i'd imagine the main reason wikipedia doesn't have an article is because its very much a case by case thing and it would be quite hard to be general in such an article.
the basic idea is to avoid re-generating stuff unnessacerally and in wikipedia there are multiple levels to this.
the level of caching closest to the users is the squids. theese only deal in complete pages and so thier main job is handling readers who aren't logged in. pages are purged from theese when they are edited so there is no need for them to hit the database when a page is viewed. iirc theese do something like 75% of regular traffic plus the vast majority of traffic surges from news sites etc.
the next level is the parser cache. This prevents having to re-parse the page so often whilst still allowing for customisations to how its rendered for logged in users.
there are also caches of some heavy queries for special pages and a seperate search engine with its own search indexes seperate from the main database.
basically it all comes down to analysing your logs and working out what is worth caching and in what state.
do you wan't to end up converting your cds many times?
also do you wan't to wait for conversions whilst doing the very manual process of ripping?
if you care about audio quality and time spent feeding in cds and hdd is reasonablly plentiful then proceed as follows (in batches if desired)
1: rip to an uncompressed format. it may be an idea to use something like cdparanoia to significantly reduce the chances of skips etc in the ripped files.
2: (optional) bulk convert to a lossless format like flac or monkeys audio when your not using the computer for anything else.
3: produce mp3 or whatever files for portable players etc but make sure you keep the lossless versions.
if you don't do this you'll be kicking yourself when a new format becomes dominant as you will have the hard choice between re-ripping everything and transcoding from one lossy format to another with the loss of quality/compressibility that doing so always brings.
your isps servers don't "sync with" the roots they make requests and get replies which they cache according to a time to live.
sure if the roots went offline your isps servers would probablly have com/org/net and many others cached but especially if they are smaller its likely that wouldn't have the full set so some lookups may well fail. Also because its time to live based if you are unlucky your isps nameservers may expire thier cache of com at any time after the roots wen't down depending on when they got thier cached copy in the first place.
ofc straight going down is unlikely the root servers are distributed to counter that
whilst the us government may control the zone file, if they tried to do anything too nasty like removing a major tld isps would find ways to bypass thier control.
lawyers exist because the law has become too complex for people to represent themselves. Especilly for smaller companies and individuls the need for them turns lawsuits into a huge financial drain regardless of who is in the right or who brought the case.
In most systems i belive you can either be awarded costs or countersue for them but you still have to have the money to pay in the first place and if your opponent goes bankrupt then you may not get it payed back even if you win.
dns is supposed to give globally uniquie identifiers and as such for it to be usable thier must be a single organisation controlling what goes into the root zone.
pretty much all of the unofficial tld systems have fallen flat for this reason. people wan't urls that everyone on the internet can use!
slow: slow to start definately, slow in actual code speed no. Unfortuately things can happen like a GC cycle at a bad time that can cause annoying slowdowns at the worst moment
bloated: the java class libraries are huge and so deploying a java environment (and you can't assume a decent java system will already be installed by default) is a huge undertaking
NOT FREE: well there are free jvms but there is still catching up to do and as they get better they will undoubtablly suffer the same issue as wine, most stuff will work but there will be slight differences that fuck over some apps.
the fact that languages like java allow poor coders to produce code that kind of works rather than total failures probablly doesn't help the languages reputation either.
and who would seriously wan't to release thier software as java bytecode when jad is arround?
i guess your trying to make a joke but just in case anyone takes this seriously.
the original figures are probablly only to about 10% tollerance. all the standard tempreature measurement systems have thier zeros close enough to absoloute zero that when considering reactor type temperatures the offsets are just not significant.
true but because its only dangerous for a relatively short period disposal is pretty easy.
ofc some stuff actually decays into stuff that is more radioactive than the original material so to start with the radioactivity of a fairly pure sample actually rises.
on what details do you base that? did you make the falacy of comparing the total output per yeah that could be made by turbines on those sites with the total consumption per year of the US?
one of the biggest problems with wind is that it provides neither base load power (constant power 24/7) nor power on demand. so adding wind power means you need more generating stations that can work well on demand which tend to be more expensive types to run.
the other alternative is energy storage but that tends to be very expensive too.
a lot depends on what exactly you mean by logging.
there are loggers in many poor rainforest contries that do a huge ammount of damage for a relatively small ammount of wood because thats the cheapest way to get certain hardwoods. (much cheaper than trying to cut just the trees you wan't and find a way to get them out).
otoh wood and especailly the fast growing woods used to make paper etc is perfectly farmable in sustainable forests where replanting is done.
i suspect when greenpeace etc talk about logging they reffer to the first of theese.
fusion as a source of usefull energy however does not. We have to put more energy into controlling the conditions than we can generate. iirc its generally belived that by going bigger the advantages of scale will make it a source of usefull energy but there is still a long way to go.
a packet inspection firewall (statefull or otherwise) is usefull for sealing up things that can either be restricted to a limited number of clients or you wan't to turn off but can't or for hiding your box from the network but it is no use whatsoever for the services you actually wan't to run for the main userbase (be that the whole of your internal network or the whole internet or whatever).
there are also application level gateways for some protocols that can filter at a higher level but i'm not sure theese exist for smb.
or to put it another way they got the system completed to allow them to switch a wiki to utf-8 without block-converting the body text with the wiki offline.
this was very important to allow converting a wiki the size of the english wikipedia with downtime for the conversion being kept to a reasonable level.
i don't know if mediawiki still allows you to run an iso-8859-1 wiki or not.
i don't think we have seen a slowdown in a month or so actually. mostly the really major slowdowns/timeouts are caused when some resource gets contended because something else is running a little slower than normal (think database locks etc)
has anyone else tried to scale a wiki to the kind of usercounts wikipedia is running on? i doubt it.
even if wp did direct link/. on the front page (which goes against thier way of doing things) i doubt it would have that much effect./. is story orientated. folks go to whatever pages are listed in the current story. wikipedia is an encyclopedia, people generally go there looking for something in particlar and might look at the other stuff on the homepage if they are boared.
the gameboys mainly relied on the fact that its difficult to make unofficial carts for them because the things are so damn small
but there wasn't much in the way of copy protection. the original GB and the GBC didn't have any at all and the GBA had the ability to use encrypted carts but i don't think any such carts were ever made.
now the DS is a totally different matter they got pretty serious on protecting that one although in the end thier systems were worked arround by the homebrew guys
the internet archive is a sufficiantly major project that i'd expect them to have competent lawyers advising them on the legality of what they are doing.
Slashdot Mirror
it could be done phone lines are actually fairly symetrical
however you would have to provide line voltage (about 50Vdc with a high AC impedance) on/off hook detection and ringing yourself
basically you'd be building a minimal phone exchange it could deffinately be done though.
i'd imagine the main reason wikipedia doesn't have an article is because its very much a case by case thing and it would be quite hard to be general in such an article.
the basic idea is to avoid re-generating stuff unnessacerally and in wikipedia there are multiple levels to this.
the level of caching closest to the users is the squids. theese only deal in complete pages and so thier main job is handling readers who aren't logged in. pages are purged from theese when they are edited so there is no need for them to hit the database when a page is viewed. iirc theese do something like 75% of regular traffic plus the vast majority of traffic surges from news sites etc.
the next level is the parser cache. This prevents having to re-parse the page so often whilst still allowing for customisations to how its rendered for logged in users.
there are also caches of some heavy queries for special pages and a seperate search engine with its own search indexes seperate from the main database.
basically it all comes down to analysing your logs and working out what is worth caching and in what state.
do you wan't to end up converting your cds many times?
also do you wan't to wait for conversions whilst doing the very manual process of ripping?
if you care about audio quality and time spent feeding in cds and hdd is reasonablly plentiful then proceed as follows (in batches if desired)
1: rip to an uncompressed format. it may be an idea to use something like cdparanoia to significantly reduce the chances of skips etc in the ripped files.
2: (optional) bulk convert to a lossless format like flac or monkeys audio when your not using the computer for anything else.
3: produce mp3 or whatever files for portable players etc but make sure you keep the lossless versions.
if you don't do this you'll be kicking yourself when a new format becomes dominant as you will have the hard choice between re-ripping everything and transcoding from one lossy format to another with the loss of quality/compressibility that doing so always brings.
your isps servers don't "sync with" the roots they make requests and get replies which they cache according to a time to live.
sure if the roots went offline your isps servers would probablly have com/org/net and many others cached but especially if they are smaller its likely that wouldn't have the full set so some lookups may well fail. Also because its time to live based if you are unlucky your isps nameservers may expire thier cache of com at any time after the roots wen't down depending on when they got thier cached copy in the first place.
ofc straight going down is unlikely the root servers are distributed to counter that
whilst the us government may control the zone file, if they tried to do anything too nasty like removing a major tld isps would find ways to bypass thier control.
true the difference with law is that someone else can force you into a case and therefore pretty much force you into buying lawyer time.
I guess the same could be said of medicine if someone decided to mug you in the street then they are basically forcing you to pay a doctor.
lawyers exist because the law has become too complex for people to represent themselves. Especilly for smaller companies and individuls the need for them turns lawsuits into a huge financial drain regardless of who is in the right or who brought the case.
In most systems i belive you can either be awarded costs or countersue for them but you still have to have the money to pay in the first place and if your opponent goes bankrupt then you may not get it payed back even if you win.
i'd read that as they negotiated for its release but it died before they actually released it so it got eaten.
dns is supposed to give globally uniquie identifiers and as such for it to be usable thier must be a single organisation controlling what goes into the root zone.
pretty much all of the unofficial tld systems have fallen flat for this reason. people wan't urls that everyone on the internet can use!
slow: slow to start definately, slow in actual code speed no. Unfortuately things can happen like a GC cycle at a bad time that can cause annoying slowdowns at the worst moment
bloated: the java class libraries are huge and so deploying a java environment (and you can't assume a decent java system will already be installed by default) is a huge undertaking
NOT FREE: well there are free jvms but there is still catching up to do and as they get better they will undoubtablly suffer the same issue as wine, most stuff will work but there will be slight differences that fuck over some apps.
the fact that languages like java allow poor coders to produce code that kind of works rather than total failures probablly doesn't help the languages reputation either.
and who would seriously wan't to release thier software as java bytecode when jad is arround?
i guess your trying to make a joke but just in case anyone takes this seriously.
the original figures are probablly only to about 10% tollerance. all the standard tempreature measurement systems have thier zeros close enough to absoloute zero that when considering reactor type temperatures the offsets are just not significant.
true but because its only dangerous for a relatively short period disposal is pretty easy.
ofc some stuff actually decays into stuff that is more radioactive than the original material so to start with the radioactivity of a fairly pure sample actually rises.
on what details do you base that? did you make the falacy of comparing the total output per yeah that could be made by turbines on those sites with the total consumption per year of the US?
one of the biggest problems with wind is that it provides neither base load power (constant power 24/7) nor power on demand. so adding wind power means you need more generating stations that can work well on demand which tend to be more expensive types to run.
the other alternative is energy storage but that tends to be very expensive too.
a lot depends on what exactly you mean by logging.
there are loggers in many poor rainforest contries that do a huge ammount of damage for a relatively small ammount of wood because thats the cheapest way to get certain hardwoods. (much cheaper than trying to cut just the trees you wan't and find a way to get them out).
otoh wood and especailly the fast growing woods used to make paper etc is perfectly farmable in sustainable forests where replanting is done.
i suspect when greenpeace etc talk about logging they reffer to the first of theese.
fusion as in fusing atoms together indeed exists
fusion as a source of usefull energy however does not. We have to put more energy into controlling the conditions than we can generate. iirc its generally belived that by going bigger the advantages of scale will make it a source of usefull energy but there is still a long way to go.
over here in the uk i belive that that card vendors were banned from cutting peoples merchant accounts for having card surcharges.
sure msis aren't evil. less evil than setup exes anyway
but that still doesn't make them an appropriate format for publishing documentation!
with different uses.
a packet inspection firewall (statefull or otherwise) is usefull for sealing up things that can either be restricted to a limited number of clients or you wan't to turn off but can't or for hiding your box from the network but it is no use whatsoever for the services you actually wan't to run for the main userbase (be that the whole of your internal network or the whole internet or whatever).
there are also application level gateways for some protocols that can filter at a higher level but i'm not sure theese exist for smb.
it *would* be the expensive thing if they weren't growing at such an insane rate.
or to put it another way they got the system completed to allow them to switch a wiki to utf-8 without block-converting the body text with the wiki offline.
this was very important to allow converting a wiki the size of the english wikipedia with downtime for the conversion being kept to a reasonable level.
i don't know if mediawiki still allows you to run an iso-8859-1 wiki or not.
i don't think we have seen a slowdown in a month or so actually. mostly the really major slowdowns/timeouts are caused when some resource gets contended because something else is running a little slower than normal (think database locks etc)
has anyone else tried to scale a wiki to the kind of usercounts wikipedia is running on? i doubt it.
even if wp did direct link /. on the front page (which goes against thier way of doing things) i doubt it would have that much effect. /. is story orientated. folks go to whatever pages are listed in the current story. wikipedia is an encyclopedia, people generally go there looking for something in particlar and might look at the other stuff on the homepage if they are boared.
sure they could see the boss got a bonus and dre out a load of cash.
but how the hell could they find out that the boss gave that CASH to a spyware company with no record.
there is this stuff called cash.
pay it as a bonus to the boss. (that gets the money out of the corp accounts) and get him to pay for the spyware in cash.
the gameboys mainly relied on the fact that its difficult to make unofficial carts for them because the things are so damn small
but there wasn't much in the way of copy protection. the original GB and the GBC didn't have any at all and the GBA had the ability to use encrypted carts but i don't think any such carts were ever made.
now the DS is a totally different matter they got pretty serious on protecting that one although in the end thier systems were worked arround by the homebrew guys
the internet archive is a sufficiantly major project that i'd expect them to have competent lawyers advising them on the legality of what they are doing.