I can't speak for other people, but here's why I don't think it's a problem (or ironic, as you say): 1. Linux is a hobbiest system, designed and programmed by hobbiests. Most Linux users (myself included) don't care whether or not Linux will make it meanstream. I couldn't care less if Linux "takes over the world" or not, as long as it works for me. Linux was not meant to be a consumer product, companies like RedHat and Mandrake, et al are trying to make it such. 2. Most of the hobbiests that use Linux are quite comfortable building applications from source. To them, binary compatibility is not necessary, as long as the source code exists. 3. Linux is all about contributions. The people who constantly bitch and moan about lack of binary compatibility are mostly people who leech from the community without giving back.
Linus Torvalds has gone on record stating that he doesn't care about binary compatibility. He has stated that while he won't go out of his way to break it, if something needs to change that does break it, well, it gets broken.
If you want easy access to applications, they pay for RedHat, Mandrake, or SuSE, and pay for people to package applications for those distributions. The distributions will never have a common, single executable for download for the applications, since the distributions all have different focuses. A lot of people expect the world for free, and then bitch when they can't get it.
The problem here is that of binary compatibility. Every distribution ships with different library versions, some patched, some vanilla. Odds are that an RPM for Mandrake will not work with Fedora.
Linux seemingly wasn't designed for binary backwards-compatibility, it's a completely different mindset from Windows (where they bend over backwards to make sure that applications still run, to the detriment of the system). I don't mind this, since I tend to stick with one distribution (I prefer Slackware) and am quite comfortable compiling my own software if necessary. That doesn't work for everybody - oh well.
I agree with the rest of your post, but driving really is a privilege, not a right. Not having a drivers' license does not impair the freedom of movement, it only takes one method of movement away. If you don't have a license, you can still take a plane, call a cab, take a bus, ride a bicycle, or worst case, walk.
I'm in favor of difficult drivers' tests with the intent of taking away drivers' licenses from those who clearly cannot safely operate a car. If you can't drive a car without presenting a risk to yourself or others, you don't deserve to operate a motor vehicle, period.
You are correct. In fact, the second article states that. But, this is what else it had to say: "Thwarting President Bush's global dollar diplomacy and its designs on breaking OPEC's oil pricing power provide additional reasons for OPEC to switch to payments in euros. This would mean that the United States would have to buy euros with dollars before it could buy OPEC oil.
The dollar would fall further -- and the euro would rise. The U.S. economy would eventually have to adjust to $5-a-gallon gasoline (the average world price).
The bad news would be a deeper U.S. recession, SUV owners would suffer while Toyota and Honda would grab more market-share with their 50-60 mpg hybrid cars. "
That said, so yes, we may be able to export more things because of a lower dollar. But, with that lower dollar, if we're no longer the dominant currency, then we have a lower exchange rate compared with the Euro. So, instead of paying one American dollar per gallon, we would pay 1 Euro per gallon (because they now want Euros, not dollars). But if 1 Euro is trading for 2 American dollars, then we're paying twice as much as we were before. In order to make up that difference, we'd have to export a lot of goods (and no, exporting jobs doesn't count!), but the U.S. (granted, I haven't looked this up) doesn't seem to export as much to the rest of the world as we import.
The oil products that we consume on a daily basis (and let's face it, there's LOTS of SUVs on the roads) will cost a lot more, because we would have a weak dollar compared to the Euro (as I explained above). And if people have to pay more for gas so that they can get to and from work everyday, they're going to have less money to spend on the extras. Let's not forget that with all of the gas increases lately, a lot of shippers are now adding a "gas tax" to the regular shipping fee - this increases the cost of goods even further.
Solution: Invade Iraq (which has a plentiful oil supply, plus a bad man running it), take over the oil supply and continue to sell it in U.S. dollars. We then don't have to worry about the Euro, since we still have our own oil supply in our own currency.
It was most definitely about the oil. But not necessarily the United States getting the oil. The U.S. just needed to stop Iraq from selling oil in Euros and devaluing the U.S. currency even further.
Not from the "mainstream" press, but excellent articles detailing of how Iraq switching from the U.S. dollar (approved by OPEC in the early 70's as the official currency for oil) to the Euro for oil could seriously harm the U.S. economy.
Social Security is a Ponzi scheme, I'm under 30, and it's highly unlikely that I'll actually see anything that I'm contributing. Currently, I contribute to my own personal savings plans, and I'd love to have the option to opt out of Social Security (I'd put that money into my own savings instead).
The FDIC is of limited use as well. If you screw up your finances, you could still lose tons of money (FDIC only covers up to $100,000), but if the economy really tanked and all banks went belly-up, it has been estimated that FDIC would cover seven cents on the dollar (no source to cite that, I received it from a friend who was subscribed to an e-mail list).
Some of the other programs you mention are useful, but some government programs are a complete waste (or benefit too few people to be considered very useful).
The 9200 is hardly a "new" card. While the designation/release of the card might be relatively new, the card is actually using a variant of the chipset used in the older 8500. Therefore, the drivers support it.
"New" cards would refer to the 9600/9700/9800 - there are no open source 3D drivers for them (although I'm using the XFree86 "radeon" driver for 2D on my 9700 right now).
If you look at http://dri.sourceforge.net/cgi-bin/moin.cgi/ATIRad eon, you'll see the breakdown of the cards. The 9200 has a RV280 core, whereas the 9600/9700/9800 have RV350/R300/R350 cores. R(v)300 and higher don't have open-source drivers, although the GATOS team claims to be working on something.
The lack of NTFS included in Fedora Core 2 Test 1 was a bit annoying. But, as a parent poster pointed out, there are RPM's available, or you can just build the NTFS module: 1. cd/usr/src/linux-2.6 2. make menuconfig 3. Find NTFS, and build it as a module ("M") 4. Save configuration 5. make modules;make modules_install
Worked for me.
I was using Fedora Core 2 Test 1, but there was lots of things wrong with it. The installer wouldn't work with my PATA to SATA converter on a Silicon Image 3112A chipset (I logged the bug, Fedora Core 1 also breaks), and for some random, unknown reason, the system would just slow down.
At the moment, I'm back to Slackware-current, which does work for me.
Here in California, 511 (so the signs/radio ads say, I haven't tried it) will give you travel information - road conditions, traffic accident reports, bus schedules, carpooling information, etc.
The thing is that in the process of doing the outsourcing, the Indian companies gain experience at our expense. In due time, we'll be competing against the companies we outsource to.
I work at the American office of a Taiwanese company that makes broadband products. At first (before I worked for them) we were just an ODM (they would make the products for other customers and slap the customer's label on it), but now with the experience they have gained from making these other products, we are now selling products under our own name.
A lot of my co-workers in Taiwan have trouble understanding some concepts at the moment (like security), but they are slowly learning from my example.
You can bet that the same thing is happening elsewhere... And once they finally get the details down, we're of no use to them anymore.
So? Most motherboards come with a video card built-in nowadays.
Sure, if you go with MicroATX. Most ATX motherboard do NOT come with onboard video (at least not according to looking at the 18 or so motherboards displayed at Fry's).
I'm sorry, but that's crap. It's not 'extremely slow'. Have you tried recent versions of the ivtv driver with the proper patches? I don't think so.
While it may be true that recent patches fix things, I have a stable MythTV machine using the IVTV RPM from Axel's site (ivtv-0.1.9-24.rhfc1.at - not the latest and greatest, but one rev back). I'm not about to apply a bunch of patches just to "maybe" get things working. With the stock ivtv driver (which is what I and many other people use), it has been reported that even on a fast Athlon machine (XP 2400+), DiVX video playback (as an example) is extremely slow.
From what I have read on the MythTV and ivtv mailing lists, the big problem is that the PVR-350 framebuffer driver does not support XVideo extensions, which means that there is no accelerated video playback, unless the video is in a format that the PVR-350 natively handles. So, outside of maybe watching TV in MythTV or watching MPEG-2 videos, the PVR-350 will be slow.
If you feel confident applying patches to get things to work, great. I feel confident as well, but my MythTV machine works for me with it's current configuration, and since already have a TV-Out card that works, I'll continue to use it until there's a new ivtv release that irons out the problems.
This is true, there are driver instability issues, particularly on this point. But they are, in my experience, very dependant on hardware (avoid VIA chipsets). I had all the trouble in the world with an older EPIA system and a PVR-350 (though there are plenty of success stories with that combination!) until I switched to an even older PIII system on a decent Abit board, and the latest version of the IVTV driver with some patches by Chris Kennedy. No more DMA issues. It works very stable now.
Again, because of the above compatibility issues, I haven't been using my PVR-350 for TV-Out, but there have also been reported problems with the OSD (On Screen Display) causing problems and using previews in the Watch Recordings screen on MythTV. While I agree with your assertion that VIA hardware and PVR-x50 cards don't seem to get along under Linux very well (I replaced my KT-400-based board with a SiS748-based board recently, it helped quite a bit), I think that's still a driver issue. Apparently there are patches in testing, but when those will get merged into a new 0.1.10 release of the ivtv driver is unknown.
It's interesting that you switched to an slower CPU and some of your problems went away. There's been some speculation on the ivtv mailing list that there's some timing issues on faster CPUs.
Right now, it all depends on how much patching you're willing to do.
I build MythTV from the RPMs that are done by Axel Thimm, which are linked to in Jarod's MythTV website. From what I have read on the MythTV mailing list, there are some issues with X/MythTV on the PVR-350 TV-Out. From Jarod's own website (which might be a bit dated):
General notes and observations from my first few days of using the PVR-350's TV-Out:
* mplayer WILL play back movies through the 350's tv-out, but the processor has to do all the decoding and the sound has to go through your sound card. However, even on an Athlon XP 2400 system, video playback isn't entirely smooth (for high-quality divx rips). Better support for the 350 as an output device is slated for a future mplayer release. * Trying to use MythGallery locked up the frontend, MythDVD suffers from the same problems as mplayer (until you use a patched mplayer, which I haven't yet tried). I haven't yet tried MythGame or MythMusic. * On many TVs, the picture is heavily overscanned. I haven't yet tried tweaking the XF86Config mode to see if I can reduce it, though I've seen indications on the mythtv-users mailing list that this may not be something one should attempt. For reference, I'm currently having to manually specify a MythTV window size of something like 632x472, 40 pixel X offset, 8 pixel Y offset, to fit it to the screen. This will vary heavily from TV to TV. * Thus far, I'm not seeing a monstrous difference between the 350's output and a GF4MX's output (both via SVid), but I would say the picture is a bit sharper in scenes with high motion (I see a soft edge to things w/the GF4MX that are very sharp w/the 350). * I think the 350's picture is a touch sharper than when I'm feeding my HDTV a progressive-scan signal via my VGA->Component video adapter, but not by a huge amount. The adapter wins hands-down for me though, because of my large divx collection (both because of 350 playback issues and resolution), the far superior readability (yes, I occasionally web browse on it, occasionally use the shell on it, etc., and a progressive-scan signal is WORLDS better for that), and stability * So far as stability, I hadn't had a single crash of my MythTV system in ages until introducing the 350. Generally, it works great, but I've froze up the frontend a number of times in a few days, caused two or three crashes of the backend (which is on an entirely different machine), and completely hard-locked the system twice (I've been hammering on it pretty good though). At the moment, I'm back to using the output on my GeForce 4 MX, because I'll take stability over a slight picture quality improvement any day. The 350 was slaughtering the high WAF (Wife Approval Factor) MythTV had been enjoying for some time.;-)
Myself, I use the IVTV driver with very few problems, using both the PVR-250 and PVR-350 as hardware encoders (but use a GeForce 4MX for TV-Out). My MythTV machine stays on 24/7, and I don't really run into too many problems. The most annoying problem right now (and I'm not sure if it's IVTV, MythTV or lirc) is that if I was recording something on card 1, and watching TV on card 2, when I change channels on card 2 (after card 1 is finished recording), then I get all sorts of stuttering problems. But, if I exit TV, and restart it, then all is fine.
If I understand things correctly, the "stable" branch of IVTV is being maintained by other developers while the main developer(s) go on with the 0.2 branch, which should fix most of the issues.
For now, my recommendation is go for it, but unless you feel like putting in patches, then use it as a PVR-350, and just wait.
I can't answer the question about DVD playback, because I haven't configured my PVR-350 for DVD playback.
I investigated myHTPC, read the forums, checked out the available software, and didn't really find what I was looking for. I would imagine that the situation has since changed.
MythTV in the beginning for me was quite challenging, and I've been using UNIX in some form or another for 10+ years (in fact, I learned UNIX before I learned Windows, turned out to be a fluke in high school computer lab scheduling). There was always documentation for MythTV that evolved, but back when I started using it (MythTV version 0.9 I think), it was still kind of spotty, and there wasn't a straight, easy to follow HOWTO. And compiling, LOTS of compiling, if you wanted everything installed.
That situation has since changed, there's plenty of RPM packages for Fedora Core 1, AptForRPM takes care of the dependences, and Jarod Wilson has written an excellent MythTV HOWTO.
I used to have a wireless NIC in my MythTV machine, but the RealTek driver (binary driver) crashes with Fedora, so I can't use that. I adminster my machine (the rare time I need to) via SSH, and use a universal remote control with the IR receiver on the PVR-250.
I ended up buying a desktop ATX case before the Overture came out (a cheap case), and when it did came out, I was kind of kicking myself. Just out of curiousity, how is it? I've heard both good and bad about it, the main thing is the cooling. Does it run really hot?
I feel that I should point out a couple of things for anybody considering getting a PVR-350 for Linux at the moment: 1. Even though the PVR-350 has TV-Out on it, AFAIK, you'll still need a regular video card in your PC (I don't think that motherboard BIOS' will recognize the PVR-350 as a video card, though I may be wrong on that). 2. PVR-350 TV-Out at the moment isn't that great. Last I heard, it's just a framebuffer, so unless you're running specific apps that will use the acceleration, like mplayer, MythTV (for watching MPEG-2 recordings only, I think), and maybe Xine, it's extremely slow. And for applications that require some sort of GL support, well, forget it (people have had numerous troubles with the MythGame module/XMAME).. 3. Apparently there's some issues regarding using the PVR-350 to record, and using the TV-Out at the same time.
However, your advice is still sound. I bought a PVR-350, and at the moment, I'm using it as a PVR-250, using my GeForce 4MX for TV-Out. Once the ivtv driver stabilizes, I'll switch to the PVR-350. There's been a lot of traffic on the ivtv mailing list lately with patches for testing and the like.
Granted, I now have a PVR-250 and a PVR-350 in my MythTV machine, but I started out with dual BT878 capture cards (a Pinnacle PCTV Rave and a Hauppauge WinTV/Radio).
Although the picture quality of the PVR-x50 cards are better, I would argue that MythTV actually runs better on the BT8x8 cards, only because the ivtv driver is a bit unstable still.
The two main reasons why everybody raves about the PVR-250 cards (and thus tends to go with them) under MythTV are: 1. The graphics quality is SO much better. Even when you crank up the resolution on the BT8x8 cards to 720x480, there's still a noticable difference in picture quality (although, this could be due to the software codecs). 2. While giving this amazing picture quality, the bulk of the work is done on the PVR-x50 card. Thus, the main CPU remains free for other things.
I can speak from first hand experience, trying to do a dual-tuner system using BT8x8 cards really chews up CPU power. If you have the settings set up too high, and you end up recording two things at once (watching Live TV counts as a recording), then you'll end up dropping frames in the second recording (stuttering). The PVR-x50 eliminates all of these worries (assuming you have a motherboard that can handle the DMA traffic).
Hell yeah, I have to agree. I could have shelled out a total of $500-$600 for a TiVo with a lifetime subscription, but I like to do things the hard way, it seems. I was seriously considering the TiVo, because it would just work.
Then, I did some investigation into do-it-yourself PVR's, and stumbled onto MythTV and Freevo. I tried Freevo first, and didn't really like it (even before using it to watch TV), but the clincher was that, at that time, it didn't support PVR functionality (rewinding through the live TV stream).
I had (still have to some extent) a bunch of older hardware kicking around, so I decided to give MythTV a shot. The hardware was somewhat under-powered, but if I scaled down the capture resolution (to 320x480, for instance), I was able to get it to work. That was enough for me to start spec'ing out some new hardware and make purchases over a period of time.
It was definitely a learning experience in putting together a MythTV machine, helping find bugs, submitting a small code patch or two. I'd do it again, but that's because I'm usually up for a good challenge. If you want something that just works, seriously, buy a TiVo, that's what they do is make devices that work.
That is quite possible, and quite probable. I was at Sega only for the Dreamcast, and even then, was not the graphics guru in the department.
Most of my graphics experience comes from 2D stuff on the Commodore 64, where I'm used to having a 320x200 visible (main screen), but under NTSC (on a C64), you actually get somewhere around 260-270 raster scanlines to play around with (I don't remember exactly how many, it was in the documentation somewhere). Of course, most of those scanlines were "off-screen" (I believe "on-screen ranged from something like 50-249), so the extra scanlines weren't really useful for drawing anything (not for me, but there were people who did turn the borders off and things like that).
Usually what I used those extra scanlines for was updating the screen so that my scrolling messages wouldn't tear. Rather than double buffer, it was quite easy (and it presented no visual artifacts) if you updated screen memory after it was drawn by the raster. So if I was doing a 16 scanline split screen (one color per scanline), but I didn't trigger the VIC-II raster interrupt for another 40 scanlines below that, if I had a routine that would fit into the time it took to draw those extra 48 scanlines, I'd call it.
Of course, this timing is crucial, since it was using interrupts, and your ISRs must be fast, or you bog the whole system down.
You can see the same thing in the Commodore 64, or almost any system where the developers relied on the NTSC or PAL timings. I guarantee that demos that I wrote on the C64 (which were done in NTSC) don't run very well in PAL - all sorts of visual artifacts, because I timed them for NTSC (the Commodore 64 in particular had the same raster timing repeated every 8 pixels, if I recall correctly, though it has been a while).
Generally, if you're doing video related work, it's easier to rely on the timing of the video, since you're generally shooting for good screen graphics (no tears, smooth lines, etc.).
The most common way that I'm familiar of for the NTSC game developers to make a PAL port of the title is to get the PAL version of the game system and a TV capable of PAL. From there, you insert some checks in the game, fix it up, and boom, you have an NTSC/PAL version.
Of course, you also have to account for the happiness of the developers too, because once you're used to NTSC, looking at PAL is maddening (and presumably the other way around) until you're used to it.:)
Almost every computer-like machine in existance has a clock in it. This clock isn't necessarily a clock like you'd use for viewing the current date/time, but is in some cases internal to the CPU.
What units this clock runs in varies from chip to chip, but most of the time, the OS that you're using provides you with a decent way of using it, in some sort of standard measure of time (vxWorks with the BSPs that I've used, for example, provides you with 60ths of a second, which is very convenient).
This is very important, since especially in cable modems, there are a lot of events that need to be synchronized, and the CPUs change in clock speed (i.e. MHz) on a regular basis.
In terms of consoles, timing things to the VSYNC is generally popular. In order to properly do double buffering, you need to swap the buffers during the point of time when the raster isn't being shown within the visible screen, or at least in the section you're drawing (otherwise, you'll get tearing).
That, and it's a fixed timing, either 50Hz (PAL) or 60Hz (NTSC).:)
Ah. But see, there are very few providers of broadband out there (not that this is a good thing).
If the customer is really bone-headed and doesn't address the issue, they'll eventually run out of providers in their area, as they keep getting kicked off. Then they'll either have to fix the problem, or be without internet service.
Of course, the bad side of this is that cable providers might start sharing information with each other and creating blacklists. It has it's advantages and disadvantages.
Plus, as a (potential) customer myself, I'd probably sign up for an ISP that could provide me extra bandwidth because the wasters aren't on the network.
I think that it's completely fair to blame the user.
Warning... Obligatory analogy to cars coming up...
Users should understand how their computers work, and the consequences of using a computer. I'm not saying that they should know how everything works down to the circuit level, but they should understand that a computer will do whatever a program that is running on it tells the machine to do. They should understand that without proper precautions, other people can "hi-jack" their machine.
I'm a tech guy, but honestly, I don't know completely how my car works. I know some of the basics, like what the buttons on the consoles do, how to check my fluid levels, and the sound that my engine makes. Granted, I can't change my own oil or brakes, but I do know who I can take the car to in order to get that done, and I do have the common sense to listen for strange sounds. If I have questions, I find the answer (whether that means reading a book, searching on Google, or talking to a mechanic).
At the same time, I also understand that while my car can get me from point A to point B, I also understand that it can be an instrument of destruction if I don't take proper precautions. If I don't have the brakes serviced regularly, I can hit something causing damage (or death). If I don't top-up/change the oil on a regular basis, the car's engine can seize up.
Oh, and if Windows Update doesn't work (funny though, it seems to work for me on the machine I have here at work, and my desktop at home), then maybe people should be bitching to Microsoft to fix it? And if Microsoft won't fix it, then maybe don't buy their products, use something else?
Computers (at the moment anyway) are not simple little tools that play games, let you send e-mails or surf the Internet. They are very complex machines that can do those things, and more. It's the "more" part that you have to realize is there. If you don't understand the "more" part, then you should be talking to people who do, even if it means buying a book, buying some software, or paying a professional for their time. Just like how I bought a repair manual for my car, and I pay a mechanic to change the brakes or the oil.
If your modem is an older one, it is quite possible that you were able to do this. That's why a lot of the hacks telling you how to do this mention that you need a specific version of the firmware.
The cause for it was that you would poison the ARP cache, and when the TFTP request went out, the modem had to then ARP for the TFTP server. If you timed it carefully enough (and your TFTP server matched the IP address of the real TFTP server), the request over Ethernet would typically come back before the request over the cable interface (your PC is right next to you, the cable operator's TFTP server is many hops away).
Simple fix: Newer firmwares just don't accept any TFTP traffic over the ethernet port during those critical stages (as the simplest, hackiest fix).
Granted, I like it that you're given options, but for the Upper House, that is just way too many choices. In my opinion, it seems that they're just asking for confusion.
From what I've seen, it's somewhat difficult to get people out to vote in the United States. It's not necessarily that they don't care, but rather, as I've seen, it's more because even though it's law that your employer must give you time off to vote, it's still somewhat inconvenient to get to the polling place and actually cast your vote.
So, I guess with that in mind (among other things), they decide to cram everything that they can on the ballot. Heck, I'm a fairly patient and reasonable person (well, most of the time;) ), but if I was faced with choosing from 100+ options, I'd be a little daunted too.
It's been a while since I've lived and voted in Canada, but when you get your ballot, it comes with an instruction card telling you to mark a huge X inside the box of the candidate that you wish to vote for. It even shows you a graphical example.
In terms of how the votes are counted, I think that if you marked your ballot incorrectly, it's just counted as spoiled (so no vote for any candidate, nor do they try to figure out who you voted for). Quite frankly, I like this system, if you can't properly fill out the ballot, your vote shouldn't count, and there shouldn't be a guess as to what you meant.
I think that part of the simplicity of Canada's voting system is that there are usually at most six candiates on the ballot, since you only vote for the representative of your riding.
Read the instructions that come with the ballot If you run into a problem, ask for assistance, that's what the volunteers are there for.
Slashdot Mirror
I can't speak for other people, but here's why I don't think it's a problem (or ironic, as you say):
1. Linux is a hobbiest system, designed and programmed by hobbiests. Most Linux users (myself included) don't care whether or not Linux will make it meanstream. I couldn't care less if Linux "takes over the world" or not, as long as it works for me. Linux was not meant to be a consumer product, companies like RedHat and Mandrake, et al are trying to make it such.
2. Most of the hobbiests that use Linux are quite comfortable building applications from source. To them, binary compatibility is not necessary, as long as the source code exists.
3. Linux is all about contributions. The people who constantly bitch and moan about lack of binary compatibility are mostly people who leech from the community without giving back.
Linus Torvalds has gone on record stating that he doesn't care about binary compatibility. He has stated that while he won't go out of his way to break it, if something needs to change that does break it, well, it gets broken.
If you want easy access to applications, they pay for RedHat, Mandrake, or SuSE, and pay for people to package applications for those distributions. The distributions will never have a common, single executable for download for the applications, since the distributions all have different focuses. A lot of people expect the world for free, and then bitch when they can't get it.
-- Joe
The problem here is that of binary compatibility. Every distribution ships with different library versions, some patched, some vanilla. Odds are that an RPM for Mandrake will not work with Fedora.
Linux seemingly wasn't designed for binary backwards-compatibility, it's a completely different mindset from Windows (where they bend over backwards to make sure that applications still run, to the detriment of the system). I don't mind this, since I tend to stick with one distribution (I prefer Slackware) and am quite comfortable compiling my own software if necessary. That doesn't work for everybody - oh well.
-- Joe
I agree with the rest of your post, but driving really is a privilege, not a right. Not having a drivers' license does not impair the freedom of movement, it only takes one method of movement away. If you don't have a license, you can still take a plane, call a cab, take a bus, ride a bicycle, or worst case, walk.
I'm in favor of difficult drivers' tests with the intent of taking away drivers' licenses from those who clearly cannot safely operate a car. If you can't drive a car without presenting a risk to yourself or others, you don't deserve to operate a motor vehicle, period.
-- Joe
You are correct. In fact, the second article states that. But, this is what else it had to say:
"Thwarting President Bush's global dollar diplomacy and its designs on breaking OPEC's oil pricing power provide additional reasons for OPEC to switch to payments in euros. This would mean that the United States would have to buy euros with dollars before it could buy OPEC oil.
The dollar would fall further -- and the euro would rise. The U.S. economy would eventually have to adjust to $5-a-gallon gasoline (the average world price).
The bad news would be a deeper U.S. recession, SUV owners would suffer while Toyota and Honda would grab more market-share with their 50-60 mpg hybrid cars. "
That said, so yes, we may be able to export more things because of a lower dollar. But, with that lower dollar, if we're no longer the dominant currency, then we have a lower exchange rate compared with the Euro. So, instead of paying one American dollar per gallon, we would pay 1 Euro per gallon (because they now want Euros, not dollars). But if 1 Euro is trading for 2 American dollars, then we're paying twice as much as we were before. In order to make up that difference, we'd have to export a lot of goods (and no, exporting jobs doesn't count!), but the U.S. (granted, I haven't looked this up) doesn't seem to export as much to the rest of the world as we import.
The oil products that we consume on a daily basis (and let's face it, there's LOTS of SUVs on the roads) will cost a lot more, because we would have a weak dollar compared to the Euro (as I explained above). And if people have to pay more for gas so that they can get to and from work everyday, they're going to have less money to spend on the extras. Let's not forget that with all of the gas increases lately, a lot of shippers are now adding a "gas tax" to the regular shipping fee - this increases the cost of goods even further.
Solution: Invade Iraq (which has a plentiful oil supply, plus a bad man running it), take over the oil supply and continue to sell it in U.S. dollars. We then don't have to worry about the Euro, since we still have our own oil supply in our own currency.
-- Joe
It was most definitely about the oil. But not necessarily the United States getting the oil. The U.S. just needed to stop Iraq from selling oil in Euros and devaluing the U.S. currency even further.
Not from the "mainstream" press, but excellent articles detailing of how Iraq switching from the U.S. dollar (approved by OPEC in the early 70's as the official currency for oil) to the Euro for oil could seriously harm the U.S. economy.
Not Oil, but Dollars vs. Euros
Iraq, the Dollar and the Euro
Social Security is a Ponzi scheme, I'm under 30, and it's highly unlikely that I'll actually see anything that I'm contributing. Currently, I contribute to my own personal savings plans, and I'd love to have the option to opt out of Social Security (I'd put that money into my own savings instead).
The FDIC is of limited use as well. If you screw up your finances, you could still lose tons of money (FDIC only covers up to $100,000), but if the economy really tanked and all banks went belly-up, it has been estimated that FDIC would cover seven cents on the dollar (no source to cite that, I received it from a friend who was subscribed to an e-mail list).
Some of the other programs you mention are useful, but some government programs are a complete waste (or benefit too few people to be considered very useful).
-- Joe
Just to clarify...
d eon, you'll see the breakdown of the cards. The 9200 has a RV280 core, whereas the 9600/9700/9800 have RV350/R300/R350 cores. R(v)300 and higher don't have open-source drivers, although the GATOS team claims to be working on something.
The 9200 is hardly a "new" card. While the designation/release of the card might be relatively new, the card is actually using a variant of the chipset used in the older 8500. Therefore, the drivers support it.
"New" cards would refer to the 9600/9700/9800 - there are no open source 3D drivers for them (although I'm using the XFree86 "radeon" driver for 2D on my 9700 right now).
If you look at http://dri.sourceforge.net/cgi-bin/moin.cgi/ATIRa
-- Joe
The lack of NTFS included in Fedora Core 2 Test 1 was a bit annoying. But, as a parent poster pointed out, there are RPM's available, or you can just build the NTFS module: /usr/src/linux-2.6
1. cd
2. make menuconfig
3. Find NTFS, and build it as a module ("M")
4. Save configuration
5. make modules;make modules_install
Worked for me.
I was using Fedora Core 2 Test 1, but there was lots of things wrong with it. The installer wouldn't work with my PATA to SATA converter on a Silicon Image 3112A chipset (I logged the bug, Fedora Core 1 also breaks), and for some random, unknown reason, the system would just slow down.
At the moment, I'm back to Slackware-current, which does work for me.
-- Joe
It differs from area to area.
Here in California, 511 (so the signs/radio ads say, I haven't tried it) will give you travel information - road conditions, traffic accident reports, bus schedules, carpooling information, etc.
-- Joe
The thing is that in the process of doing the outsourcing, the Indian companies gain experience at our expense. In due time, we'll be competing against the companies we outsource to.
I work at the American office of a Taiwanese company that makes broadband products. At first (before I worked for them) we were just an ODM (they would make the products for other customers and slap the customer's label on it), but now with the experience they have gained from making these other products, we are now selling products under our own name.
A lot of my co-workers in Taiwan have trouble understanding some concepts at the moment (like security), but they are slowly learning from my example.
You can bet that the same thing is happening elsewhere... And once they finally get the details down, we're of no use to them anymore.
-- Joe
So? Most motherboards come with a video card built-in nowadays.
Sure, if you go with MicroATX. Most ATX motherboard do NOT come with onboard video (at least not according to looking at the 18 or so motherboards displayed at Fry's).
I'm sorry, but that's crap. It's not 'extremely slow'. Have you tried recent versions of the ivtv driver with the proper patches? I don't think so.
While it may be true that recent patches fix things, I have a stable MythTV machine using the IVTV RPM from Axel's site (ivtv-0.1.9-24.rhfc1.at - not the latest and greatest, but one rev back). I'm not about to apply a bunch of patches just to "maybe" get things working. With the stock ivtv driver (which is what I and many other people use), it has been reported that even on a fast Athlon machine (XP 2400+), DiVX video playback (as an example) is extremely slow.
From what I have read on the MythTV and ivtv mailing lists, the big problem is that the PVR-350 framebuffer driver does not support XVideo extensions, which means that there is no accelerated video playback, unless the video is in a format that the PVR-350 natively handles. So, outside of maybe watching TV in MythTV or watching MPEG-2 videos, the PVR-350 will be slow.
If you feel confident applying patches to get things to work, great. I feel confident as well, but my MythTV machine works for me with it's current configuration, and since already have a TV-Out card that works, I'll continue to use it until there's a new ivtv release that irons out the problems.
This is true, there are driver instability issues, particularly on this point. But they are, in my experience, very dependant on hardware (avoid VIA chipsets). I had all the trouble in the world with an older EPIA system and a PVR-350 (though there are plenty of success stories with that combination!) until I switched to an even older PIII system on a decent Abit board, and the latest version of the IVTV driver with some patches by Chris Kennedy. No more DMA issues. It works very stable now.
Again, because of the above compatibility issues, I haven't been using my PVR-350 for TV-Out, but there have also been reported problems with the OSD (On Screen Display) causing problems and using previews in the Watch Recordings screen on MythTV. While I agree with your assertion that VIA hardware and PVR-x50 cards don't seem to get along under Linux very well (I replaced my KT-400-based board with a SiS748-based board recently, it helped quite a bit), I think that's still a driver issue. Apparently there are patches in testing, but when those will get merged into a new 0.1.10 release of the ivtv driver is unknown.
It's interesting that you switched to an slower CPU and some of your problems went away. There's been some speculation on the ivtv mailing list that there's some timing issues on faster CPUs.
-- Joe
Right now, it all depends on how much patching you're willing to do.
;-)
I build MythTV from the RPMs that are done by Axel Thimm, which are linked to in Jarod's MythTV website. From what I have read on the MythTV mailing list, there are some issues with X/MythTV on the PVR-350 TV-Out. From Jarod's own website (which might be a bit dated):
General notes and observations from my first few days of using the PVR-350's TV-Out:
* mplayer WILL play back movies through the 350's tv-out, but the processor has to do all the decoding and the sound has to go through your sound card. However, even on an Athlon XP 2400 system, video playback isn't entirely smooth (for high-quality divx rips). Better support for the 350 as an output device is slated for a future mplayer release.
* Trying to use MythGallery locked up the frontend, MythDVD suffers from the same problems as mplayer (until you use a patched mplayer, which I haven't yet tried). I haven't yet tried MythGame or MythMusic.
* On many TVs, the picture is heavily overscanned. I haven't yet tried tweaking the XF86Config mode to see if I can reduce it, though I've seen indications on the mythtv-users mailing list that this may not be something one should attempt. For reference, I'm currently having to manually specify a MythTV window size of something like 632x472, 40 pixel X offset, 8 pixel Y offset, to fit it to the screen. This will vary heavily from TV to TV.
* Thus far, I'm not seeing a monstrous difference between the 350's output and a GF4MX's output (both via SVid), but I would say the picture is a bit sharper in scenes with high motion (I see a soft edge to things w/the GF4MX that are very sharp w/the 350).
* I think the 350's picture is a touch sharper than when I'm feeding my HDTV a progressive-scan signal via my VGA->Component video adapter, but not by a huge amount. The adapter wins hands-down for me though, because of my large divx collection (both because of 350 playback issues and resolution), the far superior readability (yes, I occasionally web browse on it, occasionally use the shell on it, etc., and a progressive-scan signal is WORLDS better for that), and stability
* So far as stability, I hadn't had a single crash of my MythTV system in ages until introducing the 350. Generally, it works great, but I've froze up the frontend a number of times in a few days, caused two or three crashes of the backend (which is on an entirely different machine), and completely hard-locked the system twice (I've been hammering on it pretty good though). At the moment, I'm back to using the output on my GeForce 4 MX, because I'll take stability over a slight picture quality improvement any day. The 350 was slaughtering the high WAF (Wife Approval Factor) MythTV had been enjoying for some time.
Myself, I use the IVTV driver with very few problems, using both the PVR-250 and PVR-350 as hardware encoders (but use a GeForce 4MX for TV-Out). My MythTV machine stays on 24/7, and I don't really run into too many problems. The most annoying problem right now (and I'm not sure if it's IVTV, MythTV or lirc) is that if I was recording something on card 1, and watching TV on card 2, when I change channels on card 2 (after card 1 is finished recording), then I get all sorts of stuttering problems. But, if I exit TV, and restart it, then all is fine.
If I understand things correctly, the "stable" branch of IVTV is being maintained by other developers while the main developer(s) go on with the 0.2 branch, which should fix most of the issues.
For now, my recommendation is go for it, but unless you feel like putting in patches, then use it as a PVR-350, and just wait.
I can't answer the question about DVD playback, because I haven't configured my PVR-350 for DVD playback.
-- Joe
The way I look at it, to each their own.
I investigated myHTPC, read the forums, checked out the available software, and didn't really find what I was looking for. I would imagine that the situation has since changed.
MythTV in the beginning for me was quite challenging, and I've been using UNIX in some form or another for 10+ years (in fact, I learned UNIX before I learned Windows, turned out to be a fluke in high school computer lab scheduling). There was always documentation for MythTV that evolved, but back when I started using it (MythTV version 0.9 I think), it was still kind of spotty, and there wasn't a straight, easy to follow HOWTO. And compiling, LOTS of compiling, if you wanted everything installed.
That situation has since changed, there's plenty of RPM packages for Fedora Core 1, AptForRPM takes care of the dependences, and Jarod Wilson has written an excellent MythTV HOWTO.
I used to have a wireless NIC in my MythTV machine, but the RealTek driver (binary driver) crashes with Fedora, so I can't use that. I adminster my machine (the rare time I need to) via SSH, and use a universal remote control with the IR receiver on the PVR-250.
I ended up buying a desktop ATX case before the Overture came out (a cheap case), and when it did came out, I was kind of kicking myself. Just out of curiousity, how is it? I've heard both good and bad about it, the main thing is the cooling. Does it run really hot?
-- Joe
I feel that I should point out a couple of things for anybody considering getting a PVR-350 for Linux at the moment:
1. Even though the PVR-350 has TV-Out on it, AFAIK, you'll still need a regular video card in your PC (I don't think that motherboard BIOS' will recognize the PVR-350 as a video card, though I may be wrong on that).
2. PVR-350 TV-Out at the moment isn't that great. Last I heard, it's just a framebuffer, so unless you're running specific apps that will use the acceleration, like mplayer, MythTV (for watching MPEG-2 recordings only, I think), and maybe Xine, it's extremely slow. And for applications that require some sort of GL support, well, forget it (people have had numerous troubles with the MythGame module/XMAME)..
3. Apparently there's some issues regarding using the PVR-350 to record, and using the TV-Out at the same time.
However, your advice is still sound. I bought a PVR-350, and at the moment, I'm using it as a PVR-250, using my GeForce 4MX for TV-Out. Once the ivtv driver stabilizes, I'll switch to the PVR-350. There's been a lot of traffic on the ivtv mailing list lately with patches for testing and the like.
-- Joe
Granted, I now have a PVR-250 and a PVR-350 in my MythTV machine, but I started out with dual BT878 capture cards (a Pinnacle PCTV Rave and a Hauppauge WinTV/Radio).
Although the picture quality of the PVR-x50 cards are better, I would argue that MythTV actually runs better on the BT8x8 cards, only because the ivtv driver is a bit unstable still.
The two main reasons why everybody raves about the PVR-250 cards (and thus tends to go with them) under MythTV are:
1. The graphics quality is SO much better. Even when you crank up the resolution on the BT8x8 cards to 720x480, there's still a noticable difference in picture quality (although, this could be due to the software codecs).
2. While giving this amazing picture quality, the bulk of the work is done on the PVR-x50 card. Thus, the main CPU remains free for other things.
I can speak from first hand experience, trying to do a dual-tuner system using BT8x8 cards really chews up CPU power. If you have the settings set up too high, and you end up recording two things at once (watching Live TV counts as a recording), then you'll end up dropping frames in the second recording (stuttering). The PVR-x50 eliminates all of these worries (assuming you have a motherboard that can handle the DMA traffic).
-- Joe
Hell yeah, I have to agree. I could have shelled out a total of $500-$600 for a TiVo with a lifetime subscription, but I like to do things the hard way, it seems. I was seriously considering the TiVo, because it would just work.
Then, I did some investigation into do-it-yourself PVR's, and stumbled onto MythTV and Freevo. I tried Freevo first, and didn't really like it (even before using it to watch TV), but the clincher was that, at that time, it didn't support PVR functionality (rewinding through the live TV stream).
I had (still have to some extent) a bunch of older hardware kicking around, so I decided to give MythTV a shot. The hardware was somewhat under-powered, but if I scaled down the capture resolution (to 320x480, for instance), I was able to get it to work. That was enough for me to start spec'ing out some new hardware and make purchases over a period of time.
It was definitely a learning experience in putting together a MythTV machine, helping find bugs, submitting a small code patch or two. I'd do it again, but that's because I'm usually up for a good challenge. If you want something that just works, seriously, buy a TiVo, that's what they do is make devices that work.
-- Joe
That is quite possible, and quite probable. I was at Sega only for the Dreamcast, and even then, was not the graphics guru in the department.
Most of my graphics experience comes from 2D stuff on the Commodore 64, where I'm used to having a 320x200 visible (main screen), but under NTSC (on a C64), you actually get somewhere around 260-270 raster scanlines to play around with (I don't remember exactly how many, it was in the documentation somewhere). Of course, most of those scanlines were "off-screen" (I believe "on-screen ranged from something like 50-249), so the extra scanlines weren't really useful for drawing anything (not for me, but there were people who did turn the borders off and things like that).
Usually what I used those extra scanlines for was updating the screen so that my scrolling messages wouldn't tear. Rather than double buffer, it was quite easy (and it presented no visual artifacts) if you updated screen memory after it was drawn by the raster. So if I was doing a 16 scanline split screen (one color per scanline), but I didn't trigger the VIC-II raster interrupt for another 40 scanlines below that, if I had a routine that would fit into the time it took to draw those extra 48 scanlines, I'd call it.
Of course, this timing is crucial, since it was using interrupts, and your ISRs must be fast, or you bog the whole system down.
-- Joe
You can see the same thing in the Commodore 64, or almost any system where the developers relied on the NTSC or PAL timings. I guarantee that demos that I wrote on the C64 (which were done in NTSC) don't run very well in PAL - all sorts of visual artifacts, because I timed them for NTSC (the Commodore 64 in particular had the same raster timing repeated every 8 pixels, if I recall correctly, though it has been a while).
:)
Generally, if you're doing video related work, it's easier to rely on the timing of the video, since you're generally shooting for good screen graphics (no tears, smooth lines, etc.).
The most common way that I'm familiar of for the NTSC game developers to make a PAL port of the title is to get the PAL version of the game system and a TV capable of PAL. From there, you insert some checks in the game, fix it up, and boom, you have an NTSC/PAL version.
Of course, you also have to account for the happiness of the developers too, because once you're used to NTSC, looking at PAL is maddening (and presumably the other way around) until you're used to it.
-- Joe
Almost every computer-like machine in existance has a clock in it. This clock isn't necessarily a clock like you'd use for viewing the current date/time, but is in some cases internal to the CPU.
:)
What units this clock runs in varies from chip to chip, but most of the time, the OS that you're using provides you with a decent way of using it, in some sort of standard measure of time (vxWorks with the BSPs that I've used, for example, provides you with 60ths of a second, which is very convenient).
This is very important, since especially in cable modems, there are a lot of events that need to be synchronized, and the CPUs change in clock speed (i.e. MHz) on a regular basis.
In terms of consoles, timing things to the VSYNC is generally popular. In order to properly do double buffering, you need to swap the buffers during the point of time when the raster isn't being shown within the visible screen, or at least in the section you're drawing (otherwise, you'll get tearing).
That, and it's a fixed timing, either 50Hz (PAL) or 60Hz (NTSC).
-- Joe
Ah. But see, there are very few providers of broadband out there (not that this is a good thing).
If the customer is really bone-headed and doesn't address the issue, they'll eventually run out of providers in their area, as they keep getting kicked off. Then they'll either have to fix the problem, or be without internet service.
Of course, the bad side of this is that cable providers might start sharing information with each other and creating blacklists. It has it's advantages and disadvantages.
Plus, as a (potential) customer myself, I'd probably sign up for an ISP that could provide me extra bandwidth because the wasters aren't on the network.
-- Joe
I think that it's completely fair to blame the user.
Warning... Obligatory analogy to cars coming up...
Users should understand how their computers work, and the consequences of using a computer. I'm not saying that they should know how everything works down to the circuit level, but they should understand that a computer will do whatever a program that is running on it tells the machine to do. They should understand that without proper precautions, other people can "hi-jack" their machine.
I'm a tech guy, but honestly, I don't know completely how my car works. I know some of the basics, like what the buttons on the consoles do, how to check my fluid levels, and the sound that my engine makes. Granted, I can't change my own oil or brakes, but I do know who I can take the car to in order to get that done, and I do have the common sense to listen for strange sounds. If I have questions, I find the answer (whether that means reading a book, searching on Google, or talking to a mechanic).
At the same time, I also understand that while my car can get me from point A to point B, I also understand that it can be an instrument of destruction if I don't take proper precautions. If I don't have the brakes serviced regularly, I can hit something causing damage (or death). If I don't top-up/change the oil on a regular basis, the car's engine can seize up.
Oh, and if Windows Update doesn't work (funny though, it seems to work for me on the machine I have here at work, and my desktop at home), then maybe people should be bitching to Microsoft to fix it? And if Microsoft won't fix it, then maybe don't buy their products, use something else?
Computers (at the moment anyway) are not simple little tools that play games, let you send e-mails or surf the Internet. They are very complex machines that can do those things, and more. It's the "more" part that you have to realize is there. If you don't understand the "more" part, then you should be talking to people who do, even if it means buying a book, buying some software, or paying a professional for their time. Just like how I bought a repair manual for my car, and I pay a mechanic to change the brakes or the oil.
-- Joe
If your modem is an older one, it is quite possible that you were able to do this. That's why a lot of the hacks telling you how to do this mention that you need a specific version of the firmware.
The cause for it was that you would poison the ARP cache, and when the TFTP request went out, the modem had to then ARP for the TFTP server. If you timed it carefully enough (and your TFTP server matched the IP address of the real TFTP server), the request over Ethernet would typically come back before the request over the cable interface (your PC is right next to you, the cable operator's TFTP server is many hops away).
Simple fix: Newer firmwares just don't accept any TFTP traffic over the ethernet port during those critical stages (as the simplest, hackiest fix).
-- Joe
That is just f**kin' crazy.
Granted, I like it that you're given options, but for the Upper House, that is just way too many choices. In my opinion, it seems that they're just asking for confusion.
-- Joe
That is certainly an excellent point.
;) ), but if I was faced with choosing from 100+ options, I'd be a little daunted too.
From what I've seen, it's somewhat difficult to get people out to vote in the United States. It's not necessarily that they don't care, but rather, as I've seen, it's more because even though it's law that your employer must give you time off to vote, it's still somewhat inconvenient to get to the polling place and actually cast your vote.
So, I guess with that in mind (among other things), they decide to cram everything that they can on the ballot. Heck, I'm a fairly patient and reasonable person (well, most of the time
-- Joe
It's been a while since I've lived and voted in Canada, but when you get your ballot, it comes with an instruction card telling you to mark a huge X inside the box of the candidate that you wish to vote for. It even shows you a graphical example.
In terms of how the votes are counted, I think that if you marked your ballot incorrectly, it's just counted as spoiled (so no vote for any candidate, nor do they try to figure out who you voted for). Quite frankly, I like this system, if you can't properly fill out the ballot, your vote shouldn't count, and there shouldn't be a guess as to what you meant.
I think that part of the simplicity of Canada's voting system is that there are usually at most six candiates on the ballot, since you only vote for the representative of your riding.
Read the instructions that come with the ballot If you run into a problem, ask for assistance, that's what the volunteers are there for.
-- Joe