It's just that you also have an option to only use you finger too... which is more direct, more natural...
It's more natural for some things, but I dare you to take it to class, open up a note taking program, and take notes on the iPad then come back to me and say that's more natural than using a stylus.
The point is that a good tablet with more functionality than the iPad requires a good amount of research into how to do tablet UIs.
No it doesn't. Take the iPad and make it possible to load 3rd party apps that haven't gone through Apple's approval process. Voila, a good tablet with more functionality than the iPad.
Add a little slot for a sytlus; voila, a good tablet with (IMO) more functionality than the iPad.
If it can detect it during an offline scan, it can probably detect it during an online scan too.
Not if it's a good rootkit. If the rootkit gets loaded before your scanner, the rootkit can hide itself from your scanner. And if it's ever loaded before your scanner, it can make sure it will always be loaded before it.
Rootkits have been replacing read calls to files that it modified by returning the original, correct data for a long time now. So it wouldn't matter if your scanner hits the right file and has the signature, if the rootkit was there first it can hook the call and return the expected data. Bam... false negative.
The only way to ensure that the rootkit doesn't load is to boot from an uninfected OS.
Aaah OneNote, that is the one application that has done RIGHT in a long time.
I complain about software a lot, whine about editors, dislike every OS out there on some pretty strong points, and am generally pretty picky.
But honestly... there's almost nothing I can complain about with OneNote. There are two or three features I think it'd be awesome if it had*, but... it just works. Really damn well. It's really remarkable. (And I use it heavily for note taking in class with a tablet PC.)
(Okay, I do have one complaint: there's a way to insert some extra space into a page between things you've already written, but there *doesn't* seem to be a way to remove it. Or at least I haven't found it. Not that I've looked particularly hard.)
* The two features are: first, aimed at tablet users, searching for math symbols is basically impossible, but hard to do. But what they could do is a fuzzy drawing search: you draw what you want in a special "find" window and it finds things of approximately the same strokes. Currently you can search for text (and this works surprisingly well -- does fuzzy matching on your notes too) but not pictures.
Second, animation. I see this feature working by marking a section of the screen for an animation, and then it will record everything you draw in it. It gives you a slider so you can go forward and backward in time. Maybe it's set up so it just displays everything as you drew it, but you could also have it so you have to manually give it states to save and it just cycles between those. The point of this would be if you want to illustrate a process, you're restricted to what you have with paper -- which typically means draw multiple diagrams. Copy and paste can make this a little easier, but animation would actually let you take advantage of the fact that you're not working with paper.
Windows has tried to protect its own files when running for years; even XP had some form of protection. (E.g. replacing uxtheme.dll I think took a little handiwork so that Windows wouldn't immediately replace it with the old version.)
In order to get around this, malware already hides changes from the system; e.g. intercepting ReadFile calls to that file and returning incorrect (but expected) results would do the trick.
Oh, I forgot to mention that I would say there's no reason to do such a thorough evaluation very often. Every minor release is probably unnecessary. I've really only done that on major versions, though I've played around a small amount with the minor releases.
I also follow a couple enhancement requests on their bug tracking software, so I know whether they've gotten around to fixing a few things that are important to me. A couple times the answer has been yes; most of the time it's no. (Hence the fairly low opinion of the software.)
So, I'm asking: is it worth it to download OpenOffice, and play a bit around with it, or if I'm a "happy" Microsoft Office user, I can just "move along"?:)
Quite possibly. If you think you might be interested, then download it and actually use it for a day or two and see what you think. Go ahead and use MS Office if you really have to during that time (e.g. the PPTX import feature of Impress is not much better than useless, at least in 3.1), but try not to. The point is to use it long enough that you get used to it being different and hopefully start judging on its merits and not just have a visceral reaction "ugh this is ugly and obnoxious".
The great thing about Free software like this is you can do that -- just download it and try it out. And it's good enough that it'll probably suffice for at least your Word and Excel needs. (I doubt you'll think it even comes close to Visio, and there's nothing OneNote-like in it.)
You may decide you really like it, and switch to it primarily. You may decide that you don't mind it, but using MS Office makes things easier (esp. with the Visio/OneNote/Access use). Or you may hate it. Any of them are fine. (I fall somewhere between the last two categories.)
Let's say a production web server, while you may cd and ls a lot, you're gonna do a lot more grep and probably even more top or wget than you would make.
Oh, I totally agree. But if you're on a production web server, hopefully you're not a "Windows nube" like interval1066 was talking about.:-)
Point is that 99.9% of Windows users wouldn't use wget if they had it.
Oh yes, that's really easier that to type ls -l, ps -ef or ps -ef|grep firefox
Okay, now what's your command line for printing just the file name and it's size? (Pretend you can't use 'du' or something like that.) Or just a list of process IDs with their command lines. (Both of these may be possible -- but the point is that the means of doing so isn't discoverable, you have to read the docs. In PowerShell, these operations *are* somewhat discoverable.)
Besides, I never claimed that PowerShell was better or easier to use than the Unix shells. (I certainly don't claim it's as concise; verbosity is pretty common with Windows API names and such too.) My main claim is that it's merely different -- different enough that to say that it's a copy on anything more than a "hey look, Windows has a half decent command line" level does a disservice to what the PowerShell team did.
Sorry, but the real advantages in the *nix shells is that every output is just plain simple text. That means, I can grep it, parse it, format it what ever I like and won't be restricted to the PowerShell to do anything use full.
The fact that "every output is just plain simple text" can very much be a drawback too, because it means that a lot of the time you wind up doing some ad-hoc parsing that often works "well enough" but has problems.
For instance, take something that I did earlier today for this post: extract from my shell history file a list of the commands I have run so I could sort them and count occurrences.
My history file has lines that look like this:
: 1265787576:0;tail zsh-history
(The first number is the timestamp, the second number is duration.) Give me a command line that will return a list of command names I've run, so that I can then pipe it to "sort | uniq -c | sort -g".
No really, I'm not kidding; come up with what you would do before reading on.
My assertion is that this would be trivial in the PowerShell world, if there was a "history" command that would return a list of objects containing, e.g., a CommandPath field. Just 'get-history | select-object CommandPath'.
What did I do in Linux? This: cat zsh-history | cut "-d;" -f2 | cut "-d " -f1 This isn't so ugly... but it also has a ton of problems:
If I had quoted a command name -- say because the path had spaces -- then those quotes wouldn't have been removed for the sort step and would have been counted separately from an unquoted command. Worse, if the path had actually contained spaces, it would have only picked up the path until the first space.
Running a command by specifying a full path shows up differently than letting the shell search $PATH for it. Piping through 'basename' or something could fix this -- but at the cost of incorrectly collapsing commands that are in different directories into one entry.
I occasionally started a command with an environment variable explicitly set on the command line -- e.g. BLAH=foo cmd. In this case, the output from my pipeline would say the command is BLAH=foo.
How many of these problems did your solution have?
(I don't claim that mine is the best possible one -- but I don't know a way to do better without adding *substantial* complexity, and I'm quite comfortable at the command line and at least somewhat conversant with most of the standard Unix utilities.)
So basically it took how many years for MS to re-implement grep ?
Not as many as you're implying. (Okay, I don't when findstr showed up, but that's 'find' in an installation of MS-DOS 6.22. That's as far back as I have media for.)
It's not installed by default on either WinXP or Vista.
What default installation are you using? I just did a default installation of XP (based on a CD image with SP2 included) -- not that there was really anything in the installation that lets you choose components -- and I have find and findstr (opening the command prompt, running find and findstr, and taking that screenshot were the first things I did when I got to the desktop).
In fact, this was incorrect; I got that number by grepping my history for 'wget' and piping to 'wc -l'. One of those occurrences was 'grep wget zsh-history', issued as I was gathering these statistics, so really I've only used wget twice in the last 10,000 commands.
(Okay, I can't speak to their power, but most of the time I use grep I'm doing something that 'find' could do just fine; 99% of the rest of the time the only thing 'find' couldn't handle is a recursive search, and then 'findstr' would work.)
As for WGet, I'm surprised you selected that; I am a heavy user of both Linux and Windows + Cygwin and I use wget no more than once in a blue moon. Something like 'sed' in my mind is missing way more than wget. (For the curious, in the last ~10,000 commands stored in my shell history (dating back to July 2009, though this makes me suspicious that not everything is there), the most commonly used ones are: cd (1290), scons (1061), ls (981), python (520), make (496), cvs (484), cat (423), fg (296), git (248), and exit (196). 'wget' was used three times, which puts it about the median.)
Don't do that. Go to ctrl panel, administrative tools, services, find the "Windows Update" service (I think that's it's name) in the list, and tell it to stop.
That's obviously coming from someone who hasn't spend much time with PowerShell.
I don't make the claim that PS is better (or worse) than the Unix shells, but it does bring a bunch of things to the table that aren't in any common shell. In particular, the ability to pipe objects between processes instead of just text.
In fact, besides "a capable command line tool", there's really almost nothing that MS took from Unix with PowerShell. (In particular, in some very obnoxious ways it still behaves like cmd.exe, and it still is hosted in the god-awful terminal program that cmd is.)
For instance, here's the output of 'dir' in PowerShell:
Suppose I want just the name? Under Unix, I'd have to pass some ls-specific flag to get just the name. (Pretend ls worked the opposite it does, and gave long listings by default. This detail doesn't change what I'm saying.) In PowerShell, I just say I want the name field, with dir | select-object name:
Name
examples-v3 swyx file.dot file.dot~ ...
I want the name and time it was created? That's dir | select-object Name,CreationTime:
Name CreationTime
examples-v3 12/2/2009 4:48:55 PM swyx 11/2/2009 4:57:30 PM file.dot 11/19/2009 11:22:33 AM file.dot~ 11/19/2009 11:24:34 AM ...
The same syntax works for other commands. This is get-process | select-object Id,ProcessName:
That's because what 'dir' and 'get-process' actually output is a list of objects, which PowerShell then formats in the table it displays. 'select-object' (I don't claim it's well-named) removes unselected fields from the given objects. 'select-object' (I don't claim it's well-named) removes unselected fields from the given objects.
(lameness filter blah blah blah... using up some space blah blah blah. Hey, did you hear about the/. poster who got trolled? Oh, that's everyone right.)
Heck MS's PSS would routinely have you use his tools even before the purchase because nothing they put out internally was nearly as useful.
Around the time MS hired Russinovich a lot of people on Slashdot were worried that it would mean the death of Sysinternals's tools, but what you say is why that argument was almost ridiculous: there'd have been open revolt within MS if that went down.
(I suppose they could have kept the tools internal to MS, but that didn't seem likely.)
I believe it is also absent in 64 bit Windows 7. I wonder about 64 bit XP & 64 bit Vista.
The 16-bit subsystem is absent in all 64-bit versions of Windows.
While I'm not positive, my understanding is that this has to do with the CPU not running 16-bit code when in 64-bit mode or something like that, but don't quote me on that. I know VMWare can run a 64-bit guest on a 32-bit host and vice versa, so this can't be the whole story; perhaps VMWare changes the processor's mode between 64-bit and 32-bit when doing a world switch or something like that, and MS doesn't want to do that in order to keep around the 16-bit subsystem.
This is very possibly not workable. None of standard C's fopen, C++'s fstream, or Win32's CreateFile function will interpolate the %USERNAME% component; you have to specify the hard path. (This was tested on XP.)
As long as you're rooting around, might as well scan for any files named/password.*/, and send them back to control, along with a list of all sites with cookies.
The article is crap when it comes to actual information, but it's quite possible (I'd say more likely than not) that both of these are not possible.
Slashdot Mirror
It's just that you also have an option to only use you finger too... which is more direct, more natural...
It's more natural for some things, but I dare you to take it to class, open up a note taking program, and take notes on the iPad then come back to me and say that's more natural than using a stylus.
The point is that a good tablet with more functionality than the iPad requires a good amount of research into how to do tablet UIs.
No it doesn't. Take the iPad and make it possible to load 3rd party apps that haven't gone through Apple's approval process. Voila, a good tablet with more functionality than the iPad.
Add a little slot for a sytlus; voila, a good tablet with (IMO) more functionality than the iPad.
If it can detect it during an offline scan, it can probably detect it during an online scan too.
Not if it's a good rootkit. If the rootkit gets loaded before your scanner, the rootkit can hide itself from your scanner. And if it's ever loaded before your scanner, it can make sure it will always be loaded before it.
Rootkits have been replacing read calls to files that it modified by returning the original, correct data for a long time now. So it wouldn't matter if your scanner hits the right file and has the signature, if the rootkit was there first it can hook the call and return the expected data. Bam... false negative.
The only way to ensure that the rootkit doesn't load is to boot from an uninfected OS.
Aaah OneNote, that is the one application that has done RIGHT in a long time.
I complain about software a lot, whine about editors, dislike every OS out there on some pretty strong points, and am generally pretty picky.
But honestly... there's almost nothing I can complain about with OneNote. There are two or three features I think it'd be awesome if it had*, but... it just works. Really damn well. It's really remarkable. (And I use it heavily for note taking in class with a tablet PC.)
(Okay, I do have one complaint: there's a way to insert some extra space into a page between things you've already written, but there *doesn't* seem to be a way to remove it. Or at least I haven't found it. Not that I've looked particularly hard.)
* The two features are: first, aimed at tablet users, searching for math symbols is basically impossible, but hard to do. But what they could do is a fuzzy drawing search: you draw what you want in a special "find" window and it finds things of approximately the same strokes. Currently you can search for text (and this works surprisingly well -- does fuzzy matching on your notes too) but not pictures.
Second, animation. I see this feature working by marking a section of the screen for an animation, and then it will record everything you draw in it. It gives you a slider so you can go forward and backward in time. Maybe it's set up so it just displays everything as you drew it, but you could also have it so you have to manually give it states to save and it just cycles between those. The point of this would be if you want to illustrate a process, you're restricted to what you have with paper -- which typically means draw multiple diagrams. Copy and paste can make this a little easier, but animation would actually let you take advantage of the fact that you're not working with paper.
Windows has tried to protect its own files when running for years; even XP had some form of protection. (E.g. replacing uxtheme.dll I think took a little handiwork so that Windows wouldn't immediately replace it with the old version.)
In order to get around this, malware already hides changes from the system; e.g. intercepting ReadFile calls to that file and returning incorrect (but expected) results would do the trick.
This post does not have nearly enough funny mods.
Oh, I forgot to mention that I would say there's no reason to do such a thorough evaluation very often. Every minor release is probably unnecessary. I've really only done that on major versions, though I've played around a small amount with the minor releases.
I also follow a couple enhancement requests on their bug tracking software, so I know whether they've gotten around to fixing a few things that are important to me. A couple times the answer has been yes; most of the time it's no. (Hence the fairly low opinion of the software.)
So, I'm asking: is it worth it to download OpenOffice, and play a bit around with it, or if I'm a "happy" Microsoft Office user, I can just "move along"? :)
Quite possibly. If you think you might be interested, then download it and actually use it for a day or two and see what you think. Go ahead and use MS Office if you really have to during that time (e.g. the PPTX import feature of Impress is not much better than useless, at least in 3.1), but try not to. The point is to use it long enough that you get used to it being different and hopefully start judging on its merits and not just have a visceral reaction "ugh this is ugly and obnoxious".
The great thing about Free software like this is you can do that -- just download it and try it out. And it's good enough that it'll probably suffice for at least your Word and Excel needs. (I doubt you'll think it even comes close to Visio, and there's nothing OneNote-like in it.)
You may decide you really like it, and switch to it primarily. You may decide that you don't mind it, but using MS Office makes things easier (esp. with the Visio/OneNote/Access use). Or you may hate it. Any of them are fine. (I fall somewhere between the last two categories.)
Why are they afraid of telling us exactly what that "competetive" price is, and competetive to what?
Maybe they're still working it out.
Let's say a production web server, while you may cd and ls a lot, you're gonna do a lot more grep and probably even more top or wget than you would make.
Oh, I totally agree. But if you're on a production web server, hopefully you're not a "Windows nube" like interval1066 was talking about. :-)
Point is that 99.9% of Windows users wouldn't use wget if they had it.
Oh yes, that's really easier that to type ls -l, ps -ef or ps -ef|grep firefox
Okay, now what's your command line for printing just the file name and it's size? (Pretend you can't use 'du' or something like that.) Or just a list of process IDs with their command lines. (Both of these may be possible -- but the point is that the means of doing so isn't discoverable, you have to read the docs. In PowerShell, these operations *are* somewhat discoverable.)
Besides, I never claimed that PowerShell was better or easier to use than the Unix shells. (I certainly don't claim it's as concise; verbosity is pretty common with Windows API names and such too.) My main claim is that it's merely different -- different enough that to say that it's a copy on anything more than a "hey look, Windows has a half decent command line" level does a disservice to what the PowerShell team did.
Sorry, but the real advantages in the *nix shells is that every output is just plain simple text. That means, I can grep it, parse it, format it what ever I like and won't be restricted to the PowerShell to do anything use full.
The fact that "every output is just plain simple text" can very much be a drawback too, because it means that a lot of the time you wind up doing some ad-hoc parsing that often works "well enough" but has problems.
For instance, take something that I did earlier today for this post: extract from my shell history file a list of the commands I have run so I could sort them and count occurrences.
My history file has lines that look like this:
(The first number is the timestamp, the second number is duration.) Give me a command line that will return a list of command names I've run, so that I can then pipe it to "sort | uniq -c | sort -g".
No really, I'm not kidding; come up with what you would do before reading on.
My assertion is that this would be trivial in the PowerShell world, if there was a "history" command that would return a list of objects containing, e.g., a CommandPath field. Just 'get-history | select-object CommandPath'.
What did I do in Linux? This:
cat zsh-history | cut "-d;" -f2 | cut "-d " -f1
This isn't so ugly... but it also has a ton of problems:
How many of these problems did your solution have?
(I don't claim that mine is the best possible one -- but I don't know a way to do better without adding *substantial* complexity, and I'm quite comfortable at the command line and at least somewhat conversant with most of the standard Unix utilities.)
So basically it took how many years for MS to re-implement grep ?
Not as many as you're implying. (Okay, I don't when findstr showed up, but that's 'find' in an installation of MS-DOS 6.22. That's as far back as I have media for.)
It's not installed by default on either WinXP or Vista.
What default installation are you using? I just did a default installation of XP (based on a CD image with SP2 included) -- not that there was really anything in the installation that lets you choose components -- and I have find and findstr (opening the command prompt, running find and findstr, and taking that screenshot were the first things I did when I got to the desktop).
'wget' was used three times
In fact, this was incorrect; I got that number by grepping my history for 'wget' and piping to 'wc -l'. One of those occurrences was 'grep wget zsh-history', issued as I was gathering these statistics, so really I've only used wget twice in the last 10,000 commands.
Grep and wget.
Yeah, it's too bad that Windows doesn't come with a command line utility like Grep.
(Okay, I can't speak to their power, but most of the time I use grep I'm doing something that 'find' could do just fine; 99% of the rest of the time the only thing 'find' couldn't handle is a recursive search, and then 'findstr' would work.)
As for WGet, I'm surprised you selected that; I am a heavy user of both Linux and Windows + Cygwin and I use wget no more than once in a blue moon. Something like 'sed' in my mind is missing way more than wget. (For the curious, in the last ~10,000 commands stored in my shell history (dating back to July 2009, though this makes me suspicious that not everything is there), the most commonly used ones are: cd (1290), scons (1061), ls (981), python (520), make (496), cvs (484), cat (423), fg (296), git (248), and exit (196). 'wget' was used three times, which puts it about the median.)
Don't do that. Go to ctrl panel, administrative tools, services, find the "Windows Update" service (I think that's it's name) in the list, and tell it to stop.
You mean other than UNIX and Linux systems? I don't see any comparable functionality that is not already available on those systems.
If you're so confidant, perhaps you can answer this guy's questions.
I have 'top' aliased to 'htop' on my Linux setup, htop is that awesome.
My personal response to that is "Those who do understand Unix are condemned to think it's actually a good idea."
That's obviously coming from someone who hasn't spend much time with PowerShell.
I don't make the claim that PS is better (or worse) than the Unix shells, but it does bring a bunch of things to the table that aren't in any common shell. In particular, the ability to pipe objects between processes instead of just text.
In fact, besides "a capable command line tool", there's really almost nothing that MS took from Unix with PowerShell. (In particular, in some very obnoxious ways it still behaves like cmd.exe, and it still is hosted in the god-awful terminal program that cmd is.)
For instance, here's the output of 'dir' in PowerShell:
Suppose I want just the name? Under Unix, I'd have to pass some ls-specific flag to get just the name. (Pretend ls worked the opposite it does, and gave long listings by default. This detail doesn't change what I'm saying.) In PowerShell, I just say I want the name field, with dir | select-object name:
I want the name and time it was created? That's dir | select-object Name,CreationTime:
The same syntax works for other commands. This is get-process | select-object Id,ProcessName:
That's because what 'dir' and 'get-process' actually output is a list of objects, which PowerShell then formats in the table it displays. 'select-object' (I don't claim it's well-named) removes unselected fields from the given objects. 'select-object' (I don't claim it's well-named) removes unselected fields from the given objects.
(lameness filter blah blah blah... using up some space blah blah blah. Hey, did you hear about the /. poster who got trolled? Oh, that's everyone right.)
Heck MS's PSS would routinely have you use his tools even before the purchase because nothing they put out internally was nearly as useful.
Around the time MS hired Russinovich a lot of people on Slashdot were worried that it would mean the death of Sysinternals's tools, but what you say is why that argument was almost ridiculous: there'd have been open revolt within MS if that went down.
(I suppose they could have kept the tools internal to MS, but that didn't seem likely.)
I believe it is also absent in 64 bit Windows 7. I wonder about 64 bit XP & 64 bit Vista.
The 16-bit subsystem is absent in all 64-bit versions of Windows.
While I'm not positive, my understanding is that this has to do with the CPU not running 16-bit code when in 64-bit mode or something like that, but don't quote me on that. I know VMWare can run a 64-bit guest on a 32-bit host and vice versa, so this can't be the whole story; perhaps VMWare changes the processor's mode between 64-bit and 32-bit when doing a world switch or something like that, and MS doesn't want to do that in order to keep around the 16-bit subsystem.
C:\windows\system32\config\sam
Read-only access is all you need...
And I'm sure that'd work real well:
C:\WINDOWS\system32>echo %USERNAME%
administrator
C:\WINDOWS\system32>c:\cygwin\bin\head config\sam /usr/bin/head: cannot open `config\\sam' for reading: Device or resource busy
C:\WINDOWS\system32>type config\sam
The process cannot access the file because it is being used by another process.
How about the system doesn't allow the fecking web browser to read your personal files?
That's somewhat what protected mode does.
This is very possibly not workable. None of standard C's fopen, C++'s fstream, or Win32's CreateFile function will interpolate the %USERNAME% component; you have to specify the hard path. (This was tested on XP.)
As long as you're rooting around, might as well scan for any files named /password.*/, and send them back to control, along with a list of all sites with cookies.
The article is crap when it comes to actual information, but it's quite possible (I'd say more likely than not) that both of these are not possible.