Thanks for both of these. I've looked them over and will read them in detail later. The anandtech article covers power only on page 7 (the one linked) I think, and I was pretty disappointed to see the ratio of idle power to full power. More than 60% it seems. That doesn't bode well for my plans to have an HTPC that can stay on all the time. I'd like it to use 5% of the power when it isn't doing anything.
The xbitlabs story is longer and I couldn't find exact numbers I was looking for during a lunch break.
It appears labs compare processors by running a given program and comparing watts (which works for some things like decoding an HDTV stream), but I'd also like to see numbers from a given task and energy measurements (e.g. compressing a video) over the period of time needed. I'll try looking for that too.
I agree - I'm very concerned about idling power. If the idling power can be brought low enough on an entire system, say 5 W, I could leave it on all the time and not bother with trying to get nvram wakeup schemes to work so an HTPC can start up and record a show. I'm typing this on a Windows Centrino laptop now and it annoys the hell out of me how hot the thing gets when I'm doing nothing but scrolling through a document.
This brings up the general issue - why aren't reviewers covering all power efficiency numbers given that is what AMD and Intel are both claiming they are concerned about lately. I didn't read the entire article linked here, but I didn't see any numbers referring to "number of Joules required to transcode a file" or something. I can get lumens per watt for all types of lighting, hopefully computing devices advertising will improve in the future.
Yes, a reference please. I am highly skeptical that a good study shows that between cars of equal weight, the car with more power has fewer accidents. I would think (on the average) more yahoos are buying the car with more power and driving recklessly and that this will outweigh the event of being able to accelerate away from an accident. (Not that I'm saying this a valid reason to raise insurance rates on these cars - drivers shouldn't be presumed guilty before an accident.) But I'll drop this opinion if I see the study.
I drive a car (2005 Prius) with a power to weight ratio of 62 W/kg. It seems perfectly adequate (though probably anemic to drivers used to EV1s).
I don't yet own a (U.S.) cell phone, but I've been looking at some of the Linux smartphones as they come out. I was really hoping I could eventually get one that was like a phone attached to a general purpose computer (possibly with a small hard drive - or maybe flash will be good enough by then). In that case, I could install whatever mp3 player I want and get the phone to follow exactly the protocol I want when the call comes in (e.g., lower the volume, tell me who is calling before I pick up,...). Is this unrealistic? Does anyone have a feeling on who the first U.S. provider to offer service on this type of Linux phone will be? Anyone outside of the U.S. doing something similar now?
Though it doesn't match matter to me - I'd be only using the free service. If that service can accomplish an RMS accuracy of 5 m under moderately dense forest canopy (or if it can work at all under very dense canopy), I will definitely be looking forward to an upgrade to my Garmin etrex. Heck, maybe by then I can get something that runs Linux even - either a smart phone or a more general pocket computer.
Dara
Disclosure: I work at one of the aerospace companies involved with GPS III.
This is the best short reply I have read on Slashdot in years. I can't believe it is only at +4 right now. I will only add one thing to yog's argument:
Sun sells StarOffice for support. Sun writes most of the code for OpenOffice and this code is used in StarOffice. Obviously Sun is in exactly the same place as Microsoft in terms of wanting to minimize the number of support calls.
Also a 100 full-time Sun employees is nothing to sneeze at. I don't need a more complicated OpenOffice for 3.0, I'm happy if most of the improvement (for a while anyway) is in speed and fewer bugs. 100 employees can do a lot of code optimization.
Dara - I wish Sun had gone with SunOffice for their version and left the StarOffice name for the open source version -
I was interested to see this link since I've had in mind for a while a project to make a very low rate two way data messaging pair of radios that I could use while rock climbing. I find the existing solutions of voice radios suboptimal and sometimes they don't work even with a separation of 50 m if your line of sight has a bunch of rock in the way. I was thinking a bit rate on the order of 100 to 1000 bps would be plenty and would allow for such a large Eb/No that I could probably close a link even with a huge path attenuation using a single double AA battery device. I would definitely want to use an unlicensed band, and I imagine the lower frequency the better to curve around the rock somewhat.
I read the Wikipedia article, and I see the lowest rate/lowest freq solution of 20 kbit/s in the 868 MHz band isn't that far off from what I was thinking of. Does anyone think this protocol is a good choice for a low power 2 way messaging project? Or is there too much overhead from trying to solve something more complicated (beacons and other tricks).
The HD-3000 card was also designed and tested for Linux. The company managed to get code into the kernel so things are supposed to be pretty transparent, but nobody has written a recent HOWTO (I'm waiting for Jarod's, but anything would do). My point is that if e.g. FreeBSD and Linux swapped places in terms of popularity, the card would be designed for FreeBSD and there would be more likelihood of finding documentation for the current driver code, since BSD is more consolidated than Linux (I'm guessing most users of something like MythTV would be on FreeBSD as opposed to NetBSD or OpenBSD).
Congrats on getting your HDTV MythTV system to work. The very fact that you did means there are a set number of commands to make it so. Clearly with less variance on the OS choices, you could just publish those commands on the web and everyone could get theirs working too right?
As far as my choice of distro, I am more comfortable with Fedora than Ubuntu, Debian, KnoppMyth, Mepis, etc. only because I started with Redhat 5 or so. But I will switch to whatever I can get working first. What distro did you get the card working with?
Yes, I've played with KnoppMyth, and it got me to a MythTV config screen faster, but it doesn't have a 2.6.12 kernel yet (I don't think) and it still wasn't trivial to get my card working (and so I didn't).
Contrast this to a MyHD card I played with under Windows 2000 (not media center edition). It was a bit buggy, but everything worked more or less out of the box and the documentation didn't need any (that I remember) disclaimers about which Windows I was using.
Furthermore, I only have one Linux PC and I'd like to use it for everything - recording TV, running Octave to play with Matlab scripts I write at work, surfing the web, etc. I don't want to use a specialized distro unless I have to (and I might have to - I'm keeping my eye out for R5A17 to see if KnoppMyth makes me life a lot easier).
I'm now struggling with a MythTV, HD-3000, FC4 installation, and I absolutely believe it would be easier if more people used a common platform to get this kind of thing working. I look at MythTV documentation and it says stuff like this information is only valid if you have compiled from source. So I can't even rely on the documentation because I use smart to grab the rpms from ATRPMS? Do they think detailed information exists for each distro? This kind of problem simply does not happen under Windows or MacOS and as much as I'd prefer to use Linux, I have a much easier time getting most things to work under Windows (I'll experiment with MacOS this fall when I have access to a new iMac). I won't give up, but this issue is definitely holding Linux back, not that I have any idea what can be done about it.
Yes, the range of technology one can imagine will influence their estimate of carrying capacity. I haven't yet read anyone I considered reasonable suggest we will be able to achieve a reasonable lifestyle for > 20 billion people, but perhaps you are right. I'd prefer not to keep going in that direction to find out though.
I read some of O'Neil's work, and I'm familiar with the concept of rail guns, but it was a long time ago. When you say the earth can be completely depopulated in a few centuries, are you calculating on a basis of 10 billion or 10 billion * 200/70? I'll have to look up that calculation for future reference - do you have a link for it? Where will these people go by the way - orbiting space stations (where's the material for that come from?), or other planets (just how many people could an ideally terra-formed Mars hold?)
I have no way of knowing if longevity will be significantly expanded (say from 75 years to 750 years). I know people are doing research on this. That wouldn't change the steady state fertility rate obviously if people did eventually die, but it would cause a factor of 10 increase in population until the steady state condition is reached again.
My point 4 (governments effect on fertility rates) is arguable - I only have somewhat anecdotal information. But if a conservative religious society like Iran can make such a strong turnaround, it seems like it could be achievable in many societies. The Technology Review author even claimed that Mormon birth rates are down, but I don't know about that. Do you also disagree about the other direction - that it is possible for governments to successfully encourage fertility?
I guess I don't have much to say about the racism issue, but I am concerned about my country in particular so I am "country" centric. Even in my lifetime (42 years), I've noticed the public lands getting more crowded. For a while pollution in Los Angeles got better due to better technology, but since I got here, it's getting worse again now that population has blown out that advantage.
Population is the most important issue in politics for me, so I read the section on this topic (but skipped the rest). I'm so tired of the descriptions of "doom and gloom" that will happen with low fertility rates and a shrinking population - these authors are a mirror image of the mistakes they claim that past environmental authors have made in predicting the future.
There are some scientific facts on population that are rarely disputed:
1] The earth has a finite carrying capacity
Actual numbers will vary anywhere from 1 to 10 billion people, but it's obvious that constraints on food, water, energy, pollution sinks do constrain the number of us. My opinion is that the number is less than we are now, but we are getting by (some of us anyway) because of unsustainable oil and water use. Perhaps we could get by on renewable energy with around 2 billion people.
2] Large numbers of humans cannot leave the earth
There is no way we could move even 1/1000th the world population off the earth even if there was someplace to go. The resources/pollution needed to do this make it a non-starter for addressing population growth.
3] Adjustments need to be made to run an economy with a declining population growth
Not impossible, but obviously it is harder to operate a system that is shrinking instead of growing. Tricks like using lots of workers to support fewer retirees won't work. Any pyramid scheme seems great when you are on the growth side, but I'd prefer not to have the human race crash like a big pyramid scheme.
4] Fertility rates can be adjusted by government action
Coercive measures while espoused by some as necessary have been avoided in very successful transitions to lower fertility (e.g. Iran). We have less experience with going the other way, but some countries (e.g. Singapore) are trying incentives to raise the fertility rate. I see no reason that these rates can't be successfully adjusted if for some reason, 50 years from now, the world wide fertility rate dips down well below 2 and stays there so long that our population goes below 2 billion.
Now, back to the article:
In each country listed: Japan, Germany, Spain, Russia (I think) and Italy, they could stand to lose 30% of their population anyway. I think the U.S. is too crowded and Europe has much higher densities (and Japan is worse) in terms of population per arable land unit.
"It turns out that population decrease accelerates downward just as fiercely as population increase accelerated upward, for the same reason."
What does this mean? If you measure the increase or decrease of an exponential function (what he's talking abut here) as a percentage, then of course they have the same fierceness, but there is no concept of acceleration (percentage growth is constant). If you measure the amount in absolute numbers, then exponential increase is accelerating, but exponential decrease is always decelerating.
As far as fertility going down everywhere, we in the U.S. are now at 2.08 and this is going up (albeit slowly). We were closer to 2 about 5 years ago I think. If you look at http://www.cia.gov/cia/publications/factbook/ranko rder/2127rank.html, you will see there are still quite a few countries that have fertility rates above 2.1. (By the way, saying 2.1 is steady state assumes an average infant mortality rate that is pretty high. If you want the human race to all move into a the modern industrialized world, something under 2.05 is required). Granted, I don't have the plots of all countries fertility rates over time and some of these countries near the top may be declining, but I see absolutely no way we can declare success now. I expected better out of Technology Review, the magazine where I first learned about fuel cells for automotive use.
Unfortunately, the address book in Thunderbird is still very primitive compared to Outlook, Evolution, or Kmail. You can't use pictures, Geo locations, there is no place for Birthday, Children, etc. - only the four custom fields and Notes. There are only two email addresses that can be stored. How in the world could it import information from Outlook without losing it?
In response to another poster, I'm sorry, but.csv is not good enough for an address book anymore. There are too many fields (many of which won't be used for all entries) so viewing your book as a spreadsheet becomes tedious. Plus, there is no way to store picture information in a.csv file.
I'd prefer to use Thunderbird (or Mozilla) over Evolution or Kmail since I use both Windows and Linux, but I wish the three would get together and hammer at a way to have a really powerful address book standard, leaving no vCard info behind and storing any other info not included in vCard (e.g. Last Sort View State). And of course leaving no useful Outlook info behind either.
My experience with OpenOffice is almost all on Windows. Occasionally some interface item will bug me, but I can accept that I'm just used to doing it the MS Office way. What is most disappointing is the speed.
Speed to start, open a file, save a file, and perform certain operations is painfully slow compared to Office. I've played with the 1.9.51 branch a bit, and it doesn't seem 2.0 is going to be enough of an improvement to compete with Microsoft on the speed front.
I used to think that Moore's law will take over, but I'm now using a brand new P2.8 with 1 Gig of RAM at work, and after editing a presentation file with some large images I couldn't edit a slide with only text (don't ask me what OpenOffice was doing in the background with those pictures - it couldn't be autosave, since the problem was constant). I also used to think that OpenOffice should keep adding new features (e.g., macro recorder, which is in 1.9.51), but now I wish they would just optimize the hell out of it and add no new features for a while.
Perhaps it doesn't feel as slow on Solaris or Linux, but I doubt it - my Linux machine is pretty anemic, but it used to run Office reasonably when it had Windows on it. Now I don't even try to use OpenOffice on it as it is unbearable. When Koffice becomes file compatible, I may try to use that program on this machine.
The two free cross-platform software projects I use most are OpenOffice and Mozilla (Seamonkey or Firefox). Of course Mozilla's task is completely different, but it works reasonably fast compared to Internet Explorer (faster with some tasks, slower with others). I look forward to the day I can say the same thing about OpenOffice.
The GPS Lite mouse from deluo.com is $60. I have an older mouse from Deluo, and it works reasonably well under Windows. I've been told you can get the old one to work on Linux as well, but then there is no routing software available (let alone turn by turn voice guidance).
If you don't need turn by turn, and have a Windows laptop, you could also get a package with a GPS mouse + Streets and Trips 2005 for $85 (after rebate on amazon).
First of all, nobody is talking about doubling the existing DPI (dots per inch). Moving from 1/2 VGA to VGA is an increase in DPI of sqrt(2) not 2.
Secondly, you said it yourself, you usually look a lot closer (from my rough estimate right now), about twice as close. I agree, 15" 1600x1200 (i.e., 133 DPI) looks pretty good from a 20" laptop viewing distance. Therefore 266 DPI is about what we need for a PDA viewing at 10". VGA has this DPI at a 3" diagonal which is pretty small, (the T5 has a 3.8" screen, so the DPI if it had VGA would only be 210). To really get the resolution of the monitor you suggest from twice as close on a 3.8" screen will require 800x600 (SVGA).
We will definitely get PDA SVGA screens in the future, and we will probably have a lot more comments like yours that it is overkill. I may not need SVGA, but I'm looking forward to it. I do want VGA though, and I won't be getting a Palm T5 because it doesn't have it. (I can't get the Zaurus either since I want GPS software with voice instructions - I'll probably end up with an ASUS or Dell VGA PocketPC.)
I'm definitely getting a VGA PDA, maybe the HP, but more likely the Asus (also available now) or Dell (when it is available). I would have already purchased a Zaurus by now except for one thing - there is no GPS software for the Zaurus that does routing and voice instructions in the U.S. (when I last checked, I can't imagine this has changed). So I'm limited to Palm and Pocket PC, and as others have pointed out, Palm hasn't announced any VGA products yet.
I wrote deluo.com and asked if Routis would display optimally on a VGA PDA and they said no and didn't know when that would change. Does anyone else know of GPS software that works correctly on a VGA PDA? I've looked at Dale's excellent site (http://www.gpsinformation.org/dale/PocketPC/wince.htm), but the reviews predate the development of VGA.
By the way, GPS is one answer for what a PDA is good for. I've used a laptop w/ GPS in a car before - when you're the passenger, it's OK, but when you're the driver it sucks. You're worried its going to slide off the seat, you have to look 90 deg away from the road to see it, the screens can't handle sunlight, and the interface can't use large touch screen buttons. Phone screens are too small to be a good GPS - even a 4" PDA screen is a bit small, but I guess a 6" device is too big for most people. I guess I could also consider a dedicated GPS (maybe there exist VGA ones), but if I'm spending that kind of money, I want to use it for other things too - like a music player that can turn down the volume while giving voice instructions.
I'll consider an Apple iMac (or laptop for that matter) when they come out with the same resolution as the 23" Cinema Display (1920x1200, commonly called WUXGA). Then I can watch HDTV natively (there is a stand alone box option for Apple now) and I'll have more pixels for showing digital photos.
Dara
Can arrow key history be like Matlab's?
on
Bash 3.0 Released
·
· Score: 3, Interesting
I read the announcement and it mentions "History traversal with arrow keys", but what I would really like doesn't seem to be mentioned (but perhaps it is possible with bash-2.05, I'm not much of a shell expert). In Matlab, the up-arrow key searches the history for commands that match all the characters on the line. No characters and it acts like a normal bash arrow, if "figure, plot" is at the beginning of the line, it will quickly scroll through all plotting commands that have been entered at the shell.
What should be possible is to state that you only have CDs 1, 1-2, 1-3, or 1-4 during package selection time and the installer would only offer you the choices that are available.
I, like others here, would prefer to do a reasonable workstation installation with just CD 1 (currently disks 1 and 2 are required). And links to torrents for individual CDs should be more prominent than they are now.
Mandrake did something like this quite a while ago and it worked fine.
Dara Parsavand
Re:Don't ticket me - control my car's max speed
on
Road Marker Marks You
·
· Score: 1
Thanks for all the replies, there were some very interesting ones. I'll try to respond to some of the them, but I'll do it all in one post. I hope this format isn't too confusing for anyone still interested in this subtopic.
To everyone complaining about the temporary need to speed (for passing or avoiding an accident): as heytherefancypants said, it isn't hard to imagine a solution. I drive mostly on crowded 4 lane roads or highways and if I'm doing near the speed limit, there is almost never a possibility for me too speed and avoid an accident (since there is usually someone in front of me - L.A. traffic, ugh). And I rarely have the passing problem (the one 2 lane road I drive is so windy there is almost no passing allowed). But sure, allow temporary speeding for both reasons.
To sickmtbnutcase: I guess I'm willing to have an override for someone who has a medical problem in the car (and set up a comm link between the car and the hospital, route the best path, etc.), but not for any other reason. People will have to adjust, just like you do when you use transit or if there is a traffic jam. I wouldn't want someone racing to see me as I was dying - they are probably stressed and in even more chance of getting in an accident. As far as enjoying driving goes, we'll all have to learn how to do that with nimble handling cars and curvy roads I guess. My car (2001 Prius) has barley adequate acceleration, but it's still fun to drive in the mountains (handles way better than my old Isuzu Trooper did).
to blunte: Interesting points on the significant reasons for traffic flow problems, I agree with all of them. I don't know if any of them negate my point though. It's true that if all cars were going the same speed (assume there a level highway and no reason to go slower than the speed limit), the only way to change lanes is to put on your turn signal and drop behind the car next to you. If the highway is absolutely at capacity and everyone is rude, it's going to be hard to do this (but it's hard if you speed up too).
to 87C751: No I'm not familiar with the exact revenue stream of small towns. I do realize speeding tickets are revenue which my post addressed. Those towns better figure start thinking of other solutions because I have no doubt in the long run there will be no traffic tickets (hopefully because all the cars will be fully automated - or maybe because energy is so expensive and everywhere is so populated, we'll all be using mass transit)
to gnu-generation-one: The actual scheme of reading the speed limit from the road is possible a number of different ways: paint stripes, magnets under the pavement, RF sign posts, accurate GPS, etc. I really don't know what the best way would be.
To YrWrstNtmr: your point is well taken (repeated below)
Situation A) Speed limit of 70. Cars cannot exceed. Most drivers want to exceed. Many drivers pissed off. Situation B) Speed limit of 70. Enough drivers complain, and the limit is raised to 80. Not all drivers will go 80, leaving a disparity of speeds. Quite similar to what we have now.
I guess my reply is that 75 mph (120 kph) seems fast enough for most places that get any significant traffic around where I am (L.A.). I guess Montana can go for 85 mph or so if it wants to, and if they don't have much traffic, they won't have a big problem with speed disparity. I doubt many people would actively campaign here for more than 75 on any of our freeways, but you're right, there will be some that don't want to (or can't) do any more than 65 or so, so there will always be somewhat of a disparity issue until we all get automated cars with enough power to always do the limit.
Dara Parsavand
Don't ticket me - control my car's max speed
on
Road Marker Marks You
·
· Score: 5, Insightful
As I say every time this subject comes up, I'd much rather have my car know the max speed on a given road for a given set of conditions and not be allowed to go over the max speed, than I want fancy electronics to check to see if I go over the max speed, and if I do, take my picture, and send me a ticket. I'd rather pay higher taxes than fund police through tickets (and we wouldn't need as much traffic police either if the cars were smarter).
I claim that if no one could go over the speed limit, traffic would flow much more smoothly, and if the limit is too low (because you are expected to speed 10 mph), we will all complain loudly enough to get it changed.
Other aspects of this project sound interesting though.
I've only played with Gimp and Photoshop Elements briefly so far, but I get the feeling the Gimp is designed to do a whole lot more than touch up digital photos (crop, rotate, red-eye elimination, etc.), which is all I want to do.
Are there any open source alternatives (preferably cross-platform) that do this job just as well as the Gimp but are simpler to use?
What is with the restriction on filenames (in the migration checklist)?
I have not made the transition to Linux (and I'm not planning to for a while longer), but I have a lot of files named with upper and lower case and spaces, and when I play on Linux, I don't have any problem with them. I am not changing any filename I don't have to (unless there is a character allowed by Windows which is not allowed by Unix - probably true - but I couldn't find a reference). I'm not aware of any reason not to have spaces in Unix. The command line still works using quotes and/or tab completion and any modern file manager is going to work fine too. If a certain application doesn't work opening the file, I can find that out later and rename the file then.
This kind of advice is just hindering people from switching by putting another obstacle in front of them. What for?
I use MS Office a fair amount at work (some PowerPoint animation, some Excel VB, nothing too complicated in Word - but I like the Outline mode).
I would prefer to switch to OpenOffice eventually (though I doubt I will have any luck convincing my company), but I don't want to until the speed of opening and saving files is improved (the current development versions of 2.0 are nowhere near fast enough yet, but maybe later).
I have never had a problem with quick save in MS Office, and when I work with files that a bigger than a few megs, it sure is nice to be able to open and close them in just a few seconds.
The speed issue and some way to partially convert VB code to OO Basic (or python or whatever works with OO.o) is all I need.
Slashdot Mirror
I wasn't browsing low enough and I missed an AC post before mine giving a good link (http://www.xbitlabs.com/articles/cpu/display/amd- energy-efficient.html). There was also one after my post (http://www.anandtech.com/cpuchipsets/showdoc.aspx ?i=2795&p=7).
Thanks for both of these. I've looked them over and will read them in detail later. The anandtech article covers power only on page 7 (the one linked) I think, and I was pretty disappointed to see the ratio of idle power to full power. More than 60% it seems. That doesn't bode well for my plans to have an HTPC that can stay on all the time. I'd like it to use 5% of the power when it isn't doing anything.
The xbitlabs story is longer and I couldn't find exact numbers I was looking for during a lunch break.
It appears labs compare processors by running a given program and comparing watts (which works for some things like decoding an HDTV stream), but I'd also like to see numbers from a given task and energy measurements (e.g. compressing a video) over the period of time needed. I'll try looking for that too.
Dara
I agree - I'm very concerned about idling power. If the idling power can be brought low enough on an entire system, say 5 W, I could leave it on all the time and not bother with trying to get nvram wakeup schemes to work so an HTPC can start up and record a show. I'm typing this on a Windows Centrino laptop now and it annoys the hell out of me how hot the thing gets when I'm doing nothing but scrolling through a document.
This brings up the general issue - why aren't reviewers covering all power efficiency numbers given that is what AMD and Intel are both claiming they are concerned about lately. I didn't read the entire article linked here, but I didn't see any numbers referring to "number of Joules required to transcode a file" or something. I can get lumens per watt for all types of lighting, hopefully computing devices advertising will improve in the future.
Dara
Yes, a reference please. I am highly skeptical that a good study shows that between cars of equal weight, the car with more power has fewer accidents. I would think (on the average) more yahoos are buying the car with more power and driving recklessly and that this will outweigh the event of being able to accelerate away from an accident. (Not that I'm saying this a valid reason to raise insurance rates on these cars - drivers shouldn't be presumed guilty before an accident.) But I'll drop this opinion if I see the study.
I drive a car (2005 Prius) with a power to weight ratio of 62 W/kg. It seems perfectly adequate (though probably anemic to drivers used to EV1s).
Dara
I don't yet own a (U.S.) cell phone, but I've been looking at some of the Linux smartphones as they come out. I was really hoping I could eventually get one that was like a phone attached to a general purpose computer (possibly with a small hard drive - or maybe flash will be good enough by then). In that case, I could install whatever mp3 player I want and get the phone to follow exactly the protocol I want when the call comes in (e.g., lower the volume, tell me who is calling before I pick up, ...). Is this unrealistic? Does anyone have a feeling on who the first U.S. provider to offer service on this type of Linux phone will be? Anyone outside of the U.S. doing something similar now?
Dara
Dun Malg wrote that "high accuracy at the cm scale" has a specific meaning, and that is "accuracy at a scale of less than one meter".
l ileo_The_European_satellite_navigation_system.asp) which seems to corroborate the definition above. I have to admit that when I first read "at the cm scale", I thought of an accuracy of better than 10 cm.
I read the BBC article and poked around on the net to find the accuracy potential of Galileo. So far, I've found a figure of 10-45 cm (http://www.eurescom.de/message/messageMar2005/Ga
Though it doesn't match matter to me - I'd be only using the free service. If that service can accomplish an RMS accuracy of 5 m under moderately dense forest canopy (or if it can work at all under very dense canopy), I will definitely be looking forward to an upgrade to my Garmin etrex. Heck, maybe by then I can get something that runs Linux even - either a smart phone or a more general pocket computer.
Dara
Disclosure: I work at one of the aerospace companies involved with GPS III.
This is the best short reply I have read on Slashdot in years. I can't believe it is only at +4 right now. I will only add one thing to yog's argument:
Sun sells StarOffice for support. Sun writes most of the code for OpenOffice and this code is used in StarOffice. Obviously Sun is in exactly the same place as Microsoft in terms of wanting to minimize the number of support calls.
Also a 100 full-time Sun employees is nothing to sneeze at. I don't need a more complicated OpenOffice for 3.0, I'm happy if most of the improvement (for a while anyway) is in speed and fewer bugs. 100 employees can do a lot of code optimization.
Dara
- I wish Sun had gone with SunOffice for their version and left the StarOffice name for the open source version -
I was interested to see this link since I've had in mind for a while a project to make a very low rate two way data messaging pair of radios that I could use while rock climbing. I find the existing solutions of voice radios suboptimal and sometimes they don't work even with a separation of 50 m if your line of sight has a bunch of rock in the way. I was thinking a bit rate on the order of 100 to 1000 bps would be plenty and would allow for such a large Eb/No that I could probably close a link even with a huge path attenuation using a single double AA battery device. I would definitely want to use an unlicensed band, and I imagine the lower frequency the better to curve around the rock somewhat.
I read the Wikipedia article, and I see the lowest rate/lowest freq solution of 20 kbit/s in the 868 MHz band isn't that far off from what I was thinking of. Does anyone think this protocol is a good choice for a low power 2 way messaging project? Or is there too much overhead from trying to solve something more complicated (beacons and other tricks).
Dara Parsavand
The HD-3000 card was also designed and tested for Linux. The company managed to get code into the kernel so things are supposed to be pretty transparent, but nobody has written a recent HOWTO (I'm waiting for Jarod's, but anything would do). My point is that if e.g. FreeBSD and Linux swapped places in terms of popularity, the card would be designed for FreeBSD and there would be more likelihood of finding documentation for the current driver code, since BSD is more consolidated than Linux (I'm guessing most users of something like MythTV would be on FreeBSD as opposed to NetBSD or OpenBSD).
Congrats on getting your HDTV MythTV system to work. The very fact that you did means there are a set number of commands to make it so. Clearly with less variance on the OS choices, you could just publish those commands on the web and everyone could get theirs working too right?
As far as my choice of distro, I am more comfortable with Fedora than Ubuntu, Debian, KnoppMyth, Mepis, etc. only because I started with Redhat 5 or so. But I will switch to whatever I can get working first. What distro did you get the card working with?
Dara
Yes, I've played with KnoppMyth, and it got me to a MythTV config screen faster, but it doesn't have a 2.6.12 kernel yet (I don't think) and it still wasn't trivial to get my card working (and so I didn't).
Contrast this to a MyHD card I played with under Windows 2000 (not media center edition). It was a bit buggy, but everything worked more or less out of the box and the documentation didn't need any (that I remember) disclaimers about which Windows I was using.
Furthermore, I only have one Linux PC and I'd like to use it for everything - recording TV, running Octave to play with Matlab scripts I write at work, surfing the web, etc. I don't want to use a specialized distro unless I have to (and I might have to - I'm keeping my eye out for R5A17 to see if KnoppMyth makes me life a lot easier).
Dara
I'm now struggling with a MythTV, HD-3000, FC4 installation, and I absolutely believe it would be easier if more people used a common platform to get this kind of thing working. I look at MythTV documentation and it says stuff like this information is only valid if you have compiled from source. So I can't even rely on the documentation because I use smart to grab the rpms from ATRPMS? Do they think detailed information exists for each distro? This kind of problem simply does not happen under Windows or MacOS and as much as I'd prefer to use Linux, I have a much easier time getting most things to work under Windows (I'll experiment with MacOS this fall when I have access to a new iMac). I won't give up, but this issue is definitely holding Linux back, not that I have any idea what can be done about it.
Dara
Yes, the range of technology one can imagine will influence their estimate of carrying capacity. I haven't yet read anyone I considered reasonable suggest we will be able to achieve a reasonable lifestyle for > 20 billion people, but perhaps you are right. I'd prefer not to keep going in that direction to find out though.
I read some of O'Neil's work, and I'm familiar with the concept of rail guns, but it was a long time ago. When you say the earth can be completely depopulated in a few centuries, are you calculating on a basis of 10 billion or 10 billion * 200/70? I'll have to look up that calculation for future reference - do you have a link for it? Where will these people go by the way - orbiting space stations (where's the material for that come from?), or other planets (just how many people could an ideally terra-formed Mars hold?)
I have no way of knowing if longevity will be significantly expanded (say from 75 years to 750 years). I know people are doing research on this. That wouldn't change the steady state fertility rate obviously if people did eventually die, but it would cause a factor of 10 increase in population until the steady state condition is reached again.
My point 4 (governments effect on fertility rates) is arguable - I only have somewhat anecdotal information. But if a conservative religious society like Iran can make such a strong turnaround, it seems like it could be achievable in many societies. The Technology Review author even claimed that Mormon birth rates are down, but I don't know about that. Do you also disagree about the other direction - that it is possible for governments to successfully encourage fertility?
I guess I don't have much to say about the racism issue, but I am concerned about my country in particular so I am "country" centric. Even in my lifetime (42 years), I've noticed the public lands getting more crowded. For a while pollution in Los Angeles got better due to better technology, but since I got here, it's getting worse again now that population has blown out that advantage.
Dara
Population is the most important issue in politics for me, so I read the section on this topic (but skipped the rest). I'm so tired of the descriptions of "doom and gloom" that will happen with low fertility rates and a shrinking population - these authors are a mirror image of the mistakes they claim that past environmental authors have made in predicting the future.
o rder/2127rank.html, you will see there are still quite a few countries that have fertility rates above 2.1. (By the way, saying 2.1 is steady state assumes an average infant mortality rate that is pretty high. If you want the human race to all move into a the modern industrialized world, something under 2.05 is required). Granted, I don't have the plots of all countries fertility rates over time and some of these countries near the top may be declining, but I see absolutely no way we can declare success now. I expected better out of Technology Review, the magazine where I first learned about fuel cells for automotive use.
There are some scientific facts on population that are rarely disputed:
1] The earth has a finite carrying capacity
Actual numbers will vary anywhere from 1 to 10 billion people, but it's obvious that constraints on food, water, energy, pollution sinks do constrain the number of us. My opinion is that the number is less than we are now, but we are getting by (some of us anyway) because of unsustainable oil and water use. Perhaps we could get by on renewable energy with around 2 billion people.
2] Large numbers of humans cannot leave the earth
There is no way we could move even 1/1000th the world population off the earth even if there was someplace to go. The resources/pollution needed to do this make it a non-starter for addressing population growth.
3] Adjustments need to be made to run an economy with a declining population growth
Not impossible, but obviously it is harder to operate a system that is shrinking instead of growing. Tricks like using lots of workers to support fewer retirees won't work. Any pyramid scheme seems great when you are on the growth side, but I'd prefer not to have the human race crash like a big pyramid scheme.
4] Fertility rates can be adjusted by government action
Coercive measures while espoused by some as necessary have been avoided in very successful transitions to lower fertility (e.g. Iran). We have less experience with going the other way, but some countries (e.g. Singapore) are trying incentives to raise the fertility rate. I see no reason that these rates can't be successfully adjusted if for some reason, 50 years from now, the world wide fertility rate dips down well below 2 and stays there so long that our population goes below 2 billion.
Now, back to the article:
In each country listed: Japan, Germany, Spain, Russia (I think) and Italy, they could stand to lose 30% of their population anyway. I think the U.S. is too crowded and Europe has much higher densities (and Japan is worse) in terms of population per arable land unit.
"It turns out that population decrease accelerates downward just as fiercely as population increase accelerated upward, for the same reason."
What does this mean? If you measure the increase or decrease of an exponential function (what he's talking abut here) as a percentage, then of course they have the same fierceness, but there is no concept of acceleration (percentage growth is constant). If you measure the amount in absolute numbers, then exponential increase is accelerating, but exponential decrease is always decelerating.
As far as fertility going down everywhere, we in the U.S. are now at 2.08 and this is going up (albeit slowly). We were closer to 2 about 5 years ago I think. If you look at http://www.cia.gov/cia/publications/factbook/rank
Dara
Unfortunately, the address book in Thunderbird is still very primitive compared to Outlook, Evolution, or Kmail. You can't use pictures, Geo locations, there is no place for Birthday, Children, etc. - only the four custom fields and Notes. There are only two email addresses that can be stored. How in the world could it import information from Outlook without losing it?
.csv is not good enough for an address book anymore. There are too many fields (many of which won't be used for all entries) so viewing your book as a spreadsheet becomes tedious. Plus, there is no way to store picture information in a .csv file.
In response to another poster, I'm sorry, but
I'd prefer to use Thunderbird (or Mozilla) over Evolution or Kmail since I use both Windows and Linux, but I wish the three would get together and hammer at a way to have a really powerful address book standard, leaving no vCard info behind and storing any other info not included in vCard (e.g. Last Sort View State). And of course leaving no useful Outlook info behind either.
Dara
My experience with OpenOffice is almost all on Windows. Occasionally some interface item will bug me, but I can accept that I'm just used to doing it the MS Office way. What is most disappointing is the speed.
Speed to start, open a file, save a file, and perform certain operations is painfully slow compared to Office. I've played with the 1.9.51 branch a bit, and it doesn't seem 2.0 is going to be enough of an improvement to compete with Microsoft on the speed front.
I used to think that Moore's law will take over, but I'm now using a brand new P2.8 with 1 Gig of RAM at work, and after editing a presentation file with some large images I couldn't edit a slide with only text (don't ask me what OpenOffice was doing in the background with those pictures - it couldn't be autosave, since the problem was constant). I also used to think that OpenOffice should keep adding new features (e.g., macro recorder, which is in 1.9.51), but now I wish they would just optimize the hell out of it and add no new features for a while.
Perhaps it doesn't feel as slow on Solaris or Linux, but I doubt it - my Linux machine is pretty anemic, but it used to run Office reasonably when it had Windows on it. Now I don't even try to use OpenOffice on it as it is unbearable. When Koffice becomes file compatible, I may try to use that program on this machine.
The two free cross-platform software projects I use most are OpenOffice and Mozilla (Seamonkey or Firefox). Of course Mozilla's task is completely different, but it works reasonably fast compared to Internet Explorer (faster with some tasks, slower with others). I look forward to the day I can say the same thing about OpenOffice.
Dara
The GPS Lite mouse from deluo.com is $60. I have an older mouse from Deluo, and it works reasonably well under Windows. I've been told you can get the old one to work on Linux as well, but then there is no routing software available (let alone turn by turn voice guidance).
If you don't need turn by turn, and have a Windows laptop, you could also get a package with a GPS mouse + Streets and Trips 2005 for $85 (after rebate on amazon).
Dara
First of all, nobody is talking about doubling the existing DPI (dots per inch). Moving from 1/2 VGA to VGA is an increase in DPI of sqrt(2) not 2.
Secondly, you said it yourself, you usually look a lot closer (from my rough estimate right now), about twice as close. I agree, 15" 1600x1200 (i.e., 133 DPI) looks pretty good from a 20" laptop viewing distance. Therefore 266 DPI is about what we need for a PDA viewing at 10". VGA has this DPI at a 3" diagonal which is pretty small, (the T5 has a 3.8" screen, so the DPI if it had VGA would only be 210). To really get the resolution of the monitor you suggest from twice as close on a 3.8" screen will require 800x600 (SVGA).
We will definitely get PDA SVGA screens in the future, and we will probably have a lot more comments like yours that it is overkill. I may not need SVGA, but I'm looking forward to it. I do want VGA though, and I won't be getting a Palm T5 because it doesn't have it. (I can't get the Zaurus either since I want GPS software with voice instructions - I'll probably end up with an ASUS or Dell VGA PocketPC.)
Dara
I'm definitely getting a VGA PDA, maybe the HP, but more likely the Asus (also available now) or Dell (when it is available). I would have already purchased a Zaurus by now except for one thing - there is no GPS software for the Zaurus that does routing and voice instructions in the U.S. (when I last checked, I can't imagine this has changed). So I'm limited to Palm and Pocket PC, and as others have pointed out, Palm hasn't announced any VGA products yet.
e .htm), but the reviews predate the development of VGA.
I wrote deluo.com and asked if Routis would display optimally on a VGA PDA and they said no and didn't know when that would change. Does anyone else know of GPS software that works correctly on a VGA PDA? I've looked at Dale's excellent site (http://www.gpsinformation.org/dale/PocketPC/winc
By the way, GPS is one answer for what a PDA is good for. I've used a laptop w/ GPS in a car before - when you're the passenger, it's OK, but when you're the driver it sucks. You're worried its going to slide off the seat, you have to look 90 deg away from the road to see it, the screens can't handle sunlight, and the interface can't use large touch screen buttons. Phone screens are too small to be a good GPS - even a 4" PDA screen is a bit small, but I guess a 6" device is too big for most people. I guess I could also consider a dedicated GPS (maybe there exist VGA ones), but if I'm spending that kind of money, I want to use it for other things too - like a music player that can turn down the volume while giving voice instructions.
Dara
I'll consider an Apple iMac (or laptop for that matter) when they come out with the same resolution as the 23" Cinema Display (1920x1200, commonly called WUXGA). Then I can watch HDTV natively (there is a stand alone box option for Apple now) and I'll have more pixels for showing digital photos.
Dara
I read the announcement and it mentions "History traversal with arrow keys", but what I would really like doesn't seem to be mentioned (but perhaps it is possible with bash-2.05, I'm not much of a shell expert). In Matlab, the up-arrow key searches the history for commands that match all the characters on the line. No characters and it acts like a normal bash arrow, if "figure, plot" is at the beginning of the line, it will quickly scroll through all plotting commands that have been entered at the shell.
Any idea if this is possible?
Dara Parsavand
What should be possible is to state that you only have CDs 1, 1-2, 1-3, or 1-4 during package selection time and the installer would only offer you the choices that are available.
I, like others here, would prefer to do a reasonable workstation installation with just CD 1 (currently disks 1 and 2 are required). And links to torrents for individual CDs should be more prominent than they are now.
Mandrake did something like this quite a while ago and it worked fine.
Dara Parsavand
Thanks for all the replies, there were some very interesting ones. I'll try to respond to some of the them, but I'll do it all in one post. I hope this format isn't too confusing for anyone still interested in this subtopic.
To everyone complaining about the temporary need to speed (for passing or avoiding an accident): as heytherefancypants said, it isn't hard to imagine a solution. I drive mostly on crowded 4 lane roads or highways and if I'm doing near the speed limit, there is almost never a possibility for me too speed and avoid an accident (since there is usually someone in front of me - L.A. traffic, ugh). And I rarely have the passing problem (the one 2 lane road I drive is so windy there is almost no passing allowed). But sure, allow temporary speeding for both reasons.
To sickmtbnutcase: I guess I'm willing to have an override for someone who has a medical problem in the car (and set up a comm link between the car and the hospital, route the best path, etc.), but not for any other reason. People will have to adjust, just like you do when you use transit or if there is a traffic jam. I wouldn't want someone racing to see me as I was dying - they are probably stressed and in even more chance of getting in an accident. As far as enjoying driving goes, we'll all have to learn how to do that with nimble handling cars and curvy roads I guess. My car (2001 Prius) has barley adequate acceleration, but it's still fun to drive in the mountains (handles way better than my old Isuzu Trooper did).
to blunte: Interesting points on the significant reasons for traffic flow problems, I agree with all of them. I don't know if any of them negate my point though. It's true that if all cars were going the same speed (assume there a level highway and no reason to go slower than the speed limit), the only way to change lanes is to put on your turn signal and drop behind the car next to you. If the highway is absolutely at capacity and everyone is rude, it's going to be hard to do this (but it's hard if you speed up too).
to 87C751: No I'm not familiar with the exact revenue stream of small towns. I do realize speeding tickets are revenue which my post addressed. Those towns better figure start thinking of other solutions because I have no doubt in the long run there will be no traffic tickets (hopefully because all the cars will be fully automated - or maybe because energy is so expensive and everywhere is so populated, we'll all be using mass transit)
to gnu-generation-one: The actual scheme of reading the speed limit from the road is possible a number of different ways: paint stripes, magnets under the pavement, RF sign posts, accurate GPS, etc. I really don't know what the best way would be.
To YrWrstNtmr: your point is well taken (repeated below)
Situation A) Speed limit of 70. Cars cannot exceed. Most drivers want to exceed. Many drivers pissed off.
Situation B) Speed limit of 70. Enough drivers complain, and the limit is raised to 80. Not all drivers will go 80, leaving a disparity of speeds. Quite similar to what we have now.
I guess my reply is that 75 mph (120 kph) seems fast enough for most places that get any significant traffic around where I am (L.A.). I guess Montana can go for 85 mph or so if it wants to, and if they don't have much traffic, they won't have a big problem with speed disparity. I doubt many people would actively campaign here for more than 75 on any of our freeways, but you're right, there will be some that don't want to (or can't) do any more than 65 or so, so there will always be somewhat of a disparity issue until we all get automated cars with enough power to always do the limit.
Dara Parsavand
As I say every time this subject comes up, I'd much rather have my car know the max speed on a given road for a given set of conditions and not be allowed to go over the max speed, than I want fancy electronics to check to see if I go over the max speed, and if I do, take my picture, and send me a ticket. I'd rather pay higher taxes than fund police through tickets (and we wouldn't need as much traffic police either if the cars were smarter).
I claim that if no one could go over the speed limit, traffic would flow much more smoothly, and if the limit is too low (because you are expected to speed 10 mph), we will all complain loudly enough to get it changed.
Other aspects of this project sound interesting though.
Dara Parsavand
I've only played with Gimp and Photoshop Elements briefly so far, but I get the feeling the Gimp is designed to do a whole lot more than touch up digital photos (crop, rotate, red-eye elimination, etc.), which is all I want to do.
Are there any open source alternatives (preferably cross-platform) that do this job just as well as the Gimp but are simpler to use?
What is with the restriction on filenames (in the migration checklist)?
I have not made the transition to Linux (and I'm not planning to for a while longer), but I have a lot of files named with upper and lower case and spaces, and when I play on Linux, I don't have any problem with them. I am not changing any filename I don't have to (unless there is a character allowed by Windows which is not allowed by Unix - probably true - but I couldn't find a reference). I'm not aware of any reason not to have spaces in Unix. The command line still works using quotes and/or tab completion and any modern file manager is going to work fine too. If a certain application doesn't work opening the file, I can find that out later and rename the file then.
This kind of advice is just hindering people from switching by putting another obstacle in front of them. What for?
Dara
I use MS Office a fair amount at work (some PowerPoint animation, some Excel VB, nothing too complicated in Word - but I like the Outline mode).
I would prefer to switch to OpenOffice eventually (though I doubt I will have any luck convincing my company), but I don't want to until the speed of opening and saving files is improved (the current development versions of 2.0 are nowhere near fast enough yet, but maybe later).
I have never had a problem with quick save in MS Office, and when I work with files that a bigger than a few megs, it sure is nice to be able to open and close them in just a few seconds.
The speed issue and some way to partially convert VB code to OO Basic (or python or whatever works with OO.o) is all I need.
Dara