I've had this behavior before, but it turned out to be the D-Link firewall/router in front of the box dropping the connections. Windows XP wouldn't push the router hard enough, but a very similar program on Linux would.
When W32/Nachi came out, this was a common problem under Windows, as it would push these little routers to the point that TCP connections would routinely time out. Anyone who's DSL mysteriously "stopped working" (when it was beyond saturated) because of this virus knows exactly what I'm talking about. I had four customers call me to fix their DSL only to find out that they had viral infestations.
Firefox requires you to explicitly install extensions. Not only must you click install, it makes you read the install dialog (the "Install" button is greyed out for a period after the window pops up). Oh, and it needs to be a trusted site too.
Most spyware gets into IE by exploiting bugs so that you never see the install.
This is largely because of IE's marketing game. At some point, the web browser threatened the traditional application space. Put less kindly, all of those VB monkeys (not to be confused with programmers) would lose their jobs because of those darned web page thingies.
Microsoft threw the web people a curve ball with IE. They attacked open standards with their negligence, gave the VB monkeys a mechanism to spread their pox on the world (ActiveX), and created a new, exciting way to be tied to Microsoft (ActiveX and IE in general).
Before Microsoft, 9 out of 10 people weren't clever enough to "develop software". Of course, they aren't now either, but they can make it look decent enough. It's no wonder that people trying to make real progress have to deal with this kind of overgeneralized tripe.
More to the point, any web browser needs a plugin architecture to be extensible. All of your assertions above blithely assume this is not your problem. Regardless, the only way to allow:
a) Extensibility
and
b) Security
is to allow the user to provide the security layer. People miss this, but there is no way (and likely will never be any way) to have a computer recognize what is malicious. At some point, the user will always have to make the choice to install something that fundamentally changes the computer. At that point, there will always be this problem.
The catch is, IE often will allow installations without so much as a prompt. This is the problem. So, actually, IE *DOES* install bad code. It often does so because IE contains broken code itself.
In this case, Firefox functioned within the security model and the use instructed it to install code which was broken. That was the problem. Worlds of difference here. This is a GreaseMonkey problem, pure and simple.
Put another way, remove GreaseMonkey from the equation and there ceases to be a problem. Well, unless you continue to maintain that a browser (or any software) should magically infer intent, in which case I have some snake oil to sell you--assuming Microsoft has left you with any money.
The BSD TCP stack was as good as it was ONLY because people gave back. It's still very good ONLY because people give back. You only serve to illustrate that Open Source (even the BSD kind) is so effective that it works even when only partially adopted.
If people benefit from your product, they benefit from it being better. How does that become "The ones who care more about the benifit of everyone rather than just their project will go with BSD."?
Face it, getting code back into the project is better for them.
FYI, I've developed tons of GPL derived code without giving the code back to the project. It's false the GPL requires it. The GPL only requires it if you DISTRIBUTE the software. 90% of software development is in house. This is why most businesses don't usually give a damn about GPL vs BSD--because they don't intend to sell their core business techniques to their competitors.
The GPL only prevents someone from building a lucrative business on what you developed and let them use. BSD, on the other hand, isn't much stronger than the Public Domain. All of this "freedom to use the software" is a red herring. The GPL gives you the same freedom to use the software as the BSD, until you intend to build a business on it. At that point, why is it bad they a company has to either play the sharing game or earn their pay?
Either share or don't, but get off the BSD-licensed fence. As far as I'm concerned, only dead software (of the "maybe someone else will do something with it" type) is appropriate for BSD licensing.
I don't see why people get up in arms about the GPL "discouraging business" but somehow think that BSD "is business friendly". This only matters to business selling software--which is not most of business. I think it's obvious from the.COM Bust, rampant user "sharing", and the advance FOSS how sensible the "software as a product" business model is becoming.
The fact of the matter is this: bytes are cheap. Economics is against you. You can't sell something with no intrinsic value without grossly abusing the economic process (which is never sustainable). Don't put yourself there, it's stupid. Provide something valuable instead.
It's not just capitalism at work detering suicide bombers in the US. Keep in mind that there are many types of coercion possible in Iraq, Saudi Arabia, and Syria that just aren't so easy to pull off in the US.
Looking at the daily suicide bombing in Iraq, you'd think they show a decent pool of people willing to die.
However, many of them have been shown to be handcuffed to the steering wheel of their bomb-laden car. Most of these are then detonated by someone watching.
Most disturbing are the reports that these people are often sympathetic to the new Iraqi regime. Insurgents target these people, chain them to bombs, and give them the option to drive up to the enemy + get blown up + have your family live or not drive up to the enemy + get blown up + have your family die.
It's particularly insidious but note that most foreign fighters must be brought to the US, since it can be tough to make a whole, local family disappear here. Police are just too eager and effective to let something like that slide.
In the same vein as the parent post, all of this is pretty much the direct consequence of religion (defined as religious institutions specifically). Religion has always been about creating a control structure (and all of the coercion that comes with it). People will tell you that you can't have God without religion, but I believe that to be obviously false.
You cannot have a religion without somebody in charge (they fracture into small cults, read about the Branch Davidians). In creating this structure, someone always dictates what "the greater good" wants.
Radical militant Muslim extremests are but one of the examples, but the corrosive effects of systematized social coercion can be much more insidious. I live in the Midwest US (town of less than 350,000 people) and I've known more than one person to leave a church because of serious pressure to vote a certain way.
When God dictates it, there can be no dissent. Without dissent, there can be no democracy. I don't think that anyone's God wants a totalitarian state (theocratic or otherwise). The fallacy lies in a man dictating the word of their "God".
Not that I'm so sure about the whole "God" thing...
I mean, I know that some sites/products (notably Plone use some JavaScript to implement missing CSS properties in various browsers, but an ActiveX super-fixer-upper?
I've seen ActiveX do a lot of things (many that it shouldn't), but never fix IE rendering to the point it's pixel perfect with Photoshop. Not to mention that the biggest problem on the web is still fonts (although this may be due to legal, not technical, concerns).
Not to mention that ActiveX controls only ever make anything worse. Seriously. If you need ActiveX for a web interface, you should probably be writing a native interface instead.
I would give almost anything to watch Dick Cheney and John Kerry duke it out with red light sabers, force push, force throw, force lightning, etc. on the floor of Congress.
I'd watch CSPAN for THAT!
I'd love to see some House of Representatives lightsaber battles too!
How much petroleum goes into petroleum production? Seriously.
Don't forget that fossil fuels had massive amounts of energy going into their production. Definitely more than 6x. It's just that we didn't have to do most of that work and that it happened on a million year time scale. I mean, we're talking millions of tons of biomass, millions of years, countless joules of solar energy, tectonic forces, etc. Not to mention the later drilling, excavating, pumping, hauling, etc. As such, the fact that Ethanol can be produced with 6x the energy it outputs is amazing, considering how how that number really is for fossil fuels.
That said, this just goes to show that more work needs to go into it (or something else) to get there. I don't see anybody posting the amount of energy that goes into CREATING fossil fuels (since we don't CREATE them at all).
There's more than just efficiency of production. Ethanol hold energy better, is more portable, and expends it more consistently than, say, electricity in a battery. So if you're talking solar panels, we can't use them for what we use ethanol for (not just energy, plastics too).
I don't think it is too out of reach that we will someday use abundant fusion, tidal, solar, or thermal (the oceans) sources of electrical power to still create hydro-carbon fuels for many smaller vehicles. I also would be willing to bet we may even get less efficient by using that same power to produce antimatter as a fuel--for the same reasons we need a hydro-carbon fuel (portability, energy density, ease of reaction, exciting explosions when our machines fail, etc).
I find it ironic you can read "requires open formats" as "bad for open source". I find the statement of "outright removing a competitor" humorous. The customer demands a product, any fool knows that Microsoft can produce almost any product they want to. The issue is that they don't want to provide that product, thereby removing THEMSELVES from the market.
Were the governments of the world to declare that they wanted only proprietary software, you'd probably find Open Source not in that market. Of course, the only way to argue for this is pretty much an economic issue. At this point, I think it's pretty obvious that Open Source provides a better value than Proprietary for a large swath of needs. Also increasingly obvious is that Open Source is scales better than proprietary. Increases in proprietary players are predatory, increases in open players are additive (assuming they are truly open).
There was a time when the existence of Mail Merge in an application was the same way. Sure, you could create your own mail merge, but in the end, the word processor having this functionality was the right decision for the customers. The same goes for open formats. Lack of "vendor lock-in" is the killer feature that drives Open Source more than anything.
The fact of the matter is that the free-program interoperability that you enjoy today exists because of millions of man-hours of work to achieve it. Say what you want about a "political cudgel", the cold hard fact is that a few million spent by Microsoft creates a few hundred million dollars of work just to interoperate with it (maybe more including the outside QA that goes into the products by the time they mature).
The investment by Norway in continuing to create this problem is not justified. As for PDF, if you ever have to implement a PDF handler, you'll be yearning for the coming days of SVG-P. It's like the bastard stepchild of PostScript and an early filesystem. Seriously, it almost creates a filesystem in a file (all kinds of block lists of irregularly sized blocks) and enough complexity to make it very difficult to parse (and impossible to recover when suitably corrupted). Adobe realizes this, that's why they're already knee deep in SVG (see the Adobe SVG Viewer if you don't believe me).
No, the sad thing is that so many people are so in love with the sacred cow of Capitalism (of which Proprietary software is apparently the posterchild), that they don't realize that Capitalism is still beholden to the Free Market. Every time I see someone crying about abuse of the proprietary model, I can't help but realize a simple fact.
It's all about demand. People demand what they want. They want software as cheaply as possible. It is no longer necessary that it cost very much at all. In fact, given the support/installation/development model that a lot of FOSS uses today, it can be funded entirely from people providing those very real services (instead of billing for bytes).
Face it, selling software as a business is pretty much doomed unless it is really that complex to write or your market is pitifully small. There will come a time when people really can't be fooled anymore. Software is not valuable. Time is valuable. When people are paid for their time and the software is open, everyone saves time--and every one thus saves money. Efficiency is what makes or breaks businesses (see Walmart, evil but dead efficient).
In otherwords, I'll gleefully revel in all of this as I put Microsoft to the fire, both because they would do the same for me and because this really is a victory for the freedom of developers and users alike.
Most of Google's magic is really data mining the semantic data from the Internet.
Gmail is nothing more than an attempt at getting a massive corpus of data on which to let their algorithms loose.
I really think that, while there is potential for abuse, this is really the only way to tackle their problem space. After all, Google doesn't really rank web sites, people do. It's just that Google has some really clever ways for determining that people liked a web site.
Sometimes it relates to webs of links, sometimes it relates to combinations of words, but Google's software doesn't deal in semantics--only algorithmically generating statistics from the data generated by people.
I don't worry so much about Google, I worry about our future AI overlords. Although, if a truly scalable Artificial Intelligence ever gets Internet access, I fear it has the potential to know us better than we do.
If MS needs extensions, that's what namespaces are for.
As long as MS extensions don't change formatting functionality (this is really not rocket science, Word is not an innovator here), they can tack whatever metadata they need into the file format and still have it be portable.
If you don't believe me, look at what Inkscape has done with SVG. Psodipodi built on it, adding a namespace to provide their needed data. Inkscape did the same on top of that. It produces one file containing three XML namespaces containing with interoperable metadata for two editors--and it's viewable in stock SVG viewers.
Obviously it's up to Microsoft to "do the right thing" with their metadata, but this certainly levels the playing field so that others can do what they need to with the documents.
Now the patent on XML-to-object-mappings, that's another story...
Ummm, geeks like Google because Google employs scientists. Which mere scientists were you talking about?
Were you talking about the PhDs at universities busy teaching classes, churning out research papers to avoid being fired (an ugly numbers game some departments play), or perhaps burning time generating volumes of grant paperwork?
Oh, maybe you were talking about the scientists employed by the private sector. I'm sure the management teams wherever they work are willing to take the time and care that Google won't.
You do know how may PhDs Google employs, right? Not to mention that they won't be fighting for resources there either. No backstabbing, liquidating MBAs trashing their corporate budget. No football-crazed alumni assassinating their funding proposals either.
Also, I would remind you that "mere scientists" often come up with the needed research (there are volumes in MT alone), but rarely can afford to put in the years that it takes into a good implementation.
Geeks love Google because it is, in many respects, where the best of business meets the best of academia.
Being a Python programmer, I misremembered it as "one of those other P-languages that gives me maintenance headaches".:)
That being said, again, RT can be quite functional. I've found it runs better under PostgreSQL, but never tested it with Mod_Perl (didn't know it did that).
You would be wrong to characterize Freenet in that way.
As for being wrong to ban or be against that system, I don't think I can address that question in the abstract. I know nothing about that system I'm supposed to endorse banning. Not that a system exists that there is "only one legitimate user" and "millions of inappropriate users". Who would endorse banning something they know nothing about (except for Slashdot Trolls, perhaps).
You could have used that rationale, for instance, to justify the wholesale destruction of the Underground Railroad prior to the Civil War. Slave smuggling was wrong, immoral, and illegal for a number of Southerners.
Besides, if you knew much about Freenet you'd know that it would be a lot faster if there were that many users (legitimate or not).
Re:Great, here come the CP trolls
on
Revamping Freenet
·
· Score: 2, Insightful
Ill conceived my arse.
Take America for instance. How many children are killed by guns each year? How many children are abused in CP each year?
The guns are just as guilty of killing children as Freenet is of creating CP.
Children abuse in real life, recordings trafficked on the Internet. Apparently that means we should suppress the technology of the internet.
Thirteen times as many people are shot the same year (even taking suicides and under-reporting of CP into account, the numbers are still overwhelming) but we settle for public service announcements. Ask any police officer in a big city, it burn you up inside finding abused children every day, but never once would they prefer to find one dead. Give a parent the same option, same answer.
Blaming the technology instead of the criminal only makes doing even reasonable things illegal and puts innocent people in jail, because, let's face it, Child Abusers will find a way to do it unless we remove every last shred of freedom, privacy. and due process from our society.
No, the issue here isn't a analogies. The issue here is the fact that some people value their speech (or their guns) and some don't. I value children's lives more than most Slashdotters, but I also understand that Child Pornography is the crime, NOT transfering files over a P2P network or communicating anonymously.
In fact, I'd say that anonymous and P2P communications might actually be two of the "rights" that the children of the future may enjoy most. If the casualties of the war on child pornography (or the war on terror for that matter) are going to be the freedoms that allow democracy to exist, there will be far more victims than we can appreciate right now.
As for OpenOffice, I use central network location, see the setup guide (I think you have to run setup.exe with the -net option). I'm not sure what must be done from there to automate installation, we usually do it manually because Workstation installs of OOo (from a central network location) take seconds.
As for the question of whether the MS deal was a "good value". First, let me say that there's more to "value" than cost. Also realize that $50000 per year might be cheaper than MS's $15000. Once you figure in MSCE training for an IT team and the increased labor it takes to run a Windows network you might be surprised. Believe me, once configured, Linux machines can be dead reliable and reimaged lightning fast, I do it for a living. That said, Firefox has saved me 8 hours per week at one client that only has 10 computers.
Well, ask your purchasing department how many suppliers it has for, say, light bulbs. While more than a few places say "just one", I find universities in particular tend to have four or five suppliers solely for the purpose of leveraging one against the other for good pricing.
What's the point of my story? The point is that MS as a single supplier means you will pay as much as they want you to. Of course it will always be "a little cheaper". In a software world with real competition, that will change.
Regardless, it's worth pointing out that increasingly it is the case that people are choosing FOSS for reasons other than price:
Re:Great, here come the CP trolls
on
Revamping Freenet
·
· Score: 4, Insightful
Since we're exercising our rights, I agree with your point.
After careful consideration of your point, I realize the error of my ways.
Specifically, I realize that by running a Freenet node, I would be allowing others to use it the way you've said. I would be providing support to a system that, as a whole, exploits children.
I have since determined that telephones and video cameras are used in the production of CP. I can no longer supply the phone company or video empires with the money I pay for their products, since their products are used for bad things, as well as good.
I think we should do the same to the gun manufacturers, since we all know how guns are used.
Seriously though, the technology (Freenet) does not exploit children. People exploit children. Don't ban technology because it is effective. Freenet has NEVER exploited a single child any more or less than a telephone, the US Postal Service, and the entire line of Sony Handicam(TM)'s.
You do realize that by the time the video hits whatever distribution system that it's already too late, right? If you want to protect the children...why not actually protect the children. Don't use abused children further as an excuse because you're uncomfortable with other people expressing their rights outside of some sort of central control--because that's the object of Freenet--for better (political reform) OR worse (abuse of children for profit).
Finally, realize that the First Amendment is not what it used to be. There was a time that speech and thought could not be ultimately suppressed. Given the current state of technology, I'm not sure that will be sustained. Disturbing as it may be, someday the First Amendment may not be a nod to an unsupressable reality, but it will be that last bastion between an unscrupulous bureacratic machine and individual freedoms.
This does not complicate things. It simplifies them. They are not talking about writing a database application, they are talking about writing DATABASES! So, since your question is about how to write a database-based application when the poster is offering a solution for a database implementation, this is kind of an apples-oranges situation.
To expound on Reiser4, it addresses the database problem-space by allowing individual files to contain extended attributes that may be queried on using a boolean algebra (think a table row with an optional payload).
As the query language used is very Reiser specific, and very general purpose (and not people oriented), I won't bother to construct a query, but it would not be an extraordinary feat of computer science to write an SQL to Reiser4 query converter.
More to the point, the tabular nature of most enterprise databases is the rub here. In Reiser4, you can have path structures, you can have indexed queries across these paths, and you can have it at database speeds. Try to do a hierarchical data structure in SQL. It can be done, but performance for some operations can be...suboptimal.
Furthermore, Reiser4 not only has your standard ACID guarantees available, it also has what Hans calls a transcrash. Transcrashes are pretty much ACI but not so much D. This is very useful in a backend scenario where a cluster of machines provides the Durability (which is where I think databases are going).
Also, hard links beat any foreign key scheme I've ever fought with.
If you read what the poster means, new databases may be based on these filesystems. An SQL translator and distributed transaction processor based on Reiser4 would probably compete well in todays DB performance world. If not, I suspect Hans would make sure it did eventually.:)
I've always fancied writing an XML parser that backended in Reiser4 (containment = directory, attributes = xattrs) just to play with the indexing. When Reiser4 hits a stable tree, I might...
Actually, I'm constantly amazed by the number of people who don't know that there IS Microsoft Office.
A lot of the "civilians" that I help out have bought computers bundled with Office and have Office provided at work, so they don't understand that it's not PART OF WINDOWS (if they understand that Windows isn't a fundamental part of the computer, which lots don't).
Recently Office has been yanked from the more competitive PC options from Dell and the like, so more people I know buy a computer and freak out when they find out that they have to buy Office (and then freak out again when they see the price).
Worse yet, there's still a middle ground option with Works Suite (which bundles Word), so lots of people are still getting this broken impression that Office is Windows is computers!
At any rate, anyone considering OOo absolutely MUST download the new Beta. While the new Database component of it is certainly Beta quality, the Word Processing/Spreadsheet/Presentation stuff is still rock solid. That said, there have been light years of improvement since the last version.
The OOo beta handles layout better than 97 or 2003 for many documents. This is no small feat, since, at least in Office 97, there is quite a lot of variation even within Word as to printing layout. It may not be common knowledge, but Microsoft uses almost the same Win32 API calls to print as it does to display on the screen. This has the interesting side effect that things that mingle the two (Word) can be affected by switching PRINT DRIVERS!!! I've even had bad print drivers cause crashes when certain documents are displayed (had to change to a different printer on another machine to open it there, changing back to the same printer caused the crash, new driver version replaced that bug with other ones).
Crashing aside, switching printers often causes layout changes. Professional printing shops HATE this "feature" of word. The mantra is usually "come back with a PDF or sign off on a crappy proof". I had hoped we left this behind with WordPerfect 5.1 for DOS, but not so!
The number of possible ways to effectively accomplish something in software is limited.
A patent grants EXCLUSIVE rights to one of those ways.
Therefore, you have just created a land war. Only the rich and monied win a land war. Soon, you'll have nowhere to live. Good luck with that.
Put simply, there is no "stealing of ideas". That's ludicrous. I take your car? You have no car. I take your method of bush trimming? Both of our bushes get trimmed. That's life.
That's also the worst case. On the other hand, what if I develop a similar way of trimming bushes? Now who's stealing. I just wanted to trim my bushes, now you can take money from me! Who's the thief?
At some point the businesses of the world are being given the power to own the EXCLUSIVE right to sell something. I think we all know why that's bad. No competition == screwing people.
It's bad when the government mandates it (national telecomm companies). It's bad when monopoly enforces it (Standard Oil, Microsoft). It's bad when the people suffer it.
If I develop MY IDEA independently of you, I'll be damned if your patent should matter to me. Unfortunately, this is now my problem.
Worse, now the only people with significant patent portfolios won't be people. Instead they'll be the most morally reprehesible construct mankind has ever unleashed--corporations. Worst of all, they're pretty much designed to aggregate financial and legislative power.
Wow, someone was bitten by a "computer professional" as a child, huh?
Seriously, every segment has its opportunists, but corporate MIS "was and is" hardly the whole of "computer professionals". Not everyone had the problem of which you speak.
Also, understand that quite often what you've experienced was due to management. At the time. decent MIS people were both expensive and rare. As such, it was often difficult to get information from them because management dictated it to be so. After all, if everyone had them fielding requests, they'd be so backlogged that they couldn't run the stupid TPS reports for the VP. Seriously. I've been as much as told that I should only accept any requests from VPs or higher unless I wanted a bad performance review. Thank god I'm out of there now.
As for Visicalc, spreadsheets aren't always a win. I've had to "upsize" a lot of spreadsheets that should have never been made (mostly after Excel though). That said, they're useful (the parent didn't argue otherwise). I used sc long ago to do the exact same thing.
What's this microcomputer complaint? UNIX==microcomputer these days. Unfortunately the same opportunists as above were too cheap to port it at the time, so maybe it was an obstacle. Sounds like noise to me though.
Finally, read your last paragraph. You don't dislike UNIX. You dislike the interface. The COMMAND LINE interface at that. Condemning the OS for the interface is just dumb. Remember Mac System before UNIX? Remember DOS? Put another way, the car may have a pretty paint job and interior, but if it doesn't run very well and the parts always fail, do you want to own it? Don't be naive.
Not that it matters. We have a pretty GUI for you. I've been using X-Windows since UnixWare. It was usable then, today it's fantastic (KDE is there, GNOME is there). I have it deployed in businesses. People use it for like three weeks then quietly ask "Where's Minesweeper?". I actually have minesweeper amd solitaire clones that I install for the management.
If you want to compare UNIX, try comparing it to the VMS hack that became NT/XP. Or to the BSD in MacOS. Apples to Apples. Oranges to Oranges. You know the drill.
This is not a root exploit. This is not a buffer overflow. The worst we've got here is a settings loading problem.
The fact of the matter is that Firefox is exhibiting the correct behavior but the specification is flawed.
If you look at the sad mess that is today's encoding of SSL certificates, you can see the same thing. There are a lot of ways to encode a certificate so that it's valid but doesn't work most places. Of course, most people never have to deal with this dark art, so its not as obvious.
The fact of the matter is that this "vulnerability" isn't one. The vulnerability lies in the consequences of the IDN standard, the reliance of SSL on DNS and ASN.1, and the world-wide security standard not being based on truly verified public-key cryptography.
And as for the "war of ideology", the war is won in the trenches. See how long it takes to get a fix. If you don't see one fast enough, THAT is how the war will be lost. Not by raving lunatics on Slashdot!
Also, don't forget that this "vulnerability" comes with some benefit. Without IDN, having domain names in Chinese, Korean, Japanese, Arabic, Hebrew, et al. would simply not be possible. I think that, had this problem not been solved, we would be suffering seriously more balkanization in the internet--possibly resulting in a new asian/international browser. I think it's obvious to see why that could be bad (at least of Verisign).
If anything, what you are witnessing is an incomplete application of Unicode. IIRC, unicode specifies equivalency tables for sorting and comparing these things. Now support for unicode in SSL certificates in a whole other game. Mind you, I think if we settled on punycode encoding them in the SSL certificate, we would have a non problem.
FWIW, if you ever study classical texts on rhetoric, you'll quickly discover that there are a few simple falsehoods that are always present. I think it was the Greeks or the Romans that first codified it.
This one is called "ad hominem", I think. Basically meaning to attack the person, not the argument.
Ironically, I think politics in the USA today show the level two which the intellect of people have fallen. There was a time when people recognized these attacks for what they are. They would even be insulted by them. IIRC, a lot of the European aristocracy feared rule by the people for exactly this reason--that the people could be manipulated by the standard falsehoods but the aristocracy were trained not too. It's no coincidence that you find the same theme in the US as early as Alexander Hamilton (an aristocracy of talent).
For example, I had an argument Saturday with somebody who was basically defending "ad hominem" attacks. Essentially she was saying that she wouldn't support the right idea from the wrong person. However I was dumbfounded when she carried forward stating that she would support the wrong idea from the right person!!! Of course, in this discussion the person's rightness/wrongness was somehow correlated with religion, which may make it even more of a tragedy (sin?).
The above comment is caused by not having enough experience or knowledge. I'm not sure which.
Do you have any idea how much CS actually goes on at MS? They have more Computer Scientists on staff than you probably realize.
The lack of knowledge of real, hard Computer Science (hereafter CS) is a serious liability. I'm not talking about reading Slashdot here. I'm talking about standard algorithms, the ability to write real datastructures, and all of the stuff that makes a real CS degree basically include a math minor.
Knowing your customers and knowing CS are unrelated. Failure can be due to not knowing your customers. Failure can be due to not knowing enough CS. It doesn't take a rocket scientist to determine, from the above, that failure can be due to not knowing your customers AND not knowing enough CS--INDEPENDENTLY OF EACH OTHER.
The sad reality is that most of the developers I deal with on a daily basis know virtually no CS. CS simply gives a developer real tools and real methods for accomplishing their job. A developer without CS knowledge is like a carpenter without a table saw. They can do their job, but it will never be the same--and the difference in quality will always show.
Slashdot Mirror
I'd like to see this, actually.
Which version of the kernel? 2.4? 2.6?
I've had this behavior before, but it turned out to be the D-Link firewall/router in front of the box dropping the connections. Windows XP wouldn't push the router hard enough, but a very similar program on Linux would.
When W32/Nachi came out, this was a common problem under Windows, as it would push these little routers to the point that TCP connections would routinely time out. Anyone who's DSL mysteriously "stopped working" (when it was beyond saturated) because of this virus knows exactly what I'm talking about. I had four customers call me to fix their DSL only to find out that they had viral infestations.
Not to feed the troll, but...
Firefox requires you to explicitly install extensions. Not only must you click install, it makes you read the install dialog (the "Install" button is greyed out for a period after the window pops up). Oh, and it needs to be a trusted site too.
Most spyware gets into IE by exploiting bugs so that you never see the install.
This is largely because of IE's marketing game. At some point, the web browser threatened the traditional application space. Put less kindly, all of those VB monkeys (not to be confused with programmers) would lose their jobs because of those darned web page thingies.
Microsoft threw the web people a curve ball with IE. They attacked open standards with their negligence, gave the VB monkeys a mechanism to spread their pox on the world (ActiveX), and created a new, exciting way to be tied to Microsoft (ActiveX and IE in general).
Before Microsoft, 9 out of 10 people weren't clever enough to "develop software". Of course, they aren't now either, but they can make it look decent enough. It's no wonder that people trying to make real progress have to deal with this kind of overgeneralized tripe.
More to the point, any web browser needs a plugin architecture to be extensible. All of your assertions above blithely assume this is not your problem. Regardless, the only way to allow:
a) Extensibility
and
b) Security
is to allow the user to provide the security layer. People miss this, but there is no way (and likely will never be any way) to have a computer recognize what is malicious. At some point, the user will always have to make the choice to install something that fundamentally changes the computer. At that point, there will always be this problem.
The catch is, IE often will allow installations without so much as a prompt. This is the problem. So, actually, IE *DOES* install bad code. It often does so because IE contains broken code itself.
In this case, Firefox functioned within the security model and the use instructed it to install code which was broken. That was the problem. Worlds of difference here. This is a GreaseMonkey problem, pure and simple.
Put another way, remove GreaseMonkey from the equation and there ceases to be a problem. Well, unless you continue to maintain that a browser (or any software) should magically infer intent, in which case I have some snake oil to sell you--assuming Microsoft has left you with any money.
The BSD TCP stack was as good as it was ONLY because people gave back. It's still very good ONLY because people give back. You only serve to illustrate that Open Source (even the BSD kind) is so effective that it works even when only partially adopted.
.COM Bust, rampant user "sharing", and the advance FOSS how sensible the "software as a product" business model is becoming.
If people benefit from your product, they benefit from it being better. How does that become "The ones who care more about the benifit of everyone rather than just their project will go with BSD."?
Face it, getting code back into the project is better for them.
FYI, I've developed tons of GPL derived code without giving the code back to the project. It's false the GPL requires it. The GPL only requires it if you DISTRIBUTE the software. 90% of software development is in house. This is why most businesses don't usually give a damn about GPL vs BSD--because they don't intend to sell their core business techniques to their competitors.
The GPL only prevents someone from building a lucrative business on what you developed and let them use. BSD, on the other hand, isn't much stronger than the Public Domain. All of this "freedom to use the software" is a red herring. The GPL gives you the same freedom to use the software as the BSD, until you intend to build a business on it. At that point, why is it bad they a company has to either play the sharing game or earn their pay?
Either share or don't, but get off the BSD-licensed fence. As far as I'm concerned, only dead software (of the "maybe someone else will do something with it" type) is appropriate for BSD licensing.
I don't see why people get up in arms about the GPL "discouraging business" but somehow think that BSD "is business friendly". This only matters to business selling software--which is not most of business. I think it's obvious from the
The fact of the matter is this: bytes are cheap. Economics is against you. You can't sell something with no intrinsic value without grossly abusing the economic process (which is never sustainable). Don't put yourself there, it's stupid. Provide something valuable instead.
It's not just capitalism at work detering suicide bombers in the US. Keep in mind that there are many types of coercion possible in Iraq, Saudi Arabia, and Syria that just aren't so easy to pull off in the US.
Looking at the daily suicide bombing in Iraq, you'd think they show a decent pool of people willing to die.
However, many of them have been shown to be handcuffed to the steering wheel of their bomb-laden car. Most of these are then detonated by someone watching.
Most disturbing are the reports that these people are often sympathetic to the new Iraqi regime. Insurgents target these people, chain them to bombs, and give them the option to drive up to the enemy + get blown up + have your family live or not drive up to the enemy + get blown up + have your family die.
It's particularly insidious but note that most foreign fighters must be brought to the US, since it can be tough to make a whole, local family disappear here. Police are just too eager and effective to let something like that slide.
In the same vein as the parent post, all of this is pretty much the direct consequence of religion (defined as religious institutions specifically). Religion has always been about creating a control structure (and all of the coercion that comes with it). People will tell you that you can't have God without religion, but I believe that to be obviously false.
You cannot have a religion without somebody in charge (they fracture into small cults, read about the Branch Davidians). In creating this structure, someone always dictates what "the greater good" wants.
Radical militant Muslim extremests are but one of the examples, but the corrosive effects of systematized social coercion can be much more insidious. I live in the Midwest US (town of less than 350,000 people) and I've known more than one person to leave a church because of serious pressure to vote a certain way.
When God dictates it, there can be no dissent. Without dissent, there can be no democracy. I don't think that anyone's God wants a totalitarian state (theocratic or otherwise). The fallacy lies in a man dictating the word of their "God".
Not that I'm so sure about the whole "God" thing...
This is a troll or a joke, right?
I mean, I know that some sites/products (notably Plone use some JavaScript to implement missing CSS properties in various browsers, but an ActiveX super-fixer-upper?
I've seen ActiveX do a lot of things (many that it shouldn't), but never fix IE rendering to the point it's pixel perfect with Photoshop. Not to mention that the biggest problem on the web is still fonts (although this may be due to legal, not technical, concerns).
Not to mention that ActiveX controls only ever make anything worse. Seriously. If you need ActiveX for a web interface, you should probably be writing a native interface instead.
And if the (RIAA|MPAA|Terrorists|United States|SCO Corporation|Republicans|Democrats|.*) gets their way, there'll be no free oxygen on Earth either!
RMS! Woohoo!
</sarcasm>
I would give almost anything to watch Dick Cheney and John Kerry duke it out with red light sabers, force push, force throw, force lightning, etc. on the floor of Congress.
I'd watch CSPAN for THAT!
I'd love to see some House of Representatives lightsaber battles too!
How much petroleum goes into petroleum production? Seriously.
Don't forget that fossil fuels had massive amounts of energy going into their production. Definitely more than 6x. It's just that we didn't have to do most of that work and that it happened on a million year time scale. I mean, we're talking millions of tons of biomass, millions of years, countless joules of solar energy, tectonic forces, etc. Not to mention the later drilling, excavating, pumping, hauling, etc. As such, the fact that Ethanol can be produced with 6x the energy it outputs is amazing, considering how how that number really is for fossil fuels.
That said, this just goes to show that more work needs to go into it (or something else) to get there. I don't see anybody posting the amount of energy that goes into CREATING fossil fuels (since we don't CREATE them at all).
There's more than just efficiency of production. Ethanol hold energy better, is more portable, and expends it more consistently than, say, electricity in a battery. So if you're talking solar panels, we can't use them for what we use ethanol for (not just energy, plastics too).
I don't think it is too out of reach that we will someday use abundant fusion, tidal, solar, or thermal (the oceans) sources of electrical power to still create hydro-carbon fuels for many smaller vehicles. I also would be willing to bet we may even get less efficient by using that same power to produce antimatter as a fuel--for the same reasons we need a hydro-carbon fuel (portability, energy density, ease of reaction, exciting explosions when our machines fail, etc).
I find it ironic you can read "requires open formats" as "bad for open source". I find the statement of "outright removing a competitor" humorous. The customer demands a product, any fool knows that Microsoft can produce almost any product they want to. The issue is that they don't want to provide that product, thereby removing THEMSELVES from the market.
Were the governments of the world to declare that they wanted only proprietary software, you'd probably find Open Source not in that market. Of course, the only way to argue for this is pretty much an economic issue. At this point, I think it's pretty obvious that Open Source provides a better value than Proprietary for a large swath of needs. Also increasingly obvious is that Open Source is scales better than proprietary. Increases in proprietary players are predatory, increases in open players are additive (assuming they are truly open).
There was a time when the existence of Mail Merge in an application was the same way. Sure, you could create your own mail merge, but in the end, the word processor having this functionality was the right decision for the customers. The same goes for open formats. Lack of "vendor lock-in" is the killer feature that drives Open Source more than anything.
The fact of the matter is that the free-program interoperability that you enjoy today exists because of millions of man-hours of work to achieve it. Say what you want about a "political cudgel", the cold hard fact is that a few million spent by Microsoft creates a few hundred million dollars of work just to interoperate with it (maybe more including the outside QA that goes into the products by the time they mature).
The investment by Norway in continuing to create this problem is not justified. As for PDF, if you ever have to implement a PDF handler, you'll be yearning for the coming days of SVG-P. It's like the bastard stepchild of PostScript and an early filesystem. Seriously, it almost creates a filesystem in a file (all kinds of block lists of irregularly sized blocks) and enough complexity to make it very difficult to parse (and impossible to recover when suitably corrupted). Adobe realizes this, that's why they're already knee deep in SVG (see the Adobe SVG Viewer if you don't believe me).
No, the sad thing is that so many people are so in love with the sacred cow of Capitalism (of which Proprietary software is apparently the posterchild), that they don't realize that Capitalism is still beholden to the Free Market. Every time I see someone crying about abuse of the proprietary model, I can't help but realize a simple fact.
It's all about demand. People demand what they want. They want software as cheaply as possible. It is no longer necessary that it cost very much at all. In fact, given the support/installation/development model that a lot of FOSS uses today, it can be funded entirely from people providing those very real services (instead of billing for bytes).
Face it, selling software as a business is pretty much doomed unless it is really that complex to write or your market is pitifully small. There will come a time when people really can't be fooled anymore. Software is not valuable. Time is valuable. When people are paid for their time and the software is open, everyone saves time--and every one thus saves money. Efficiency is what makes or breaks businesses (see Walmart, evil but dead efficient).
In otherwords, I'll gleefully revel in all of this as I put Microsoft to the fire, both because they would do the same for me and because this really is a victory for the freedom of developers and users alike.
Most of Google's magic is really data mining the semantic data from the Internet.
Gmail is nothing more than an attempt at getting a massive corpus of data on which to let their algorithms loose.
I really think that, while there is potential for abuse, this is really the only way to tackle their problem space. After all, Google doesn't really rank web sites, people do. It's just that Google has some really clever ways for determining that people liked a web site.
Sometimes it relates to webs of links, sometimes it relates to combinations of words, but Google's software doesn't deal in semantics--only algorithmically generating statistics from the data generated by people.
I don't worry so much about Google, I worry about our future AI overlords. Although, if a truly scalable Artificial Intelligence ever gets Internet access, I fear it has the potential to know us better than we do.
Ummmm, hello people. This is an XML format.
If MS needs extensions, that's what namespaces are for.
As long as MS extensions don't change formatting functionality (this is really not rocket science, Word is not an innovator here), they can tack whatever metadata they need into the file format and still have it be portable.
If you don't believe me, look at what Inkscape has done with SVG. Psodipodi built on it, adding a namespace to provide their needed data. Inkscape did the same on top of that. It produces one file containing three XML namespaces containing with interoperable metadata for two editors--and it's viewable in stock SVG viewers.
Obviously it's up to Microsoft to "do the right thing" with their metadata, but this certainly levels the playing field so that others can do what they need to with the documents.
Now the patent on XML-to-object-mappings, that's another story...
Ummm, geeks like Google because Google employs scientists. Which mere scientists were you talking about?
Were you talking about the PhDs at universities busy teaching classes, churning out research papers to avoid being fired (an ugly numbers game some departments play), or perhaps burning time generating volumes of grant paperwork?
Oh, maybe you were talking about the scientists employed by the private sector. I'm sure the management teams wherever they work are willing to take the time and care that Google won't.
You do know how may PhDs Google employs, right? Not to mention that they won't be fighting for resources there either. No backstabbing, liquidating MBAs trashing their corporate budget. No football-crazed alumni assassinating their funding proposals either.
Also, I would remind you that "mere scientists" often come up with the needed research (there are volumes in MT alone), but rarely can afford to put in the years that it takes into a good implementation.
Geeks love Google because it is, in many respects, where the best of business meets the best of academia.
Uhhh, that was a test?
:)
Oops, my bad Perl.
Being a Python programmer, I misremembered it as "one of those other P-languages that gives me maintenance headaches".
That being said, again, RT can be quite functional. I've found it runs better under PostgreSQL, but never tested it with Mod_Perl (didn't know it did that).
You would be wrong to characterize Freenet in that way.
As for being wrong to ban or be against that system, I don't think I can address that question in the abstract. I know nothing about that system I'm supposed to endorse banning. Not that a system exists that there is "only one legitimate user" and "millions of inappropriate users". Who would endorse banning something they know nothing about (except for Slashdot Trolls, perhaps).
You could have used that rationale, for instance, to justify the wholesale destruction of the Underground Railroad prior to the Civil War. Slave smuggling was wrong, immoral, and illegal for a number of Southerners.
Besides, if you knew much about Freenet you'd know that it would be a lot faster if there were that many users (legitimate or not).
Ill conceived my arse.
r =326
Take America for instance. How many children are killed by guns each year? How many children are abused in CP each year?
The guns are just as guilty of killing children as Freenet is of creating CP.
Bottom line:
http://www.ichv.org/Statistics.htm
Over 30000 gun deaths in a year.
http://www.nch.org.uk/information/index.php?i=77&
2234 Child Pornography busts in the same year.
Children abuse in real life, recordings trafficked on the Internet. Apparently that means we should suppress the technology of the internet.
Thirteen times as many people are shot the same year (even taking suicides and under-reporting of CP into account, the numbers are still overwhelming) but we settle for public service announcements. Ask any police officer in a big city, it burn you up inside finding abused children every day, but never once would they prefer to find one dead. Give a parent the same option, same answer.
Blaming the technology instead of the criminal only makes doing even reasonable things illegal and puts innocent people in jail, because, let's face it, Child Abusers will find a way to do it unless we remove every last shred of freedom, privacy. and due process from our society.
No, the issue here isn't a analogies. The issue here is the fact that some people value their speech (or their guns) and some don't. I value children's lives more than most Slashdotters, but I also understand that Child Pornography is the crime, NOT transfering files over a P2P network or communicating anonymously.
In fact, I'd say that anonymous and P2P communications might actually be two of the "rights" that the children of the future may enjoy most. If the casualties of the war on child pornography (or the war on terror for that matter) are going to be the freedoms that allow democracy to exist, there will be far more victims than we can appreciate right now.
A lot of people have seemed to think this question was about going totally Linux (and many claiming that the MS deal was a good "value").
0 92929216
In case the question was about using FOSS on a Windows network (for the time being), the following might help.
This tool is fairly useful for deploying Firefox on a network:
http://firefox.dbltree.com/
As for OpenOffice, I use central network location, see the setup guide (I think you have to run setup.exe with the -net option). I'm not sure what must be done from there to automate installation, we usually do it manually because Workstation installs of OOo (from a central network location) take seconds.
As for the question of whether the MS deal was a "good value". First, let me say that there's more to "value" than cost. Also realize that $50000 per year might be cheaper than MS's $15000. Once you figure in MSCE training for an IT team and the increased labor it takes to run a Windows network you might be surprised. Believe me, once configured, Linux machines can be dead reliable and reimaged lightning fast, I do it for a living. That said, Firefox has saved me 8 hours per week at one client that only has 10 computers.
Well, ask your purchasing department how many suppliers it has for, say, light bulbs. While more than a few places say "just one", I find universities in particular tend to have four or five suppliers solely for the purpose of leveraging one against the other for good pricing.
What's the point of my story? The point is that MS as a single supplier means you will pay as much as they want you to. Of course it will always be "a little cheaper". In a software world with real competition, that will change.
Regardless, it's worth pointing out that increasingly it is the case that people are choosing FOSS for reasons other than price:
http://www.groklaw.net/article.php?story=20050426
Since we're exercising our rights, I agree with your point.
After careful consideration of your point, I realize the error of my ways.
Specifically, I realize that by running a Freenet node, I would be allowing others to use it the way you've said. I would be providing support to a system that, as a whole, exploits children.
I have since determined that telephones and video cameras are used in the production of CP. I can no longer supply the phone company or video empires with the money I pay for their products, since their products are used for bad things, as well as good.
I think we should do the same to the gun manufacturers, since we all know how guns are used.
Seriously though, the technology (Freenet) does not exploit children. People exploit children. Don't ban technology because it is effective. Freenet has NEVER exploited a single child any more or less than a telephone, the US Postal Service, and the entire line of Sony Handicam(TM)'s.
You do realize that by the time the video hits whatever distribution system that it's already too late, right? If you want to protect the children...why not actually protect the children. Don't use abused children further as an excuse because you're uncomfortable with other people expressing their rights outside of some sort of central control--because that's the object of Freenet--for better (political reform) OR worse (abuse of children for profit).
Finally, realize that the First Amendment is not what it used to be. There was a time that speech and thought could not be ultimately suppressed. Given the current state of technology, I'm not sure that will be sustained. Disturbing as it may be, someday the First Amendment may not be a nod to an unsupressable reality, but it will be that last bastion between an unscrupulous bureacratic machine and individual freedoms.
Well, it can be ugly at times and there are certainly some rough edges, but Request Tracker will probably do the trick.
Mind you, you'll probably need a Linux or BSD server running Apache, PHP, and an SQL engine (MySQL or PostgreSQL, we use Postgres).
Look at http://www.bestpractical.com/rt/.
This does not complicate things. It simplifies them. They are not talking about writing a database application, they are talking about writing DATABASES! So, since your question is about how to write a database-based application when the poster is offering a solution for a database implementation, this is kind of an apples-oranges situation.
:)
To expound on Reiser4, it addresses the database problem-space by allowing individual files to contain extended attributes that may be queried on using a boolean algebra (think a table row with an optional payload).
As the query language used is very Reiser specific, and very general purpose (and not people oriented), I won't bother to construct a query, but it would not be an extraordinary feat of computer science to write an SQL to Reiser4 query converter.
More to the point, the tabular nature of most enterprise databases is the rub here. In Reiser4, you can have path structures, you can have indexed queries across these paths, and you can have it at database speeds. Try to do a hierarchical data structure in SQL. It can be done, but performance for some operations can be...suboptimal.
Furthermore, Reiser4 not only has your standard ACID guarantees available, it also has what Hans calls a transcrash. Transcrashes are pretty much ACI but not so much D. This is very useful in a backend scenario where a cluster of machines provides the Durability (which is where I think databases are going).
Also, hard links beat any foreign key scheme I've ever fought with.
If you read what the poster means, new databases may be based on these filesystems. An SQL translator and distributed transaction processor based on Reiser4 would probably compete well in todays DB performance world. If not, I suspect Hans would make sure it did eventually.
I've always fancied writing an XML parser that backended in Reiser4 (containment = directory, attributes = xattrs) just to play with the indexing. When Reiser4 hits a stable tree, I might...
Actually, I'm constantly amazed by the number of people who don't know that there IS Microsoft Office.
A lot of the "civilians" that I help out have bought computers bundled with Office and have Office provided at work, so they don't understand that it's not PART OF WINDOWS (if they understand that Windows isn't a fundamental part of the computer, which lots don't).
Recently Office has been yanked from the more competitive PC options from Dell and the like, so more people I know buy a computer and freak out when they find out that they have to buy Office (and then freak out again when they see the price).
Worse yet, there's still a middle ground option with Works Suite (which bundles Word), so lots of people are still getting this broken impression that Office is Windows is computers!
At any rate, anyone considering OOo absolutely MUST download the new Beta. While the new Database component of it is certainly Beta quality, the Word Processing/Spreadsheet/Presentation stuff is still rock solid. That said, there have been light years of improvement since the last version.
The OOo beta handles layout better than 97 or 2003 for many documents. This is no small feat, since, at least in Office 97, there is quite a lot of variation even within Word as to printing layout. It may not be common knowledge, but Microsoft uses almost the same Win32 API calls to print as it does to display on the screen. This has the interesting side effect that things that mingle the two (Word) can be affected by switching PRINT DRIVERS!!! I've even had bad print drivers cause crashes when certain documents are displayed (had to change to a different printer on another machine to open it there, changing back to the same printer caused the crash, new driver version replaced that bug with other ones).
Crashing aside, switching printers often causes layout changes. Professional printing shops HATE this "feature" of word. The mantra is usually "come back with a PDF or sign off on a crappy proof". I had hoped we left this behind with WordPerfect 5.1 for DOS, but not so!
The number of possible ways to effectively accomplish something in software is limited.
A patent grants EXCLUSIVE rights to one of those ways.
Therefore, you have just created a land war. Only the rich and monied win a land war. Soon, you'll have nowhere to live. Good luck with that.
Put simply, there is no "stealing of ideas". That's ludicrous. I take your car? You have no car. I take your method of bush trimming? Both of our bushes get trimmed. That's life.
That's also the worst case. On the other hand, what if I develop a similar way of trimming bushes? Now who's stealing. I just wanted to trim my bushes, now you can take money from me! Who's the thief?
At some point the businesses of the world are being given the power to own the EXCLUSIVE right to sell something. I think we all know why that's bad. No competition == screwing people.
It's bad when the government mandates it (national telecomm companies). It's bad when monopoly enforces it (Standard Oil, Microsoft). It's bad when the people suffer it.
If I develop MY IDEA independently of you, I'll be damned if your patent should matter to me. Unfortunately, this is now my problem.
Worse, now the only people with significant patent portfolios won't be people. Instead they'll be the most morally reprehesible construct mankind has ever unleashed--corporations. Worst of all, they're pretty much designed to aggregate financial and legislative power.
Someday, this may cause a revolution...I hope.
Wow, someone was bitten by a "computer professional" as a child, huh?
Seriously, every segment has its opportunists, but corporate MIS "was and is" hardly the whole of "computer professionals". Not everyone had the problem of which you speak.
Also, understand that quite often what you've experienced was due to management. At the time. decent MIS people were both expensive and rare. As such, it was often difficult to get information from them because management dictated it to be so. After all, if everyone had them fielding requests, they'd be so backlogged that they couldn't run the stupid TPS reports for the VP. Seriously. I've been as much as told that I should only accept any requests from VPs or higher unless I wanted a bad performance review. Thank god I'm out of there now.
As for Visicalc, spreadsheets aren't always a win. I've had to "upsize" a lot of spreadsheets that should have never been made (mostly after Excel though). That said, they're useful (the parent didn't argue otherwise). I used sc long ago to do the exact same thing.
What's this microcomputer complaint? UNIX==microcomputer these days. Unfortunately the same opportunists as above were too cheap to port it at the time, so maybe it was an obstacle. Sounds like noise to me though.
Finally, read your last paragraph. You don't dislike UNIX. You dislike the interface. The COMMAND LINE interface at that. Condemning the OS for the interface is just dumb. Remember Mac System before UNIX? Remember DOS? Put another way, the car may have a pretty paint job and interior, but if it doesn't run very well and the parts always fail, do you want to own it? Don't be naive.
Not that it matters. We have a pretty GUI for you. I've been using X-Windows since UnixWare. It was usable then, today it's fantastic (KDE is there, GNOME is there). I have it deployed in businesses. People use it for like three weeks then quietly ask "Where's Minesweeper?". I actually have minesweeper amd solitaire clones that I install for the management.
If you want to compare UNIX, try comparing it to the VMS hack that became NT/XP. Or to the BSD in MacOS. Apples to Apples. Oranges to Oranges. You know the drill.
Well, I'd say that might be dramatizing a bit.
This is not a root exploit. This is not a buffer overflow. The worst we've got here is a settings loading problem.
The fact of the matter is that Firefox is exhibiting the correct behavior but the specification is flawed.
If you look at the sad mess that is today's encoding of SSL certificates, you can see the same thing. There are a lot of ways to encode a certificate so that it's valid but doesn't work most places. Of course, most people never have to deal with this dark art, so its not as obvious.
The fact of the matter is that this "vulnerability" isn't one. The vulnerability lies in the consequences of the IDN standard, the reliance of SSL on DNS and ASN.1, and the world-wide security standard not being based on truly verified public-key cryptography.
And as for the "war of ideology", the war is won in the trenches. See how long it takes to get a fix. If you don't see one fast enough, THAT is how the war will be lost. Not by raving lunatics on Slashdot!
Also, don't forget that this "vulnerability" comes with some benefit. Without IDN, having domain names in Chinese, Korean, Japanese, Arabic, Hebrew, et al. would simply not be possible. I think that, had this problem not been solved, we would be suffering seriously more balkanization in the internet--possibly resulting in a new asian/international browser. I think it's obvious to see why that could be bad (at least of Verisign).
If anything, what you are witnessing is an incomplete application of Unicode. IIRC, unicode specifies equivalency tables for sorting and comparing these things. Now support for unicode in SSL certificates in a whole other game. Mind you, I think if we settled on punycode encoding them in the SSL certificate, we would have a non problem.
Comments?
FWIW, if you ever study classical texts on rhetoric, you'll quickly discover that there are a few simple falsehoods that are always present. I think it was the Greeks or the Romans that first codified it.
This one is called "ad hominem", I think. Basically meaning to attack the person, not the argument.
Ironically, I think politics in the USA today show the level two which the intellect of people have fallen. There was a time when people recognized these attacks for what they are. They would even be insulted by them. IIRC, a lot of the European aristocracy feared rule by the people for exactly this reason--that the people could be manipulated by the standard falsehoods but the aristocracy were trained not too. It's no coincidence that you find the same theme in the US as early as Alexander Hamilton (an aristocracy of talent).
For example, I had an argument Saturday with somebody who was basically defending "ad hominem" attacks. Essentially she was saying that she wouldn't support the right idea from the wrong person. However I was dumbfounded when she carried forward stating that she would support the wrong idea from the right person!!! Of course, in this discussion the person's rightness/wrongness was somehow correlated with religion, which may make it even more of a tragedy (sin?).
But I digress, since I'm way off topic.
The above comment is caused by not having enough experience or knowledge. I'm not sure which.
Do you have any idea how much CS actually goes on at MS? They have more Computer Scientists on staff than you probably realize.
The lack of knowledge of real, hard Computer Science (hereafter CS) is a serious liability. I'm not talking about reading Slashdot here. I'm talking about standard algorithms, the ability to write real datastructures, and all of the stuff that makes a real CS degree basically include a math minor.
Knowing your customers and knowing CS are unrelated. Failure can be due to not knowing your customers. Failure can be due to not knowing enough CS. It doesn't take a rocket scientist to determine, from the above, that failure can be due to not knowing your customers AND not knowing enough CS--INDEPENDENTLY OF EACH OTHER.
The sad reality is that most of the developers I deal with on a daily basis know virtually no CS. CS simply gives a developer real tools and real methods for accomplishing their job. A developer without CS knowledge is like a carpenter without a table saw. They can do their job, but it will never be the same--and the difference in quality will always show.