Funny, he didn't mention that he "wrote" the book in HTML, just that he "published" it in HTML.
"It is now possible, even feasible, to use HTML as the document format for books." (Granted, that's two links off the/. summary.) But "To prove how powerful it can be, the authors decided to use CSS in the production process" is following only one link.
That PDF posted above was generated entirely from an HTML + CSS document.
html/css sucks at MANY things - how about a self-updating TOC? (don't even try to say some javascript parsing the DOM for header tags with certain IDs to generate it dynamically!)
This would have to be done by the tool displaying it, same as a self-updating TOC in a Word or OpenOffice Writer document. The information is present in a correctly-structured HTML document in the form of Hx tags.
Hell, how can you even tell the page numbers in a html "document" anyways?
The same way you would in a Word document. It doesn't make sense if you're looking at it as a web page in your browser, but if your editor used HTML it would work the same way. (This also partially alleviates the rendering issues.)
Instead of taking one of two specifications created just for rich document formats, he suggests making a brand new specification by extending CSS/HTML to do something it doesn't yet seem ready to do.
Did you RTFA? He's not suggesting making NEW specifications because THEY ARE ALREADY MADE. There is a proof-of-concept "book" linked from the page that demonstrates that the technology to do this at least reasonably well already exists.
...i've always thought, why doesn't updating this style make all text with that style update? Why do I always have to change the same thing over and over again?
Both OpenOffice and Word have this, they're called styles. Word has had it since at least version 6 and maybe before, though at least before 2007 (which I haven't used so can't comment on) they haven't done much to bring attention to the feature. OO has had it since the beginning, and puts rather more emphasis on it.
Not really when it's rendered into, say, a PDF formatted like any book. Based on that sample, I'd say that if that's not a (proof-of-concept) book possibly nothing you download is. It looks pretty reasonable.
BTW, the above is XP Pro pre-SP2. (It's a really old laptop I don't really use. Doesn't even have a network connection, which makes it harder to update, and it's just not worth it. In fact, I don't know if there is any third-party software on there since I've reinstalled the OS last.)
Also, it strikes me that the above isn't a very good test, and the same result could have achieved with just ACLs. If you actually care and for some reason still don't believe me, I'll boot into Linux and see if I can open the files as root with Linux's NTFS drivers.
That's because it decrypts it when you copy because FAT doesn't support encryption, so Windows decrypts it as it copies. It's a feature of NTFS and isn't supported on FAT, but NTFS is a part of Windows and it IS supported there.
Don't believe me?
1. Create two new user accounts, user1 and user2. Make user1 an admin, and user2 limited. 2. Log on as user1. Create a file, c:\user1.txt (assuming C:\ NTFS). Put some known text in it. Encrypt it. 3. Repeat step 2 with user2 (except save in my documents instead of c:\) 4. Log on as your normal admin account. Try to open user1.txt and user2.txt. Oh look! You can't!
Wow, the domain's still available, even an hour or two* after your post. This surprises me. And tempts me a bit, but I don't think I'll follow through.;-)
(I forget if/. displays times to me in CST or EST)
Hugh Thompson, who was interviewed by Slashdot on the dangers of e-voting, now has a cool blog entry on how he was able to bring down the gaming/movie console on an airplane
Except that your stripped down OS won't do one-tenth the things that VMWare does. Would you be able to migrate machines? Checkpoint them? Run Windows on a non-VT chip? From what VMWare says, run Windows as fast as they do even on a VT chip?
Even Virtual PC which MS bought and is the basis for this is noticeably behind VMWare.
Well, judging by the headline, its coolant failed and it went all Chernobyl.
Re:Static analysis is the sh-t
on
PMD Applied
·
· Score: 1
You aren't. But I doubt a static analyzer is going to catch that either.
There are tools that are powerful enough to find if small programs (such as drivers) have these bugs. They are limited by assumptions they make so they can't catch all of them, but they will usually either prove a program free of these bugs up to their assumptions or find a concrete counterexample.
(In general such a problem is of course undecidable, but for the tiny subsets of possible programs that represent what we actually write, these conclusions can often be made.)
By abstracting away the mutation. You can do this with C++ RAII, or Monadically, or any number of other ways.
The above was just an example. I agree that you can remove most of these bugs with C++ for instance, but I don't think you'll get a language that doesn't have some similar mistakes you can make, at least until processors or compilers get around to implementing DWIM.
Like I said, as long as there are bugs in programs, there will be bugs in programs that static analysis tools can find and help you prevent. And aside from DWIM, I don't think you'll get a language that will change that.
Re:Static analysis is the sh-t
on
PMD Applied
·
· Score: 1
As an aside though, I question the sanity of such tools. Why does the language allow you to use constructs that an automated tool could tell you were screwy?!? Shouldn't such practices the tool catches be compiler errors, or at least glaring warnings? It seems to me that the underlying problems lie in the language specifications and the tools are only there as a band-aid.
If you're writing a device driver, how are you going to have a language that tells you not to call a sleeping function when the IRQ level is too high? How are you going to have a language that will prevent you from writing 'close(fd);' twice in a row?
Fact is, many of the things these some of tools are looking for at are a MUCH higher level than PL syntax. Making the compiler reject or complain about them would require essentially putting the static analysis into the compiler.
As long as there are buggy programs, there are going to be bugs that you can detect automatically via static analysis.
Re:Static Analyzer Run != Code Review
on
PMD Applied
·
· Score: 1
(I'm not denigrating code reviews either in the slightest; you're very right that they are nearly complimentary. But I do think that they can do more than you acknowledge.)
they = static analysis tools
Got my antecedents confused...
Re:Static Analyzer Run != Code Review
on
PMD Applied
·
· Score: 1
Static analyzers are very good at catching *certain* *limited* *kinds* of obvious and likely or conceivable bugs...
To be fair, good static analysis tools can catch a LOT of non-obvious bugs. MS Research's SLAM project found a bug in the parallel port driver, a race condition that was triggered if a program was closing the port at the same time the port was being physically removed from the computer. (Hey, it could happen if you have a laptop docking station.) This bug went undetected for like a decade to be found automatically.
You might debate the value of finding that particular bug, but static analysis can do more than find obvious bugs. (In fact, in some cases, they can prove a program free of a particular class of bugs. Let's see you do that with code reviews.)
(I'm not denigrating code reviews either in the slightest; you're very right that they are nearly complimentary. But I do think that they can do more than you acknowledge.)
Re:Static analysis is the sh-t
on
PMD Applied
·
· Score: 1
WTF? MS has invested HEAVILY in static analysis. MS Research's SLAM project, which has been semi-commercialized in the static driver verifier for driver development, is one of the most important projects in the field.
An idea if you can find a source code formatter that you can customize to your liking is to have a canonical format for storing in the repository, then on update run it through a formatter to convert it to the canonical format and on checkout/update run it through your formatter.
Sure, if every user had one instead of making it a system-wide thing.
As the other poster replying to your post said, there is a per-user part of the registry. The system-wide configuration is in HKEY_LOCAL_MACHINE mostly (often abbreviated to HKLM), and the per-user configuration is in HKEY_CURRENT_USER.
IMO, the biggest problem for the registry is that it lumps all configuration data together into a non-portable database. I can't really backup my settings for programs that store things there. But if I want to back up my Gaim configuration, that's easy; just zip up ~/.gaim.
For a Windows system, it seems to me there should be relatively few 3rd-party apps installed system-wide.
Huh? Why? In what environment?
Is the same true of Unix in your opinion? If not, why the difference?
At our company, we users don't have Admin privileges for our own machines (and I don't really blame IT for it). As a result, if I'm trying to try out open-source math/science apps, I can probably install 1 out of 5 that don't install to windows/system or some other protected directory. Almost impossible to maintain an environment where machines are secure and usable.
Agreed, that's a problem. How much is MS's fault for encouraging that sort of model and how much should be blamed on 3rd parties, I don't know.
Hey, why couldn't we use USB2, wouldn't ~400mbits be enough?
No.
DVI supports 1.65 Gbits/sec. It's got resolution to spare for most people, but not all that much. Drop the resolution by half in each dimension (which would be required to make 400 mbps assuming bandwidth is proportional to area) and you're at about 950x600, which would be downright unbearable IMO.
And I know LCDs are all the rage, but my CRT wouldn't run at the refresh rate and resolution I run mine at over DVI. (1600x1200, 75hz.)
Slashdot Mirror
File | Export to Web? Am I missing something here?
/. summary.) But "To prove how powerful it can be, the authors decided to use CSS in the production process" is following only one link.
Yes, the fact that he used a program called Prince to generate a reasonably professional-looking "book". Not "printed web page". Book.
Funny, he didn't mention that he "wrote" the book in HTML, just that he "published" it in HTML.
"It is now possible, even feasible, to use HTML as the document format for books." (Granted, that's two links off the
That PDF posted above was generated entirely from an HTML + CSS document.
html/css sucks at MANY things - how about a self-updating TOC? (don't even try to say some javascript parsing the DOM for header tags with certain IDs to generate it dynamically!)
This would have to be done by the tool displaying it, same as a self-updating TOC in a Word or OpenOffice Writer document. The information is present in a correctly-structured HTML document in the form of Hx tags.
Hell, how can you even tell the page numbers in a html "document" anyways?
The same way you would in a Word document. It doesn't make sense if you're looking at it as a web page in your browser, but if your editor used HTML it would work the same way. (This also partially alleviates the rendering issues.)
Instead of taking one of two specifications created just for rich document formats, he suggests making a brand new specification by extending CSS/HTML to do something it doesn't yet seem ready to do.
Did you RTFA? He's not suggesting making NEW specifications because THEY ARE ALREADY MADE. There is a proof-of-concept "book" linked from the page that demonstrates that the technology to do this at least reasonably well already exists.
Hmm, looks like I was beaten to the punch. Feel free to mod me redundant.
...i've always thought, why doesn't updating this style make all text with that style update? Why do I always have to change the same thing over and over again?
Both OpenOffice and Word have this, they're called styles. Word has had it since at least version 6 and maybe before, though at least before 2007 (which I haven't used so can't comment on) they haven't done much to bring attention to the feature. OO has had it since the beginning, and puts rather more emphasis on it.
Not really when it's rendered into, say, a PDF formatted like any book. Based on that sample, I'd say that if that's not a (proof-of-concept) book possibly nothing you download is. It looks pretty reasonable.
Did you RTFA?
Amen. If you want art, go to an art gallery.
What about the website of an art gallery?
And why should these aesthetic presence and ease-of-use conflict?
BTW, the above is XP Pro pre-SP2. (It's a really old laptop I don't really use. Doesn't even have a network connection, which makes it harder to update, and it's just not worth it. In fact, I don't know if there is any third-party software on there since I've reinstalled the OS last.)
Also, it strikes me that the above isn't a very good test, and the same result could have achieved with just ACLs. If you actually care and for some reason still don't believe me, I'll boot into Linux and see if I can open the files as root with Linux's NTFS drivers.
That's because it decrypts it when you copy because FAT doesn't support encryption, so Windows decrypts it as it copies. It's a feature of NTFS and isn't supported on FAT, but NTFS is a part of Windows and it IS supported there.
Don't believe me?
1. Create two new user accounts, user1 and user2. Make user1 an admin, and user2 limited.
2. Log on as user1. Create a file, c:\user1.txt (assuming C:\ NTFS). Put some known text in it. Encrypt it.
3. Repeat step 2 with user2 (except save in my documents instead of c:\)
4. Log on as your normal admin account. Try to open user1.txt and user2.txt. Oh look! You can't!
Current estimates are 6.6 billion. Where'd you get 8 from? And who's the dumbass?
Wow, the domain's still available, even an hour or two* after your post. This surprises me. And tempts me a bit, but I don't think I'll follow through. ;-)
/. displays times to me in CST or EST)
(I forget if
Hugh Thompson, who was interviewed by Slashdot on the dangers of e-voting, now has a cool blog entry on how he was able to bring down the gaming/movie console on an airplane
/. to it?
What, did they link
Except that your stripped down OS won't do one-tenth the things that VMWare does. Would you be able to migrate machines? Checkpoint them? Run Windows on a non-VT chip? From what VMWare says, run Windows as fast as they do even on a VT chip?
Even Virtual PC which MS bought and is the basis for this is noticeably behind VMWare.
Well, judging by the headline, its coolant failed and it went all Chernobyl.
You aren't. But I doubt a static analyzer is going to catch that either.
There are tools that are powerful enough to find if small programs (such as drivers) have these bugs. They are limited by assumptions they make so they can't catch all of them, but they will usually either prove a program free of these bugs up to their assumptions or find a concrete counterexample.
(In general such a problem is of course undecidable, but for the tiny subsets of possible programs that represent what we actually write, these conclusions can often be made.)
By abstracting away the mutation. You can do this with C++ RAII, or Monadically, or any number of other ways.
The above was just an example. I agree that you can remove most of these bugs with C++ for instance, but I don't think you'll get a language that doesn't have some similar mistakes you can make, at least until processors or compilers get around to implementing DWIM.
Like I said, as long as there are bugs in programs, there will be bugs in programs that static analysis tools can find and help you prevent. And aside from DWIM, I don't think you'll get a language that will change that.
As an aside though, I question the sanity of such tools. Why does the language allow you to use constructs that an automated tool could tell you were screwy?!? Shouldn't such practices the tool catches be compiler errors, or at least glaring warnings? It seems to me that the underlying problems lie in the language specifications and the tools are only there as a band-aid.
If you're writing a device driver, how are you going to have a language that tells you not to call a sleeping function when the IRQ level is too high? How are you going to have a language that will prevent you from writing 'close(fd);' twice in a row?
Fact is, many of the things these some of tools are looking for at are a MUCH higher level than PL syntax. Making the compiler reject or complain about them would require essentially putting the static analysis into the compiler.
As long as there are buggy programs, there are going to be bugs that you can detect automatically via static analysis.
(I'm not denigrating code reviews either in the slightest; you're very right that they are nearly complimentary. But I do think that they can do more than you acknowledge.)
they = static analysis tools
Got my antecedents confused...
Static analyzers are very good at catching *certain* *limited* *kinds* of obvious and likely or conceivable bugs...
To be fair, good static analysis tools can catch a LOT of non-obvious bugs. MS Research's SLAM project found a bug in the parallel port driver, a race condition that was triggered if a program was closing the port at the same time the port was being physically removed from the computer. (Hey, it could happen if you have a laptop docking station.) This bug went undetected for like a decade to be found automatically.
You might debate the value of finding that particular bug, but static analysis can do more than find obvious bugs. (In fact, in some cases, they can prove a program free of a particular class of bugs. Let's see you do that with code reviews.)
(I'm not denigrating code reviews either in the slightest; you're very right that they are nearly complimentary. But I do think that they can do more than you acknowledge.)
WTF? MS has invested HEAVILY in static analysis. MS Research's SLAM project, which has been semi-commercialized in the static driver verifier for driver development, is one of the most important projects in the field.
An idea if you can find a source code formatter that you can customize to your liking is to have a canonical format for storing in the repository, then on update run it through a formatter to convert it to the canonical format and on checkout/update run it through your formatter.
Maybe he's just an uber-techie who always counts from zero, ever think of that?
;-)
Sheesh, you Bush bashers, always so quick to the trigger.
Really? I would have said most of the country was settled with horses in mind, but that's just me.
(I know what you're trying to say, but I don't know how much cheap gas has to do with our settlement patterns.)
Maybe we can reduce that lag if we install broadband. I hear ping times are noticeably faster than dial-up...
Sure, if every user had one instead of making it a system-wide thing.
As the other poster replying to your post said, there is a per-user part of the registry. The system-wide configuration is in HKEY_LOCAL_MACHINE mostly (often abbreviated to HKLM), and the per-user configuration is in HKEY_CURRENT_USER.
IMO, the biggest problem for the registry is that it lumps all configuration data together into a non-portable database. I can't really backup my settings for programs that store things there. But if I want to back up my Gaim configuration, that's easy; just zip up ~/.gaim.
For a Windows system, it seems to me there should be relatively few 3rd-party apps installed system-wide.
Huh? Why? In what environment?
Is the same true of Unix in your opinion? If not, why the difference?
At our company, we users don't have Admin privileges for our own machines (and I don't really blame IT for it). As a result, if I'm trying to try out open-source math/science apps, I can probably install 1 out of 5 that don't install to windows/system or some other protected directory. Almost impossible to maintain an environment where machines are secure and usable.
Agreed, that's a problem. How much is MS's fault for encouraging that sort of model and how much should be blamed on 3rd parties, I don't know.
Hey, why couldn't we use USB2, wouldn't ~400mbits be enough?
No.
DVI supports 1.65 Gbits/sec. It's got resolution to spare for most people, but not all that much. Drop the resolution by half in each dimension (which would be required to make 400 mbps assuming bandwidth is proportional to area) and you're at about 950x600, which would be downright unbearable IMO.
And I know LCDs are all the rage, but my CRT wouldn't run at the refresh rate and resolution I run mine at over DVI. (1600x1200, 75hz.)