It's a spreadsheet, damnit. While I might think it was cool to be able to enter cell formulas in Java, or C, or even APL (now, that would be cool), most people who use spreadsheets just want what Excel and OOCalc and GNUmeric already give them.
They don't want to reimplement the whole fricking document in Javascript or Java (or anything else) and run it in a web browser (which is apparently what Opera's CTO wants them to do), they just want to be able to save their work and read it later, and exchange it with others.
Spinning or not spinning is hardly going to make it hard to detect being off course by hundreds of thousands of miles.
True enough, but 3-axis stabilization (not spinning) implies some kind of reaction control system (*) that may be cancelling this unknown effect as it stabilizes the spacecraft, i.e. it also corrects the course.
(* Even if the primary attitude control is via momentum wheel, you still some kind of thruster-based RCS system to periodically dump the momentum you're building up in those wheels.)
This is Microsoft we're talking about. What do you think? They don't do anything without considering the strategic marketing impact. (Well, okay, there was Bob...)
Remember in order for there to be developers someone somewhere has to make money selling software.
Nope. In fact most developers work for companies that do not make money selling software. Now, aside from those few that are losing money selling software (grin), I mean those companies (and other organizations - governments, universities, etc) whose primary product(s) is/are something other than software. (Take your Fortune 500 -- how many of them make most (or any) of their money selling software? How many employ developers?)
Besides which, that's totally irrelevant to open document formats -- just because the document format is open doesn't mean the application software has to be either open or free, any more than standardizing on a character code of ASCII or Unicode does. (EBCDIC, you're on your own.)
I had been thinking that ODF was "obviously" a good thing until I read the rant by Opera's CTO about how shit both standards are (a memory dump between angle brackets), and how the correct way would be to go for XHTML with CSS formatting.
So how do I do a spreadsheet in XHTML with CSS formatting? And I mean a serious computational spreadsheet, perhaps with some charts thrown in, not just some data layed out in a table.
ODF is not just for pretty text documents, its for the product of all kinds of office apps, including spreadsheets and presentations.
As for "memory dump between angle brackets" -- yeah, that's a pretty fair description of OOXML, but doesn't really explain the dozen or more different apps out there that use ODF. Is he trying to tell us that KWord, AbiWord, OOWriter and Google Docs (to name a few that use ODF) all use the same memory layout?
Opera's CTO isn't worthy of his title, if that's the kind of crap he spews.
I'm concerned that standardising on ODF will come to bite us,
Just because it's an ISO standard doesn't mean it's immutable. Standards can be changed, and frequently are to keep up with technology improvements. Also, the language in the California legislation doesn't specify ODF (or it's ISO number) per se, just requires that whatever document spec is used meets the requirements of being open, freely implementable, fully specified, multi-vendor etc that only ODF meets at the moment (and OOXML doesn't). If some new Whizbang Document Format (WDF) is invented and it meets those qualifications, the IT world is free to move to it.
But surely if OOXML is as open as the wikipedia page (and everything I've heard) makes it sound,
Have you heard that Microsoft hired it's own wikipedia contributer to (try to) control the spin on the OOXML and ODF pages?
And I guess you haven't heard about the parts of the OOXML "spec" that say something ot the effect of: "Word95Spacing - This tag means that document spacing should conform to that produced by Word95. That's too complicated to go into here, see Word95 for details."
No, I didn't think you were claiming to be a forensics expert, just echoing something you heard. But what you may have heard (or thought you heard) is not quite correct. (Sorry if you thought I was attacking you.)
There reaches a point after a certain number of rewrites where any theoretical residual magnetic signature is down in the quantum noise level, and few than that its in the noise level of the detection gear. Where that particular point is depends on a lot of things: the recovery technology used, the time between when the original data was written and when it was written over (allowing more opportunity for difference in head tracking), the specific recording technology, and so on. (This even assumes you're overwriting the blocks you want overwritten, not all filesystems necessarily do that.)
The FSF program "shred" overwrites data a default of 25 times with different patterns. That's reasonably secure against any recovery, but is bloody slow. (And if the r/w head has been misaligned enough since the original data was written (but not so much as to render the drive failed) then there might still be recoverable data on the edge of the tracks.)
So yeah, if you really want to be sure, nuke the drive, but it's also not possible to retrieve everything that was ever on the disk, unless the disk hasn't been used much. (Ie, given blocks haven't been reused much.)
Some apps deliberately use the IP of the NAT host rather than the private IP when informing other nodes of their existence, so that there'll be a usable return address.
Some NAT software recognizes certain well-known P2P apps and translates IPs in the payload as well as in the header, for similar reasons.
A matching IP in payload and header suggests, but in no way proves, that the originating computer and the NAT host are the same.
Heck, I've got an old Sun IPC that's perfectly capable of saturating my DSL line, but the CMOS battery died a long time ago. Until I replaced it (it's built into the clock chip, not something you can pick up at Batteries-Plus) I had to manually enter the MAC every time I rebooted. Usually something with c0:ff:ee in it..
(b) the application could have been custom-hacked to lie about its private IP address,
"Custom hack" doesn't seem to be necessary. Apparently there are a number of apps out there that do this to present a publicly accessible IP when running NAT'd, possibly including some versions of KaZaa (personally I wouldn't know, I've never used it), and also some application-smart NAT software that will translate IP's in the payload of certain known packets as well as the header.
Unless Verizon screwed up, (a) seems out.
You seem to have more faith in phone companies than I do -- I've worked with some. But faults with (a) are not limited to Verizon, they assume that the time stamps presented in the MediaSentry output match (within the window of a DHCP lease at least, which can be very short) the time stamps in Verizon's logs (and that the lookups were done properly -- very questionable if there's a human element). Unless the two different computer systems in question (Verizon's and MediaSentry's) are both running frequent NTP syncs from time servers that can both trace their chains of time synchronization back to the same time server (or time severs guaranteed to be in sync with each other to atomic-clock accuracy), there's no guarantee of that at all.
(For full confidence, you'd have to be able to prove that the NTP sync was being maintained at the specific time the logs were created, too, not merely that the servers were configured that way. My servers are monitored to alarm if ntpd stops running, because sometimes it does, and clock drift happens.)
Anyone who has done a bit of forensics work will tell you that, short of physical destruction of the drive, you can reconstruct every last bit of a computer.
Anyone who tells you that is spouting nonsense, and no forensics expert. If he's calling himself a forensics expert, he's bullshitting.
Yes, you can -- with the right tools -- recover a hell of a lot of what has been deleted (especially on Windows) and in some cases even overwritten, but the latter requires very sophisticated hardware and access to the original drive. Most forensics investigations work off of a bit-for-bit copy of the original drive, not the original itself.
And for stuff that has been written over several times, you'd need the equipment of an advanced physics lab to even begin to have a hope of recovering it, but you'd probably fail.
Security-sensitive outfits mandate physical destruction of drives because it's usually faster, cheaper, the drive's obsolete anyway, and prospective enemies probably do have access to advanced physics labs and aren't too worried about the expense.
Heck, that doesn't even make sense, unless you're talking about a system for distributing CDs, tapes, or floppies. Contrary to Marshall McLuhan, the medium is not the message.
I thought it was near on impossible to write error free software these days.
No, it's just expensive, and slow. There are standard processes for doing this, but they require a lot of up-front specification and design, and thorough documentation thereof, and not just for the end result but of the entire tool chain. (Gotta certify that the compiler actually produces correct binaries from correct source, for example.) Think of things like avionics software or the Shuttle software.
And yes, even then mistakes are sometimes made, but that happens with civil engineering too.
(The problem of these processes being so expensive (time and labor intensive) is that it encourages procurement agencies to go with "cheap off the shelf" solutions where they can get away with. That leads to things like Windows on warships, {shudder}. That's like using potmetal counterfeit bolts in bridge because they're cheaper.)
More like 1st = 5-7, 2nd = 6-8, and so on, although some school districts do have a policy against admitting younger (and smarter) kids even if they can do the work.
I know quite a few people who started even younger than that, or skipped grades. Starting high school (9th grade) at age 12 is admittedly a couple of sigmas beyond the mean, but I personally know several people in that category.
after several recent compromises, sensitive government installations opted to completely replace their equipment.
That presumes that the replacement hardware hasn't had its BIOS and/or firmware flashed with malware itself. Do they checksum the ROMs and compare that against what the vendor tells them? Can they trust the vendor, or were those parts from offshore? Etc, etc....
Sometimes I wonder if sensitive government installations are paranoid enough. Probably only the wrong ones are.
True enough, I was just bemoaning that it doesn't anymore. Just a quick look at/sbin on this Linux box shows about six statically linked binaries - those that deal with kernel modules, init, ldconfig, and sln.
Unfortunately, while/sbin is used for system binaries or binaries normally only used by the super-user, they aren't always statically-linked.
I say "unfortunately" because it's a real pain to try to recover a running system whose libc has been hosed if the binary you need isn't statically linked. (Been there both ways, in the latter case there aren't a lot of options beyond rebooting to a recovery disc, in the former you can just copy in a new libc.)
(Note to self: Don't trust anything from "linuxbasics.org", they're too stupid to know that/etc means exactly what it says: "etc." - "and the rest".)
And/usr no more stands for "Unix System Resources" than/tmp stands for "Transient Meta Partition". Hint: before/home existed (ie, the first fifteen or twenty years of UNIX's existence), user home directories were kept under/usr.
Interesting but wrong, as numerous above posts in different threads have pointed out.
It just stands for "et cetera". They could have used "misc" (for miscellaneous) but the original Unix developers were both more classically educated and had a preference for 3-letter abbreviations./bin is for binaries, it's not a bin as in container,/usr is for user, it originally held user's home directories (before the invention of/home) and/usr/bin was binaries more likely to be used by end users./tmp is clearly temporary (and some systems emptied it on boot)./etc means exactly what etc. means in english: et cetera - "and the rest".
And yes, I'm an old UNIX guy from the early 80s who used Version 7 Unix, although my first contact was in college in the late 70s with 6th edition.
"Editable Text Configuration" is a joke played on newbies who ask.
NASA has been studying inflatables for moonbases or temporary shelters since the Apollo era, although the design then had the tube horizontal rather than vertical (which seems to make more sense to me).
The largely forgettable 1989 movie "Moontrap" featured an inflatable shelter, which gives astronaut Walter Koenig a chance to get the moon-babe's (Leigh Lombardi) clothes off.
Microbarn doesn't have the selection of rapidly-obsolete gear (like hard drives and system boards) that somewher like Newegg might, but their prices are great for the stuff they do carry, especially for things that should be cheap, like cables.
But yeah, all the retail places these days are carrying high mark-up stuff like Belkin. (Price aside, I won't touch Belkin because of the stunt they pulled a few years ago where their routers would periodically hijack HTTP requests.)
Slashdot Mirror
But of course! ;-)
"whooosh"
It's a spreadsheet, damnit. While I might think it was cool to be able to enter cell formulas in Java, or C, or even APL (now, that would be cool), most people who use spreadsheets just want what Excel and OOCalc and GNUmeric already give them.
They don't want to reimplement the whole fricking document in Javascript or Java (or anything else) and run it in a web browser (which is apparently what Opera's CTO wants them to do), they just want to be able to save their work and read it later, and exchange it with others.
Criminy!
Spinning or not spinning is hardly going to make it hard to detect being off course by hundreds of thousands of miles.
True enough, but 3-axis stabilization (not spinning) implies some kind of reaction control system (*) that may be cancelling this unknown effect as it stabilizes the spacecraft, i.e. it also corrects the course.
(* Even if the primary attitude control is via momentum wheel, you still some kind of thruster-based RCS system to periodically dump the momentum you're building up in those wheels.)
So basically you're saying I'd have to write my own spreadsheet program in Javascript.
For some reason the words "fucking insane" come to mind.
I wonder whether the intended the confusion.
This is Microsoft we're talking about. What do you think? They don't do anything without considering the strategic marketing impact. (Well, okay, there was Bob...)
Remember in order for there to be developers someone somewhere has to make money selling software.
Nope. In fact most developers work for companies that do not make money selling software. Now, aside from those few that are losing money selling software (grin), I mean those companies (and other organizations - governments, universities, etc) whose primary product(s) is/are something other than software. (Take your Fortune 500 -- how many of them make most (or any) of their money selling software? How many employ developers?)
Besides which, that's totally irrelevant to open document formats -- just because the document format is open doesn't mean the application software has to be either open or free, any more than standardizing on a character code of ASCII or Unicode does. (EBCDIC, you're on your own.)
I had been thinking that ODF was "obviously" a good thing until I read the rant by Opera's CTO about how shit both standards are (a memory dump between angle brackets), and how the correct way would be to go for XHTML with CSS formatting.
So how do I do a spreadsheet in XHTML with CSS formatting? And I mean a serious computational spreadsheet, perhaps with some charts thrown in, not just some data layed out in a table.
ODF is not just for pretty text documents, its for the product of all kinds of office apps, including spreadsheets and presentations.
As for "memory dump between angle brackets" -- yeah, that's a pretty fair description of OOXML, but doesn't really explain the dozen or more different apps out there that use ODF. Is he trying to tell us that KWord, AbiWord, OOWriter and Google Docs (to name a few that use ODF) all use the same memory layout?
Opera's CTO isn't worthy of his title, if that's the kind of crap he spews.
I'm concerned that standardising on ODF will come to bite us,
Just because it's an ISO standard doesn't mean it's immutable. Standards can be changed, and frequently are to keep up with technology improvements. Also, the language in the California legislation doesn't specify ODF (or it's ISO number) per se, just requires that whatever document spec is used meets the requirements of being open, freely implementable, fully specified, multi-vendor etc that only ODF meets at the moment (and OOXML doesn't). If some new Whizbang Document Format (WDF) is invented and it meets those qualifications, the IT world is free to move to it.
But surely if OOXML is as open as the wikipedia page (and everything I've heard) makes it sound,
Have you heard that Microsoft hired it's own wikipedia contributer to (try to) control the spin on the OOXML and ODF pages?
And I guess you haven't heard about the parts of the OOXML "spec" that say something ot the effect of: "Word95Spacing - This tag means that document spacing should conform to that produced by Word95. That's too complicated to go into here, see Word95 for details."
This is a spec? This is open?
No, I didn't think you were claiming to be a forensics expert, just echoing something you heard. But what you may have heard (or thought you heard) is not quite correct. (Sorry if you thought I was attacking you.)
There reaches a point after a certain number of rewrites where any theoretical residual magnetic signature is down in the quantum noise level, and few than that its in the noise level of the detection gear. Where that particular point is depends on a lot of things: the recovery technology used, the time between when the original data was written and when it was written over (allowing more opportunity for difference in head tracking), the specific recording technology, and so on. (This even assumes you're overwriting the blocks you want overwritten, not all filesystems necessarily do that.)
The FSF program "shred" overwrites data a default of 25 times with different patterns. That's reasonably secure against any recovery, but is bloody slow. (And if the r/w head has been misaligned enough since the original data was written (but not so much as to render the drive failed) then there might still be recoverable data on the edge of the tracks.)
So yeah, if you really want to be sure, nuke the drive, but it's also not possible to retrieve everything that was ever on the disk, unless the disk hasn't been used much. (Ie, given blocks haven't been reused much.)
Some apps deliberately use the IP of the NAT host rather than the private IP when informing other nodes of their existence, so that there'll be a usable return address.
Some NAT software recognizes certain well-known P2P apps and translates IPs in the payload as well as in the header, for similar reasons.
A matching IP in payload and header suggests, but in no way proves, that the originating computer and the NAT host are the same.
Heck, I've got an old Sun IPC that's perfectly capable of saturating my DSL line, but the CMOS battery died a long time ago. Until I replaced it (it's built into the clock chip, not something you can pick up at Batteries-Plus) I had to manually enter the MAC every time I rebooted. Usually something with c0:ff:ee in it..
Lots of hardware will let you change the MAC.
(b) the application could have been custom-hacked to lie about its private IP address,
"Custom hack" doesn't seem to be necessary. Apparently there are a number of apps out there that do this to present a publicly accessible IP when running NAT'd, possibly including some versions of KaZaa (personally I wouldn't know, I've never used it), and also some application-smart NAT software that will translate IP's in the payload of certain known packets as well as the header.
Unless Verizon screwed up, (a) seems out.
You seem to have more faith in phone companies than I do -- I've worked with some. But faults with (a) are not limited to Verizon, they assume that the time stamps presented in the MediaSentry output match (within the window of a DHCP lease at least, which can be very short) the time stamps in Verizon's logs (and that the lookups were done properly -- very questionable if there's a human element). Unless the two different computer systems in question (Verizon's and MediaSentry's) are both running frequent NTP syncs from time servers that can both trace their chains of time synchronization back to the same time server (or time severs guaranteed to be in sync with each other to atomic-clock accuracy), there's no guarantee of that at all.
(For full confidence, you'd have to be able to prove that the NTP sync was being maintained at the specific time the logs were created, too, not merely that the servers were configured that way. My servers are monitored to alarm if ntpd stops running, because sometimes it does, and clock drift happens.)
Anyone who has done a bit of forensics work will tell you that, short of physical destruction of the drive, you can reconstruct every last bit of a computer.
Anyone who tells you that is spouting nonsense, and no forensics expert. If he's calling himself a forensics expert, he's bullshitting.
Yes, you can -- with the right tools -- recover a hell of a lot of what has been deleted (especially on Windows) and in some cases even overwritten, but the latter requires very sophisticated hardware and access to the original drive. Most forensics investigations work off of a bit-for-bit copy of the original drive, not the original itself.
And for stuff that has been written over several times, you'd need the equipment of an advanced physics lab to even begin to have a hope of recovering it, but you'd probably fail.
Security-sensitive outfits mandate physical destruction of drives because it's usually faster, cheaper, the drive's obsolete anyway, and prospective enemies probably do have access to advanced physics labs and aren't too worried about the expense.
like "Media Distribution System"?
Heck, that doesn't even make sense, unless you're talking about a system for distributing CDs, tapes, or floppies. Contrary to Marshall McLuhan, the medium is not the message.
I thought it was near on impossible to write error free software these days.
No, it's just expensive, and slow. There are standard processes for doing this, but they require a lot of up-front specification and design, and thorough documentation thereof, and not just for the end result but of the entire tool chain. (Gotta certify that the compiler actually produces correct binaries from correct source, for example.) Think of things like avionics software or the Shuttle software.
And yes, even then mistakes are sometimes made, but that happens with civil engineering too.
(The problem of these processes being so expensive (time and labor intensive) is that it encourages procurement agencies to go with "cheap off the shelf" solutions where they can get away with. That leads to things like Windows on warships, {shudder}. That's like using potmetal counterfeit bolts in bridge because they're cheaper.)
More like 1st = 5-7, 2nd = 6-8, and so on, although some school districts do have a policy against admitting younger (and smarter) kids even if they can do the work.
I know quite a few people who started even younger than that, or skipped grades. Starting high school (9th grade) at age 12 is admittedly a couple of sigmas beyond the mean, but I personally know several people in that category.
after several recent compromises, sensitive government installations opted to completely replace their equipment.
That presumes that the replacement hardware hasn't had its BIOS and/or firmware flashed with malware itself. Do they checksum the ROMs and compare that against what the vendor tells them? Can they trust the vendor, or were those parts from offshore? Etc, etc....
Sometimes I wonder if sensitive government installations are paranoid enough. Probably only the wrong ones are.
/sbin really did mean static-bin.
/sbin on this Linux box shows about six statically linked binaries - those that deal with kernel modules, init, ldconfig, and sln.
True enough, I was just bemoaning that it doesn't anymore. Just a quick look at
Unfortunately, while /sbin is used for system binaries or binaries normally only used by the super-user, they aren't always statically-linked.
I say "unfortunately" because it's a real pain to try to recover a running system whose libc has been hosed if the binary you need isn't statically linked. (Been there both ways, in the latter case there aren't a lot of options beyond rebooting to a recovery disc, in the former you can just copy in a new libc.)
(Note to self: Don't trust anything from "linuxbasics.org", they're too stupid to know that /etc means exactly what it says: "etc." - "and the rest".)
/usr no more stands for "Unix System Resources" than /tmp stands for "Transient Meta Partition". Hint: before /home existed (ie, the first fifteen or twenty years of UNIX's existence), user home directories were kept under /usr.
And
Interesting but wrong, as numerous above posts in different threads have pointed out.
/bin is for binaries, it's not a bin as in container, /usr is for user, it originally held user's home directories (before the invention of /home) and /usr/bin was binaries more likely to be used by end users. /tmp is clearly temporary (and some systems emptied it on boot). /etc means exactly what etc. means in english: et cetera - "and the rest".
It just stands for "et cetera". They could have used "misc" (for miscellaneous) but the original Unix developers were both more classically educated and had a preference for 3-letter abbreviations.
And yes, I'm an old UNIX guy from the early 80s who used Version 7 Unix, although my first contact was in college in the late 70s with 6th edition.
"Editable Text Configuration" is a joke played on newbies who ask.
NASA has been studying inflatables for moonbases or temporary shelters since the Apollo era, although the design then had the tube horizontal rather than vertical (which seems to make more sense to me).
The largely forgettable 1989 movie "Moontrap" featured an inflatable shelter, which gives astronaut Walter Koenig a chance to get the moon-babe's (Leigh Lombardi) clothes off.
... when you can invoke the War Measures Act?
That's how Canada dealt with (domestic) terrorists the last time.
Yep, and if you're going to order, might as well go somewhere like Microbarn:
14 foot patch cable, $1.99
Microbarn doesn't have the selection of rapidly-obsolete gear (like hard drives and system boards) that somewher like Newegg might, but their prices are great for the stuff they do carry, especially for things that should be cheap, like cables.
But yeah, all the retail places these days are carrying high mark-up stuff like Belkin. (Price aside, I won't touch Belkin because of the stunt they pulled a few years ago where their routers would periodically hijack HTTP requests.)
There's only so many basic settings like "people crash on a mysterious island";
Well, you said it yourself: Lost is ripping off Jules Verne's Mysterious Island. (If you're going to steal, steal from the masters.)