> When I think of a red cross, I immediately > think of emergency relief, supplies, etc.
That's what I thought of as a kid. Now, I know the history previous to that. Now, I think of the Crusades and the terrible things done in the name of Christianity during them. That's where the red cross on a white background originated.
The Knights Templar (Temple Mount Knights, who took a vow of poverty (communal property, as the term meant at that time) and allegiance to the order, who originally protected the pilgrims during a period when most other knights were ignoring them because there wasn't profit in their protection, who were a time the most powerful Military Christian order after so many donated so much in turn, who set up the first European and Crusader banking system, had as their naval flag the original white skull and crossbones on a black background later associated with pirates, who traded with the Muslims and introduced gauze (from Gaza, the city), muslin (from Muslim), and candy (from al-kandiq) these were one of the orders that flew the red cros on a white background, the original symbol of the crusades.
Similarly with the Knights Hospitaller, whose order founded and staffed the medical "hospitals" the pilgrims and Crusaders used. This is actually the origin of both the word "hospital" and the association of the red cross with medicine. Of course, while they were a welcome sight to many a wounded Christian Crusader or pilgrim, obviously, the Muslims had a rather different viewpoint! No WONDER they couldn't tolerate the Red Cross as a medical symbol! To them it meant the rape and pillage, the savagery of the Crusades! So they went with the Red Crescent.
However, the Red Crescent is also a religious symbol, so today there's a third symbol coming into use, designed to be religiously neutral while still being unmistakable for anything else. It is often used in international contexts, particularly in the mid-east. This is the Red Crystal, a diamond shape.
So, ideally, the Red Crystal will eventually come to have the positive associations of the Red Cross and Red Crescent today, without the sectarian and cultural negatives. It's also worth noting that Cross, Crescent, and Crystal, all start with C, so the ARC (and other "Red Cross and Red Crescent" organizations worldwide) could adopt the Crystal without even changing initials.
Wikipedia is a place to start, anyway. There are links from there elsewhere, and I've included another informative link on the Knights Templar as well as a Google link, below.
... Except that... for copyright law to be abolished, there'd have to be a sea change in conventional viewpoint.
If sharing became the accepted norm (not just legally accepted fact), as it very well could be if the conventional view on the issue changed sharply enough to make it legally acceptable (copyrights abolished), then those who dared go against that society accepted norm, producing proprietaryware (which could then be seen as slaveryware, see my sig), could very well be viewed with the same contempt as most of society holds malware authors and spammers today. Even if it were/legal/ to go proprietary, to do so would be considered an indication of something to hide, being shady in one's practical dealings regardless of whether it were actually legal or not. Legality in that case wouldn't matter that much, and while there would still be people at the fringe who'd do it much as there are malware/spyware authors and spammers today, some "legal", some not, it'd be considered a scorge on society, something no educated person would admit to running, just as it's not acceptable in educated portions of society today to admit that one actually makes purchases based on spam.
It would thus likely follow any sea change in public viewpoint to the degree necessary to abolish copyright, that the legal enforcement mechanisms of the GPL as we know it today would no longer be necessary in practice, even as they fail to be legal enforcement mechanisms. This I say as a strong copyleft supporter today (to the point I don't seriously consider the BSDs as they don't actively support user rights, only take the agnostic view, making them possible... and to the point I chose the sig I did).
Of course, both the copyright abolished scenario and the implications that would follow are so far out of practical reality today that it doesn't matter a whole lot, except as an epistemological (which is I contend what real science fiction, and many non-scifi novels are, as well, the biggest reason I enjoy them). It may well matter at some point in the future, but it's not likely in the immediate future and may indeed not occur for many generations, if ever. Still, it's worth exploring, both for the implications should it happen then, and for those it has on present reality and our approach to it.
One of the better ways to block ads -- and some malware -- is to use a "personal proxy" such as (for MS Windows) the Proxomitron or (for *ix and MS Windows both) Privoxy. I used the former years ago on MS, and now use the latter, Privoxy, based on the old Junkbusters proxy code but vastly improved since then.
These "personal proxies" normally run on your own machine, listening on the loopback interface (127.0.0.1) for connections from various HTTP clients. As proxies, they are client agnostic (so Firefox is fine, so is Konqueror, so is IE), as long as said client can be configured in some way to use a proxy.
They work as a rewriting (as opposed to caching) proxy, using a configurable ruleset to intercept and rewrite or block both outgoing and incoming headers and content as desired. For ad-blocking, a double-layer approach is taken. First, as a page is downloaded, many references to ads will be rewritten, either deleted from the page, often with a tiny-text note saying what filter did it (this is what I do with many of my own filters, as it aids in debugging if I get it wrong), or rewritten to replace the ad with a placholder (safer for the general case as it doesn't disturb the layout of the page), often with an "override, go there anyway" type link (just in case). Since the placeholders are local and the the deletions obviously aren't going to have the browser expending additional bandwidth downloading content that has been entirely erased from the page the browser sees, that's all saved bandwidth.
Second, the proxy can be configured to block access to certain sites (say anything doubleclick related, to use one common example) entirely, and to block but with override possible in other cases (say anything with *ad* in it, since that's pretty broad when you have "road", "adsl", etc, altho the default filters are already smart enough to not block those real obvious things). Thus, even if an ad reference gets thru, when the browser attempts to load it, either because it's part of the page, or because you perhaps accidentally clicked where the ad would have been, it'll block (with or without override option) the fetch, saving more bandwidth.
The most obvious malware this blocks in a lot of the tracker stuff, which is often interwoven with ads to the point they're the same thing anyway. However, quite apart from any browser based scripting and cookie and popup blocking (for instance), the proxy has its own configurable settings. Of course, it can be set to block other stuff too, some of which it does by default (with for privoxy, a low, medium, and aggressive setting, for those not wanting to fool with individual rules), some of which you'd create your own custom rules to deal with, as appropriate. It's sort of like Firefox's greasemonkey, but more powerful. The fact that it's all browser agnostic is pretty sweet too, for those of us that use multiple browsers.
If you want to see what a page looks like "unfiltered", or just want to browse without the filter, that's possible too. The proxies have a "bypass" mode that bypasses the normal rulesets, simply forwarding the requests and replies as they come. Most browsers also have a proxy bypass toggle as well. It's also easy enough to set some things to use the proxy, without setting everything to use it. For instance, while the proxy could be configured to ignore them (and would generally do so after a bit anyway, as the stream exceeded the max-size-filter setting), you probably don't want it proxying streaming audio or video, as that's just a waste of resources. Just don't setup those clients to use the proxy. OTOH, back on MS Windows, I DID find it handy to run my spinner.com streaming client thru the proxy, since it killed the ads that way.
As an example of some of the other stuff it can do unrelated to the above, I personally prefer light text on dark backgrounds, while most of the web is setup in the reverse, light/white background, dark/black text. While it is possible to set preferences in most browsers, those who
The operating system is a component of the computer, just like the processor, video card, sound card, network card, monitor, keyboard, mouse,... Just because it's not hardware doesn't make it less necessary for using the end product. But... in many cases one can actually choose the CPU, video card, sound card, etc, and except for laptops, monitors are actually sold separately much of the time. Keyboard and mouse, not so much, but the ones that come with a machine are basically lowest cost disposable quality anyway, and work decently for backups if you decide to get (or already have) something different.
Even the OS is often customizable, but unfortunately generally only between MS variants. That it can be switched out is self-evident both from the fact that it can be customized and by the wide availability of LiveCDs/DVD/Thumbdrive distributions that don't even touch the hard drive installation. It's even further self-evident from the EULA, which says send it back to your machine dealer for a refund if you don't agree to it.
Which is what people try to do. It's nice to see some places actually enforcing the EULA MS expects to be enforced for it, for the consumer as well, and the dealer having to eat the cost, as they should having chosen to distribute it knowing the EULA and its conditions obligating them to refund upon turn-in.
The US has similar anti-bundling laws. If the powers that be had any concern for the consumer, they'd enforce them as well. Europe seems rather ahead of us in that regard, among several others (software patents... tho behind in still others, like UK's vid monitoring and Germany's anti-security-testing-software laws).
OK, this is like two weeks old by now, but I just got it in metamod, and marked the off-topic as unfair. I read thru all four pages of comments and there were others discussing config and stuff, tho nothing quite like this one. Still, it wasn't OT given the context of the other threads, and the fact it WAS about Linux desktops (maybe the mods didn't consider a laptop close enough to a desktop?).
FWIW... make sure you're comparing the same probe temps. At least my AMD (desktop/workstation, so may be different) system has cpu core temps, and cpu case temps. The core temps run higher (typically ~5 degrees C), but due to volatility are *NOT* the temps AMD tells its system vendors to standardize on for system thermal regulation. It's thus entirely possible that (even if they are labeled core) the MS Windows reported temps are the case temps that the system is regulating on, while it's likely the actual core temps that you have chosen to have reported on Linux.
Of course, if your thermal regulation isn't working and the fans aren't coming on or speeding up when they should, it'll hit warning territory pretty fast regardless, but just in case you weren't aware of the possibility of confusing the two readings. Of course if you only have one of them...
That being said, Solaris scales on single nodes like its nobodies business if you follow the best practices and know what your doing. I've seen big Sun boxes hit loads in excess of 300 and keep chugging. I *still* havent seen a Solaris box that was well maintained crash. (cant say that about Linux -- if anything its getting less stable). While I can't argue with the crash angle (likely in large part due to the fact that it's my personal desktop/workstation and I actually prefer running -RC kernels and a beta/testing/unstable distribution level general system), I'd certainly argue the load average angle.
As I posted up-page in the slashdotting discussion, I like to run compiles in -j unlimited "jobserver" mode just to see how far I can push the hardware. Back before I upgraded from the dual-single-cores Opteron 242s (1.6 GHz clock), jobserver-mode (Linux) kernel compiles would run up 480 on the 1-minute load average before the (KDE/X) ksysguard graph would freeze up and I couldn't tell how much higher it'd go.
After upgrading to top-of-the-line-for-socket-940 dual dual-core Opteron 290s (2.8 GHz), with the user based priority/scheduling in kernel 2.6.24+, I no longer see those GUI freezes so continue to get load average updates, but the system is fast enough it apparently retires the first jobs before the last are loaded, so now job-server mode kernel compiles max out at only ~180 or so on the 1-minute load average. (FWIW, both the dual 242 480+ load and the 290 180+ loads are with 8 gigs RAM, the 480+ load would run into swap, but the 180 peak I manage on the 290s never touches swap.)
However, the nice bit is that with the per-user or control-group (I use the former) scheduling available in kernel 2.6.24+, I can run practically unlimited load average compiling as a different user, as long as I don't hit swap too hard, while all the while maintaining hitch-free net radio playback and nearly hitch-free visualization (it's slightly jerky updating but I don't miss many frames). The limit isn't on the load average aka CPU cycles, but rather on real memory vs swap. If I'm not streaming, I can maintain "acceptable" if somewhat jerky interactivity several gigs into swap, altho that's certainly partly due to the fact that swap is 4-way-RAID-0/striped.
So the 300 load average on Solaris isn't really all that special on Linux either, as I used to routinely do 480+ load average "just for fun" compiling the kernel, and that was before the recent very dramatic scheduler improvements. And be sure to check out that 2.6.24+ per-user scheduling as it's certainly worth it.
I've never seen the load on a Linux machine rise above like 6, and by then its unresponsive to anything.
How's it get up to 80? I've seen a eight core server hit hundreds.
Mind you that was after I made a mistake on a production server which essentially caused all the site's traffic to turn in to a DDoS attack.:) Just for fun, running my old dual single-core Opteron (242, 1.6 GHz) with lots (8 gig) of memory and swap on a 4-disk RAID-0 as a personal desktop/workstation I used to compile the (Linux) kernel with a patch to the makefile putting it in jobserver (-j) mode. The kernel compile is apparently VERY parallelizable! =8^) I'd run nearly 500 (480-ish) 1-min load average and a gig or two into swap before ksysguard quite responding and I couldn't tell how much higher it'd go.
Since then I've upgraded to the top of the line for socket 940, dual dual-core Opteron 290s (2.8 GHz), and with the recent per-user scheduling options turned on in the kernel, can watch the 1 min load average run to 180 or so (the highest it gets) without ever going unresponsive or stopping the 2-second ksysguard updates. Apparently with the extra speed and cores, the first jobs get worked thru and retired before it ever loads the last ones, so load average (and memory usage) never get anywhere as high.
What's great tho is as I mentioned the new (well, as of kernel 2.6.24) per-user scheduling/priority options. Set properly, you can have one user running multi-hundred load averages and it barely affects usability for other users at all, because they get equal CPU time (and with the proper I/O scheduler, I use CFQ which by default prioritizes I/O to match CPU priority, I/O to match) if they've got runnable processes, even with only fractional load averages. This is certainly one of my favorite new kernel features and one I regularly use to continue almost hitch-free internet radio playback and even visualization, plus of course general system responsiveness, even when compiling at multiple jobs per core load average. (I normally keep MAKEOPTS to "-j -l15" for general purpose compiling, which ends up running about a 16 load or 4 per core, more to limit memory usage to not run into swap (compiling with the workdir on tmpfs) than load average, tho.)
With a 120-ish day uptime... well, maybe it's a 2.6.24 kernel, but probably not. However, with that kernel or better, an 80 load average, or even a 500 load average, really shouldn't be unmanageable at all, as long as the scheduler options are set right, and if user-based scheduling groups are turned on, apache is running as a user other than root and other than the remote login shell would be running. As such, by default, root would get double the CPU cycles allowed other users (and appropriate I/O, assuming CFQ or similar prioritized I/O scheduling with priorities similar to the CPU priorities), and no matter what the apache user load average, the remote login shell should get equal cycles (assuming equal process priorities). It should thus be nearly impossible to DOS the machine to the point remote administration to get in and correct the problem becomes difficult or impossible.
Just in case it's not apparent by now, I *REALLY* like that aspect of kernel scheduling, new at least to the Linus kernel with 2.6.24. =8^)
Taking a look at the pic... From the detail I could see at 2X magnification, it appears there's no actual "breaker" there -- the magnetic core that's supposed to "trip" was simply replaced by a plain wire. As best I could tell, and this part I couldn't be sure of from the picture but it seems logical, they had it rigged up to switch off if manually turned off, so it does look like it should function as a regular switch, thus effectively deceiving the user until the moment of truth, but the bulk of the thing, the part that should trip it automatically on overload, simply ISN'T THERE!
The wire looks a bit heavy for it but depending on the rated amps and the composition of the wire, if a buyer is/lucky/ (and the counterfeiter has/some/ conscience), it might fuse out, vaporizing the wire (or another component) rather than continuing to short and starting a fire or whatever, but even then, it'd be a (switched) fuse, not a resettable circuit breaker, dead on first use.
More likely it'd cause the main breaker -- hopefully there is one and it's not similarly rigged -- to trip, cutting off not just the single circuit but the entire panel along with all its circuits. Considering, at least over here in the US, that occasionally, for circuits such as alarm panels, security systems, and relay timing circuits, it's apparently accepted and not uncommon to see their breakers blocked ON, thus allowing serious overload to the point the load gets enough to trip the main breaker, this isn't/too/ unacceptable, or at least I can imagine the counterfeiter rationalizing such to himself. (I still have trouble rationalizing the on-block to myself, but whatever, I've seen it enough places, and with blocks obviously marketed for the purpose, to know it's gotta be accepted and legal, regardless of what they tell us about the safety of the system. Of course, we have 125V standard mains. It's possible such things aren't allowed at all on 250V standard mains, which I suppose they have where those were discovered.)
Snail mail spam does allow a couple ways to get rid of it upstream.
1) At least here in the US, while the postal service doesn't do traditional blocks (but see below), I believe the regulations require advertisers to remove one's name from their lists if asked to do so. The mail must be individually addressed to make that possible, altho it can be addressed to "occupant", but it must have an individual street or whatever address. I believe bulk rate must also have something (often the small card with the individual to address; small cards often get lost and are in any case hard to find among the newsprint) listing the sender, as well, so a customer may send a removal request.
I've asked for removal from several regular mail spammers around here, and they've done their part. The problem however is that the post office doesn't always do/their/ part. First, as mentioned, those little address cards often get separated from the bulk of the newsprint and etc that forms most of the ad bundle. Technically, this is a violation of USPS regulations, the mail must be delivered in its entirety including the address card, and you can call your postmaster to complain.
Second, with bulk mail, the tendency of the carrier is often to just shove it into each box sequentially, never looking at the address on the card. After you get your name removed from the marketer's list (as I mentioned, I've had no problems with that part), you'll often still get it -- if you look at the address however, it was to the neighbor next in line (or two or three down, if several have had their addresses removed). Thus, taking your name off lists doesn't always help. However, again, and this is a more serious issue due to the privacy and mail integrity implications, delivering mail to the wrong address is against USPS regulations. This one's therefore DEFINITELY worth calling the postmaster and complaining about.
Unfortunately I've not yet followed thru on that, so can't say how effective it might be. Also note that if your complaint causes the local postmaster to come down hard on your local carrier, as it SHOULD for the regularly getting the neighbor's mail case, you could end up on not so friendly terms with said mail carrier. One would HOPE they'd be professional about it, but one would HOPE they'd not have to be told to deliver the mail to the correct address, too.
2) I've not tried this, but someone mentioned it somewhere I was reading, and it sounded reasonable, so I'm passing it on entirely without vouching for the authenticity thereof. Apparently, there's a regulation allowing people to file to block mail they consider obscene. Apparently, some folks took to using this to block advertising. Apparently, when the USPS didn't respond in such advertising cases, someone took them to court. Apparently, the resulting court decision included a ruling that only the individual can decide what's obscene to them, and thus, if they fill out the appropriate paperwork, the USPS is obligated to block mail from that sender to that recipient -- with serious legal repercussions if they fail to do so!
If this is correct and as should be plain from the language above I've no idea, just passing on what I read in the hope that it's useful, this method should allow you to block certain mail entirely, with legal action possible should the USPS fail to do so. Naturally, the USPS isn't entirely happy about such obligations and doesn't make it easy to find this information. However, a stop at the post office to request the obscene mail blocking form should get you on your way, and verify one way or another whether this is correct, in the process.
Hope that's useful and that postal services in other countries have something similar if not better. Of course, if anyone could verify the obscene mail block bit, preferably with a pointer to an authoritative source, it'd help me as well. =8^)
> [F]or some reason recycled paper doesn't make > a good final product and they have to pump > fresh paper pulp in to maintain quality.
AFAIK, that has to do with fibre length. As paper is used and then recycled, the fibres break and shorten, leading to "crumbly" paper if reused as-is. They add new fibre to maintain a minimum ratio of long to short fibres, thereby allowing coherent sheet production again.
You may have also noticed the "post-consumer" vs. "pre-consumer" content labeling. Pre-consumer recycling is stuff like trimming scraps and misprints. This material never reaches the end-user and as such has generally been held in known and controlled conditions. It hasn't been subject to the abuses of ordinary use and thus, while the fibre will be broken down to some degree, it won't be to anywhere near the same extent as post-consumer content. It's thus possible to substitute more pre-consumer recycled content for less all-new content. I don't know what the ratios are and they will certainly vary by target purpose, but if it's possible to use say 20% pre-consumer and 75% post-consumer recycled content with 5% new, it's likely that they can get the coveted 100% recycled label by using say 40/60 pre/post-consumer recycled content. Less post-consumer content, but now 100% recycled.
Duncan (not in the paper industry so numbers out of my ass, but they illustrate the situation as I understand it)
Thanks. I think the xhost bit was the piece of the puzzle I didn't have. I'm looking at putting it in place here now (but perhaps with setuid/setgid instead of sudo, so it's not possible to forget and invoke it without the sudo).
Talking about which, as I had thought of it awhile ago and saw mention of again recently, how hard is it to get Mozilla running as a low-priv alternate user, either nobody, or perhaps with slightly more rights, but only to its own dedicated home dir, for saving and the like? How does one manage the X sockets and permissions, etc, so it still displays and takes input fine in the regular logged in user's setup, but can't read/write to anything there but the screen(s) and sound device(s)? Can anyone point out a tutorial/HOWTO on setting up X based apps (such as firefox) to run as different users in a normal user's X/KDE session?
Yeah, XP=eXPrivacy to me. I had been looking forward to upgrading from '98 (which I had been in line at midnight for, after running IE4 from the public betas) to a full 32-bit kernel, until I read about the anti-privacy/privacy thing.
It was at that point I decided W98 was the end of the MS road for me, and got serious about switching to Linux. I installed Linux (Mandrake 8.1) the week eXPrivacy came out. As a power user not content with a default install, it took me ~3 months to learn how to configure build and install my own kernel, how to setup xf86config to handle triple monitors on dual cards, how to configure LILO to behave how I wanted re multiple boot drives, choose my apps for mail/news/web/media-player/etc, and in general get more comfortable on Linux than I was on MS.
After that, however, it was easily a power-of-two backoff rate in how often I booted what I soon termed MSWormOS. I remember booting to it after about six weeks to delete my local copy of all the power-applets I'd downloaded off the net over the years, go thru a bunch of stuff, move some to where I'd be able to easily find it from Linux, then delete the originals and a bunch of other stuff I no longer needed. I remember after having done so, taking a look at the desktop and wondering what else I could do on it, as I had on Linux when I first installed. Then I rebooted back to where I was comfortable, Linux.
About three months later, I repeated the experience, this time drastically shrinking the partition. About six months after that, I finally booted MSWormOS for the last time, did my last cleanup, rebooted to Linux and moved the last transferred stuff to its new Linux location, and deleted the MSWormOS partitions. I did keep a copy of the install dirs around for a couple more years, just in case, but never needed them. I finally deleted them after I upgraded the hardware to amd64, as I made room for my first test 64-bit install.
So really, I've MS to thank for finally pushing me over. After a decade on MS, I had been loath to just dump all that accumulated experience, but the anti-privacy stuff crossed a line I simply wasn't going to cross. Had Linux not been there, I'd have ironically been forced to piracy by the very thing that was supposed to stop it. So I'm very glad it was, and I'm very glad MS did that final shove, as who knows if I'd have ever actually made the jump otherwise.
Now I live in the land of software freedom, a defector still having friends and relatives back in the old country, ever willing to help them make the jump too if they decide to, but knowing he can and will never himself go back, not unless there's a revolution introducing freedom there as well. The sig is fitting, and I no longer allow such would-be slave-owners to be master over me.
Even if we keep getting exponential growth of transistors per dollar in the coming years, the question is what to do with them. Arranging them in useful circuits is increasingly difficult because at a certain point adding cache and execution units to a processor just isn't very helpful (hence multi-core).
I disagree with at least part of the above.
The problem is that you're not thinking outside the current box we're in. It's not that we have too many transistors and too much "cache", but rather that we have too few, and will continue to have too few for a few Moore's law generations yet.
Consider the effect once that "cache" reaches the half-gig and higher level (and consider that the current cache sizes are per-core, so multiply by the number of cores to get the total per chip size, since that's what we're now talking). Z-RAM http://en.wikipedia.org/wiki/Z-RAM is an interesting technology in this regard. At a half-gig, special applications (I'm assuming operating memory of say 8 gigs is common for general purpose by that point) will begin to make use of this memory as main memory. About 3-4 generations later, on-chip memory capacities will get very reasonable for general purpose use.
Currently, program design and compiler optimization must account for the fact that main memory is off-chip somewhere, running at order of magnitude lower clock rates compared to on-chip CPU speeds. Almost certainly, on-chip memory will still run at fractions of the CPU clock, but getting it on-chip will reverse the trend of the last decade plus of higher and higher CPU to memory speed ratios.
Remember the effect of moving L2 cache from mainboard to CPU daugher-board with the slot CPU generation, and then directly on-chip as transistor budgets allowed with the move back to sockets? Of course, other adaptations were happening at the same time, as main memory got even slower as compared to the CPU. That's where a lot of the out of order and speculative execution stuff that we just take for granted now days came from -- it was forced by the ever increasing speed of the CPU as opposed to RAM, but enabled by the larger/faster L1 and L2 cache sizes.
Now imagine being able to throw most of that away once again, as memory speeds actually catch up several doublers worth because they are now on-chip. That's what we're looking at in a few years, if Moore's law continues to hold and on-chip transistor budgets continue to increase.
The problem currently is that we're in a kind of lukewarm middle ground, increasing cache size doesn't bring so much bang for the transistor budget any more, but the budget is still a few generations too limited to fully integrate main-memory on-chip. As a result, we're devising all these other things to make use of those available transistors with. However, at some point, the transistor budget is going to get big enough to consider integrating main memory, and at that point, I suspect a lot of these other technologies will be at least temporarily dropped.
Yet another angle on the same thing is the fact that a decent share of current chip real estate and transistor budget goes to the whole speculative and out of order execution thing, as forced by the speed differential between CPU compute-cores and main memory. If memory speeds suddenly increase several-fold, the effect of sticking it on-chip, a good share of that speculative logic will no longer be needed as the cost of main memory access, both in direct latency and in branch prediction failure penalty, will be much smaller. Off-hand, I'd guess that'll cut a generation off the transistor budget wait right there, thus bringing it that much closer.
One thing's for sure, there'll be quite a bit of optimization un-learning to go thru once this occurs. Throw in the 1-T Z-RAM tech (assuming that still next-gen tech is actually viable both technically when fused with CPU foundary tech, and that it's even commercially viable at all), and
I could believe PB would chew close to 93K a month in costs if they have 10 million users. if 1/2 their registered users visit You apparently don't quite appreciate the magnitude of what they just announced, as mentioned in both TFS and TFA.
From the articles:
Today, The Pirate Bay asserts itself as the self-proclaimed "World's Largest Tracker" by topping over 10 million peers, while managing over 1 million torrents.
Let's consider these staggering numbers. 10 million simultaneous users represents a number never duplicated by any file-sharing entity
"Peers" actively bittorrenting are what a tracker tracks. Thus, they had 10 million accounts all active at the same time! So your "if half their [...] users visit" was only half of what you should have been considering! 10 million users all at the same time... using your estimate, that'd be 20 million regular users, and I'd consider that a low estimate, given the activity figures of say 1 in 10 at any given time that ISPs tend to use. Even if it's one in five, that's 50 million regular users, which I'd consider more likely than your one in two.
Looked at accurately, therefore, it's even more amazing. Pirate Bay doesn't even have to consider defining "regular user", as their active at any given instant numbers are so incredibly huge, they have no need to maximize or inflate the numbers using any other metric.
NB: I wonder how many of those peers are comparable bittorrent newbies, downloading that 17 gigs of myspaceprivatepicstorrent featured earlier here on/. I'm certainly in that group, having downloaded only two sets of files via bittorrent (ktorrent, FWIW) previously, Linux stuff not thru a public torrent site like PB, but having my curiosity get the best of me upon seeing that/. storry earlier, and started it up. I'm only on a cable connection, limited to 64 KB/sec (half a megabit) upload, and only averaging 70 KB/sec download, but I just passed 50% a bit ago, and expect to be done downloading some 32-ish hours from now. So I was one of those 10 million, and a new one too! =8^)
32-bit x86 Linux supports up to 64 GB of memory. It's actually given as a three-way choice in the kernel config, CONFIG_NOHIGHMEM (up to a gig of real memory, "off"), CONFIG_HIGHMEM4G, and CONFIG_HIGHMEM64G. Here's what the help text (from arch/x86/Kconfig, as in kernel 2.6.24-rc4) has to say:
Linux can use up to 64 Gigabytes of physical memory on x86 systems. However, the address space of 32-bit x86 processors is only 4 Gigabytes large. That means that, if you have a large amount of physical memory, not all of it can be "permanently mapped" by the kernel. The physical memory that's not permanently mapped is called "high memory".
If you are compiling a kernel which will never run on a machine with more than 1 Gigabyte total physical RAM, answer "off" here (default choice and suitable for most users). This will result in a "3GB/1GB" split: 3GB are mapped so that each process sees a 3GB virtual memory space and the remaining part of the 4GB virtual memory space is used by the kernel to permanently map as much physical memory as possible.
If the machine has between 1 and 4 Gigabytes physical RAM, then answer "4GB" here.
If more than 4 Gigabytes is used then answer "64GB" here. This selection turns Intel PAE (Physical Address Extension) mode on. PAE implements 3-level paging on IA32 processors. PAE is fully supported by Linux, PAE mode is implemented on all recent Intel processors (Pentium Pro and better). NOTE: If you say "64GB" here, then the kernel will not boot on CPUs that don't support PAE!
Also note that there's another config option that allows one to change that 3G/1G split for NOHIGHMEM mode, if desired. It's normally hidden, but available if one activates EXPERIMENTAL and I believe EMBEDDED.
That 3-level paging above 4 gig is a bit of a performance hit, as the kernel shifts its 4 gig window around in that 64 gig frame, tho if one runs the sort of apps that actually use that sort of memory, it's less of a performance hit than going to swap would be. Still, going 64-bit Linux isn't such a big deal any more, if your CPU supports it, and it's MUCH more efficient since multiple terabytes of memory can be directly accessed.
There's another factor at play here as well, however, and this applies to ALL OSs on 32-bit x86 and most or all on 64-bit x86 as well. It's a PCI hardware issue more than a software issue. Many old PCI devices were designed for 32-bit only operation, and their hardware can't address memory above the 32-bit 4 gig memory barrier. When memory was running less than a gig, this didn't matter much, and it became customary to reserve the virtual space at the top of the 32-bit address pool, 3.5-4 GB, for PCI device i/o access. As real memory expanded into that area, it runs into the reserved area, and the real memory behind it can't ordinarily be directly accessed.
Folks who've been around for awhile will likely recall a similar issue back at the 1 MB barrier, and how it was resolved using a "memory hole". The same technique is used here. With a BIOS setup to do so, one can configure a "memory hole" at the 3.5-4GB location, and the BIOS will remap the affected memory up above the 4 GB barrier.
This explains the complaints about Apple and MS platforms also having 4 GB look like 3 or 3.5 GB. I'm not sure if their 32-bit kernels can cope with that remapping or not -- they won't be able to if they can't address more than the 4 gigs anyway, but even if they can, the BIOS must be configured to map the hole as well.
Meanwhile, while addressing memory above the 4 gig line shouldn't be a problem for 64-bit kernels, the BIOS must still be able to do the remapping as well -- and the kernel must understand and deal with the hole. 64-bit Linux has suitable config options to do so, but I've not the foggiest how binary platform systems shipping a single binary kernel for all users deals with this. Primarily binary Linux distributions generally ship a number of different kernels, including an enterprise
... which is the whole point of Free(dom) software, of course, and why some people strongly prefer to work in a Freedom software based environment, where as long as they stay within the compatible license community, cross-company and cross-product code reuse is not only allowed but encouraged. In fact, some people strongly prefer it to the point where they won't do closed source, and pass up jobs that would require them to keep their work closed.
Funny/ironic* maybe, but as it's outsourced to a third party web survey company, and that web survey company happens to run MS platform, still understandable in this day and age of outsourcing.
Still, one would think they could find another solution provider running on a more suitable if only for appearances platform.
The problem is, this is apparently the first year the survey has made it to the popular media, and it shows in a multitude of ways. The server platform is one. That they seem to have not considered the possibility of actual home Linux users, designing the survey for business use only, even tho the announcement said nothing of that fact and was carried widely in the more general Linux-tech media, is perhaps the bigger one. The results are heavily skewed as a result. See comments I and others make elsewhere.
* Being one who sees language as dynamic, I'm a descriptivist, and see the examples in the song "Ironic" as just that, because that's the popular usage of the term, making it correct by definition, in dynamic language terms.
While the word I'd use is "skewed", I'd basically agree with you.
There was an original announcement of the survey in the Linux-covering media, and I looked at it but didn't take the survey then, as it seemed only interested in business use. Later, there was additional coverage, asking where all the North American users were, as there had been relatively few such responses to the survey at that point. Most were and still are European, altho the North American response percentage increased from about 10% to about 40%.
Anyway, since they wanted N. Am. responses, I went ahead and filled it out then, doing the "choose the best fitting answer" routine, even where the obvious answer for home users was entirely missing. The best available fit classification was small-office/home-office use, with 1-100 machines. I then had to pick an annual income range, with I think the low end being a million or less. Well, millionaires aren't as uncommon as they used to be, but that's still rather ridiculous for a low end choice on individual income. Again, it's obvious they think all Linux users are corporate, since the announcements asking for participation said nothing about business use only, only Linux use. Later, they asked a question of what my primary business was. Using best-fit logic, I think I originally chose health care, as I reasoned as an individual, the most basic purpose will be to maintain my own health and survive. That ended up conflicting with an answer later, so I went back and picked something else.
The survey is therefore incredibly skewed, because it makes an invalid assumption, that all Linux users are business users, and/or that the home-only user response will be so tiny (due to discouragement based on the obvious slant if nothing else) as to be ignorable.
The non-biz response may have been low enough that the home-user response assumption may have actually been the case in times past, due both to coverage and to obvious slant, but I think this year's was obviously skewed by the second round of coverage, asking for more North American user responses. Given the coverage I saw (including user comments on the stories), I believe it reasonable to assume that at least half of that response rate increase for N America alone was home users. That would work out to ~20% home users, minimum, dramatically skewing the results since it's incredibly obvious they were essentially ignoring the home user when they designed the survey, and didn't intend for home users to respond. Well, then why/ask/ them to respond in general community coverage?
IMO, they therefore got the skewed results they asked for and that could have been predicted given where and how the thing was publicized. No WONDER their SOHO segment jumped so dramatically! No WONDER some of the results don't follow previous trends!
Maybe next year they'll include a home user option and home user appropriate options to the further questions as well. Even if they are entirely uninterested in that segment (it doesn't spend enough money, maybe, because much of it simply downloads the free versions, and doesn't spend on the ISVs either), providing options for it would be wise, as doing so would then allow clean separation of what they consider "noise" from the signal they are really interested in, the big-money corporate accounts.
BTW, I was one of the Gentoo respondents, and put in the comments something else the ISVs etc they are apparently targeting the results for aren't likely to like -- that as far as I'm concerned, if it's not freedomware, if it doesn't allow full use of the four freedoms equally to all users and potential users, it's not a solution I can or will consider. So much for the Adobes of the world and their Linux ports. They might as well be un-wine-able MS-platform only, at a price of a trillion dollars a seat, for all I care -- if they aren't freedomware, they are that far out of usable-here solution scope. (It's a legal matter, see. I no longer sign away my rights, including the right to redress for security issues if I can't see and use the source for software running on my machine, so I can't agree to the EULAs, and the software is therefore not a legally viable solution, even if I DID wish to use it.)
Two good resources I found were the list of providers at Wikipedia, and the VoIP forums at BroadbandReports.com. Many on the the Wikipedia list won't apply to you as they're limited area or whatever, but the list is long enough, there's likely to be some in your area anyway.
You can also google some of those you find there and/or VoIP and/or SIP in general, for more resources. Take your time deciding.
Of some interest, Vyke is a Norway HQed dialout-only provider, but at 4 cents (US) for call setup, no per minute fee, it/could/ be be cheaper than going with a full two-way provider. I tried it but being European based, my US credit cards wouldn't clear, so I've honestly no idea how well it works. The parallel to it for inbound would be CallCentric.com's inbound plans, particularly if you are in an area where you can get a Dirt Cheap number from them. The drawback of using separate sources for each is that services like conference calling, etc, can be harder to manage than integrated plans. I didn't try CallCentric either, however, as I intended to setup the Vyke thing for outgoing first, and when that fell thru, I went with an integrated instead.
Of course, Vonage is one of the big, heavily advertised providers, but purely for that reason (and because their prices seem to run ~$5/mo higher than much of the competition), I had a negative gut reaction to them and never seriously considered them. Other than the price difference, that's purely personal/emotional, so feel particularly free to discount my opinion there. As they say, YMMV.
If you go looking for hardware, the big two that seem to have hardware on the retail shelf are Vonage, and Skype. Skype is SIP based but has a proprietary element, meaning anything you buy to work with them is locked to them. That's not my thing, but they do seem to charge rather less. Another lockin-only provider is Lingo. I was actually somewhat tempted there, and they have some quite reasonable International plans, so if you have contacts in Asia or Western Europe, despite their lockin, they could be worth checking out. Just realize that any hardware you get to work with them may or may not be unlockable, if you decide they aren't your style. Lingo is BTW a Primus company. I understand they are more known in Canada than here, and not all favorably. I won't say more as I'm not in a position to judge one way or the other from here, but it may be of significance one way or the other to you.
The Gizmo Project is associated with SIPphone. They're claim to fame is a software phone that does free calls (with catchs), but I only do Free Software and it's not, so it was out much as Skype was.
Wengo is active in the Linux community with Wengophone, which IS free software. Their setup didn't seem to match my needs, however (IIRC they were mostly European or something? don't remember for sure), or I'd have probably preferred them.
Sunrocket WAS a pretty big provider, but just went belly up recently. Particularly if you choose to prepay, you'll want to ensure whoever you choose has enough corporate resources either directly or because they are owned or backed by someone bigger, that's not likely to happen.
Offering the type of computer independent no-lockin hardware based service (tho if you get the hardware from them, it's likely to be locked) I was after, Packet-8, CallCentric (other plans then mentioned above), ViaVoice.com, etc. Similar hardware based service but not lockin free is Lingo as I mentioned, and Vonage (they have locked stuff widely available, don't honestly know if they are standard based enough to work with unlocked or not).
As I said, going rate is roughly $25-30/mo bottom-line, or ~$200/year prepaid (plus fees brings it to ~$230-250/year, right around $20/mo. That's for "unlimited" but consumer level calli
Seriously, consider VoIP. Because VoIP is actually a competitive market, all the extras the telcos charge an arm and a leg for are normally thrown in at the base price. Voicemail, caller-id, conference-calling, selective-call-block, etc, all included. Of course, US/Canada calling is all included too. Some offer additional features like scheduled do-not-disturb times (with selective punch-thru for your kids or mom or whoever, if you prefer), while others offer wider international calling (15 nation or so, more available but at additional cost), as part of their differentiated feature-set. I chose a company and plan with a few more features, since the US/Canada covers my calling area needs.
I have it setup here so blocked caller-id or out-of-area gets routed thru a prompt to enter a series of numbers, thus eliminating the machine calls while encouraging folks to unblock, but I could set it to totally block them if I preferred. No additional cost, of course.
Now while I did choose not to port my number, as my old regular telco number was getting the usual number of undesired calls, guess how many such calls I've gotten on the VoIP since I switched. None. Nada. Zilch! =8^)
Call quality is comparable to the former bell telco in the area, or to cell, but rather worse than the digiphone service I had from the cableco previously. (That was the best call quality I expect I'll ever have, but at comparable to telco costs, which is much higher than VoIP.)
Cost, US Dollars, ~$25/mo BOTTOM LINE INCLUDING TAXES AND FEES depending on provider, or ~$20/mo ($200/yr plus a couple bucks a month taxes) annually prepaid. (Renewals can often be had for ~$150 annual prepaid.) I'm deliberately not mentioning the VoIP provider I chose as this isn't an ad, but those are the going market rates, so pretty close to what's available from several VoIP providers. (Note that all incoming or all outgoing can be had for less, this is for both, and including full US/Canada bundled calling area. Skype's less too, but locked provider and not standardized, so I avoided that route.)
What's nice is that while I had been skipping out on the extra fee caller-ID as I had found it just wasn't effective for what I wanted it for and it cost extra, now that it's bundled, I upgraded my phone (I chose a VoIP adapter that a normal phone plugs into, tho dedicated VoIP phones are available) to a voice-announcing caller-ID phone, with distinctive ring as well. Thus, after the first standard ring, it announces who is calling based on the (no extra cost) caller ID, then switches to distinctive-ring if I have that ID programmed for it. No more looking at the caller ID, or waiting for the answerer to pickup and listening to it to find who's calling, before I decide whether it's worth picking up! Thus, even if a marketer does call (none have), it'd just announce caller out of area or whatever, and I'd just let the voicemail pickup.
Yes, you wanted to (well, as much as anyone "wants" to pay anything, for anything, anyway). After all, they didn't call up your work and hold your check hostage unless you paid them $115/mo, nor were they withholding your food, water, air, or even your shelter, if you quit paying. The service was self-evidently worth at least that $115/mo. to you or you wouldn't have been paying it, TV or no TV.
After all, TVs aren't like food, water and shelter, or even like phone service or a car with gas, insurance and maintenance. It's arguable people need those, tho plenty of folks do get by without the latter set anyway. I've yet to see anyone demonstrate they/need/ TV, so it was and is a luxury... a luxury you were willing to spend at least $115/mo. on.
Note that I haven't had TV in years, so it's certainly doable. Computers took up the time I used to spend watching TV. I "didn't want" to pay the $1100+ I just paid for the pair of dual-core Opteron 290 upgrades I have coming in this morning, either, but it turns out I wanted them more than the money, and more than other stuff I would have otherwise purchased with that money, after all! =8^) It would appear you likewise wanted the service more than that $115/mo, or anything else you could have done with that money, too, so yes, you wanted it, were willing to pay for it, and thru negligence if nothing else, put yourself in a situation where you WOULD be paying for it! All they did is see a mark more than willing to pay that price, and take advantage of that fact.
Thanks for the reply. Great to be able to actually get the view from the trenches, so to speak. I honestly do appreciate your taking the time and trouble, as I am seeing things in your comments I hadn't considered before. I do hope I'm returning the favor. =8^)
Pretty much all shoots have a fixed fee.
The time messing with paperwork is a waste of time, an invasion of my privacy and a big hassle for all parties involved. [snipped tho I appreciated it]
Add in [other hassles] and the government is slowly wasting half my life for what is essentially a hobby - cuz porn doesn't pay well at all for performers.
Fixed fee, yes, but what I was arguing was that the paperwork time is in effect built into that fee, or you'd not be doing it, unless of course it was for charity.
Perhaps I should have added "or as a non/low-profit hobby". I don't know what the industry pay rate is (nor am I implying you need to say as it is literally not my business, tho if you wish to, I'd be interested), but for argument's sake, let's say it amounts to reimbursement of expenses and perhaps a small stipend, similar to the work many "volunteers" do. That'd be effectively a hobby or charity.
However, the larger point remains, for any activity someone continues to engage in, they either consider it "worth it" in the end, or they don't, and if they don't, they don't continue engaging in it for long. This is regardless of what form the "pay" takes, whether it's in dollars or other reward (entertainment, satisfaction of some ideal as a volunteer, whatever). If you continue in the activity of being an "adult entertainment performer", you obviously consider it rewarding enough, that is, more personally rewarding than other alternatives open to you to spend your time. Maybe that reward is more than the money. Maybe it is a hobby, where the money is actually second place or in fact you spend money on it. Heh... that's what a hobby is for most folks, anyway, something they spend money on. So if you love it enough to consider it a hobby, and get at least partial reimbursement for your expenses, consider yourself lucky. Not so many people get that with their hobbies (altho one of the nice things about the FLOSS community is that many actually do end up getting paid for their hobby... tho I'm not one of them).
The point still stands, however. Taken as a whole, that time filling out paperwork is in fact paid, if indirectly, because it's a necessary requirement of the activity you obviously get reward for (money or otherwise) enough to continue, and if you were in fact not getting paid enough to justify that time as well, you'd find something else to do with your time.
Actually, this is something I as a "victim" had to learn the hard way. For some time I had a problem repeatedly falling into classic "victim" mode, where, if you are familiar with psychology, reality gets distorted and the "victim" simply cannot see what to others look like obvious ways out. It took a crisis (basically a mental breakdown) but eventually I realized that many people, not only victims, fail to see and appreciate most of the choices they have, and thus to proactively act to make the best of them.
Since I realized that, I've consistently endeavored to personally take the proactive route, being deliberately assertive in the choices I make. I've personally been much happier as a result, and thus have tried to confront the "no choices" aka "victim" mode when I see it in others, to hopefully challenge a few assumptions and get them out of "victim" mode and into "assertive life enjoyment mode" if I can as well.
The example I'm usually arguing is the broadband ISP example. People often say they have no choice due to monopoly conditions. That is in fact false, as can be seen by the thought experiment question of what would they do if that provider failed or pulled out, no longer offering that product in that market. All of a sudden,
Slashdot Mirror
> When I think of a red cross, I immediately
> think of emergency relief, supplies, etc.
That's what I thought of as a kid. Now, I know the history previous to that. Now, I think of the Crusades and the terrible things done in the name of Christianity during them. That's where the red cross on a white background originated.
The Knights Templar (Temple Mount Knights, who took a vow of poverty (communal property, as the term meant at that time) and allegiance to the order, who originally protected the pilgrims during a period when most other knights were ignoring them because there wasn't profit in their protection, who were a time the most powerful Military Christian order after so many donated so much in turn, who set up the first European and Crusader banking system, had as their naval flag the original white skull and crossbones on a black background later associated with pirates, who traded with the Muslims and introduced gauze (from Gaza, the city), muslin (from Muslim), and candy (from al-kandiq) these were one of the orders that flew the red cros on a white background, the original symbol of the crusades.
Similarly with the Knights Hospitaller, whose order founded and staffed the medical "hospitals" the pilgrims and Crusaders used. This is actually the origin of both the word "hospital" and the association of the red cross with medicine. Of course, while they were a welcome sight to many a wounded Christian Crusader or pilgrim, obviously, the Muslims had a rather different viewpoint! No WONDER they couldn't tolerate the Red Cross as a medical symbol! To them it meant the rape and pillage, the savagery of the Crusades! So they went with the Red Crescent.
However, the Red Crescent is also a religious symbol, so today there's a third symbol coming into use, designed to be religiously neutral while still being unmistakable for anything else. It is often used in international contexts, particularly in the mid-east. This is the Red Crystal, a diamond shape.
So, ideally, the Red Crystal will eventually come to have the positive associations of the Red Cross and Red Crescent today, without the sectarian and cultural negatives. It's also worth noting that Cross, Crescent, and Crystal, all start with C, so the ARC (and other "Red Cross and Red Crescent" organizations worldwide) could adopt the Crystal without even changing initials.
Wikipedia is a place to start, anyway. There are links from there elsewhere, and I've included another informative link on the Knights Templar as well as a Google link, below.
Wikipedia Red Cross/Crescent/Crystal (and some others)
http://en.wikipedia.org/wiki/Red_Cross_(symbol)
Wikipedia Knights Templar
http://en.wikipedia.org/wiki/Knights_Templar
Wikipedia Knights Hospitaller
http://en.wikipedia.org/wiki/Knights_Hospitaller
Also see Wikipedia Teutonic Knights
http://en.wikipedia.org/wiki/Teutonic_Knights
A rather informative (altho obvious viewpoint taken) history of the Knights Templar (on first page of the below google)
http://www.electricscotland.com/HISTORY/kt1.htm
Google on: "red cross" history crusades
http://www.google.com/search?q=%22red+cross%22+history+crusades
... Except that... for copyright law to be abolished, there'd have to be a sea change in conventional viewpoint.
/legal/ to go proprietary, to do so would be considered an indication of something to hide, being shady in one's practical dealings regardless of whether it were actually legal or not. Legality in that case wouldn't matter that much, and while there would still be people at the fringe who'd do it much as there are malware/spyware authors and spammers today, some "legal", some not, it'd be considered a scorge on society, something no educated person would admit to running, just as it's not acceptable in educated portions of society today to admit that one actually makes purchases based on spam.
If sharing became the accepted norm (not just legally accepted fact), as it very well could be if the conventional view on the issue changed sharply enough to make it legally acceptable (copyrights abolished), then those who dared go against that society accepted norm, producing proprietaryware (which could then be seen as slaveryware, see my sig), could very well be viewed with the same contempt as most of society holds malware authors and spammers today. Even if it were
It would thus likely follow any sea change in public viewpoint to the degree necessary to abolish copyright, that the legal enforcement mechanisms of the GPL as we know it today would no longer be necessary in practice, even as they fail to be legal enforcement mechanisms. This I say as a strong copyleft supporter today (to the point I don't seriously consider the BSDs as they don't actively support user rights, only take the agnostic view, making them possible... and to the point I chose the sig I did).
Of course, both the copyright abolished scenario and the implications that would follow are so far out of practical reality today that it doesn't matter a whole lot, except as an epistemological (which is I contend what real science fiction, and many non-scifi novels are, as well, the biggest reason I enjoy them). It may well matter at some point in the future, but it's not likely in the immediate future and may indeed not occur for many generations, if ever. Still, it's worth exploring, both for the implications should it happen then, and for those it has on present reality and our approach to it.
One of the better ways to block ads -- and some malware -- is to use a "personal proxy" such as (for MS Windows) the Proxomitron or (for *ix and MS Windows both) Privoxy. I used the former years ago on MS, and now use the latter, Privoxy, based on the old Junkbusters proxy code but vastly improved since then.
These "personal proxies" normally run on your own machine, listening on the loopback interface (127.0.0.1) for connections from various HTTP clients. As proxies, they are client agnostic (so Firefox is fine, so is Konqueror, so is IE), as long as said client can be configured in some way to use a proxy.
They work as a rewriting (as opposed to caching) proxy, using a configurable ruleset to intercept and rewrite or block both outgoing and incoming headers and content as desired. For ad-blocking, a double-layer approach is taken. First, as a page is downloaded, many references to ads will be rewritten, either deleted from the page, often with a tiny-text note saying what filter did it (this is what I do with many of my own filters, as it aids in debugging if I get it wrong), or rewritten to replace the ad with a placholder (safer for the general case as it doesn't disturb the layout of the page), often with an "override, go there anyway" type link (just in case). Since the placeholders are local and the the deletions obviously aren't going to have the browser expending additional bandwidth downloading content that has been entirely erased from the page the browser sees, that's all saved bandwidth.
Second, the proxy can be configured to block access to certain sites (say anything doubleclick related, to use one common example) entirely, and to block but with override possible in other cases (say anything with *ad* in it, since that's pretty broad when you have "road", "adsl", etc, altho the default filters are already smart enough to not block those real obvious things). Thus, even if an ad reference gets thru, when the browser attempts to load it, either because it's part of the page, or because you perhaps accidentally clicked where the ad would have been, it'll block (with or without override option) the fetch, saving more bandwidth.
The most obvious malware this blocks in a lot of the tracker stuff, which is often interwoven with ads to the point they're the same thing anyway. However, quite apart from any browser based scripting and cookie and popup blocking (for instance), the proxy has its own configurable settings. Of course, it can be set to block other stuff too, some of which it does by default (with for privoxy, a low, medium, and aggressive setting, for those not wanting to fool with individual rules), some of which you'd create your own custom rules to deal with, as appropriate. It's sort of like Firefox's greasemonkey, but more powerful. The fact that it's all browser agnostic is pretty sweet too, for those of us that use multiple browsers.
If you want to see what a page looks like "unfiltered", or just want to browse without the filter, that's possible too. The proxies have a "bypass" mode that bypasses the normal rulesets, simply forwarding the requests and replies as they come. Most browsers also have a proxy bypass toggle as well. It's also easy enough to set some things to use the proxy, without setting everything to use it. For instance, while the proxy could be configured to ignore them (and would generally do so after a bit anyway, as the stream exceeded the max-size-filter setting), you probably don't want it proxying streaming audio or video, as that's just a waste of resources. Just don't setup those clients to use the proxy. OTOH, back on MS Windows, I DID find it handy to run my spinner.com streaming client thru the proxy, since it killed the ads that way.
As an example of some of the other stuff it can do unrelated to the above, I personally prefer light text on dark backgrounds, while most of the web is setup in the reverse, light/white background, dark/black text. While it is possible to set preferences in most browsers, those who
Even the OS is often customizable, but unfortunately generally only between MS variants. That it can be switched out is self-evident both from the fact that it can be customized and by the wide availability of LiveCDs/DVD/Thumbdrive distributions that don't even touch the hard drive installation. It's even further self-evident from the EULA, which says send it back to your machine dealer for a refund if you don't agree to it.
Which is what people try to do. It's nice to see some places actually enforcing the EULA MS expects to be enforced for it, for the consumer as well, and the dealer having to eat the cost, as they should having chosen to distribute it knowing the EULA and its conditions obligating them to refund upon turn-in.
The US has similar anti-bundling laws. If the powers that be had any concern for the consumer, they'd enforce them as well. Europe seems rather ahead of us in that regard, among several others (software patents... tho behind in still others, like UK's vid monitoring and Germany's anti-security-testing-software laws).
OK, this is like two weeks old by now, but I just got it in metamod, and marked the off-topic as unfair. I read thru all four pages of comments and there were others discussing config and stuff, tho nothing quite like this one. Still, it wasn't OT given the context of the other threads, and the fact it WAS about Linux desktops (maybe the mods didn't consider a laptop close enough to a desktop?).
FWIW... make sure you're comparing the same probe temps. At least my AMD (desktop/workstation, so may be different) system has cpu core temps, and cpu case temps. The core temps run higher (typically ~5 degrees C), but due to volatility are *NOT* the temps AMD tells its system vendors to standardize on for system thermal regulation. It's thus entirely possible that (even if they are labeled core) the MS Windows reported temps are the case temps that the system is regulating on, while it's likely the actual core temps that you have chosen to have reported on Linux.
Of course, if your thermal regulation isn't working and the fans aren't coming on or speeding up when they should, it'll hit warning territory pretty fast regardless, but just in case you weren't aware of the possibility of confusing the two readings. Of course if you only have one of them...
So you get a "Titanicized" yacht... with a nice big hole sliced in it, and no way to fix the hole...
Hmm... why am I hummming the Giligan's Island theme? =8^)
As I posted up-page in the slashdotting discussion, I like to run compiles in -j unlimited "jobserver" mode just to see how far I can push the hardware. Back before I upgraded from the dual-single-cores Opteron 242s (1.6 GHz clock), jobserver-mode (Linux) kernel compiles would run up 480 on the 1-minute load average before the (KDE/X) ksysguard graph would freeze up and I couldn't tell how much higher it'd go.
After upgrading to top-of-the-line-for-socket-940 dual dual-core Opteron 290s (2.8 GHz), with the user based priority/scheduling in kernel 2.6.24+, I no longer see those GUI freezes so continue to get load average updates, but the system is fast enough it apparently retires the first jobs before the last are loaded, so now job-server mode kernel compiles max out at only ~180 or so on the 1-minute load average. (FWIW, both the dual 242 480+ load and the 290 180+ loads are with 8 gigs RAM, the 480+ load would run into swap, but the 180 peak I manage on the 290s never touches swap.)
However, the nice bit is that with the per-user or control-group (I use the former) scheduling available in kernel 2.6.24+, I can run practically unlimited load average compiling as a different user, as long as I don't hit swap too hard, while all the while maintaining hitch-free net radio playback and nearly hitch-free visualization (it's slightly jerky updating but I don't miss many frames). The limit isn't on the load average aka CPU cycles, but rather on real memory vs swap. If I'm not streaming, I can maintain "acceptable" if somewhat jerky interactivity several gigs into swap, altho that's certainly partly due to the fact that swap is 4-way-RAID-0/striped.
So the 300 load average on Solaris isn't really all that special on Linux either, as I used to routinely do 480+ load average "just for fun" compiling the kernel, and that was before the recent very dramatic scheduler improvements. And be sure to check out that 2.6.24+ per-user scheduling as it's certainly worth it.
How's it get up to 80? I've seen a eight core server hit hundreds.
Mind you that was after I made a mistake on a production server which essentially caused all the site's traffic to turn in to a DDoS attack.
Since then I've upgraded to the top of the line for socket 940, dual dual-core Opteron 290s (2.8 GHz), and with the recent per-user scheduling options turned on in the kernel, can watch the 1 min load average run to 180 or so (the highest it gets) without ever going unresponsive or stopping the 2-second ksysguard updates. Apparently with the extra speed and cores, the first jobs get worked thru and retired before it ever loads the last ones, so load average (and memory usage) never get anywhere as high.
What's great tho is as I mentioned the new (well, as of kernel 2.6.24) per-user scheduling/priority options. Set properly, you can have one user running multi-hundred load averages and it barely affects usability for other users at all, because they get equal CPU time (and with the proper I/O scheduler, I use CFQ which by default prioritizes I/O to match CPU priority, I/O to match) if they've got runnable processes, even with only fractional load averages. This is certainly one of my favorite new kernel features and one I regularly use to continue almost hitch-free internet radio playback and even visualization, plus of course general system responsiveness, even when compiling at multiple jobs per core load average. (I normally keep MAKEOPTS to "-j -l15" for general purpose compiling, which ends up running about a 16 load or 4 per core, more to limit memory usage to not run into swap (compiling with the workdir on tmpfs) than load average, tho.)
With a 120-ish day uptime... well, maybe it's a 2.6.24 kernel, but probably not. However, with that kernel or better, an 80 load average, or even a 500 load average, really shouldn't be unmanageable at all, as long as the scheduler options are set right, and if user-based scheduling groups are turned on, apache is running as a user other than root and other than the remote login shell would be running. As such, by default, root would get double the CPU cycles allowed other users (and appropriate I/O, assuming CFQ or similar prioritized I/O scheduling with priorities similar to the CPU priorities), and no matter what the apache user load average, the remote login shell should get equal cycles (assuming equal process priorities). It should thus be nearly impossible to DOS the machine to the point remote administration to get in and correct the problem becomes difficult or impossible.
Just in case it's not apparent by now, I *REALLY* like that aspect of kernel scheduling, new at least to the Linus kernel with 2.6.24. =8^)
Taking a look at the pic... From the detail I could see at 2X magnification, it appears there's no actual "breaker" there -- the magnetic core that's supposed to "trip" was simply replaced by a plain wire. As best I could tell, and this part I couldn't be sure of from the picture but it seems logical, they had it rigged up to switch off if manually turned off, so it does look like it should function as a regular switch, thus effectively deceiving the user until the moment of truth, but the bulk of the thing, the part that should trip it automatically on overload, simply ISN'T THERE!
/lucky/ (and the counterfeiter has /some/ conscience), it might fuse out, vaporizing the wire (or another component) rather than continuing to short and starting a fire or whatever, but even then, it'd be a (switched) fuse, not a resettable circuit breaker, dead on first use.
/too/ unacceptable, or at least I can imagine the counterfeiter rationalizing such to himself. (I still have trouble rationalizing the on-block to myself, but whatever, I've seen it enough places, and with blocks obviously marketed for the purpose, to know it's gotta be accepted and legal, regardless of what they tell us about the safety of the system. Of course, we have 125V standard mains. It's possible such things aren't allowed at all on 250V standard mains, which I suppose they have where those were discovered.)
The wire looks a bit heavy for it but depending on the rated amps and the composition of the wire, if a buyer is
More likely it'd cause the main breaker -- hopefully there is one and it's not similarly rigged -- to trip, cutting off not just the single circuit but the entire panel along with all its circuits. Considering, at least over here in the US, that occasionally, for circuits such as alarm panels, security systems, and relay timing circuits, it's apparently accepted and not uncommon to see their breakers blocked ON, thus allowing serious overload to the point the load gets enough to trip the main breaker, this isn't
Duncan
Snail mail spam does allow a couple ways to get rid of it upstream.
/their/ part. First, as mentioned, those little address cards often get separated from the bulk of the newsprint and etc that forms most of the ad bundle. Technically, this is a violation of USPS regulations, the mail must be delivered in its entirety including the address card, and you can call your postmaster to complain.
1) At least here in the US, while the postal service doesn't do traditional blocks (but see below), I believe the regulations require advertisers to remove one's name from their lists if asked to do so. The mail must be individually addressed to make that possible, altho it can be addressed to "occupant", but it must have an individual street or whatever address. I believe bulk rate must also have something (often the small card with the individual to address; small cards often get lost and are in any case hard to find among the newsprint) listing the sender, as well, so a customer may send a removal request.
I've asked for removal from several regular mail spammers around here, and they've done their part. The problem however is that the post office doesn't always do
Second, with bulk mail, the tendency of the carrier is often to just shove it into each box sequentially, never looking at the address on the card. After you get your name removed from the marketer's list (as I mentioned, I've had no problems with that part), you'll often still get it -- if you look at the address however, it was to the neighbor next in line (or two or three down, if several have had their addresses removed). Thus, taking your name off lists doesn't always help. However, again, and this is a more serious issue due to the privacy and mail integrity implications, delivering mail to the wrong address is against USPS regulations. This one's therefore DEFINITELY worth calling the postmaster and complaining about.
Unfortunately I've not yet followed thru on that, so can't say how effective it might be. Also note that if your complaint causes the local postmaster to come down hard on your local carrier, as it SHOULD for the regularly getting the neighbor's mail case, you could end up on not so friendly terms with said mail carrier. One would HOPE they'd be professional about it, but one would HOPE they'd not have to be told to deliver the mail to the correct address, too.
2) I've not tried this, but someone mentioned it somewhere I was reading, and it sounded reasonable, so I'm passing it on entirely without vouching for the authenticity thereof. Apparently, there's a regulation allowing people to file to block mail they consider obscene. Apparently, some folks took to using this to block advertising. Apparently, when the USPS didn't respond in such advertising cases, someone took them to court. Apparently, the resulting court decision included a ruling that only the individual can decide what's obscene to them, and thus, if they fill out the appropriate paperwork, the USPS is obligated to block mail from that sender to that recipient -- with serious legal repercussions if they fail to do so!
If this is correct and as should be plain from the language above I've no idea, just passing on what I read in the hope that it's useful, this method should allow you to block certain mail entirely, with legal action possible should the USPS fail to do so. Naturally, the USPS isn't entirely happy about such obligations and doesn't make it easy to find this information. However, a stop at the post office to request the obscene mail blocking form should get you on your way, and verify one way or another whether this is correct, in the process.
Hope that's useful and that postal services in other countries have something similar if not better. Of course, if anyone could verify the obscene mail block bit, preferably with a pointer to an authoritative source, it'd help me as well. =8^)
Duncan
> [F]or some reason recycled paper doesn't make
> a good final product and they have to pump
> fresh paper pulp in to maintain quality.
AFAIK, that has to do with fibre length. As paper is used and then recycled, the fibres break and shorten, leading to "crumbly" paper if reused as-is. They add new fibre to maintain a minimum ratio of long to short fibres, thereby allowing coherent sheet production again.
You may have also noticed the "post-consumer" vs. "pre-consumer" content labeling. Pre-consumer recycling is stuff like trimming scraps and misprints. This material never reaches the end-user and as such has generally been held in known and controlled conditions. It hasn't been subject to the abuses of ordinary use and thus, while the fibre will be broken down to some degree, it won't be to anywhere near the same extent as post-consumer content. It's thus possible to substitute more pre-consumer recycled content for less all-new content. I don't know what the ratios are and they will certainly vary by target purpose, but if it's possible to use say 20% pre-consumer and 75% post-consumer recycled content with 5% new, it's likely that they can get the coveted 100% recycled label by using say 40/60 pre/post-consumer recycled content. Less post-consumer content, but now 100% recycled.
Duncan (not in the paper industry so numbers out of my ass, but they illustrate the situation as I understand it)
Thanks. I think the xhost bit was the piece of the puzzle I didn't have. I'm looking at putting it in place here now (but perhaps with setuid/setgid instead of sudo, so it's not possible to forget and invoke it without the sudo).
Talking about which, as I had thought of it awhile ago and saw mention of again recently, how hard is it to get Mozilla running as a low-priv alternate user, either nobody, or perhaps with slightly more rights, but only to its own dedicated home dir, for saving and the like? How does one manage the X sockets and permissions, etc, so it still displays and takes input fine in the regular logged in user's setup, but can't read/write to anything there but the screen(s) and sound device(s)? Can anyone point out a tutorial/HOWTO on setting up X based apps (such as firefox) to run as different users in a normal user's X/KDE session?
Duncan
Yeah, XP=eXPrivacy to me. I had been looking forward to upgrading from '98 (which I had been in line at midnight for, after running IE4 from the public betas) to a full 32-bit kernel, until I read about the anti-privacy/privacy thing.
It was at that point I decided W98 was the end of the MS road for me, and got serious about switching to Linux. I installed Linux (Mandrake 8.1) the week eXPrivacy came out. As a power user not content with a default install, it took me ~3 months to learn how to configure build and install my own kernel, how to setup xf86config to handle triple monitors on dual cards, how to configure LILO to behave how I wanted re multiple boot drives, choose my apps for mail/news/web/media-player/etc, and in general get more comfortable on Linux than I was on MS.
After that, however, it was easily a power-of-two backoff rate in how often I booted what I soon termed MSWormOS. I remember booting to it after about six weeks to delete my local copy of all the power-applets I'd downloaded off the net over the years, go thru a bunch of stuff, move some to where I'd be able to easily find it from Linux, then delete the originals and a bunch of other stuff I no longer needed. I remember after having done so, taking a look at the desktop and wondering what else I could do on it, as I had on Linux when I first installed. Then I rebooted back to where I was comfortable, Linux.
About three months later, I repeated the experience, this time drastically shrinking the partition. About six months after that, I finally booted MSWormOS for the last time, did my last cleanup, rebooted to Linux and moved the last transferred stuff to its new Linux location, and deleted the MSWormOS partitions. I did keep a copy of the install dirs around for a couple more years, just in case, but never needed them. I finally deleted them after I upgraded the hardware to amd64, as I made room for my first test 64-bit install.
So really, I've MS to thank for finally pushing me over. After a decade on MS, I had been loath to just dump all that accumulated experience, but the anti-privacy stuff crossed a line I simply wasn't going to cross. Had Linux not been there, I'd have ironically been forced to piracy by the very thing that was supposed to stop it. So I'm very glad it was, and I'm very glad MS did that final shove, as who knows if I'd have ever actually made the jump otherwise.
Now I live in the land of software freedom, a defector still having friends and relatives back in the old country, ever willing to help them make the jump too if they decide to, but knowing he can and will never himself go back, not unless there's a revolution introducing freedom there as well. The sig is fitting, and I no longer allow such would-be slave-owners to be master over me.
Duncan
Even if we keep getting exponential growth of transistors per dollar in the coming years, the question is what to do with them. Arranging them in useful circuits is increasingly difficult because at a certain point adding cache and execution units to a processor just isn't very helpful (hence multi-core).
I disagree with at least part of the above.
The problem is that you're not thinking outside the current box we're in. It's not that we have too many transistors and too much "cache", but rather that we have too few, and will continue to have too few for a few Moore's law generations yet.
Consider the effect once that "cache" reaches the half-gig and higher level (and consider that the current cache sizes are per-core, so multiply by the number of cores to get the total per chip size, since that's what we're now talking). Z-RAM http://en.wikipedia.org/wiki/Z-RAM is an interesting technology in this regard. At a half-gig, special applications (I'm assuming operating memory of say 8 gigs is common for general purpose by that point) will begin to make use of this memory as main memory. About 3-4 generations later, on-chip memory capacities will get very reasonable for general purpose use.
Currently, program design and compiler optimization must account for the fact that main memory is off-chip somewhere, running at order of magnitude lower clock rates compared to on-chip CPU speeds. Almost certainly, on-chip memory will still run at fractions of the CPU clock, but getting it on-chip will reverse the trend of the last decade plus of higher and higher CPU to memory speed ratios.
Remember the effect of moving L2 cache from mainboard to CPU daugher-board with the slot CPU generation, and then directly on-chip as transistor budgets allowed with the move back to sockets? Of course, other adaptations were happening at the same time, as main memory got even slower as compared to the CPU. That's where a lot of the out of order and speculative execution stuff that we just take for granted now days came from -- it was forced by the ever increasing speed of the CPU as opposed to RAM, but enabled by the larger/faster L1 and L2 cache sizes.
Now imagine being able to throw most of that away once again, as memory speeds actually catch up several doublers worth because they are now on-chip. That's what we're looking at in a few years, if Moore's law continues to hold and on-chip transistor budgets continue to increase.
The problem currently is that we're in a kind of lukewarm middle ground, increasing cache size doesn't bring so much bang for the transistor budget any more, but the budget is still a few generations too limited to fully integrate main-memory on-chip. As a result, we're devising all these other things to make use of those available transistors with. However, at some point, the transistor budget is going to get big enough to consider integrating main memory, and at that point, I suspect a lot of these other technologies will be at least temporarily dropped.
Yet another angle on the same thing is the fact that a decent share of current chip real estate and transistor budget goes to the whole speculative and out of order execution thing, as forced by the speed differential between CPU compute-cores and main memory. If memory speeds suddenly increase several-fold, the effect of sticking it on-chip, a good share of that speculative logic will no longer be needed as the cost of main memory access, both in direct latency and in branch prediction failure penalty, will be much smaller. Off-hand, I'd guess that'll cut a generation off the transistor budget wait right there, thus bringing it that much closer.
One thing's for sure, there'll be quite a bit of optimization un-learning to go thru once this occurs. Throw in the 1-T Z-RAM tech (assuming that still next-gen tech is actually viable both technically when fused with CPU foundary tech, and that it's even commercially viable at all), and
Duncan
From the articles:
"Peers" actively bittorrenting are what a tracker tracks. Thus, they had 10 million accounts all active at the same time! So your "if half their [...] users visit" was only half of what you should have been considering! 10 million users all at the same time... using your estimate, that'd be 20 million regular users, and I'd consider that a low estimate, given the activity figures of say 1 in 10 at any given time that ISPs tend to use. Even if it's one in five, that's 50 million regular users, which I'd consider more likely than your one in two.
Looked at accurately, therefore, it's even more amazing. Pirate Bay doesn't even have to consider defining "regular user", as their active at any given instant numbers are so incredibly huge, they have no need to maximize or inflate the numbers using any other metric.
NB: I wonder how many of those peers are comparable bittorrent newbies, downloading that 17 gigs of myspaceprivatepicstorrent featured earlier here on
Duncan
Also note that there's another config option that allows one to change that 3G/1G split for NOHIGHMEM mode, if desired. It's normally hidden, but available if one activates EXPERIMENTAL and I believe EMBEDDED.
That 3-level paging above 4 gig is a bit of a performance hit, as the kernel shifts its 4 gig window around in that 64 gig frame, tho if one runs the sort of apps that actually use that sort of memory, it's less of a performance hit than going to swap would be. Still, going 64-bit Linux isn't such a big deal any more, if your CPU supports it, and it's MUCH more efficient since multiple terabytes of memory can be directly accessed.
There's another factor at play here as well, however, and this applies to ALL OSs on 32-bit x86 and most or all on 64-bit x86 as well. It's a PCI hardware issue more than a software issue. Many old PCI devices were designed for 32-bit only operation, and their hardware can't address memory above the 32-bit 4 gig memory barrier. When memory was running less than a gig, this didn't matter much, and it became customary to reserve the virtual space at the top of the 32-bit address pool, 3.5-4 GB, for PCI device i/o access. As real memory expanded into that area, it runs into the reserved area, and the real memory behind it can't ordinarily be directly accessed.
Folks who've been around for awhile will likely recall a similar issue back at the 1 MB barrier, and how it was resolved using a "memory hole". The same technique is used here. With a BIOS setup to do so, one can configure a "memory hole" at the 3.5-4GB location, and the BIOS will remap the affected memory up above the 4 GB barrier.
This explains the complaints about Apple and MS platforms also having 4 GB look like 3 or 3.5 GB. I'm not sure if their 32-bit kernels can cope with that remapping or not -- they won't be able to if they can't address more than the 4 gigs anyway, but even if they can, the BIOS must be configured to map the hole as well.
Meanwhile, while addressing memory above the 4 gig line shouldn't be a problem for 64-bit kernels, the BIOS must still be able to do the remapping as well -- and the kernel must understand and deal with the hole. 64-bit Linux has suitable config options to do so, but I've not the foggiest how binary platform systems shipping a single binary kernel for all users deals with this. Primarily binary Linux distributions generally ship a number of different kernels, including an enterprise
... which is the whole point of Free(dom) software, of course, and why some people strongly prefer to work in a Freedom software based environment, where as long as they stay within the compatible license community, cross-company and cross-product code reuse is not only allowed but encouraged. In fact, some people strongly prefer it to the point where they won't do closed source, and pass up jobs that would require them to keep their work closed.
Funny/ironic* maybe, but as it's outsourced to a third party web survey company, and that web survey company happens to run MS platform, still understandable in this day and age of outsourcing.
Still, one would think they could find another solution provider running on a more suitable if only for appearances platform.
The problem is, this is apparently the first year the survey has made it to the popular media, and it shows in a multitude of ways. The server platform is one. That they seem to have not considered the possibility of actual home Linux users, designing the survey for business use only, even tho the announcement said nothing of that fact and was carried widely in the more general Linux-tech media, is perhaps the bigger one. The results are heavily skewed as a result. See comments I and others make elsewhere.
* Being one who sees language as dynamic, I'm a descriptivist, and see the examples in the song "Ironic" as just that, because that's the popular usage of the term, making it correct by definition, in dynamic language terms.
While the word I'd use is "skewed", I'd basically agree with you.
/ask/ them to respond in general community coverage?
There was an original announcement of the survey in the Linux-covering media, and I looked at it but didn't take the survey then, as it seemed only interested in business use. Later, there was additional coverage, asking where all the North American users were, as there had been relatively few such responses to the survey at that point. Most were and still are European, altho the North American response percentage increased from about 10% to about 40%.
Anyway, since they wanted N. Am. responses, I went ahead and filled it out then, doing the "choose the best fitting answer" routine, even where the obvious answer for home users was entirely missing. The best available fit classification was small-office/home-office use, with 1-100 machines. I then had to pick an annual income range, with I think the low end being a million or less. Well, millionaires aren't as uncommon as they used to be, but that's still rather ridiculous for a low end choice on individual income. Again, it's obvious they think all Linux users are corporate, since the announcements asking for participation said nothing about business use only, only Linux use. Later, they asked a question of what my primary business was. Using best-fit logic, I think I originally chose health care, as I reasoned as an individual, the most basic purpose will be to maintain my own health and survive. That ended up conflicting with an answer later, so I went back and picked something else.
The survey is therefore incredibly skewed, because it makes an invalid assumption, that all Linux users are business users, and/or that the home-only user response will be so tiny (due to discouragement based on the obvious slant if nothing else) as to be ignorable.
The non-biz response may have been low enough that the home-user response assumption may have actually been the case in times past, due both to coverage and to obvious slant, but I think this year's was obviously skewed by the second round of coverage, asking for more North American user responses. Given the coverage I saw (including user comments on the stories), I believe it reasonable to assume that at least half of that response rate increase for N America alone was home users. That would work out to ~20% home users, minimum, dramatically skewing the results since it's incredibly obvious they were essentially ignoring the home user when they designed the survey, and didn't intend for home users to respond. Well, then why
IMO, they therefore got the skewed results they asked for and that could have been predicted given where and how the thing was publicized. No WONDER their SOHO segment jumped so dramatically! No WONDER some of the results don't follow previous trends!
Maybe next year they'll include a home user option and home user appropriate options to the further questions as well. Even if they are entirely uninterested in that segment (it doesn't spend enough money, maybe, because much of it simply downloads the free versions, and doesn't spend on the ISVs either), providing options for it would be wise, as doing so would then allow clean separation of what they consider "noise" from the signal they are really interested in, the big-money corporate accounts.
BTW, I was one of the Gentoo respondents, and put in the comments something else the ISVs etc they are apparently targeting the results for aren't likely to like -- that as far as I'm concerned, if it's not freedomware, if it doesn't allow full use of the four freedoms equally to all users and potential users, it's not a solution I can or will consider. So much for the Adobes of the world and their Linux ports. They might as well be un-wine-able MS-platform only, at a price of a trillion dollars a seat, for all I care -- if they aren't freedomware, they are that far out of usable-here solution scope. (It's a legal matter, see. I no longer sign away my rights, including the right to redress for security issues if I can't see and use the source for software running on my machine, so I can't agree to the EULAs, and the software is therefore not a legally viable solution, even if I DID wish to use it.)
Two good resources I found were the list of providers at Wikipedia, and the VoIP forums at BroadbandReports.com. Many on the the Wikipedia list won't apply to you as they're limited area or whatever, but the list is long enough, there's likely to be some in your area anyway.
/could/ be be cheaper than going with a full two-way provider. I tried it but being European based, my US credit cards wouldn't clear, so I've honestly no idea how well it works. The parallel to it for inbound would be CallCentric.com's inbound plans, particularly if you are in an area where you can get a Dirt Cheap number from them. The drawback of using separate sources for each is that services like conference calling, etc, can be harder to manage than integrated plans. I didn't try CallCentric either, however, as I intended to setup the Vyke thing for outgoing first, and when that fell thru, I went with an integrated instead.
http://en.wikipedia.org/wiki/List_of_commercial_voice_over_IP_network_providers
You can also google some of those you find there and/or VoIP and/or SIP in general, for more resources. Take your time deciding.
Of some interest, Vyke is a Norway HQed dialout-only provider, but at 4 cents (US) for call setup, no per minute fee, it
Of course, Vonage is one of the big, heavily advertised providers, but purely for that reason (and because their prices seem to run ~$5/mo higher than much of the competition), I had a negative gut reaction to them and never seriously considered them. Other than the price difference, that's purely personal/emotional, so feel particularly free to discount my opinion there. As they say, YMMV.
If you go looking for hardware, the big two that seem to have hardware on the retail shelf are Vonage, and Skype. Skype is SIP based but has a proprietary element, meaning anything you buy to work with them is locked to them. That's not my thing, but they do seem to charge rather less. Another lockin-only provider is Lingo. I was actually somewhat tempted there, and they have some quite reasonable International plans, so if you have contacts in Asia or Western Europe, despite their lockin, they could be worth checking out. Just realize that any hardware you get to work with them may or may not be unlockable, if you decide they aren't your style. Lingo is BTW a Primus company. I understand they are more known in Canada than here, and not all favorably. I won't say more as I'm not in a position to judge one way or the other from here, but it may be of significance one way or the other to you.
The Gizmo Project is associated with SIPphone. They're claim to fame is a software phone that does free calls (with catchs), but I only do Free Software and it's not, so it was out much as Skype was.
Wengo is active in the Linux community with Wengophone, which IS free software. Their setup didn't seem to match my needs, however (IIRC they were mostly European or something? don't remember for sure), or I'd have probably preferred them.
Sunrocket WAS a pretty big provider, but just went belly up recently. Particularly if you choose to prepay, you'll want to ensure whoever you choose has enough corporate resources either directly or because they are owned or backed by someone bigger, that's not likely to happen.
Offering the type of computer independent no-lockin hardware based service (tho if you get the hardware from them, it's likely to be locked) I was after, Packet-8, CallCentric (other plans then mentioned above), ViaVoice.com, etc. Similar hardware based service but not lockin free is Lingo as I mentioned, and Vonage (they have locked stuff widely available, don't honestly know if they are standard based enough to work with unlocked or not).
As I said, going rate is roughly $25-30/mo bottom-line, or ~$200/year prepaid (plus fees brings it to ~$230-250/year, right around $20/mo. That's for "unlimited" but consumer level calli
Seriously, consider VoIP. Because VoIP is actually a competitive market, all the extras the telcos charge an arm and a leg for are normally thrown in at the base price. Voicemail, caller-id, conference-calling, selective-call-block, etc, all included. Of course, US/Canada calling is all included too. Some offer additional features like scheduled do-not-disturb times (with selective punch-thru for your kids or mom or whoever, if you prefer), while others offer wider international calling (15 nation or so, more available but at additional cost), as part of their differentiated feature-set. I chose a company and plan with a few more features, since the US/Canada covers my calling area needs.
I have it setup here so blocked caller-id or out-of-area gets routed thru a prompt to enter a series of numbers, thus eliminating the machine calls while encouraging folks to unblock, but I could set it to totally block them if I preferred. No additional cost, of course.
Now while I did choose not to port my number, as my old regular telco number was getting the usual number of undesired calls, guess how many such calls I've gotten on the VoIP since I switched. None. Nada. Zilch! =8^)
Call quality is comparable to the former bell telco in the area, or to cell, but rather worse than the digiphone service I had from the cableco previously. (That was the best call quality I expect I'll ever have, but at comparable to telco costs, which is much higher than VoIP.)
Cost, US Dollars, ~$25/mo BOTTOM LINE INCLUDING TAXES AND FEES depending on provider, or ~$20/mo ($200/yr plus a couple bucks a month taxes) annually prepaid. (Renewals can often be had for ~$150 annual prepaid.) I'm deliberately not mentioning the VoIP provider I chose as this isn't an ad, but those are the going market rates, so pretty close to what's available from several VoIP providers. (Note that all incoming or all outgoing can be had for less, this is for both, and including full US/Canada bundled calling area. Skype's less too, but locked provider and not standardized, so I avoided that route.)
What's nice is that while I had been skipping out on the extra fee caller-ID as I had found it just wasn't effective for what I wanted it for and it cost extra, now that it's bundled, I upgraded my phone (I chose a VoIP adapter that a normal phone plugs into, tho dedicated VoIP phones are available) to a voice-announcing caller-ID phone, with distinctive ring as well. Thus, after the first standard ring, it announces who is calling based on the (no extra cost) caller ID, then switches to distinctive-ring if I have that ID programmed for it. No more looking at the caller ID, or waiting for the answerer to pickup and listening to it to find who's calling, before I decide whether it's worth picking up! Thus, even if a marketer does call (none have), it'd just announce caller out of area or whatever, and I'd just let the voicemail pickup.
I've been very happy with it! =8^)
Yes, you wanted to (well, as much as anyone "wants" to pay anything, for anything, anyway). After all, they didn't call up your work and hold your check hostage unless you paid them $115/mo, nor were they withholding your food, water, air, or even your shelter, if you quit paying. The service was self-evidently worth at least that $115/mo. to you or you wouldn't have been paying it, TV or no TV.
/need/ TV, so it was and is a luxury... a luxury you were willing to spend at least $115/mo. on.
After all, TVs aren't like food, water and shelter, or even like phone service or a car with gas, insurance and maintenance. It's arguable people need those, tho plenty of folks do get by without the latter set anyway. I've yet to see anyone demonstrate they
Note that I haven't had TV in years, so it's certainly doable. Computers took up the time I used to spend watching TV. I "didn't want" to pay the $1100+ I just paid for the pair of dual-core Opteron 290 upgrades I have coming in this morning, either, but it turns out I wanted them more than the money, and more than other stuff I would have otherwise purchased with that money, after all! =8^) It would appear you likewise wanted the service more than that $115/mo, or anything else you could have done with that money, too, so yes, you wanted it, were willing to pay for it, and thru negligence if nothing else, put yourself in a situation where you WOULD be paying for it! All they did is see a mark more than willing to pay that price, and take advantage of that fact.
Duncan
Fixed fee, yes, but what I was arguing was that the paperwork time is in effect built into that fee, or you'd not be doing it, unless of course it was for charity.
Perhaps I should have added "or as a non/low-profit hobby". I don't know what the industry pay rate is (nor am I implying you need to say as it is literally not my business, tho if you wish to, I'd be interested), but for argument's sake, let's say it amounts to reimbursement of expenses and perhaps a small stipend, similar to the work many "volunteers" do. That'd be effectively a hobby or charity.
However, the larger point remains, for any activity someone continues to engage in, they either consider it "worth it" in the end, or they don't, and if they don't, they don't continue engaging in it for long. This is regardless of what form the "pay" takes, whether it's in dollars or other reward (entertainment, satisfaction of some ideal as a volunteer, whatever). If you continue in the activity of being an "adult entertainment performer", you obviously consider it rewarding enough, that is, more personally rewarding than other alternatives open to you to spend your time. Maybe that reward is more than the money. Maybe it is a hobby, where the money is actually second place or in fact you spend money on it. Heh... that's what a hobby is for most folks, anyway, something they spend money on. So if you love it enough to consider it a hobby, and get at least partial reimbursement for your expenses, consider yourself lucky. Not so many people get that with their hobbies (altho one of the nice things about the FLOSS community is that many actually do end up getting paid for their hobby... tho I'm not one of them).
The point still stands, however. Taken as a whole, that time filling out paperwork is in fact paid, if indirectly, because it's a necessary requirement of the activity you obviously get reward for (money or otherwise) enough to continue, and if you were in fact not getting paid enough to justify that time as well, you'd find something else to do with your time.
Actually, this is something I as a "victim" had to learn the hard way. For some time I had a problem repeatedly falling into classic "victim" mode, where, if you are familiar with psychology, reality gets distorted and the "victim" simply cannot see what to others look like obvious ways out. It took a crisis (basically a mental breakdown) but eventually I realized that many people, not only victims, fail to see and appreciate most of the choices they have, and thus to proactively act to make the best of them.
Since I realized that, I've consistently endeavored to personally take the proactive route, being deliberately assertive in the choices I make. I've
personally been much happier as a result, and thus have tried to confront the "no choices" aka "victim" mode when I see it in others, to hopefully challenge a few assumptions and get them out of "victim" mode and into "assertive life enjoyment mode" if I can as well.
The example I'm usually arguing is the broadband ISP example. People often say they have no choice due to monopoly conditions. That is in fact false, as can be seen by the thought experiment question of what would they do if that provider failed or pulled out, no longer offering that product in that market. All of a sudden,