yes java lacks the ability to put any form of structure or array on the stack and also lacks parameter passing by reference.
so how do you propose to return more than a single primitive from a function without creating a temporary object on the heap?
same goes for string handling. making a string an immutable class is pretty much the only reasonable thing to do given the basic environment of java but it means every string you generate and return is at least two more objects (the string itself and its internal array) and quite possiblly more thrown on the garbage pile.
well with windows i can build an app with delphi and even if i elect not to use packages the resulting executable (for a small utility app or similar) is still small enough not to bother even modem users and it will run on pretty much any 32 bit version of windows.
I haven't done much gui stuff for linux but i know for non-graphical apps free pascal can once again produce an executable that will run on pretty much any linux system.
And java apps just seem more bloated in general both in how they respond themselves and the affect they have on system performance in general not to mention the fact that few modem users are going to sit through a JVM download unless they have absoloutey no choice.
It shouldn't even be required to use that password to install software for just yourself. shouldn't be but knowing the apathy of windows software vendors probablly will be. Even running as non-admin is an achivement for some software vendors!
a distrorted a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y, z, space or punctionation mark provided it can still be recognised correctly can be reproduced as a perfectly formed new a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y, z, space or punctionation mark
the big difference is we can read the ones and zeros on a hard drive with an error rate that can be considered to be effectively zero (partly due to error correction systems built into the drive) whereas we cannot recognise the letters in a book by machine with anywhere near that level of accuracy.
the problem with digital data is it tends to die quite quick if noone cares to maintain it any longer (i have loads of old floppies and quite a few of them are unreadable due to media aging or new drives or whatever). Books routinely last hundreds of years no current digital storage media has been arround anything like that long!
For really long term storage (say from the collapse of a civilisation to the development of the next one) reproducing text on a fairly durable medium at a small size (so it can be safely stored in a small place huge libraries are extremely vulnerable) some form of etching like that used by the rosetta disc is probablly the best soloution.
drm for CDs relies on hiding itslef from the legitimate user of the PC so in a very real sense it relies on being a rootkit to function at all.
Re:How much difference between Java and C++?
on
OpenOffice Bloated?
·
· Score: 1
i'm not sure on the details but i belive that gcj can be set up to act as a caching jit for gij then the cache can be preseeded so that known dependencies don't even have to be built on demand.
All i'm really interested in is the vanilla kernels (if i use a distro kernel i expect the distro to look after it). Is there any site that tracks what versions of the vanilla kernel fixed what security holes?
from what i've heared things like LUFS and FUSE suck performance wise compared to real drivers in the kernel. This is not so imporant for a special purpose FS or a network mapper but you don't particularlly wan't your main partitions on them.
can grub do a boot configuration once like lilo can.
boot configuration once is a very handy feature if you have a remote powerswitch but no other way of getting at the machine until/unless networking comes up.
that track which versions of the linux kernel are affected by which vunerabilities are fixed in what kernel versions? i'd like to know what kernel versions need to be upgraded or patched (and what to patch them with) to maintain a secure system.
yes it has its also highly confusing to those not in on the joke to see such things moderated interesting or insightfull so imo its something that does need to be pointed out.
Scotty provided them with the formulae for transparent aluminium and they provided him with material for his whale enclosure. IIRC they didn't make if it was plexiglass or the new transparent aluminium but the timescales would indeed imply the former.
people put down paypal for being expensive but disagree with this at least for small international transactions.
I enquired about direct transfers from my (UK) bank account to someone in germany and the MINIMUM charge was £14!
now admittedly that charge is flat for transfers right up to E4000 (btw why does slashdot still not support unicode) but it makes direct international transfers from my bank account unworkable for small transactions.
I've also used bidpay once (i bid on an ebay auction in the USA where the seller didn't take paypal and i had to pay him somehow) and not only were thier fees quite high (i think about $5 and bidpay charge the buyer not the seller) i had to pay a £1.50 charge on top of that for using my debit card on a US site.
Do you have any suggestions for payment handlers that are easy to use for INTERNATIONAL online transactions and have reasonable charges for small transactions other than paypal.
what do you americans use to prove your ID at job interviews at the like if you aren't yet old enough to drive and don't have a passport. I know its very tricky here in the uk to provide good proof of your identity if you don't have either a passport or a driving license.
the exact details of what was agreed to will probablly never be public
one possible condition could be moving some of the peering to other locations so level3 has to do less work and cognet has to do more to get the traffic between the desired endpoints. I belive depeerings have caused changes like that in the past.
another possiblity is the peering is theoretically setlement free but cognet may end up paying some of the "fines" mentioned.
yet another possibility is as you suggest level 3's customers said enough and they backed down but put in some secret conditions to try and save face and make it look like not thier fault.
sure they could, the formatting may be a little screwed up but afaict that can apply just as much between different versions of office as between office and openoffice.
indeed i belive some people keep openoffice installed just to read broken ms office documents that crash ms office.
umm i don't see how "ratio of supply to demand" is really that relavent to price.
what matters more is the elasticity of the market. oil is very inelastic on both the supply and demand sides so small changes in quantity availible mean large changes in price.
they use it for the public machines at the university of manchester (most machines with only a single user are generally not centrally managed at least in the department i'm in)
PROS: you can use most (there are a few exceptions for licensing reasons but most stuff is either free, site licesed, or licensed on a concurrent user basis using licenseing servers) of your departmental applications anywhere on campus
software can be added reasonablly quickly
the same image can be used throughout the whole university in both departmental and public clusters.
CONS: The login times are long due to the extreme size of the zenworks tree, on some of the slower machines/networks (both machine speed and net speed seem to affect login times) the login time can be as much as 5 minuites.
the university puts a LOT of man hours into creating the annual images (they are well done though theese guys know how to keep the system secure without resorting to cripling the user interface like so many other places i've been do) and the packages
some application objects take a long time to deploy. This seems to be made worse by a braindead virus scanner setup (is there really any need to scan stuff thats being downloaded from your main deployment servers?!). For others they try to run too much over network shares with resulting poor performance of the app.
P.S. our departmental cluster now seems to be making a clever use of lilos boot configuration once feature to allow them to re-image machines on the next boot (the linux based zenworks imaging system boots first and then runs lilo to tell it to boot windows once and reboots the pc).
Slashdot Mirror
hmm that page doesn't explain what covariance,contravariance and erasure are at all just rants a bit.
by erasure you mean the compiler turns them into non-generic classes and casts (which are always checked in java) at compile time right?
yes java lacks the ability to put any form of structure or array on the stack and also lacks parameter passing by reference.
so how do you propose to return more than a single primitive from a function without creating a temporary object on the heap?
same goes for string handling. making a string an immutable class is pretty much the only reasonable thing to do given the basic environment of java but it means every string you generate and return is at least two more objects (the string itself and its internal array) and quite possiblly more thrown on the garbage pile.
well with windows i can build an app with delphi and even if i elect not to use packages the resulting executable (for a small utility app or similar) is still small enough not to bother even modem users and it will run on pretty much any 32 bit version of windows.
I haven't done much gui stuff for linux but i know for non-graphical apps free pascal can once again produce an executable that will run on pretty much any linux system.
And java apps just seem more bloated in general both in how they respond themselves and the affect they have on system performance in general not to mention the fact that few modem users are going to sit through a JVM download unless they have absoloutey no choice.
i was under the impression that the bytecode verifier was only used on untrusted code (e.g. browser applets).
It shouldn't even be required to use that password to install software for just yourself.
shouldn't be but knowing the apathy of windows software vendors probablly will be. Even running as non-admin is an achivement for some software vendors!
a distrorted a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y, z, space or punctionation mark provided it can still be recognised correctly can be reproduced as a perfectly formed new a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y, z, space or punctionation mark
the big difference is we can read the ones and zeros on a hard drive with an error rate that can be considered to be effectively zero (partly due to error correction systems built into the drive) whereas we cannot recognise the letters in a book by machine with anywhere near that level of accuracy.
the problem with digital data is it tends to die quite quick if noone cares to maintain it any longer (i have loads of old floppies and quite a few of them are unreadable due to media aging or new drives or whatever). Books routinely last hundreds of years no current digital storage media has been arround anything like that long!
For really long term storage (say from the collapse of a civilisation to the development of the next one) reproducing text on a fairly durable medium at a small size (so it can be safely stored in a small place huge libraries are extremely vulnerable) some form of etching like that used by the rosetta disc is probablly the best soloution.
drm for CDs relies on hiding itslef from the legitimate user of the PC so in a very real sense it relies on being a rootkit to function at all.
i'm not sure on the details but i belive that gcj can be set up to act as a caching jit for gij then the cache can be preseeded so that known dependencies don't even have to be built on demand.
All i'm really interested in is the vanilla kernels (if i use a distro kernel i expect the distro to look after it). Is there any site that tracks what versions of the vanilla kernel fixed what security holes?
from your description of one pc at a time i'm guessing you mean behind a linux nat.
from what i've heared things like LUFS and FUSE suck performance wise compared to real drivers in the kernel. This is not so imporant for a special purpose FS or a network mapper but you don't particularlly wan't your main partitions on them.
can grub do a boot configuration once like lilo can.
boot configuration once is a very handy feature if you have a remote powerswitch but no other way of getting at the machine until/unless networking comes up.
that track which versions of the linux kernel are affected by which vunerabilities are fixed in what kernel versions? i'd like to know what kernel versions need to be upgraded or patched (and what to patch them with) to maintain a secure system.
yes it has its also highly confusing to those not in on the joke to see such things moderated interesting or insightfull so imo its something that does need to be pointed out.
Scotty provided them with the formulae for transparent aluminium and they provided him with material for his whale enclosure. IIRC they didn't make if it was plexiglass or the new transparent aluminium but the timescales would indeed imply the former.
people put down paypal for being expensive but disagree with this at least for small international transactions.
I enquired about direct transfers from my (UK) bank account to someone in germany and the MINIMUM charge was £14!
now admittedly that charge is flat for transfers right up to E4000 (btw why does slashdot still not support unicode) but it makes direct international transfers from my bank account unworkable for small transactions.
I've also used bidpay once (i bid on an ebay auction in the USA where the seller didn't take paypal and i had to pay him somehow) and not only were thier fees quite high (i think about $5 and bidpay charge the buyer not the seller) i had to pay a £1.50 charge on top of that for using my debit card on a US site.
Do you have any suggestions for payment handlers that are easy to use for INTERNATIONAL online transactions and have reasonable charges for small transactions other than paypal.
what do you americans use to prove your ID at job interviews at the like if you aren't yet old enough to drive and don't have a passport. I know its very tricky here in the uk to provide good proof of your identity if you don't have either a passport or a driving license.
the exact details of what was agreed to will probablly never be public
one possible condition could be moving some of the peering to other locations so level3 has to do less work and cognet has to do more to get the traffic between the desired endpoints. I belive depeerings have caused changes like that in the past.
another possiblity is the peering is theoretically setlement free but cognet may end up paying some of the "fines" mentioned.
yet another possibility is as you suggest level 3's customers said enough and they backed down but put in some secret conditions to try and save face and make it look like not thier fault.
sure they could, the formatting may be a little screwed up but afaict that can apply just as much between different versions of office as between office and openoffice.
indeed i belive some people keep openoffice installed just to read broken ms office documents that crash ms office.
umm i don't see how "ratio of supply to demand" is really that relavent to price.
what matters more is the elasticity of the market. oil is very inelastic on both the supply and demand sides so small changes in quantity availible mean large changes in price.
sounds like a good form of demand side load management.
they use it for the public machines at the university of manchester (most machines with only a single user are generally not centrally managed at least in the department i'm in)
PROS:
you can use most (there are a few exceptions for licensing reasons but most stuff is either free, site licesed, or licensed on a concurrent user basis using licenseing servers) of your departmental applications anywhere on campus
software can be added reasonablly quickly
the same image can be used throughout the whole university in both departmental and public clusters.
CONS:
The login times are long due to the extreme size of the zenworks tree, on some of the slower machines/networks (both machine speed and net speed seem to affect login times) the login time can be as much as 5 minuites.
the university puts a LOT of man hours into creating the annual images (they are well done though theese guys know how to keep the system secure without resorting to cripling the user interface like so many other places i've been do) and the packages
some application objects take a long time to deploy. This seems to be made worse by a braindead virus scanner setup (is there really any need to scan stuff thats being downloaded from your main deployment servers?!). For others they try to run too much over network shares with resulting poor performance of the app.
P.S. our departmental cluster now seems to be making a clever use of lilos boot configuration once feature to allow them to re-image machines on the next boot (the linux based zenworks imaging system boots first and then runs lilo to tell it to boot windows once and reboots the pc).
the other thing lcds give is very sharp pixel edges which is imo a good thing on monitors but bad on TVs.
the disadvantage of that is you have 3 distros that need different admin tools and generally do things slightly differently.