for by big companies. Sure the code then gets shared openly, but that's another matter.
Innovation
Marshall tries to argue that OSS is merely a matter of copying what has gone before and is incapable of innovation. In a lot of ways this is more to do with catching up than to do with a significant lack innovative capability. In every "Linux isn't ready for the desktop" article since the late 90's there have been people crying "well I would switch bt there isn't an equivalent of X", and so OSS developers have been endeavouring to provide said equivalent. There's a lot of ground to make up. Why are so many of the equivalent applications so similar? Because it is easier to bring users over that way - look at how many people won't use GIMP because it's interface isn't a clone of Photoshop, and other similar cases. OSS does innovate, it just tends to do so in the areas where it is strongest and is playing catchup the least: in security with SELinux, in networking with Stateless Linux, and other network services, in scripting and programming with new languages like Perl and Python and Ruby, and so on. Those aren't major desktop applications, so they tend to be less visible to the average consumer, but innovation is happening. As OSS catches up in other areas innovation will start becoming more obvious there too.
A lot of the "OSS can't innovate" sentiment stems from a belief that the only motivation is money
Wellit seemsZetahasbeen kickingaroundfor some time. The earliest of those being from 2002 when it was first announced that yellowTab had picked up some rights to BeOS. I gather (from comments in those many stories) that Zeta has been on sale, at least in Germany, for quite some time now, and went 1.0 in July. The reviews have been lukewarm, and it really hasn't raised much (if any) mainstream attention.
Will Yellowtab raise BeOS from the ashes and inflame public interest in the OS?"
I find that rather doubtful. BeOS was a fine OS in its day, but while the rest of the world has been improving (MacOS, for instance, now actually has something decent to offer) BeOS has been mostly treading water as yellowTab try and modernise it where possible and get support for modern hardware. It's not that Zeta is bad - it looks like quite a nice OS - it's just that it certainly isn't revolutionary or particularly interesting for any reasons other than BeOS nostalgia... and these days you need to manage to stand out in some way or other as an OS to attract enough application developers. Without applications your OS is just going to slowly stagnate and die unless you can find and fill a niche. Given that Zeta is aiming at the general desktop... I just don't see them managing to get enough strng application support to really pull that off.
It seems very silly to hold them to a double standard. Microsoft is under no obligation to cater to Novell and their associates. Just as you would not expect groups associated with Linux to be under any obligation to cater for Microsoft.
Sure, no one is complaining that the Mono session wasn't allowed, it is more about the manner in which it was done. "Birds of a Feather" was supposed to be a set of open meetings - anyone could suggest a topic, the topics would get listed, then voted on, an the top n topics would be chosen for the PDC. Miguel requested a slot for Mono, the request was accepted and then... well somehow it didn't get listed for voting, and all his queries from then on were ignored. Naturally enough, never having even been listed for voting, the mono session got no votes and was dropped.
If there was a Linux conference that advertised open sessions with topics to be voted on and Microsoft requested a slot then I would fully expect their topic to be put up for voting. If it got voted down, fine, if it got voted on I expect that would get honored.
It's not the being ignored, its the dishonesty that irks. Although some would say Miguel should have expected as much from Microsoft, it would be nice to think that we shouldn't epect such petty churlish behaviour.
Jedidiah.
Re:Python is nice but consider LUA for game script
on
Game Scripting With Python
·
· Score: 4, Informative
The point of using Stackless Python (as opposed to ordinary Python) is that it provides a particularly good system for handling multiple threads and communcation for threads via tasklets and channels. If your game engine works by creating actors as threads then using a scripting language that has a simple to use, efficient, and platform independent threading model is likely of great importance, and Stackless Python offers that.
If you're generally interested in better threading models, and being able to think and reason about threads and their interactions more easily then you really ought to check out CSP. Multithreading is actually easy if you do it right - it's just that most languages don't.
Slashdot Mirror
For example, a good word processor provides tools to help with writing the document, such as outline modes or views. All the great typesetting in the world is useless if the words you are typesetting are poor.
And a good LaTeX editor (and there are more than one) provides a lot of the same sorts of thing, including structure views (an outline view), spell checking, section folding and so on. A good typesetting program takes care of all the finer points of layout and formatting (in prepared documentclasses) elegantly so you can get on with worrying about what you are writing rather than how you're presenting it. All the lovely WYSIWYG options are useless if they just keep you distracted from the actual content of what you are writing.
Jedidiah.
And word processors are pretty fucking useless when you can't even expect to be able to give a saved file to another user and have them open it without problem. At least LaTeX gets around this problem by using a plain text format.
How is opening a raw LaTeX file really any different from opening the XML zipped inside an OpenDocument file? In practice both use a program to render the raw file into a more presentable format. if you don't have the program then you can't get the presentable format. Both are readable in the raw if that's required though.
Just because MS made a particularly fucked up closed binary format for their word processor doesn't mean that word processors are evil, it simply means MS isn't particularly good at making portable formats.
Jedidiah.
Why even bother with word processors these days when LaTeX is more than capable of the vast majority of document typesetting needs? It does take a bit longer to learn that Word, but everyone I know who has learned it has become far more efficient and can produce documents that are far more professional.
This is, quite simply, a remarkably stupid comment. I use LaTeX. For pretty much all my documents and presentations. I write my own document classes. Previously I have written LaTeX document classes reproducing the format of company Word and Powerpoint templates so I could produce my documents and presentations in LaTeX instead of MS Office - and yes, I did get that cleared with marketing. I am quite intimately familiar with all the power, flexibility and benefits that LaTeX has to offer. The fact remains that word processors are remarkably fast efficient and easy to use and entirely suitable for the majority of users. Most of the real benefits of LaTeX simply aren't of sufficient importance for most casual and business needs to bother - and it's not like word processors these days don't have their on benefits (usually relating to integration with the rest of an "Office Suite" package.
LaTeX is truly wonderful, and if you know how then by all means use it. But don't pretend that it's a replacement for a word processor - they are really filling different niches, and have quite different areas at which they excel, and at which they are weak. The right tool for the job and all that.
Jedidiah.
"no cheap and lazy 'holodeck' plot contrivances"
Well you had space ships with wings that banked like airplanes, I do not remember how they solved that whole FTL thing. And if I remember right they often had the fighter jocks land there space planes and become ground pounders.
It may have been enjoyable but no cheap and lazy contrivances?
Less than a lot of shows. It compares pretty well to the new BSG actually. The ships actually had control jets front and back (similar to BSG) and while they did have a tendency to do too much airplane style maneuvering, they did at least have some acknowledgement of the sorts of control systems required. They solved the whole FTL thing the same way everyone else does - they cheated and had some sort of hyperspace "jump". Find me an SF TV show that has some careful considered and well explained FTL system. And as for the last point - they were supposed to be "marines" thus mixing space and land combat - it's not an entirely unreasonable suggestion and, for instance, BSG does much the same thing: the pilots often find themselves acting and land troops.
Which is not to say it wasn't without plot contrivances, but they weren't of the "cheap lazy" form that all the holodeck and alternate dimension adn so on episodes of Trek.
Jedidiah.
Next best is an oldie: The Prisoner.
Indeed, how that went missing is a bit of a puzzle. Strangely they managed to pick up Nowhere Man which was an interesting series with huge debts to The Prisoner, yet completely missed the superior source material.
Jedidiah.
I have to agree. As someone else said of Farscape, there are better SF shows, but there certainly aren't 50 of them! How it could be left off (along with Farscape and Red Dwarf apparently) is a mystery to me. Sure, it started poorly, but the last 2/3rds of the season were great. It had a lot more darkness and grit than many SF shows.
Also a little cnfusing is that while they managed to pick up Nowhere Man (which was a surprisingly good show all things considered) they somehow neglected The Prisoner (to which Nowhere Man owes a great deal).
And how exactly does Firefly finish behind Voyager and Xena?
Seems like a real slap dash hodge podge to me.
Jedidiah.
Thirdly, a key is a REALLY good idea - something analogous to the Rosetta Stone.
Not exactly replying to your post as simply having my memory spurred with regard to something relevant: if you're really interested in storing information for future generations then The Rosetta Project is an interesting on. They seek to have as many distinct languages as possible printed on a small disk, beginning in large print but decreasing in size as it spirals inwards to the point where it is micro-etched. It's easy enough to figure out how to read it, and as long as you cna build tools to magnify it you can read everything on it.
Jedidiah.
And everybody gets cheaper goods in return. What's the difference to the American economy if money goes to some rich guy in China as opposed to some rich guy/company in the US?
Because the US pays for those goods with US dollars. A rich guy in China can't spend US dollars on goods where he lives, so those dollars need to get sold. In the end those dollars need to make their way back to US because they can only really buy goods in the US. Normally the money gets back to the US when it is used to buy US goods, and that's how the global economy works. The problem is that if the US is spending more US dollars on imports than it is buying back via exports then there are a lot of excess US dollars floating around out there. What good are those extra US dollars exactly? Well they are good for buying US goods, but there are more dollars piling up than there is value of US goods to buy. This tends to cause depreciation of the US dollar on global markets: people are less keen to buy US dollars because, having already bought all the US goods they want, they really aren't worth anything. As the US dollar depreciates it costs more and more in US dollars to buy foreign goods. That, in turn, promotes inflation in the US. If it gets bad enough total chaos ensues and you end up like Argentina with rampant run away inflation and total economic meltdown.
There are of course, many other factors involved, and many reasons why such economic chaos is quite avoidable, but fundamentally you simply can't keep sending all your money to China via a trade deficit, it just won't work.
Jedidiah.
What an excellent effort at misrepresenting what he was trying to say. The web is indeed document-centric. That doesn't mean you are restricted to reading, you can have natural extensions to interactive documents such as basic forms, which covers forums, online banking, interactive maps,and whatever else. Trying to cram entire programs with complicated interfaces down a document-centric pipe is not just a simple extension however. His point was that for those purposes there are other systems and other ports specifically designed for transferring complex GUIs and other things that may require significant client side computation. Trying to shoehorn everything into port 80 makes as much sense trying to do all your cooking in a microwave. Microwaves are great, but sometimes frying, roasting, baking, or grilling produces better results for some dishes.
Jedidiah.
Is it just me, or is slashdot actually using CSS now? It seems a little spotty (I get unformatted pages being rendered occasionally) but, um, otherwise it looks good... well, the same, but using style sheets. Way to go slashdot crew - I hope the kinks get worked out shortly.
Jedidiah.
It will be interesting to see how this pans out because it will have a lot to say about how copyright and intellectual property are being interpreted in the courts.
On the one had the authors do have a point: regardless of how little of the copyrighted works Google exposes to people searching, the fact is that Google itself is copying and making use of the whole work. Google is a for profit enterprise, and making books available for searching is part of that endeavour, so having a copy of the text is worth something to Google, yet they haven't sought any agreement with the authors to do so.
On the other hand, this is just stupid! What the fuck are they thinking? Google is effectively providing free advertising for them. Moreover such a service is obviously invaluable to the wider public, making it much easier for them to find (and then buy) the information they want.
Jedidiah.
Well I just had a look at AucTex and it has improved since I last used it, but things that would be nice are: automatic refernce and citation completion, folding that doesn't suck, a more integrated dvi/ps/pdf/html viewer, document structure, project and file, and reference/figure/table tag trees, a symbol picker (for those more obscure symbols), automated project archiving, context sensitive integrated LaTeX help, and a clean looking environemnt with antialised fonts.
Kile offers spellchecking and dynamic abbreviations. You don't have an entire lisp environment and email, webbrowser, newsreader and kitchensink, but then I'm more interested in editing LaTeX.
Look, I love Emacs. You wouldn't believe how much I used to use (pretty much everything), but in the end different tools were all slightly better at different specific tasks and Emacs attempt to e everything to everybody left it looking like an ugly stepchild. If you honestly think the speedbar is an elegant and beautiful solution, you need help.
Jedidiah.
Alternatively you can use
M-:
(use-global-map (make-sparse-keymap))
It's not quite as complete an emulation as Viper mode, but it is a surprisingly close approximation.
Jedidiah.
I like Emacs and its brethren, however, I don't do software development in Emacs anymore. Full blown IDEs like Eclipse are really worth the power, especially since they are able to semantically understand source code.
I used to be a huge Emacs user and had all manner of interesting custom macros and functions written specialised for some of the things I did. In practice though the full scale IDEs have plowed ahead while Emacs is still wed to its way of doing things. Eclipse simply offers more features and a much more pleasant interface for editing code. Kile is so much better for editing LaTeX documents than the Emacs modes that it isn't funny. If I just want to edit config files then vi or nano are so much faster to load that it isn't worth breaking out Emacs.
Emacs works great though when I have to interact in a complex way with the shell. For instance, I find it very useful when used in conjunction with command line SQL clients for Postgres, Mysql, or Oracle.
This is the one area Emacs still stands out as being a better option than most. It works nicely for R and a number of other similar applications as well. I think, however, that such a lead is going to diminish. Already things like JEdit offer similar sorts of tet editor extensibility features that make customised functions and macros easily available in a more modern editing environment.
Jedidiah.
I like LaTeX, but you know the default presentation in the standard document classes was only meant to be a quick demo, right? It was assumed that serious writers/publications would all create their own classes using sensible typesetting preferences.
I do tend to write my own documentclasses, though often for reasons other than just to fiddle with margins. And as much of a quick demo as it may have been, it remains a remarkably good default all things considered, and is certainly better than what Word has to offer as its defaults.
Jedidiah.
That's what I meant. The more text you cram on the page, the less readable, so you're trying to maximise the function of "readable + amount of text on page".
Jedidiah.
Personally I prefer LaTeX and send pdf files. That works ok till I am working alone. But if we have to work and interact, keeping track of changes is not the easiest thing to do in LaTeX.
Keeping track of changes is as easy as RCS/Subversion/version control system of choice (I've even used Visual SourceSafe when I was in an MS shop). Sharing changes can be done easily enough via PDF annotations, or LaTeXdiff depending on what tools you have available.
LaTeX also offers possiblities that simply aren't available in word processors like MS Word and OO.o Writer. Using packages like xcomment it is possible to write a single document that is both a paper report and slide presentation - just change the document class and recompile. I've written document classes such that I have a couple of extra environments available: \begin{summary} and \begin{shared}. Anything in a summary environment is included in the presentation, but not in the report, and anything in shared is in both report and presentation. Anything not in either environment is left out of the presentation. With that done it is easy enough to start writing your report, adding a little set of bullet points summarising each paragraph in a summary environment as you go (and sharing any equations and diagrams as needed) and once you're done you've got your presentation complete as well as your report. You've also go the whole package encapsulated in a single file: any changes are easy to propogate from report to presentation of vice-versa, and maintenance is far easier. Try that with your standard office suite.
Jedidiah.
So, were your margins a bit too big? Or do you mean you sent him the latex source and he printed that out? Just because it's Latex doesn't mean you can't make shitty looking documents with it.
I suspect the OP was referring to the fact that MS Word has what are, by typesetting standards, very narrow margins that make for long lines of text. In practice narrow text actually proves to easier to read, requiring less left-right scanning with the eyes and making the end-of-line to start-of-next-line shift much easier and less prone to error. Professional typesetters are not idiots and have been studying and refining such things for a very long time. LaTeX defaults to the same margins you'll find in professionally typeset books and other publications - the same margins professional typesetters have come to use after years and years of experience and refinement. They look large if you're used to MS Word documents, but are by most other measures, the margin size that maximise readability and amount of text on the page.
Jedidiah.
In the past, I've made the argument that forcing people to declare types lowers productivity and decreases legibility.
The one good rebuttal I got was that it really helps to declare types on function parameters; it serves as a kind of documentation of the intended use and operation of a function. Plus, of course, it allows the compiler to catch errors where the function you write doesn't have the type you say it does.
I really wish using contracts (preconditions and postconditions) for this purpose would catch on. It provides better more exacting documentation of the function (not just the type on inputs and outputs, but specifics of what they must be, and what sorts of results you can expect in return), with an OO language that supports it (like Eiffel or D) it fits very naturally with inheritance (you cna weaken inherited preconditions, but not strengthen and vice versa for post-conditions), and it automatically provides an extremely powerful test harness for all your code by simply setting a switch to check all contracts at runtime.
If this sounds interesting and you program in Java, please take a look at JML and related tools. It lets you specify contracts via a JavaDoc style notation that
(1) Gets added to your javadoc documentation.
(2) Can be used to automatically generate a unit testing framework.
(3) Allows you to use extra static checking tools (like ESC/Java2) that can catch far more errors than static types alone - all prior to runtime.
If you're bothering to document your functions why not formalise the documentation a little further to allow generation of unit tests and extra static checks? It just makes sense.
Jedidiah.
So because the function says it takes an integer, and I fed it some integer, everything is going to be fine by definition? I think not. You are still going to need to exercise every line of code in runtime tests.
Static types are just a very weak form of formal specification. A stronger form would be something like contracts (preconditions, postconditions and invariants) as used in things like Eiffel and D and SPARK, which allow you to specify not just the type of the inputs and outputs, but specific properties that are required of them. Another step up the formality chain again requires you to be very explicit in your specification allowing the contracts to be statically checked prior to compilation - that covers all the code paths and doesn't require runtime checking - and properties of the software to be formally proved.
Static typing finds maybe 10% of the possible potential errors in your program at the expense of doubling the effort to write and understand it. In some cases static typing may be a necessary evil to attain the required execution speed, but as a "safety" feature any benefits it provides are outweighed by the complexity it adds to your code.
Whether the benefits you gain from any level of formal specification, be it static typing or full formal methods, are worth it depends on how badly you need to be sure the system works and how costly any errors are. For some systems it really isn't that important - having a prototype or working code is more important that catching every last glitch. For other systems (security, avionics, banking and finance software etc.) any sort of glitch, bug or exploit is sufficiently costly that the extra effort to catch as many errors statically as possible is going to be worth it. There is also a spectrum of different needs in between.
As another note, static types needn't complicate your code as much as you might imagine. Try looking at a language with good static types (SML or Haskell come to mind) and you'll see that static types can be quite easy and require very little extra work (in comaprison to static types in other languages).
In my experience static vs. dynamic typing is really a matter of how well I've managed to nail down exactly what sorts of data structures I'm going to use. If I'm still doing exploratory/prototype/research code then dynamic types win out every time because they provide incredible flexibility. If I have nailed down what data structures I want to use then static types aren't any extra work and provide and extra buffer of safety.
Jedidiah.
C# is strongly typed, Ruby and Python (not sure about Delphi) are dynamically typed.
First off it's static typing that C# has and Python and Ruby don't. Python and Ruby have strong typing, while Perl and C have weak typing. Delphi has strong static typing.
As a side note, it looks as if Python is going to get optional static typing with type inference in future versions, which could make for an interesting and flexible language.
Jedidiah.
I think a certain amount of his argument comes via mischaracterising how most popular and prevalent open source software is actually developed. He ants to envision it as the collection of 1000 lonely teenage hackers all chipping away at the code in their parents basement. Admittedly OSS has promoted itself that way in the past, but that doesn't seem to be how things really work.
Conceptual Integrity
Marshall tries to make the argument that because OSS projects are a pure design by committee hodge podge effort from hundreds of developers they have no overriding conceptual design to create a solid architecture. While that is true, to some extent, of a whole OSS OS, utlities and applications (such as a Linux distribution) it isn't really true of most of the major OSS projects. Sure there are plenty of little projects that may work that way, but most of the big ones like Linux, OpenBSD, Firefox, Samba, GIMP, most anything that's actually achieved significant mindshare, tends to have an iconic leader, or a small team of gatekeepers that head up the project and concern themselves with conceptual design and conceptual integrity. It is very rare indeed to find a major open source project that is truly an open free-for-all with no small group dictating the overall design. Hell, there seem to be a lot of people bitching abotu the fact that GNOME is run by a small group of people who are hewing tightly to their particular conceptual design. This doesn't seem to be anything like the issue Marshall makes it out to be.
Professionalism
Marshall tries to argue that OSS is like the games industry of the early 80's with a plethora of basement hackers turning out a fine array of crap. Again, in some sense this is true: trawl through Freshmeat or Sourceforge and you will find no end of half-assed poorly written barely functional open source projects. That's mostly because anyone can write something and call it open source. Take a look at the world of shareware Windows applications and you'll see the same thing.
If you take a look at the major software and applications in the OSS world, from the Linux kernel to OpenOffice, a large amount of the work done is done by professionals working for major companies. You see, major companies are interested in having a good kernel, or office suite, or desktop environment, or whatever - and they are often willing to pay people to work on those things. Think of it as an opportunity for IBM, Sun and Novell to work together on an Office Suite that they can all get to use. Increasingly OSS development is professional development paid for by big companies. Sure the code then gets shared openly, but that's another matter.
Innovation
Marshall tries to argue that OSS is merely a matter of copying what has gone before and is incapable of innovation. In a lot of ways this is more to do with catching up than to do with a significant lack innovative capability. In every "Linux isn't ready for the desktop" article since the late 90's there have been people crying "well I would switch bt there isn't an equivalent of X", and so OSS developers have been endeavouring to provide said equivalent. There's a lot of ground to make up. Why are so many of the equivalent applications so similar? Because it is easier to bring users over that way - look at how many people won't use GIMP because it's interface isn't a clone of Photoshop, and other similar cases. OSS does innovate, it just tends to do so in the areas where it is strongest and is playing catchup the least: in security with SELinux, in networking with Stateless Linux, and other network services, in scripting and programming with new languages like Perl and Python and Ruby, and so on. Those aren't major desktop applications, so they tend to be less visible to the average consumer, but innovation is happening. As OSS catches up in other areas innovation will start becoming more obvious there too.
A lot of the "OSS can't innovate" sentiment stems from a belief that the only motivation is money
Well it seems Zeta has been kicking around for some time. The earliest of those being from 2002 when it was first announced that yellowTab had picked up some rights to BeOS. I gather (from comments in those many stories) that Zeta has been on sale, at least in Germany, for quite some time now, and went 1.0 in July. The reviews have been lukewarm, and it really hasn't raised much (if any) mainstream attention.
Will Yellowtab raise BeOS from the ashes and inflame public interest in the OS?"
I find that rather doubtful. BeOS was a fine OS in its day, but while the rest of the world has been improving (MacOS, for instance, now actually has something decent to offer) BeOS has been mostly treading water as yellowTab try and modernise it where possible and get support for modern hardware. It's not that Zeta is bad - it looks like quite a nice OS - it's just that it certainly isn't revolutionary or particularly interesting for any reasons other than BeOS nostalgia... and these days you need to manage to stand out in some way or other as an OS to attract enough application developers. Without applications your OS is just going to slowly stagnate and die unless you can find and fill a niche. Given that Zeta is aiming at the general desktop... I just don't see them managing to get enough strng application support to really pull that off.
Jedidiah.
It seems very silly to hold them to a double standard. Microsoft is under no obligation to cater to Novell and their associates. Just as you would not expect groups associated with Linux to be under any obligation to cater for Microsoft.
Sure, no one is complaining that the Mono session wasn't allowed, it is more about the manner in which it was done. "Birds of a Feather" was supposed to be a set of open meetings - anyone could suggest a topic, the topics would get listed, then voted on, an the top n topics would be chosen for the PDC. Miguel requested a slot for Mono, the request was accepted and then... well somehow it didn't get listed for voting, and all his queries from then on were ignored. Naturally enough, never having even been listed for voting, the mono session got no votes and was dropped.
If there was a Linux conference that advertised open sessions with topics to be voted on and Microsoft requested a slot then I would fully expect their topic to be put up for voting. If it got voted down, fine, if it got voted on I expect that would get honored.
It's not the being ignored, its the dishonesty that irks. Although some would say Miguel should have expected as much from Microsoft, it would be nice to think that we shouldn't epect such petty churlish behaviour.
Jedidiah.
The point of using Stackless Python (as opposed to ordinary Python) is that it provides a particularly good system for handling multiple threads and communcation for threads via tasklets and channels. If your game engine works by creating actors as threads then using a scripting language that has a simple to use, efficient, and platform independent threading model is likely of great importance, and Stackless Python offers that.
If you're generally interested in better threading models, and being able to think and reason about threads and their interactions more easily then you really ought to check out CSP. Multithreading is actually easy if you do it right - it's just that most languages don't.
Jedidiah.