the ability to copy reliably is equivalent to imortality. Copyright violation as a basis for religion? Well, it's as good as any other, I suppose, and would vilify the usual demons - RIAA, MPAA, publishers, etc.
Or did you manage to usefully run X11 on a 486 PC with 8 MB of RAM? On a 16MHz 386 "laptop" (Toshiba T5200) with 8MB and orange plasma VGA. After an upgrade to 14MB, linux really flew. I think I used either fvwm or twm as the window manager.
This "laptop" also booted to OS/2, which could run X11 as a separate GUI simultaneously with the Win3.1 and OS/2 GUIs and a bunch of virtual DOS machines. One of the DOS VMs often ran the GEM GUI because I used GEM Draw quite a lot in those days. It also had OS/2's NFS client+server. Four different GUIs with multitasked applications and daemons, all snappy enough in 14MB RAM with a 100MB disk.
Bloat Sucks. Windows seems always to have had more of it than the alternatives.
It could be 6,116,545 with abstract:
"A method is provided for using a lunar flyby maneuver to transfer a satellite from a quasi-geosynchronous transfer orbit having a high inclination to a final geosynchronous orbit having a low inclination. The invention may be used to take the inclination of a final geosynchronous orbit of a satellite to zero, resulting in a geostationary orbit, provided that the satellite is launched in March or September."
Alternatively, it might be 6,149,103 which does not have the March/September dependency.
Both of these patents were originally assigned to Hughes. Boeing and GM own nearly all of what was Hughes, and Boeing got the space and satellite parts (including bits via Raytheon as an intermediate owner).
Notice how the EU is all dark orange, except for parts of central Spain. ...and most of Sweden and Finland, which make up a significant fraction of the EU's area.
I live in central Finland (yellow on that map), in the countryside, and have fiber to the house. 100/10 Mbps unlimited and unthrottled costs 75euro per month. That price also includes TV and telephone.
I think the main point in broadband that people just don't get is that the US is huge while many smaller countries are the size of one of the US's states, its is expensive to get broadband. I live in Finland. Compared to some of the larger US states, Finland is (i) slightly larger than Arizona with a somewhat smaller population, (ii) twice the area of Florida with one third its population, or (iii) half the area of Texas with one quarter its population.
I don't live in Helsinki or any other large city; in fact, I live in the countryside outside a small city a few hundred km north of Helsinki. A 100/10Mbps fiber connection here costs 75euro per month, with NO capacity limits or throttling. That price also includes telephone and a basic TV package. Wherever a new house is built in my area, the ISP puts down fiber to it (2km fiber to reach me and 4 neighbours). They stopped putting down copper a couple of years ago, and are progressively replacing existing copper with fiber. On fiber, they give you two choices: 20/2 or 100/10 Mbps, unlimited and unthrottled.
In places with existing infrastructure (cable or decent telephone lines), I can understand ISPs preferring to "extract the value" from their assets rather than add fiber beside the existing lines. But from much commentary on/. and elsewhere, it seems that in many parts of the US the existing infrastructure is overwhelmed at the local level, lacking the capacity to support even a fraction of the bandwidth that has been sold to connected customers. The highways get repaved or widened when the circumstances demand it; communication infrastructure also needs occasional improvements.
Posts and sub-threads like this make me really appreciative of living in Finland. ...or most of Finland, anyway. Living around Helsinki involves a degree of suffering and stress in commuting and traffic jams. Tampere & Turku are better, but nothing beats living in or near the smaller cities. I used to commute 5 minutes each direction by bike when we lived in the city, but then we moved to a bigger place in the countryside, and our commute became almost 15 minutes by car.
... of the same scale... The scales are quite different, as you can see by looking at the scale bars in the lower left of the images. One has z=3 while the other has z=4 in the URL. In fact, even if you set the same z value in both URLs, the scale bars will differ significantly. This is partly due to the projection used, and the different latitudes the views are centered on.
BTW, the US+Canada have a combined land area of 19 million sq.km or so [depending whether Puerto Rico, Guam, etc. are included]. By contrast, the EU has an area of 4½ million sq.km [includes new members and those bits of France located in the Caribbean/Pacific/etc.]. The total area of Europe is 10 million sq.km [includes Iceland, Faroes, and Russia west of the Urals; excludes Greenland & Turkey].
So, in approximate areas: US = 2*EU; Canada=Europe.
If you have 1,000 HD's running, and it takes 40 hours before one fails, that's a 40,000 hr MTBF. Manufacturers employ a more sophisticated method to arrive at unreliable numbers for reliability.
The manufacturer tests a population of drives, and waits for a significant fraction of the drives under test to fail, recoding the failure time for each. In this way, it is possible to separate "infant mortality" failures from "random event" failures. Typically, the failure times are fitted to a Weibull distribution http://en.wikipedia.org/wiki/Weibull_distribution. This process also provides a value for the post-manufacture burn-in time which will kill most units which are prone to "infant mortality" type failure. The MTBF is estimated based on the "random event" failure rate, giving absurdly large MTBF values. Unfortunately, the test rarely lasts long enough to identify the third type of failure: "wear out", which determines the end of life, and which is often less than the MTBF.
Think of estimating human life expectancy in the U.S. as MTBF, using data from http://www.data360.org/dsg.aspx?Data_Set_Group_Id=587. A small fraction (683 out of 100000) die in the first year. Death rates are low for the next 60 years, then climb to a very steep peak. The 85+ category is not subdivided, because very few live long enough to make separate 85-94 and 95+ categories worthwhile. However, if life expectancy were calculated for humans in the same way as MTBF for disk drives, then only deaths between ages 1 and 24 would be used. Since from the 99317 who survived infancy, only 126 die in that time, the MTBF for humans would be estimated at several centuries. If only we didn't wear out...
I'm buying one (or two - must think of mom) Asus eee PCs. I've never felt so good about buying a computer in many years. I was very close to buying it online the past week but finally I decided I'll buy it locally in Helsinki. Where are you planning to buy, and for what price? The only place I've seen an EEEPC advertised in Finland, it was euro999 with 7" screen, 512MB, and 4GB SSD. This looks like obscene price gouging since an ASUS 14" laptop with core duo, 2GB RAM and 160GB disk with Vista Home Premium can be got for euro789, while an ASUS 17" laptop with Athlon, 2GB RAM, 160GB disk and Vista Home Premium can be got for euro 729 (prices in Finland, from search at mbnet.fi).
$150 is about euro100. For 50/6 Mbps in an urban area, where there is a potential customer for every 20-30 meters of fiber, that must include a bloated profit margin!
In my rural area, FTTH 100/10 Mbps costs euro75 per month, and that includes basic TV over IP. It also has no throttling or monthly quotas. The local ISP considers this attractive enough to lay a couple of kilometers of fiber to reach a handful of widely separated houses.
Well, I don't know about your Internet connection but the only place I notice congestion is on the first few hops (and possibly the last few hops if we're talking a single host and not P2P). Beyond that on the big backbone lines I at least don't notice it, though I suppose it could be different for the computer. Well, a slow connection at the other end is always a bottleneck, but in my experience there are also bottlenecks in the "backbones", probably due to load on routers or deliberate QOS throttling. From your description it sounds as if either (i) you've got a pig slow connection to your ISP or (ii) your ISP has a pig slow connection to its upstream supplier.
I've got 20Mbps (actually FTTH 100Mbps raw, but throttled) and I notice speed differences based on where I connect to. The speed limits come from either a routing bottleneck or QOS throttling or just a pig slow connection at the other end.
There are many bandwidth testing sites around the world. If I try one at my ISP (kpyverkot.fi in Kuopio, about 20km away, 3 hops), then I usually get a speed far above what I'm paying for, typically 45-50Mbps, and I suspect my ancient SMC router is the reason it's not higher. Testing with my ISP's upstream supplier (kponet.fi in Helsinki, about 400km away, 7 hops) then I typically get 19-20Mbps, which is the throttled speed I pay for. I get similar speeds on ftp to sites in Finland with fast links, such as funet.fi. However, bandwidth tests give lower speeds for connections to other parts of the world, often below 10Mbps, even for destinations with fast links, such as arxiv.org. For instance, if I try with a site in Denmark (gemal.dk, 16 hops), the speed is only about 2Mbps.
I see no reason to upgrade from 20Mbps to 100Mbps until the bottlenecks between Finland and other parts of the world (North America and the rest of Europe) are addressed.
Well, if you can tolerate a plug-it-together solution using CF cards and a PCI adapter, you might check out http://www.addonics.com/products/flash_memory_reader/ad4cfprj.asp. You can put up to 4 CF cards into it, and have them configured as RAID 0/1/10. Supports Linux as well as that inferior-brand OS. You could have a 64GB (raw capacity, less with RAID) SSD for $350 (based on $75 for a 16GB CF).
I also got a giggle out of US 6,368,227 B1 (issued April 2002), and added it to my collection of wierd patents. I'm pretty certain I and my friends all employed that method of swinging on swings decades earlier.
A re-examination request was filed the month following issuance, and all of its claims were cancelled on re-examination (re-issued as US 6,368,227 C1 in July 2003).
So sometimes, the twisted parts of the patent process can be straightened out without much delay.
But there has been no solution to this short of raising prices and charging users more so the isp can afford additional bandwidth. Interestingly, for the princely sum of eur55 per month, my ISP gives me 20Mbit/sec unthrottled with no limits on weekly/monthly/whatever usage, with TV and telephone service. All come via the optical fiber they installed to the house (we're in the countryside). For an extra eur20, we could get 100Mbit/sec, but at present, we don't need it. They installed fiber going past most of the houses in the area, and some of our neighbours have also switched from ADSL to the fiber service.
The only bandwidth limit is that in any one second, I cannot download more than 20Mbit. I download a lot of linux liveCD & liveDVD distros, and my kids are on the 'net from their PCs for many hours each week, so we go well beyond 100Gbytes in a month and have never encountered a throttle or slowdown of any sort.
The investment in laying 2½km of fiber to reach the 4 houses in our patch of forest was apparently worth it to one ISP.
I used Norman firewall and Norman Antivirus about 4 years ago, when I still had one Windows PC machine connected to the net. It was a decent enough anti-virus and firewall setup, although it did not support stealth mode for ports at that time. Apparently this was fixed in a later version. Thanks to a fubar by my ISP, I got a perpetual license for it, with version updates and no expiry date (net security was part of the ISP's service deal, included in the monthly bill). All of our PCs are now on linux, so we don't use Norman any more. http://www.norman.com/
The fact that IBM's lawyers are colloquially known as 'Nazgul' should probably be more worrying to them. Nazgul is a Middle-Eastern name meaning "Shy Rose" or "Delicate Flower" http://en.wikipedia.org/wiki/Nazgul. IBM's lawyers might not take kindly to being called a bunch of pansies, and are possibly preparing a number of defamation suits.
One of our home PCs is a 10 year old Dell XPS T450, which has a 450MHz Pentium III and its original 20GB ATA disk (we also have a couple of new PCs). Its principal use is probably running Stepmania 4 with a couple of USB dance pads, which is why it's in the kids' play room. It also works just fine for web browsing, playing media from our server, and so forth. Starting gimp or inkscape or other large-ish applications takes longer than on the newer machines, but once started, they seem to work OK.
This PC currently has Ubuntu, but has also had PCLinuxOS. It was upgraded from its original 128MB of RAM to 384MB, which made a significant difference to graphics editing, and use of OpenOffice, but made little difference to web-related activities.
Slashdot Mirror
In other news: beatings will continue until morale improves.
... the obligatory Beowulf cluster of these?
Well, it's as good as any other, I suppose, and would vilify the usual demons - RIAA, MPAA, publishers, etc.
This "laptop" also booted to OS/2, which could run X11 as a separate GUI simultaneously with the Win3.1 and OS/2 GUIs and a bunch of virtual DOS machines. One of the DOS VMs often ran the GEM GUI because I used GEM Draw quite a lot in those days. It also had OS/2's NFS client+server. Four different GUIs with multitasked applications and daemons, all snappy enough in 14MB RAM with a 100MB disk.
Bloat Sucks. Windows seems always to have had more of it than the alternatives.
...with the disruptive technology of your lower-case-only keyboard?
It could be 6,116,545 with abstract:
"A method is provided for using a lunar flyby maneuver to transfer a satellite from a quasi-geosynchronous transfer orbit having a high inclination to a final geosynchronous orbit having a low inclination. The invention may be used to take the inclination of a final geosynchronous orbit of a satellite to zero, resulting in a geostationary orbit, provided that the satellite is launched in March or September."
Alternatively, it might be 6,149,103 which does not have the March/September dependency.
Both of these patents were originally assigned to Hughes. Boeing and GM own nearly all of what was Hughes, and Boeing got the space and satellite parts (including bits via Raytheon as an intermediate owner).
I live in central Finland (yellow on that map), in the countryside, and have fiber to the house. 100/10 Mbps unlimited and unthrottled costs 75euro per month. That price also includes TV and telephone.
I don't live in Helsinki or any other large city; in fact, I live in the countryside outside a small city a few hundred km north of Helsinki. A 100/10Mbps fiber connection here costs 75euro per month, with NO capacity limits or throttling. That price also includes telephone and a basic TV package. Wherever a new house is built in my area, the ISP puts down fiber to it (2km fiber to reach me and 4 neighbours). They stopped putting down copper a couple of years ago, and are progressively replacing existing copper with fiber. On fiber, they give you two choices: 20/2 or 100/10 Mbps, unlimited and unthrottled.
In places with existing infrastructure (cable or decent telephone lines), I can understand ISPs preferring to "extract the value" from their assets rather than add fiber beside the existing lines. But from much commentary on
I used to commute 5 minutes each direction by bike when we lived in the city, but then we moved to a bigger place in the countryside, and our commute became almost 15 minutes by car.
... of the same scaleIn fact, even if you set the same z value in both URLs, the scale bars will differ significantly. This is partly due to the projection used, and the different latitudes the views are centered on.
BTW, the US+Canada have a combined land area of 19 million sq.km or so [depending whether Puerto Rico, Guam, etc. are included]. By contrast, the EU has an area of 4½ million sq.km [includes new members and those bits of France located in the Caribbean/Pacific/etc.]. The total area of Europe is 10 million sq.km [includes Iceland, Faroes, and Russia west of the Urals; excludes Greenland & Turkey].
So, in approximate areas: US = 2*EU; Canada=Europe.
The manufacturer tests a population of drives, and waits for a significant fraction of the drives under test to fail, recoding the failure time for each. In this way, it is possible to separate "infant mortality" failures from "random event" failures. Typically, the failure times are fitted to a Weibull distribution http://en.wikipedia.org/wiki/Weibull_distribution. This process also provides a value for the post-manufacture burn-in time which will kill most units which are prone to "infant mortality" type failure. The MTBF is estimated based on the "random event" failure rate, giving absurdly large MTBF values. Unfortunately, the test rarely lasts long enough to identify the third type of failure: "wear out", which determines the end of life, and which is often less than the MTBF.
Think of estimating human life expectancy in the U.S. as MTBF, using data from http://www.data360.org/dsg.aspx?Data_Set_Group_Id=587. A small fraction (683 out of 100000) die in the first year. Death rates are low for the next 60 years, then climb to a very steep peak. The 85+ category is not subdivided, because very few live long enough to make separate 85-94 and 95+ categories worthwhile. However, if life expectancy were calculated for humans in the same way as MTBF for disk drives, then only deaths between ages 1 and 24 would be used. Since from the 99317 who survived infancy, only 126 die in that time, the MTBF for humans would be estimated at several centuries. If only we didn't wear out...
The only place I've seen an EEEPC advertised in Finland, it was euro999 with 7" screen, 512MB, and 4GB SSD. This looks like obscene price gouging since an ASUS 14" laptop with core duo, 2GB RAM and 160GB disk with Vista Home Premium can be got for euro789, while an ASUS 17" laptop with Athlon, 2GB RAM, 160GB disk and Vista Home Premium can be got for euro 729 (prices in Finland, from search at mbnet.fi).
$150 is about euro100. For 50/6 Mbps in an urban area, where there is a potential customer for every 20-30 meters of fiber, that must include a bloated profit margin!
In my rural area, FTTH 100/10 Mbps costs euro75 per month, and that includes basic TV over IP. It also has no throttling or monthly quotas. The local ISP considers this attractive enough to lay a couple of kilometers of fiber to reach a handful of widely separated houses.
disk 1 http://www.freedb.org/freedb/misc/be120d1f
disk 2 http://www.freedb.org/freedb/classical/fd111f21
disk 3 http://www.freedb.org/freedb/misc/fb0b4e14
disk 4 http://www.freedb.org/freedb/classical/d70ea820
They'll all know who did it, of course, since I use the Minaret and Trio movement of the Pervertimento for Bagpipes, Bicycle and Balloons (on disk 4) as the main ring tone in my phone.
I've got 20Mbps (actually FTTH 100Mbps raw, but throttled) and I notice speed differences based on where I connect to. The speed limits come from either a routing bottleneck or QOS throttling or just a pig slow connection at the other end.
There are many bandwidth testing sites around the world. If I try one at my ISP (kpyverkot.fi in Kuopio, about 20km away, 3 hops), then I usually get a speed far above what I'm paying for, typically 45-50Mbps, and I suspect my ancient SMC router is the reason it's not higher. Testing with my ISP's upstream supplier (kponet.fi in Helsinki, about 400km away, 7 hops) then I typically get 19-20Mbps, which is the throttled speed I pay for. I get similar speeds on ftp to sites in Finland with fast links, such as funet.fi. However, bandwidth tests give lower speeds for connections to other parts of the world, often below 10Mbps, even for destinations with fast links, such as arxiv.org. For instance, if I try with a site in Denmark (gemal.dk, 16 hops), the speed is only about 2Mbps.
I see no reason to upgrade from 20Mbps to 100Mbps until the bottlenecks between Finland and other parts of the world (North America and the rest of Europe) are addressed.
Well, if you can tolerate a plug-it-together solution using CF cards and a PCI adapter, you might check out http://www.addonics.com/products/flash_memory_reader/ad4cfprj.asp. You can put up to 4 CF cards into it, and have them configured as RAID 0/1/10. Supports Linux as well as that inferior-brand OS. You could have a 64GB (raw capacity, less with RAID) SSD for $350 (based on $75 for a 16GB CF).
And if you don't need RAID, then check out the CF-SATA adapter which lets you use a CF card for a 2.5" SATA drive http://www.addonics.com/products/flash_memory_reader/adsahdcf.asp. A 16GB SSD in a notebook for about $110.
BTW, I have no connection with addonics - google dropped them on me in a recent search.
I also got a giggle out of US 6,368,227 B1 (issued April 2002), and added it to my collection of wierd patents. I'm pretty certain I and my friends all employed that method of swinging on swings decades earlier.
A re-examination request was filed the month following issuance, and all of its claims were cancelled on re-examination (re-issued as US 6,368,227 C1 in July 2003).
So sometimes, the twisted parts of the patent process can be straightened out without much delay.
The only bandwidth limit is that in any one second, I cannot download more than 20Mbit. I download a lot of linux liveCD & liveDVD distros, and my kids are on the 'net from their PCs for many hours each week, so we go well beyond 100Gbytes in a month and have never encountered a throttle or slowdown of any sort.
The investment in laying 2½km of fiber to reach the 4 houses in our patch of forest was apparently worth it to one ISP.
I used Norman firewall and Norman Antivirus about 4 years ago, when I still had one Windows PC machine connected to the net. It was a decent enough anti-virus and firewall setup, although it did not support stealth mode for ports at that time. Apparently this was fixed in a later version. Thanks to a fubar by my ISP, I got a perpetual license for it, with version updates and no expiry date (net security was part of the ISP's service deal, included in the monthly bill). All of our PCs are now on linux, so we don't use Norman any more.
http://www.norman.com/
Or did you mean Nazgûl? http://en.wikipedia.org/wiki/Nazg%C3%BBl
One of our home PCs is a 10 year old Dell XPS T450, which has a 450MHz Pentium III and its original 20GB ATA disk (we also have a couple of new PCs). Its principal use is probably running Stepmania 4 with a couple of USB dance pads, which is why it's in the kids' play room. It also works just fine for web browsing, playing media from our server, and so forth. Starting gimp or inkscape or other large-ish applications takes longer than on the newer machines, but once started, they seem to work OK.
This PC currently has Ubuntu, but has also had PCLinuxOS. It was upgraded from its original 128MB of RAM to 384MB, which made a significant difference to graphics editing, and use of OpenOffice, but made little difference to web-related activities.
The original story, with images etc. is at
http://hubblesite.org/newscenter/archive/releases/2008/04
Of course, this site lacks the amusing comments in the OP's linked site.