I would be a big fan of Adium, except that it doesn't do Address Book integration. As far as I'm concerned, this is the one 'killer feature' that makes iChat indispensable (premise: I use mostly AIM). I would love for some other client to offer anything like what iChat does for AIM for multiple protocols.
I'm not sure who to blame for the lack of this -- I wonder whether Apple is using some secret internal APIs or something, or if nobody else has seen fit to do it.
Actually what would really be neat is if somebody could do something like the OS X address book in Linux; it's probably the biggest thing that keeps me tied to my Mac. One Address Book for iChat, Mail, my label printer, synchronized across all my computers and to my cell phone. All automagically (for $99 a year).
I can see your point regarding cars purchased with a loan -- if you own the vehicle, then you should be legally entitled to have such a device disabled or removed if you choose, since it's your car.
However where I think there definitely could be a place for these boxes, like them or not, is on leased vehicles. With them it's more like renting an apartment than buying a house: you don't ever hold the title IIRC, the leasing company does and lets you drive it in return for payments; at the end of the lease you relinquish the vehicle.
In that scenario I think the argument could be made in their favor, since you are essentially borrowing someone else's property they can put conditions on it (like 'don't remove the black box').
I could imagine that a company which did this could provide cars even to people who currently are unable to buy or lease ones because their credit is so bad (e.g. the unemployed).
Huh, that's a slick little number for $600. I wonder how much the toner refills cost, and whether the purchase price includes real (full capacity) carts or whether they do what Samsung does with their low-end lasers and include a "starter cartridge" that only is about 20% full. I hope not the latter.
I've actually always had a soft spot for Minolta, dating back to when I used their SLR film cameras. And when I used to work in sales, their reps were some of the most knowledgeable and friendly around. I still have piles of literature sitting around about their first generation digital cameras -- maybe someday they'll be collectible or something.:)
Wow, good find. I had no idea that you could get 1U Xeon servers that cheap. The last time I looked at rackmount machines, the chassis alone cost three times what they're charging.
This will pretty much be the only good thing that I've gotten so far out of the dot-com bubble...
I don't know; I said netatalk/afp was difficult because I've never actually gotten it to work myself on my Debian Sarge box and small home network of Macs (mostly 10.4 installs), for no small amount of trying. The configuration method isn't necessarily that complicated, in that you're right, but when it doesn't work, there wasn't much that I could work with.
In my case the server would appear on a Mac, but logging in (entering a username and pass, clicking OK) to see the available shares/volumes would just cause the Mac to hang indefinitely until you killed the Finder (spinning-beachball effect). I got the whole thing to work exactly once, and that was by disabling all the security and encryption packages and overriding the Mac's security settings to allow for plaintext passwords, which I consider unacceptable. It's been suggested to me that the problem I'm having might be because of something broken in the Debian/stable package's crypto libraries, but other than that I didn't get much in the way of suggestions from the help lists.
Given that my network setup is about as dead simple as you can get, and the only other thing running on the Linux box was Samba, it was a rather frustrating experience. (Actually it's possibly the most frustrating networking experience I've ever had to date.) I'm sure it's something I could puzzle out given enough time, but eventually it got to the point where I just canned the project and kept my iTunes stored locally, and use CLI tools for Documents backups.
This was my experience with it, I feel like I gave it a more than fair shot with otherwise well-supported hardware on a decidedly simple network, and couldn't get it working in less time than it took to find a workaround (rsync). I sincerely hope that not everyone has this much trouble, but this was why I think that a NAS box that "just worked" with Macs would really be in demand.
> I got a TB of storage... 650MB after RAID 5 formatting
With that kind of overhead, I hope it's really fault tolerant.:)
Anyway it does seem like a really cool device, though. I have been DIYing something similar, using an old OfficeDepot Compaq $500 special running a minimal install of Debian with Samba and AFP, but AFP is really a bitch to configure sometimes. It would be really nice to have something that just works out of the box. And the streaming media features are similarly something that, although a person could probably do it themselves with a suitable investment of time, are great to not have to fiddle around too much with. Assuming everything works as advertised, of course.
By "off lease" do you mean 'from the back of a van downtown'? And do they scratch the serial numbers off and wipe the drives, or is that part of the self installation?
Seriously, where does one go to get a deal like that? I'd take one for each foot.
I guess there aren't enough dying dot-coms where I am or something.
I'm going to guess that a "toy printer" is anything that:
1) uses ink costing more by weight than gold 2) has space for less than a ream of blank paper 3) has a memory-card slot 4) costs less to purchase initially than two of its consumable refills do
So yes, basically, I think he's referring to the crummy home "hey it came free with my Dell" inkjets which are the bane of anyone who doesn't use Windows' existence.
Generally speaking, Linux has great support for high-end peripheral hardware because that's the stuff that uses open standards most predictably. For whatever reason, it's never the $20k devices that require some Windows-only proprietary driver; it's the $99 ones from Best Buy that do.
I'm going to take a stab at answering this question, or at least my theory as to the answer.
I think it's because -- despite the stereotype of Linux developers being high school or college kids -- the people who actually work on the Linux codebase aren't gamers. Or, more to the point, most gamers aren't interested in writing code and working to make stuff run on Linux. It's a different mindset: one group sees a computer as something to play with, not just a means but an end in itself; the other group sees it as something to play on, a means only.
Maybe I'm biased and maybe I'm using a bad sample, but the majority of computer geeks and actual software developers I know really don't spend that much time playing PC games (or they use consoles), and the people who are really wed to their game machines aren't really interested in playing around with a new OS for the heck of it. It's just two groups of people which, while I'm sure there's some overlap, don't have as much in common as you might think on the surface. Most Linux users are honestly interested in their computer's OS; to a gamer it's just a launchpad that lets them play whatever they want, and the open/closed source controversy is basically irrelevant. The only advantage Linux might have is that it's free as in beer (but a lot of gamers that I know have pirated versions of Windows anyway, so it's not much of a selling point).
I'm sure this will get a lot of people responding with "but I'm a gamer and I use Linux," but before you do, realize that you're a statistical anomaly in the big picture just by virtue of reading Slashdot and caring about the issue either way.
Quite possibly it's not Emperor's fault, I have a feeling that IBM/Lenovo may not sell ANY ThinkPads without Windows (especially sad considering that IBM ought to be the one place you could get a Linux machine, if anywhere) and thus you're really paying for a copy of Windows plus Emperor's overhead and whatever it costs them for the support contract (which might be worth something to some people, but not $600!). But at the end of the day, that's a hell of a premium.
The way Microsoft has the hardware market twisted around their finger right now, it's basically impossible to get a quality, name-brand laptop without buying Windows. (I know there are some white-box machines available out there, but put one of them down next to a ThinkPad and there's really no comparison.)
The only thing I disagree with is this: "I know it's unreasonable to treat every Microsoft proposal as suspect."
At this point I think it's entirely reasonable, even prudent, to treat every Microsoft proposal not just as suspect, but as outright hostile, until definitively proven otherwise. I can't remember the last time I saw something come out of that company that didn't have an angle or a hook buried in it somewhere, or give me the vague feeling that they were out to screw me as a consumer down the road.
Microsoft is a shark, and we are all minnows. Just because we play along today doesn't mean they won't try and have us for breakfast tomorrow. No matter what they do, or what they say, it doesn't change what they are.
(And for the record I don't even think they're 'playing along' today -- this open formats business is a ridiculous farce, as even a blind man should be able to see; what remains now is whether the politicians will realize it.)
I think the second part of your comment got you modded down, but I wanted to say that I thought you hit the nail on the head with the first part:
It's also important that MS is only judge of what is and is not conformant with the spec because they get to decide what is "complete". This means that at any time they can decide that an open source project is not "complete" and then sue them if they continue to insist that it is.
The additional concern is that MS is so big and known to be legally aggressive, the fact that they have the judgmental authority over whether an implemetation is conformant will be enough to keep a lot of smaller companies from touching the spec. Implement it the wrong way, or in a way that pisses MS off, and you'll find yourself getting sued into the ground. Sure they may not win, but they'll have just enough of a case to keep it in court for a while and bleed you to death.
The result here is that MS can say they have an open format, but the licensing will effectively keep most of the other implementations rather trivial; my guess is that other companies will write "converters" or plugins, rather than actually use the format natively, and then MS can use this lack of native support as a case for continued use of their own products.
> 2. WMF is well understood and many Mac apps support it.
Would you like to support that statement with some evidence, perhaps?
I have yet to see any Mac application that provides good WMF support. As other people have pointed out, even MS' own implementation (used in MS Word Mac and Powerpoint Mac) sucks terribly, and has a tendency to display garbage/white boxes/etc. in place of graphics.
If there are other software packages out there that have better support for it, I'd love to know what they are. I'm not aware of any though.
This is why I'm glad to live in a place where the gun ownership and concealed-carry rate is very high. Assaulting a random stranger would be basically akin to playing russian roulette.
You don't see this kind of crap happening except in places where you can guarantee that your victim is going to be unarmed.
The kicker: despite the fact that the author of the site is ostensibly a Mac user (by virtue of the host), all of the dialogs appear to be Windows. Guess Apple isn't on par yet either.
Why would you want Javascripts to do any of those things? I can do them just as easily with the built-in functions of my browser. Back = back button, refresh = refresh button, new window = right click - new window/tab. And using a script to open a new window is a little excessive, as I recall it's been possible to make a link open in a new window since roughly the time the Earth's crust was cooling by doing TARGET="_blank".
I hate those little scripts most of all, since they're nothing but unnecessary bloat. The most common use I've seen of Javascripts is creating popup windows that lack the proper controls across the top, which I find infuriating and is probably the number one reason I browse with it disabled.
I understand there are probably situations where doing something like that with scripts is necessary (GMail, for instance), but I'm sorry to say that they're horribly overused and rarely do anything for the viewer.
I agree that story (bug in Sun's Java) should have merited a story, but it's not quite the same level of seriousness as the IE issue. First, I don't think there are nearly as many users of JRE as there are of IE (although I suppose that's arguable since it's preinstalled on a lot of systems), but more importantly, I don't think that the Java hole was published and unpatched for any significant length of time. The article you linked to, after all, is talking about the fix being released and the original hole in the same sentence.
The IE one is important because it's a pretty gaping flaw in a very popular (or at least often-used, perhaps less liked) piece of software which has remained open and basically exploitable for a very long time since it's been published and everyone and their cousin made aware of it. I even heard a quick mention of it on one of the cable news networks, so there's really no way to argue that the cat's not out of the bag and into the hands of the crackers, and once they're done, script kiddies.
Yeah I was looking at the same thing also. They have some OSS stuff there for download (and helpful instructions in case you want to upload your changes -- I mean who wouldn't want to contribute at no charge to their friendly local defense contractor?) but no source.
This by itself doesn't mean anything though. Remember that the GPL doesn't require you to make your code available online to just anyone, it just says that you have to distribute it along with the software. So it could be something as simple as a/src directory on systems they sell, or on the CDs or however they distribute it.
They can't prohibit you from buying the product and redistributing the source code, but it doesn't mean they have to go out of their way to make it particularly easy for you to do so. The only real way to tell if they're breaking the license would be for somebody to buy RedHawk Linux in whatever form is closest to a "retail box" and see if it comes with source, or a written offer to provide it. I believe those are the requirements of the GPL.
Exactly -- I was going to say the same thing, but you beat me to it.
The flip side to having Free software is that you have to be prepared to let some people that you might not exactly like or agree with use your work. On the other hand, people who might not like you very much have to do the same.
The community would really be shooting itself in the foot if they put some sort of clause like that into OSS licenses, because basically it would be throwing away all the millions of dollars of R&D money that government contractors can and do put into software development, and which might trickle down into OSS projects. Plus, there are a lot of companies out there that you wouldn't think would be affected, that might shy away from OSS after a restriction like that was added, just because they don't want to limit their future business. IBM, for instance, probably wouldn't want to write itself out of the defense sector because some hippies thought it would be cool to ban military uses of Linux, and they're one of the biggest contributors. The NSA, if I remember correctly, has a secure distribution of Linux that's provided some useful things back to the community, and you'd be throwing that away as well. With the exception of RedHat and a few other Linux-only shops, a lot of money comes from companies that sill have proprietary OS options sitting around. If they start to feel that Linux is a hindrance instead of an advantage, there's nothing stopping them from going back and pulling the plug on their OSS ventures.
Overall, it's a foolish and dangerous idea to consider such limitations in licensing, and I think it's important that we put that idea down quickly before it gains any mindshare. It would quite literally be the end of the corporate deep-pocket support of Linux that we've enjoyed until today.
You do realize I hope that this system, THAAD, is a theater missile-defense system, not a global "Star Wars" hemisphere-defense one.
It's designed to protect strategic assets from medium range, single warhead ballistic missiles, which are exactly the thing that China, N. Korea, and other ex-Soviet client states have in spades (and are significantly easier to put together than an ICBM). The intended use is to place them as a spot defense over a high-value target, as the farthest reaches of a layered system that includes short-range defenses like Patriot, etc.
This is not an ICBM shield for the U.S. in the manner that I think you are thinking it is, that would really have any effect in the event of a global thermonuclear war. The preventative measure against that is still MAD. However when you step down from that scenario (and terrorism), the next most likely case of a nuclear weapon being used against us is with a theater ballistic missile against a strategic target like a foreign city, aircraft carrier, or air base. In a situation like that, a defensive system like this makes a lot more sense.
Actually, this isn't the DoD, it's Lockheed Martin, a private company which does business as a government contractor. Although I claim no knowledge of their finances, I'm willing to bet however that their budget is definitely not "unlimited," although it might be large compared to what you and I are used to. Especially if they are on a fixed-price contract, where their total revenue has been specified in advance, the employees are probably under a lot of pressure to keep costs down since every dollar they spend eats into their profit.
I think in this case though there wasn't much competition, at least of the Linux/Windows variety, since Windows doesn't really do real-time applications very well apparently. I'm not actually sure what the other options would have been besides Linux for a project like this: I can only assume that there are some highly proprietary, niche-market OSes for this sort of thing, but that in general the COTS field is pretty limited.
I guess I made a bad assumption and had you mistakenly pegged as one of the "gah, installers suck, everyone knows that the only way to do Linux is to download everything as an undocumented source tarball" crowd.:-) My error. (Obviously if you were, you'd have responded with some ad hominem attacks, and insinuations about my my sexuality.) I've been here too long, I guess; I had forgotten not everyone is engaged in some vicious Debian vs. Gentoo vs. Slackware OSS philosophy war, and it's gone to my head.;)
I have a particular soft spot for single-purpose "appliance installations" perhaps because I have an inordinate amount of non-functioning white boxes sitting around, but also because I just like the concept of computer-as-interchangable-part on which you can pop in a CD, reboot, and have a functioning device for some particular purpose. It's the antithesis of most people's conceptions of computers as finicky, temperamental, and highly individualized. Of course, the downside to this is that if you want something that's highly customized and application-specific, you're probably not going to be happy. I guess I'm just lucky that in many cases I'm pretty close to that mythical "average user."
At any rate, you did make a good point regarding SmoothWall; it's clearly set up for "ease of use," but only if you match their intended setup: someone with a reasonably modern spare machine and complete access to it (at least for the initial steps). I suppose this is just the necessary tradeoff for what they're trying to do.
Interesting about the article. I was not aware of that method of production.
Regarding your truck, where do you get your bio-diesel? Do you roll your own, or have someone make it for you? I have seen it made in a small batch before, but the amount of lye required seemed rather large and the "washing" process labor-intensive. I've heard rumors that there is somebody out there who has homebuilt a continuous-production system, rather than a batch-based one; I don't have any information on it though. But it's always something that's intrigued me though, and hopefully as the technology matures it'll become less resource and labor-intensive.
LTSP does look pretty close to what I was talking about, it is a pity though that it won't work for your architecture. I'll have to read into their docs/wiki a little and see if it does what I'm hoping for. The Achilles heel of many projects that I've checked out is that they're very insecure: not only is the x-windows client/server communication unencrypted, but it requires a large number of ports left open on the client machines in some cases. Although I understand that a lot of people are going to use these on internal/dedicated subnets and behind firewalls, it just seems like a dangerous practice in general to build a system without regard to security these days. (Suppose you want to use one client wirelessly, and your wifi hub is located outside your firewall?) Hopefully LTSP isn't like that.
The OpenMosix system is something I was not aware of at all -- at least not in the way they're using it. (In short they're taking a thin-client/server combination and using it as a cluster, so that it "load balances" computationally across the various machines. Pretty slick, if it works.)
At risk of sounding like a shill, I have to disagree with you about SmoothWall. Although I haven't actually tried to install their software yet (it's on my list of things to do), unless their manual is a complete pack of lies (which I don't think it is), I'm not sure how they could have made it simpler. I think perhaps what you were trying to do is something that's just outside the intended use. They're pretty upfront about the fact that the CD-Rom installation option is the easiest way to go, and that if you can't do this the network or floppy installs are more complicated.
Personally I much prefer their method of distribution and philosophy to some other OSS projects' who are inclined to give the user a tarball and vague instructions to "figure it out." While that's fine for a hobby project, it's a crummy way to make a product -- even one you're giving away for free.
SmoothWall isn't a piece of software in the same way a regular application is; it's really an operating system / distro that turns an old box into a dedicated appliance. So the distribution methods that might be convienient for software aren't necessarily as suited to it. One of the reasons I bookmarked SmoothWall's page though is specifically because of the way they give you a straight install ISO, no screwing around with compiling or configuring anything. If it was a tarball, I wouldn't have given it a second glance, such has been my experience installing software from source. (I'm actually pretty sure that it is available as a tarball also, here.)
I understand this probably boils down to some very basic disagreement over the philosophy of software and OSS in particular, but I think StoneWall is an example of how a project ought to operate: producing a relatively easy-to-use, robust, single-purpose product, which is both 'Free' and 'for free,' to the public.
It's something of a pity that they charge money for the version that lets you install over a RS232 serial connection, so you can't do a true headless install with the free one, but I guess they have to support themselves somehow. (Also it doesn't do load balancing even in the corporate version, which I found somewhat surprising, although it will do a fallback from a primary to a secondary connection.)
Slashdot Mirror
I would be a big fan of Adium, except that it doesn't do Address Book integration. As far as I'm concerned, this is the one 'killer feature' that makes iChat indispensable (premise: I use mostly AIM). I would love for some other client to offer anything like what iChat does for AIM for multiple protocols.
I'm not sure who to blame for the lack of this -- I wonder whether Apple is using some secret internal APIs or something, or if nobody else has seen fit to do it.
Actually what would really be neat is if somebody could do something like the OS X address book in Linux; it's probably the biggest thing that keeps me tied to my Mac. One Address Book for iChat, Mail, my label printer, synchronized across all my computers and to my cell phone. All automagically (for $99 a year).
I can see your point regarding cars purchased with a loan -- if you own the vehicle, then you should be legally entitled to have such a device disabled or removed if you choose, since it's your car.
However where I think there definitely could be a place for these boxes, like them or not, is on leased vehicles. With them it's more like renting an apartment than buying a house: you don't ever hold the title IIRC, the leasing company does and lets you drive it in return for payments; at the end of the lease you relinquish the vehicle.
In that scenario I think the argument could be made in their favor, since you are essentially borrowing someone else's property they can put conditions on it (like 'don't remove the black box').
I could imagine that a company which did this could provide cars even to people who currently are unable to buy or lease ones because their credit is so bad (e.g. the unemployed).
Huh, that's a slick little number for $600. I wonder how much the toner refills cost, and whether the purchase price includes real (full capacity) carts or whether they do what Samsung does with their low-end lasers and include a "starter cartridge" that only is about 20% full. I hope not the latter.
:)
I've actually always had a soft spot for Minolta, dating back to when I used their SLR film cameras. And when I used to work in sales, their reps were some of the most knowledgeable and friendly around. I still have piles of literature sitting around about their first generation digital cameras -- maybe someday they'll be collectible or something.
Wow, good find. I had no idea that you could get 1U Xeon servers that cheap. The last time I looked at rackmount machines, the chassis alone cost three times what they're charging.
This will pretty much be the only good thing that I've gotten so far out of the dot-com bubble...
I don't know; I said netatalk/afp was difficult because I've never actually gotten it to work myself on my Debian Sarge box and small home network of Macs (mostly 10.4 installs), for no small amount of trying. The configuration method isn't necessarily that complicated, in that you're right, but when it doesn't work, there wasn't much that I could work with.
In my case the server would appear on a Mac, but logging in (entering a username and pass, clicking OK) to see the available shares/volumes would just cause the Mac to hang indefinitely until you killed the Finder (spinning-beachball effect). I got the whole thing to work exactly once, and that was by disabling all the security and encryption packages and overriding the Mac's security settings to allow for plaintext passwords, which I consider unacceptable. It's been suggested to me that the problem I'm having might be because of something broken in the Debian/stable package's crypto libraries, but other than that I didn't get much in the way of suggestions from the help lists.
Given that my network setup is about as dead simple as you can get, and the only other thing running on the Linux box was Samba, it was a rather frustrating experience. (Actually it's possibly the most frustrating networking experience I've ever had to date.) I'm sure it's something I could puzzle out given enough time, but eventually it got to the point where I just canned the project and kept my iTunes stored locally, and use CLI tools for Documents backups.
This was my experience with it, I feel like I gave it a more than fair shot with otherwise well-supported hardware on a decidedly simple network, and couldn't get it working in less time than it took to find a workaround (rsync). I sincerely hope that not everyone has this much trouble, but this was why I think that a NAS box that "just worked" with Macs would really be in demand.
> I got a TB of storage... 650MB after RAID 5 formatting
:)
With that kind of overhead, I hope it's really fault tolerant.
Anyway it does seem like a really cool device, though. I have been DIYing something similar, using an old OfficeDepot Compaq $500 special running a minimal install of Debian with Samba and AFP, but AFP is really a bitch to configure sometimes. It would be really nice to have something that just works out of the box. And the streaming media features are similarly something that, although a person could probably do it themselves with a suitable investment of time, are great to not have to fiddle around too much with. Assuming everything works as advertised, of course.
By "off lease" do you mean 'from the back of a van downtown'? And do they scratch the serial numbers off and wipe the drives, or is that part of the self installation?
Seriously, where does one go to get a deal like that? I'd take one for each foot.
I guess there aren't enough dying dot-coms where I am or something.
I'm going to guess that a "toy printer" is anything that:
1) uses ink costing more by weight than gold
2) has space for less than a ream of blank paper
3) has a memory-card slot
4) costs less to purchase initially than two of its consumable refills do
So yes, basically, I think he's referring to the crummy home "hey it came free with my Dell" inkjets which are the bane of anyone who doesn't use Windows' existence.
Generally speaking, Linux has great support for high-end peripheral hardware because that's the stuff that uses open standards most predictably. For whatever reason, it's never the $20k devices that require some Windows-only proprietary driver; it's the $99 ones from Best Buy that do.
I'm going to take a stab at answering this question, or at least my theory as to the answer.
I think it's because -- despite the stereotype of Linux developers being high school or college kids -- the people who actually work on the Linux codebase aren't gamers. Or, more to the point, most gamers aren't interested in writing code and working to make stuff run on Linux. It's a different mindset: one group sees a computer as something to play with, not just a means but an end in itself; the other group sees it as something to play on, a means only.
Maybe I'm biased and maybe I'm using a bad sample, but the majority of computer geeks and actual software developers I know really don't spend that much time playing PC games (or they use consoles), and the people who are really wed to their game machines aren't really interested in playing around with a new OS for the heck of it. It's just two groups of people which, while I'm sure there's some overlap, don't have as much in common as you might think on the surface. Most Linux users are honestly interested in their computer's OS; to a gamer it's just a launchpad that lets them play whatever they want, and the open/closed source controversy is basically irrelevant. The only advantage Linux might have is that it's free as in beer (but a lot of gamers that I know have pirated versions of Windows anyway, so it's not much of a selling point).
I'm sure this will get a lot of people responding with "but I'm a gamer and I use Linux," but before you do, realize that you're a statistical anomaly in the big picture just by virtue of reading Slashdot and caring about the issue either way.
I hate to dump on any company that's selling Linux, but those laptops are outrageously expensive.
Linux:
IBM/Lenovo ThinkPad T42: 1.7GHz, 14" display, 512MB RAM, 40GB disk, CD-RW; $2175.00
Windows:
IBM/Lenovo Thinkpad T42: 1.8GHz, 14" display, 512MB RAM, 60GB disk, CD-RW; $1499.00
Quite possibly it's not Emperor's fault, I have a feeling that IBM/Lenovo may not sell ANY ThinkPads without Windows (especially sad considering that IBM ought to be the one place you could get a Linux machine, if anywhere) and thus you're really paying for a copy of Windows plus Emperor's overhead and whatever it costs them for the support contract (which might be worth something to some people, but not $600!). But at the end of the day, that's a hell of a premium.
The way Microsoft has the hardware market twisted around their finger right now, it's basically impossible to get a quality, name-brand laptop without buying Windows. (I know there are some white-box machines available out there, but put one of them down next to a ThinkPad and there's really no comparison.)
Bingo.
The only thing I disagree with is this: "I know it's unreasonable to treat every Microsoft proposal as suspect."
At this point I think it's entirely reasonable, even prudent, to treat every Microsoft proposal not just as suspect, but as outright hostile, until definitively proven otherwise. I can't remember the last time I saw something come out of that company that didn't have an angle or a hook buried in it somewhere, or give me the vague feeling that they were out to screw me as a consumer down the road.
Microsoft is a shark, and we are all minnows. Just because we play along today doesn't mean they won't try and have us for breakfast tomorrow. No matter what they do, or what they say, it doesn't change what they are.
(And for the record I don't even think they're 'playing along' today -- this open formats business is a ridiculous farce, as even a blind man should be able to see; what remains now is whether the politicians will realize it.)
The additional concern is that MS is so big and known to be legally aggressive, the fact that they have the judgmental authority over whether an implemetation is conformant will be enough to keep a lot of smaller companies from touching the spec. Implement it the wrong way, or in a way that pisses MS off, and you'll find yourself getting sued into the ground. Sure they may not win, but they'll have just enough of a case to keep it in court for a while and bleed you to death.
The result here is that MS can say they have an open format, but the licensing will effectively keep most of the other implementations rather trivial; my guess is that other companies will write "converters" or plugins, rather than actually use the format natively, and then MS can use this lack of native support as a case for continued use of their own products.
> 2. WMF is well understood and many Mac apps support it.
Would you like to support that statement with some evidence, perhaps?
I have yet to see any Mac application that provides good WMF support. As other people have pointed out, even MS' own implementation (used in MS Word Mac and Powerpoint Mac) sucks terribly, and has a tendency to display garbage/white boxes/etc. in place of graphics.
If there are other software packages out there that have better support for it, I'd love to know what they are. I'm not aware of any though.
This is why I'm glad to live in a place where the gun ownership and concealed-carry rate is very high. Assaulting a random stranger would be basically akin to playing russian roulette.
You don't see this kind of crap happening except in places where you can guarantee that your victim is going to be unarmed.
Linux also doesn't have an online hall of fame of Worst-Designed Dialog Boxes Ever.
The kicker: despite the fact that the author of the site is ostensibly a Mac user (by virtue of the host), all of the dialogs appear to be Windows. Guess Apple isn't on par yet either.
Why would you want Javascripts to do any of those things? I can do them just as easily with the built-in functions of my browser. Back = back button, refresh = refresh button, new window = right click - new window/tab. And using a script to open a new window is a little excessive, as I recall it's been possible to make a link open in a new window since roughly the time the Earth's crust was cooling by doing TARGET="_blank".
I hate those little scripts most of all, since they're nothing but unnecessary bloat. The most common use I've seen of Javascripts is creating popup windows that lack the proper controls across the top, which I find infuriating and is probably the number one reason I browse with it disabled.
I understand there are probably situations where doing something like that with scripts is necessary (GMail, for instance), but I'm sorry to say that they're horribly overused and rarely do anything for the viewer.
I agree that story (bug in Sun's Java) should have merited a story, but it's not quite the same level of seriousness as the IE issue. First, I don't think there are nearly as many users of JRE as there are of IE (although I suppose that's arguable since it's preinstalled on a lot of systems), but more importantly, I don't think that the Java hole was published and unpatched for any significant length of time. The article you linked to, after all, is talking about the fix being released and the original hole in the same sentence.
The IE one is important because it's a pretty gaping flaw in a very popular (or at least often-used, perhaps less liked) piece of software which has remained open and basically exploitable for a very long time since it's been published and everyone and their cousin made aware of it. I even heard a quick mention of it on one of the cable news networks, so there's really no way to argue that the cat's not out of the bag and into the hands of the crackers, and once they're done, script kiddies.
Yeah I was looking at the same thing also. They have some OSS stuff there for download (and helpful instructions in case you want to upload your changes -- I mean who wouldn't want to contribute at no charge to their friendly local defense contractor?) but no source.
/src directory on systems they sell, or on the CDs or however they distribute it.
This by itself doesn't mean anything though. Remember that the GPL doesn't require you to make your code available online to just anyone, it just says that you have to distribute it along with the software. So it could be something as simple as a
They can't prohibit you from buying the product and redistributing the source code, but it doesn't mean they have to go out of their way to make it particularly easy for you to do so. The only real way to tell if they're breaking the license would be for somebody to buy RedHawk Linux in whatever form is closest to a "retail box" and see if it comes with source, or a written offer to provide it. I believe those are the requirements of the GPL.
Exactly -- I was going to say the same thing, but you beat me to it.
The flip side to having Free software is that you have to be prepared to let some people that you might not exactly like or agree with use your work. On the other hand, people who might not like you very much have to do the same.
The community would really be shooting itself in the foot if they put some sort of clause like that into OSS licenses, because basically it would be throwing away all the millions of dollars of R&D money that government contractors can and do put into software development, and which might trickle down into OSS projects. Plus, there are a lot of companies out there that you wouldn't think would be affected, that might shy away from OSS after a restriction like that was added, just because they don't want to limit their future business. IBM, for instance, probably wouldn't want to write itself out of the defense sector because some hippies thought it would be cool to ban military uses of Linux, and they're one of the biggest contributors. The NSA, if I remember correctly, has a secure distribution of Linux that's provided some useful things back to the community, and you'd be throwing that away as well. With the exception of RedHat and a few other Linux-only shops, a lot of money comes from companies that sill have proprietary OS options sitting around. If they start to feel that Linux is a hindrance instead of an advantage, there's nothing stopping them from going back and pulling the plug on their OSS ventures.
Overall, it's a foolish and dangerous idea to consider such limitations in licensing, and I think it's important that we put that idea down quickly before it gains any mindshare. It would quite literally be the end of the corporate deep-pocket support of Linux that we've enjoyed until today.
You do realize I hope that this system, THAAD, is a theater missile-defense system, not a global "Star Wars" hemisphere-defense one.
It's designed to protect strategic assets from medium range, single warhead ballistic missiles, which are exactly the thing that China, N. Korea, and other ex-Soviet client states have in spades (and are significantly easier to put together than an ICBM). The intended use is to place them as a spot defense over a high-value target, as the farthest reaches of a layered system that includes short-range defenses like Patriot, etc.
This is not an ICBM shield for the U.S. in the manner that I think you are thinking it is, that would really have any effect in the event of a global thermonuclear war. The preventative measure against that is still MAD. However when you step down from that scenario (and terrorism), the next most likely case of a nuclear weapon being used against us is with a theater ballistic missile against a strategic target like a foreign city, aircraft carrier, or air base. In a situation like that, a defensive system like this makes a lot more sense.
Actually, this isn't the DoD, it's Lockheed Martin, a private company which does business as a government contractor. Although I claim no knowledge of their finances, I'm willing to bet however that their budget is definitely not "unlimited," although it might be large compared to what you and I are used to. Especially if they are on a fixed-price contract, where their total revenue has been specified in advance, the employees are probably under a lot of pressure to keep costs down since every dollar they spend eats into their profit.
I think in this case though there wasn't much competition, at least of the Linux/Windows variety, since Windows doesn't really do real-time applications very well apparently. I'm not actually sure what the other options would have been besides Linux for a project like this: I can only assume that there are some highly proprietary, niche-market OSes for this sort of thing, but that in general the COTS field is pretty limited.
I guess I made a bad assumption and had you mistakenly pegged as one of the "gah, installers suck, everyone knows that the only way to do Linux is to download everything as an undocumented source tarball" crowd. :-) My error. (Obviously if you were, you'd have responded with some ad hominem attacks, and insinuations about my my sexuality.) I've been here too long, I guess; I had forgotten not everyone is engaged in some vicious Debian vs. Gentoo vs. Slackware OSS philosophy war, and it's gone to my head. ;)
I have a particular soft spot for single-purpose "appliance installations" perhaps because I have an inordinate amount of non-functioning white boxes sitting around, but also because I just like the concept of computer-as-interchangable-part on which you can pop in a CD, reboot, and have a functioning device for some particular purpose. It's the antithesis of most people's conceptions of computers as finicky, temperamental, and highly individualized. Of course, the downside to this is that if you want something that's highly customized and application-specific, you're probably not going to be happy. I guess I'm just lucky that in many cases I'm pretty close to that mythical "average user."
At any rate, you did make a good point regarding SmoothWall; it's clearly set up for "ease of use," but only if you match their intended setup: someone with a reasonably modern spare machine and complete access to it (at least for the initial steps). I suppose this is just the necessary tradeoff for what they're trying to do.
Interesting about the article. I was not aware of that method of production.
Regarding your truck, where do you get your bio-diesel? Do you roll your own, or have someone make it for you? I have seen it made in a small batch before, but the amount of lye required seemed rather large and the "washing" process labor-intensive. I've heard rumors that there is somebody out there who has homebuilt a continuous-production system, rather than a batch-based one; I don't have any information on it though. But it's always something that's intrigued me though, and hopefully as the technology matures it'll become less resource and labor-intensive.
LTSP does look pretty close to what I was talking about, it is a pity though that it won't work for your architecture. I'll have to read into their docs/wiki a little and see if it does what I'm hoping for. The Achilles heel of many projects that I've checked out is that they're very insecure: not only is the x-windows client/server communication unencrypted, but it requires a large number of ports left open on the client machines in some cases. Although I understand that a lot of people are going to use these on internal/dedicated subnets and behind firewalls, it just seems like a dangerous practice in general to build a system without regard to security these days. (Suppose you want to use one client wirelessly, and your wifi hub is located outside your firewall?) Hopefully LTSP isn't like that.
The OpenMosix system is something I was not aware of at all -- at least not in the way they're using it. (In short they're taking a thin-client/server combination and using it as a cluster, so that it "load balances" computationally across the various machines. Pretty slick, if it works.)
Anyway, thanks for the link.
At risk of sounding like a shill, I have to disagree with you about SmoothWall. Although I haven't actually tried to install their software yet (it's on my list of things to do), unless their manual is a complete pack of lies (which I don't think it is), I'm not sure how they could have made it simpler. I think perhaps what you were trying to do is something that's just outside the intended use. They're pretty upfront about the fact that the CD-Rom installation option is the easiest way to go, and that if you can't do this the network or floppy installs are more complicated.
Personally I much prefer their method of distribution and philosophy to some other OSS projects' who are inclined to give the user a tarball and vague instructions to "figure it out." While that's fine for a hobby project, it's a crummy way to make a product -- even one you're giving away for free.
SmoothWall isn't a piece of software in the same way a regular application is; it's really an operating system / distro that turns an old box into a dedicated appliance. So the distribution methods that might be convienient for software aren't necessarily as suited to it. One of the reasons I bookmarked SmoothWall's page though is specifically because of the way they give you a straight install ISO, no screwing around with compiling or configuring anything. If it was a tarball, I wouldn't have given it a second glance, such has been my experience installing software from source. (I'm actually pretty sure that it is available as a tarball also, here.)
I understand this probably boils down to some very basic disagreement over the philosophy of software and OSS in particular, but I think StoneWall is an example of how a project ought to operate: producing a relatively easy-to-use, robust, single-purpose product, which is both 'Free' and 'for free,' to the public.
It's something of a pity that they charge money for the version that lets you install over a RS232 serial connection, so you can't do a true headless install with the free one, but I guess they have to support themselves somehow. (Also it doesn't do load balancing even in the corporate version, which I found somewhat surprising, although it will do a fallback from a primary to a secondary connection.)