It requires explicit support from the bios manufacturer, thus it's not just a software application.
However, and making assumptions here... It uses an internet connection configured by the running OS, and therefore only works with certain systems, if you were to run linux on the laptop there's not much it could do. It would be a lot smarter if it tried to dial/dhcp/whatever in bios mode. Also, it must contain a copy of the software in some writable area of the bios that it reinstalls after a drive format, which means that: Thieves only need to become aware of this tool, and obtain a program to wipe this area of the bios. The fact that a single system is explicitely supported by hardware manufacturers makes it easier for thieves, because they now only need to know how to defeat this single system. Malware authors could use this "feature" to automatically reinstall viruses... This will have the added consequence that anti-virus programs will have the ability to erase this area built in to them in future.
From the bios? So it overwrites your flash bios? That sounds nasty, and surely that would make it very machine specific... Also wouldnt it be defeated by reflashing the bios anyway?
That's complete FUD... Sun have no more control over ODF than any other OASIS member does, the format was designed and is maintained by a committee of people, including developers from KDE, Sun, IBM, Corel and others. Sun would need to get agreement from the other OASIS members to do much with ODF. Sun really only have an implementation of ODF, and theirs wasn't even the first implementation to surface. ODF also doesn't depend on java in any way. Sun's implementation of ODF uses Java for some functions, but also isn't dependent on it.
I had a similar conversation with a KDE developer regarding kioslaves...
The kioslaves are used by kde apps to access files, and there are ioslaves for all kinds of different things, such that you can browse an scp server of sql database using konqueror as if it were a local directory. The problem here is that you need to use the kde apis before you can access these kioslaves. Then you have fuse, which aims to do a similar thing, but which hooks into the system letting you mount something like an scp server under the existing filesystem hierarchy. Consequently, any program can access a fuse device without needing any kind of modification.
All the effort that's gone into kioslaves, is only useable by kde apps. I believe there is a gnome equivalent called gnome-vfs too, causing further duplication of effort. Really all of this should be merged into fuse, so that it can work with any applications.
// Notes is like Linux. Secure, Complex, and sometimes ugly.
I wouldnt say Linux was especially complex, the unix philosophy of everything as a file makes unix systems far less complex than other systems... VMS and z/OS seem a lot more complex, tho this could be down to my lack of knowledge of these systems (people with a lack of unix knowledge often claim unix to be complex)... Windows however, is a lot more complex than any unix, but most users don't realise this because they never really look beyond the frontend interface.
OO also tends to be more intuitive to people who learned on systems other than msoffice (eg wordperfect)... The only people who prefer msoffice, tend to have never really used anything else. A lot of people try openoffice very briefly, before giving up on it. Not giving it a proper chance to get used to it. Conversely, most msoffice users are forced to at work, so they have to use it long enough to get used to its quirks.
Not to mention the reputation for inaccuracy excel has... And unfortunately, openoffice seems to copy (i assume for compatibility, people will assume openoffice to be wrong if it returns a different answer to excel regardless of which one is in the wrong). Anyone using excel for scientific purposes is compromising the integrity of their research.
Has anyone ever been caught due to tracking software on a stolen laptop? Most thieves will just sell the laptop on to some innocent buyer without even powering it on... And slightly smarter thieves will format the machine first, completely removing the tracking software. Only the legit user will keep the tracking software running, making *them* trackable.
And if tracking software like this becomes common, then _all_ thieves will format the machines as soon as they receive them, and the process of formatting will become just another part of the theft process.
So really, don't waste your money on tracking software... Instead, get the laptop insured against theft, and make sure your important data is encrypted and backed up regularly. If your lucky, the insurance will replace it with a newer model. If you really want to track your laptop, go for a hardware solution, that way it will kick in as soon as its powered on regardless of what/ifany software is running.
The row limit is 65535, and excel has the same limitation i believe. OO is in an unfortunate position that, if they were to have a higher limit, people would use the extra cells and then try to save their files in ms formats, resulting in them not loading. And this would be blamed in OO for having poor compatibility, rather than MS for having the 65535 rows limit.
Doesnt calc support writing of macros in several languages? From a quick look, it supports: openoffice basic python javascript beanshell (what is this?)
Now, openoffice basic is probably a dead end, it seems stupid to use a language specific to open/star office as it's unlikely to be supported by anything else, and i'm not quite sure what beanshell is but i imagine it's some java based thing.... Python and Javascript tho, are good well known cross platform languages, and there are many many people out there who are already experienced with these languages. Another plus, is that if you learn python/javascript for the purposes of scripting openoffice, you will be able to reuse your new skills for other purposes.
Also, you can easily write external programs in any language to parse the ODF files outside of any application.
If you have problems with calc, can you file them as bugs and get them fixed? List them for me so i can take a read of the bug reports and keep track of their progress... Or, if there are already bugs filed, go and vote for them to show that there's interest in getting them fixed.
OpenOffice calc likes to replicate excel bugs... Gnumeric goes for correctness... I use gnumeric, because i require the answers from my spreadsheets to be correct and accurate, and i won't consider using anything less accurate regardless of what features it has.... That said, openoffice should give you the choice between "compatibility with excel" and "accuracy/correctness"
An incomplete unratified standard yes, but one that is being worked on by multiple parties to ensure it meets the needs of all of those parties. Anyone is free to join and express their needs too. The difference is that OOXML has a quickly hacked together specification for formulae which is flawed in many ways, which only has input from and only satisfied the requirements of one organisation. A ratified flawed standard is a bad thing, because it will remain in that flawed state, that why the ODF people are trying to get the formula format correct. If the ODF spreadsheet specification doesn't suit your needs, why dont you read the spec thoroughly and suggest changes?
In what way? What do old versions of msoffice do that cannot be represented by odf? Most of the so called "compatibility" present in ooxml is poorly conceived, and hinges on adding extra complexity to the format when it would be much easier to simply mimick the behaviour in the conversion program and store the results in a standard way.
Modding save games has very little to do with online play... Typically for an online game, your "save game" will be stored on the server so you can't edit it anyway. Editing single player save games would have no effect on online play... To prevent cheating with online games, you want to prevent modification of the game data itself, and modification of the network traffic. However this all gives a false sense of security, because people will still always find a way to cheat.
I have a compilation DVD that runs on a modded gamecube or wii, it includes an emulator and a whole heap of NES/SNES (and sega consoles) rom files... I believe there's an xbox version of this DVD too. You could use that, and if you really feel bad about piracy just play the games where you also posess the physical cart.
No matter how efficient the crypto, it still detracts from the end user experience... If it's software based, then its using some cpu cycles that could otherwise be used for gaming... If it's hardware based then the cost of the hardware increases to cover it (or other features are removed to reduce costs back)... Users can no longer edit save games themselves, i often found it very useful to download save games to get past sticking points... And some games stored the saves in plain text, enabling users to make the changes themselves.
All this paranoia about piracy, the wii has already been cracked far enough to play pirated games, but it still won't run homebrew. All the legit users are losing out wether they just want to play legit bought games, or want to run legit homebrew, the only people benefitting are those who want to pirate games.
Windows gives you ACLs, Linux gives you standard unix permissions *AND* ACLs... ACLs are complex, to the point that many windows admins dont bother with them. Unix permissions are simple enough to master but lack some of the flexibility. However, for most purposes permissions are more than adequate, and you also have ACLs if you need more.
But wtf is this about network security? Linux has iptables by default, ssh for communications between machines, NFSv4 for file sharing... Compare that to windows file sharing, which is vulnerable to reflection attacks (see metasploit) and will automatically send your authentication details when you connect to a remove server! Not to mention all the stuff windows has open by default (rpc, netbios, netbios-ns, and more), and which is difficult to turn off. Linux boxes, unless horribly misconfigured, will only listen on the services which are required, with unnecessary services turned off rather than kludgily filtered.
It's a game, and the gameplay of it involves combat...
I was playing quake a few days ago, just minding my own business wandering around a dungeon when some asshole decided to come along and shoot me!
And to bring up a real life example.... Would you go to a paintball arena to conduct a wedding? No, that would be stupid, you'd get hit with paint because THATS WHAT PAINTBALL ARENAS ARE FOR. Similarly, if you want a peaceful virtual wedding don't do it in a game in which combat plays a significant role.
Actually, wedding ceremonies are usually held on private land and thus the land owner or his appointed representative can demand that you leave, and call the police if you don't. Also, there are laws governing unruly conduct, harassment and threatening behaviour anyway.
But in a game like everquest, where casting spells and killing things is part of the game, you've no cause for complaint. Perhaps something less fighting-oriented like second life would be more appropriate.
Browser: Firefox 2.0.0.6 Hardware: Macbook Pro (CPU Intel Core 2 T7400 2.16GHz) OS: Gentoo Linux x86 (32-bit) Kernel 2.6.22.1
MD5 Benchmark took 2.175 seconds for 3000 hashes (1379 hashes/second) MD4 Benchmark took 1.563 seconds for 2700 hashes (1727 hashes/second) SHA1 Benchmark took 2.359 seconds for 1900 hashes (805 hashes/second)
Same hardware, but running 64bit gentoo linux:
firefox 2.0.0.6
MD5 Benchmark took 1.936 seconds for 3000 hashes (1550 hashes/second) MD4 Benchmark took 1.543 seconds for 2700 hashes (1750 hashes/second) SHA1 Benchmark took 2.366 seconds for 1900 hashes (803 hashes/second)
konqueror 3.5.7
MD5 Benchmark took 9.149 seconds for 3000 hashes (328 hashes/second) MD4 Benchmark took 5.473 seconds for 2700 hashes (493 hashes/second) SHA1 Benchmark took 9.228 seconds for 1900 hashes (206 hashes/second)
There also appears to be quite some fluctuation in results if you run it several times on the same browser (keep hitting refresh)
Yeah, it's not a difinitive benchmark... I happened to know of this benchmark already, and was just curious how well different browsers performed in light of the claims about opera being fast.
Slashdot Mirror
Having read up on this from their website...
It requires explicit support from the bios manufacturer, thus it's not just a software application.
However, and making assumptions here...
It uses an internet connection configured by the running OS, and therefore only works with certain systems, if you were to run linux on the laptop there's not much it could do. It would be a lot smarter if it tried to dial/dhcp/whatever in bios mode.
Also, it must contain a copy of the software in some writable area of the bios that it reinstalls after a drive format, which means that:
Thieves only need to become aware of this tool, and obtain a program to wipe this area of the bios. The fact that a single system is explicitely supported by hardware manufacturers makes it easier for thieves, because they now only need to know how to defeat this single system.
Malware authors could use this "feature" to automatically reinstall viruses... This will have the added consequence that anti-virus programs will have the ability to erase this area built in to them in future.
From the bios? So it overwrites your flash bios? That sounds nasty, and surely that would make it very machine specific...
Also wouldnt it be defeated by reflashing the bios anyway?
That's complete FUD...
Sun have no more control over ODF than any other OASIS member does, the format was designed and is maintained by a committee of people, including developers from KDE, Sun, IBM, Corel and others. Sun would need to get agreement from the other OASIS members to do much with ODF.
Sun really only have an implementation of ODF, and theirs wasn't even the first implementation to surface.
ODF also doesn't depend on java in any way. Sun's implementation of ODF uses Java for some functions, but also isn't dependent on it.
I had a similar conversation with a KDE developer regarding kioslaves...
The kioslaves are used by kde apps to access files, and there are ioslaves for all kinds of different things, such that you can browse an scp server of sql database using konqueror as if it were a local directory. The problem here is that you need to use the kde apis before you can access these kioslaves.
Then you have fuse, which aims to do a similar thing, but which hooks into the system letting you mount something like an scp server under the existing filesystem hierarchy. Consequently, any program can access a fuse device without needing any kind of modification.
All the effort that's gone into kioslaves, is only useable by kde apps. I believe there is a gnome equivalent called gnome-vfs too, causing further duplication of effort. Really all of this should be merged into fuse, so that it can work with any applications.
I'm sure there are other areas of duplication.
// Notes is like Linux. Secure, Complex, and sometimes ugly.
I wouldnt say Linux was especially complex, the unix philosophy of everything as a file makes unix systems far less complex than other systems...
VMS and z/OS seem a lot more complex, tho this could be down to my lack of knowledge of these systems (people with a lack of unix knowledge often claim unix to be complex)...
Windows however, is a lot more complex than any unix, but most users don't realise this because they never really look beyond the frontend interface.
OO also tends to be more intuitive to people who learned on systems other than msoffice (eg wordperfect)...
The only people who prefer msoffice, tend to have never really used anything else. A lot of people try openoffice very briefly, before giving up on it. Not giving it a proper chance to get used to it. Conversely, most msoffice users are forced to at work, so they have to use it long enough to get used to its quirks.
Not to mention the reputation for inaccuracy excel has... And unfortunately, openoffice seems to copy (i assume for compatibility, people will assume openoffice to be wrong if it returns a different answer to excel regardless of which one is in the wrong).
Anyone using excel for scientific purposes is compromising the integrity of their research.
Format painter? It's in the default toolbar, i use it quite a lot.
That documentation is obsolete...
OpenOffice since version 2.0 has supported 65535 rows, version 1.x and earlier supported 32768 rows.
Has anyone ever been caught due to tracking software on a stolen laptop?
Most thieves will just sell the laptop on to some innocent buyer without even powering it on... And slightly smarter thieves will format the machine first, completely removing the tracking software.
Only the legit user will keep the tracking software running, making *them* trackable.
And if tracking software like this becomes common, then _all_ thieves will format the machines as soon as they receive them, and the process of formatting will become just another part of the theft process.
So really, don't waste your money on tracking software... Instead, get the laptop insured against theft, and make sure your important data is encrypted and backed up regularly. If your lucky, the insurance will replace it with a newer model. If you really want to track your laptop, go for a hardware solution, that way it will kick in as soon as its powered on regardless of what/ifany software is running.
The row limit is 65535, and excel has the same limitation i believe.
OO is in an unfortunate position that, if they were to have a higher limit, people would use the extra cells and then try to save their files in ms formats, resulting in them not loading. And this would be blamed in OO for having poor compatibility, rather than MS for having the 65535 rows limit.
Doesnt calc support writing of macros in several languages?
From a quick look, it supports:
openoffice basic
python
javascript
beanshell (what is this?)
Now, openoffice basic is probably a dead end, it seems stupid to use a language specific to open/star office as it's unlikely to be supported by anything else, and i'm not quite sure what beanshell is but i imagine it's some java based thing....
Python and Javascript tho, are good well known cross platform languages, and there are many many people out there who are already experienced with these languages.
Another plus, is that if you learn python/javascript for the purposes of scripting openoffice, you will be able to reuse your new skills for other purposes.
Also, you can easily write external programs in any language to parse the ODF files outside of any application.
If you have problems with calc, can you file them as bugs and get them fixed? List them for me so i can take a read of the bug reports and keep track of their progress... Or, if there are already bugs filed, go and vote for them to show that there's interest in getting them fixed.
OpenOffice calc likes to replicate excel bugs...
Gnumeric goes for correctness...
I use gnumeric, because i require the answers from my spreadsheets to be correct and accurate, and i won't consider using anything less accurate regardless of what features it has....
That said, openoffice should give you the choice between "compatibility with excel" and "accuracy/correctness"
An incomplete unratified standard yes, but one that is being worked on by multiple parties to ensure it meets the needs of all of those parties. Anyone is free to join and express their needs too.
The difference is that OOXML has a quickly hacked together specification for formulae which is flawed in many ways, which only has input from and only satisfied the requirements of one organisation. A ratified flawed standard is a bad thing, because it will remain in that flawed state, that why the ODF people are trying to get the formula format correct.
If the ODF spreadsheet specification doesn't suit your needs, why dont you read the spec thoroughly and suggest changes?
In what way?
What do old versions of msoffice do that cannot be represented by odf?
Most of the so called "compatibility" present in ooxml is poorly conceived, and hinges on adding extra complexity to the format when it would be much easier to simply mimick the behaviour in the conversion program and store the results in a standard way.
I only download 12,999,999 emails with 500mb attachments per month, so i should be fine!
Modding save games has very little to do with online play... Typically for an online game, your "save game" will be stored on the server so you can't edit it anyway.
Editing single player save games would have no effect on online play...
To prevent cheating with online games, you want to prevent modification of the game data itself, and modification of the network traffic. However this all gives a false sense of security, because people will still always find a way to cheat.
I have a compilation DVD that runs on a modded gamecube or wii, it includes an emulator and a whole heap of NES/SNES (and sega consoles) rom files... I believe there's an xbox version of this DVD too. You could use that, and if you really feel bad about piracy just play the games where you also posess the physical cart.
No matter how efficient the crypto, it still detracts from the end user experience...
If it's software based, then its using some cpu cycles that could otherwise be used for gaming...
If it's hardware based then the cost of the hardware increases to cover it (or other features are removed to reduce costs back)...
Users can no longer edit save games themselves, i often found it very useful to download save games to get past sticking points... And some games stored the saves in plain text, enabling users to make the changes themselves.
All this paranoia about piracy, the wii has already been cracked far enough to play pirated games, but it still won't run homebrew. All the legit users are losing out wether they just want to play legit bought games, or want to run legit homebrew, the only people benefitting are those who want to pirate games.
Windows gives you ACLs, Linux gives you standard unix permissions *AND* ACLs...
ACLs are complex, to the point that many windows admins dont bother with them. Unix permissions are simple enough to master but lack some of the flexibility. However, for most purposes permissions are more than adequate, and you also have ACLs if you need more.
But wtf is this about network security? Linux has iptables by default, ssh for communications between machines, NFSv4 for file sharing...
Compare that to windows file sharing, which is vulnerable to reflection attacks (see metasploit) and will automatically send your authentication details when you connect to a remove server!
Not to mention all the stuff windows has open by default (rpc, netbios, netbios-ns, and more), and which is difficult to turn off. Linux boxes, unless horribly misconfigured, will only listen on the services which are required, with unnecessary services turned off rather than kludgily filtered.
If i can install Safari 3 alongside 2.x, where is the link and i'l install it now and post benchmarks...
It's a game, and the gameplay of it involves combat...
I was playing quake a few days ago, just minding my own business wandering around a dungeon when some asshole decided to come along and shoot me!
And to bring up a real life example....
Would you go to a paintball arena to conduct a wedding? No, that would be stupid, you'd get hit with paint because THATS WHAT PAINTBALL ARENAS ARE FOR.
Similarly, if you want a peaceful virtual wedding don't do it in a game in which combat plays a significant role.
Actually, wedding ceremonies are usually held on private land and thus the land owner or his appointed representative can demand that you leave, and call the police if you don't.
Also, there are laws governing unruly conduct, harassment and threatening behaviour anyway.
But in a game like everquest, where casting spells and killing things is part of the game, you've no cause for complaint. Perhaps something less fighting-oriented like second life would be more appropriate.
Had a couple of other people here try it...
Browser: Firefox 2.0.0.6
Hardware: Macbook Pro (CPU Intel Core 2 T7400 2.16GHz)
OS: Gentoo Linux x86 (32-bit) Kernel 2.6.22.1
MD5 Benchmark took 2.175 seconds for 3000 hashes (1379 hashes/second)
MD4 Benchmark took 1.563 seconds for 2700 hashes (1727 hashes/second)
SHA1 Benchmark took 2.359 seconds for 1900 hashes (805 hashes/second)
Same hardware, but running 64bit gentoo linux:
firefox 2.0.0.6
MD5 Benchmark took 1.936 seconds for 3000 hashes (1550 hashes/second)
MD4 Benchmark took 1.543 seconds for 2700 hashes (1750 hashes/second)
SHA1 Benchmark took 2.366 seconds for 1900 hashes (803 hashes/second)
konqueror 3.5.7
MD5 Benchmark took 9.149 seconds for 3000 hashes (328 hashes/second)
MD4 Benchmark took 5.473 seconds for 2700 hashes (493 hashes/second)
SHA1 Benchmark took 9.228 seconds for 1900 hashes (206 hashes/second)
There also appears to be quite some fluctuation in results if you run it several times on the same browser (keep hitting refresh)
Yeah, it's not a difinitive benchmark... I happened to know of this benchmark already, and was just curious how well different browsers performed in light of the claims about opera being fast.