Furthermore, the university should be protecting these students by threatening to end the contract
Maybe things have changed since I went to school, but I would expect the University's view to be that students should attend class and take their own notes. Why would you think they would want to encourage students to skip class and buy notes from a third party?
I just had a chance to take a look at the OOXML spec for this, and it does no such thing. It specifies that you can use MD2, MD4, MD5, RIPEMD-128, RIPEMD-160, SHA-1, SHA-256, SHA-384, SHA-512, or Whirlpool.
If you want tight, fast code and your project doesn't need OOP, use C
C++ doesn't require you to do OOP. If you want tight, fast code, still use C++, but just use it like C with a couple of nice enhancements (constructors and destructors on structs, std::string, exceptions, and STL), and you'll get code that is just as good as your C code, but clearer, and more likely correct.
I'm still astounded that a Computer Science curriculum includes any in-depth teaching of a programming language. Does the physics curriculum include courses on car repair? Does the biology curriculum include courses for the female students on how to land a good husband? Does the Literature curriculum include an in-depth study of calligraphy?
OK, I'm exaggerating a bit for effect, but seriously, most of computer science doesn't even require a computer, let alone an in-depth knowledge of any particular computer programming language. Some universities seem to have CS curriculum that would be more at home at DeVry's than at a university.
When you read Knuth, are you sitting in front of a computer? Gack, I bet that's what the kids nowadays do. The right way to read and study Knuth is sitting in a very fine leather chair, in front of a fireplace with fire, with a large dog sleeping at your feet, a drink in one hand, and classical music playing softly--and it should take along walk across a moor to reach the nearest computer.
What makes you think libel laws abridge freedom of speech? And even if they did, what makes you think a restriction on Congressional legislative power would affect State libel law?
It seems most of the commentators here have not looked at the filings in this case, or read the numerous articles about it on various legal blogs. It is not quite as is being reported here. The two lawyers who are suing him were representing the other side in a case against Cisco. On his blog, he accused them of altering the filing date on some court documents. That's a very serious accusation--if they did it, it would be both a breach of legal ethics and a felony.
In general, it is very very bad for a lawyer to publicly accuse another lawyer of committing a felony unless the accuser has some pretty damned convincing evidence.
And it is a zillion times worse when the accusing lawyer is counsel for the other side in a case the accused lawyer is working on.
Troll Tracker screwed up big time here (heck, commenting AT ALL on a case involving Cisco is a bit shocking), and is probably going to have to cough up a public apology and a wad of cash. It's not about free speech, as some other posters have suggested--it is a plain, old-fashioned "if you accuse someone of a serious crime, you'd better be able to back it up" case.
There's an informative post about the case here. That's a handy blog about activity in the Eastern District of Texas. There was also good coverage in Business Week.
See blackboard software and their patent trolling vs. others. See all the patent trolls vs. Microsoft (like that DirectX patent troll posted last year on slashdot). See RIM vs. patent trolls. Instead of innovation, patents breed FUD in the software world
Neither the Blackboard case nor the RIM case involved patent trolls. You probably should learn what terms mean before using them.
As far as I know IBM has never threatened Linux developers with ambiguous claims of patent infringement
On the other hand, Microsoft has never ACTUALLY sued anyone over patents. IBM has.
And not just hardware patents. They've sued over software patents.
Assuming that IBM is just automatically going to be nice, and so letting them get away with a patent license you think is bogus, does not strike me as wise.
In a standard, you have to be careful in specifying "what to do" as opposed to specifying "how to do it". In the case of password hashing, ODF does not specify which method you should use. It leaves that up to each implementation because each country has different standards. ie. Japan: (MD5, RIPEMD-160, SHA1, SHA256, SHA384, and SHA512),US:(SHA1, SHA224, SHA256,SHA384, and SHA512)
So, given a document from Japan and one from the US, both of which contain passwords, how do I know which hash was ACTUALLY used in each document? The standard should tell me how to determine that.
I take it you haven't looked at the questions raised during ODF's speedy trip through ISO, and how many were addressed by putting them off for a future version?
OK, this is probably just nuts, and a result of posting in the morning before having any caffeine to boot my brain, but a weird theory just came to mind. I'll present it for the general amusement and ridicule of the Slashdot crowd.
A lot of complaints about OOXML are over things that it did the same as, or better than, ODF. For example, one of the complaints was that it did not fully specify how to do password hashing. But ODF is even less forthcoming in this regard. It just says that you should hash any password you store in the file. It doesn't say what hash you should use, or tell you how to record in the file what hash is used so others can figure out how to process the file. But because of these complaints about OOXML, it now specifies password hashing in enough detail that you can implement it from the spec and referenced documents.
A second example: calendars. OOXML was dinged for not giving a precise reference for each supported it calendar. It just had a list of calendars, and for each a short description. But ODF was even terser. It just, in one sentence, gives a list of the names of the supported calendars, with no reference at all. But because OOXML was dinged for inadequate calendar specs, OOXML now for each gives a precise reference. For example, all you'll find in ODF about the hijri calendar is that one word in that list (and I think there is one example document fragment where it has that word in it). In OOXML now, it says this about the hijri calendar:
Specifies that the Hijri lunar calendar, as described by the Kingdom of Saudi Arabia, Ministry of Islamic Affairs, Endowments, Da'wah and Guidance (Kingdom of Saudi Arabia, Ministry of Islamic Affairs, Endowments, Da'wah and Guidance n.d.), shall be used. Net result: another place where you have to guess in ODF, but are told in the spec what to do in OOXML.
There are a lot of things like this in there--things where OOXML is now specified much more precisely than ODF. Places where you can figure out from the spec itself what to do, whereas if you are implementing ODF, you have to fall back to looking at OpenOffice source code to see what they did and match it.
And this has had a very predictable effect. A lot of third party programs and sites are starting to support OOXML, whereas ODF doesn't seem to be growing much beyond OpenOffice and the other free office suites. Reading blog entries from people who have tried to implement OOXML and ODF, I see that the OOXML ones are having an easier time. The ODF ones are more likely to run into something that is underspecified or ambiguous (at least if they are sticking to the standard, rather than working from 1.2, which is not a standard yet)
The net result of OOXML being required to clear a much higher bar than any previous document standard is that OOXML has become the most useful document specification. And how did this happen?
The surprising answer: IBM. There has been a lot of technical criticism of OOXML, but also a lot of FUD. And when you look at the FUD, and trace it back, a heck of a lot of it ends up coming from IBM.
Why? The net result of this is to make OOXML better, and to ultimately harm ODF. But IBM wouldn't want to harm ODF...
Or would they? Sun effectively controls ODF. Sun and IBM are competitors, to a greater extent than IBM and Microsoft are competitors. Could IBM have decided that they are not comfortable with Sun's tight reins over ODF? So they started a FUD effort against OOXML, knowing that it would result in (1) OOXML becoming a better spec than ODF, and (2) prompting Microsoft to turn more control over to ECMA than Sun has turned over to OASIS?
To put it succinctly, is IBM trying, in a very roundabout way, to kill ODF?
IBM has only promised not to sue over their ODF patents. Sun has only promised not to sue over their ODF patents. If you actually compare the language of all the patent pledged involved with OOXML, and those with ODF, you'll find that they are pretty much the same. Microsoft's and IBM's, in particular, are remarkably close in what they allow and what they disallow. Here are the relevant licenses on one page, side-by-side, for convenient comparison.
I did not read the OOXML proposal, I understand that there has been numerous remarks on the technical ground only. Those comments have been partly addressed by Microsoft and therefore another round of discussion on this proposal is in order
Why should the rules be different for OOXML? Other standards, including ODF, were approved with numerous technical comments only partly addressed. That's the normal procedure. The remaining issues get addressed in future versions, after people have experience using the standard.
The major thing OO is missing for me on the word processor front is good outliner support. There was a note from the developers posted on their forums a while back where they acknowledged that adding this is important, and that the navigator stuff is not a substitute. So, the good news is, OO will get good outliner support. The bad news is that it is going to be a lot of work, so it might not be soon.:-(
It's important to keep in mind the reasons we oppose the OpenXML format
Two of your first three reasons apply equally to ODF. Do you oppose that too?
First, ODF is very incomplete. To actually implement it, and have a chance of your documents interoperating with others, you have to base your implementation on reverse engineering OpenOffice. You cannot base it just on the spec and have any hope of interoperability.
Second, there are no binary blobs in OOXML that aren't also in ODF. One wonders if you ever actually have downloaded and looked at the ODF spec?
Anti-OOXML arguments would be a lot more persuasive if they didn't rely on declaring things as bad that work exactly the same way in ODF!
First off, I disagree that "digital products" have "outgrown" anything. On the contrary, I fail to see any new issues that digital text brings to the table
With digital products that are not tied to a physical medium, one major new issue they bring to the table is that a used copy is as good as a new copy.
Think about that a minute. If I buy a copy of a normal book, and read it, and sell it to a used bookstore, that copy is going to be a little worn. And if you buy that used copy, read it, and sell it, it is a little more worn, and so on.
Since people like new books better than used books, this puts a limit on the secondary market. Not so with a digital book.
Let's do some numbers, to get a feel for this. Say a new digital book sells for $10. Someone opens a used digital book site, which buys the used books for $8, and sells them for $9. I have no idea if $8/$9 would be what the market will settle on, but the exact numbers don't matter.
As a reader, I have no reason to buy my books from the publisher for $10, if I can find them for $9 at the used site.
Furthermore, if I think I'm going to read a book 9 or less times, I have no reason to keep it. Buy it for $9 the first time, and sell it back for $8 when I'm done. If I want to read it again, I can buy it again, and sell it back when done again. It effectively costs me $1 to rent the book for a single read, versus $9 to buy it.
When a new book is released, you'll have the first few readers buy it from the publisher. As the faster readers finish it, and sell to the used site, some new readers will pick it up there. So what you have is a rapidly growing used market, which, unlike with physical books, is a perfect substitute for the new books from the publisher.
Essentially, the original publisher's market size is the maximum number of people that want to be reading the book simultaneously.
And note that because the book is purely a digital file, the storage costs for the used store are insignificant, as are the transaction costs for buying and selling. In fact, the used bookstore site could be set up to pretty much run itself. It could make a profit with the spread between their buy price and sell price being much lower than the $1 I hypothesized in my example. And there will be competition between different used sites, which will drive the spread down. I think the spread will be pennies, not dollars.
Consider what that means. If you can sell back a used digital book for $0.10 less than what you paid for it, it now might take a hundred reads to make it worth keeping the book, instead of buying it temporarily whenever you want to read it. This ensures that most people will immediately sell back any book they read, which makes sure the used market always has a good supply of every book.
The good news for the reader is that this greatly lowers the cost of reading. The bad news for the publishers is that this means that when they publish some new book, a lot more of us readers will be willing to wait--we have a bunch of other material from the used stores to read. We can wait for the new book to show up used. The number of people who will want to simultaneously read a new book will be much lower than it is now.
So, what digital text brings to the table is a complete and fundamental change in how the book market works
The real problem is Microsoft's waffling on making the standard open. If they had unequivocally placed the standard and all necessary patents in the public domain and committed to keeping it stable, more people might vote for it.
That hasn't been a problem for any prior standard. Off the top of your head, can you name any other comparable standards (or any ISO standards at all?) for which that has been done?
Slashdot Mirror
Have you considered actually reading the story?
Maybe things have changed since I went to school, but I would expect the University's view to be that students should attend class and take their own notes. Why would you think they would want to encourage students to skip class and buy notes from a third party?
I just had a chance to take a look at the OOXML spec for this, and it does no such thing. It specifies that you can use MD2, MD4, MD5, RIPEMD-128, RIPEMD-160, SHA-1, SHA-256, SHA-384, SHA-512, or Whirlpool.
If I had meant the people, not the habitat type, I would have said moop, not moor.
C++ doesn't require you to do OOP. If you want tight, fast code, still use C++, but just use it like C with a couple of nice enhancements (constructors and destructors on structs, std::string, exceptions, and STL), and you'll get code that is just as good as your C code, but clearer, and more likely correct.
I'm still astounded that a Computer Science curriculum includes any in-depth teaching of a programming language. Does the physics curriculum include courses on car repair? Does the biology curriculum include courses for the female students on how to land a good husband? Does the Literature curriculum include an in-depth study of calligraphy?
OK, I'm exaggerating a bit for effect, but seriously, most of computer science doesn't even require a computer, let alone an in-depth knowledge of any particular computer programming language. Some universities seem to have CS curriculum that would be more at home at DeVry's than at a university.
When you read Knuth, are you sitting in front of a computer? Gack, I bet that's what the kids nowadays do. The right way to read and study Knuth is sitting in a very fine leather chair, in front of a fireplace with fire, with a large dog sleeping at your feet, a drink in one hand, and classical music playing softly--and it should take along walk across a moor to reach the nearest computer.
What makes you think libel laws abridge freedom of speech? And even if they did, what makes you think a restriction on Congressional legislative power would affect State libel law?
The 1st Amendment was not intended to, and did not, eliminate the tort of libel.
It seems most of the commentators here have not looked at the filings in this case, or read the numerous articles about it on various legal blogs. It is not quite as is being reported here. The two lawyers who are suing him were representing the other side in a case against Cisco. On his blog, he accused them of altering the filing date on some court documents. That's a very serious accusation--if they did it, it would be both a breach of legal ethics and a felony.
In general, it is very very bad for a lawyer to publicly accuse another lawyer of committing a felony unless the accuser has some pretty damned convincing evidence.
And it is a zillion times worse when the accusing lawyer is counsel for the other side in a case the accused lawyer is working on.
Troll Tracker screwed up big time here (heck, commenting AT ALL on a case involving Cisco is a bit shocking), and is probably going to have to cough up a public apology and a wad of cash. It's not about free speech, as some other posters have suggested--it is a plain, old-fashioned "if you accuse someone of a serious crime, you'd better be able to back it up" case.
There's an informative post about the case here. That's a handy blog about activity in the Eastern District of Texas. There was also good coverage in Business Week.
Neither the Blackboard case nor the RIM case involved patent trolls. You probably should learn what terms mean before using them.
On the other hand, Microsoft has never ACTUALLY sued anyone over patents. IBM has.
And not just hardware patents. They've sued over software patents.
Assuming that IBM is just automatically going to be nice, and so letting them get away with a patent license you think is bogus, does not strike me as wise.
So, given a document from Japan and one from the US, both of which contain passwords, how do I know which hash was ACTUALLY used in each document? The standard should tell me how to determine that.
I take it you haven't looked at the questions raised during ODF's speedy trip through ISO, and how many were addressed by putting them off for a future version?
OK, this is probably just nuts, and a result of posting in the morning before having any caffeine to boot my brain, but a weird theory just came to mind. I'll present it for the general amusement and ridicule of the Slashdot crowd.
A lot of complaints about OOXML are over things that it did the same as, or better than, ODF. For example, one of the complaints was that it did not fully specify how to do password hashing. But ODF is even less forthcoming in this regard. It just says that you should hash any password you store in the file. It doesn't say what hash you should use, or tell you how to record in the file what hash is used so others can figure out how to process the file. But because of these complaints about OOXML, it now specifies password hashing in enough detail that you can implement it from the spec and referenced documents.
A second example: calendars. OOXML was dinged for not giving a precise reference for each supported it calendar. It just had a list of calendars, and for each a short description. But ODF was even terser. It just, in one sentence, gives a list of the names of the supported calendars, with no reference at all. But because OOXML was dinged for inadequate calendar specs, OOXML now for each gives a precise reference. For example, all you'll find in ODF about the hijri calendar is that one word in that list (and I think there is one example document fragment where it has that word in it). In OOXML now, it says this about the hijri calendar:
Specifies that the Hijri lunar calendar, as described by the Kingdom of Saudi Arabia, Ministry of Islamic Affairs, Endowments, Da'wah and Guidance (Kingdom of Saudi Arabia, Ministry of Islamic Affairs, Endowments, Da'wah and Guidance n.d.), shall be used. Net result: another place where you have to guess in ODF, but are told in the spec what to do in OOXML.There are a lot of things like this in there--things where OOXML is now specified much more precisely than ODF. Places where you can figure out from the spec itself what to do, whereas if you are implementing ODF, you have to fall back to looking at OpenOffice source code to see what they did and match it.
And this has had a very predictable effect. A lot of third party programs and sites are starting to support OOXML, whereas ODF doesn't seem to be growing much beyond OpenOffice and the other free office suites. Reading blog entries from people who have tried to implement OOXML and ODF, I see that the OOXML ones are having an easier time. The ODF ones are more likely to run into something that is underspecified or ambiguous (at least if they are sticking to the standard, rather than working from 1.2, which is not a standard yet)
The net result of OOXML being required to clear a much higher bar than any previous document standard is that OOXML has become the most useful document specification. And how did this happen?
The surprising answer: IBM. There has been a lot of technical criticism of OOXML, but also a lot of FUD. And when you look at the FUD, and trace it back, a heck of a lot of it ends up coming from IBM.
Why? The net result of this is to make OOXML better, and to ultimately harm ODF. But IBM wouldn't want to harm ODF...
Or would they? Sun effectively controls ODF. Sun and IBM are competitors, to a greater extent than IBM and Microsoft are competitors. Could IBM have decided that they are not comfortable with Sun's tight reins over ODF? So they started a FUD effort against OOXML, knowing that it would result in (1) OOXML becoming a better spec than ODF, and (2) prompting Microsoft to turn more control over to ECMA than Sun has turned over to OASIS?
To put it succinctly, is IBM trying, in a very roundabout way, to kill ODF?
IBM has only promised not to sue over their ODF patents. Sun has only promised not to sue over their ODF patents. If you actually compare the language of all the patent pledged involved with OOXML, and those with ODF, you'll find that they are pretty much the same. Microsoft's and IBM's, in particular, are remarkably close in what they allow and what they disallow. Here are the relevant licenses on one page, side-by-side, for convenient comparison.
Why should the rules be different for OOXML? Other standards, including ODF, were approved with numerous technical comments only partly addressed. That's the normal procedure. The remaining issues get addressed in future versions, after people have experience using the standard.
The major thing OO is missing for me on the word processor front is good outliner support. There was a note from the developers posted on their forums a while back where they acknowledged that adding this is important, and that the navigator stuff is not a substitute. So, the good news is, OO will get good outliner support. The bad news is that it is going to be a lot of work, so it might not be soon. :-(
Won't happen. There are positions that require 20 moves.
Two of your first three reasons apply equally to ODF. Do you oppose that too?
First, ODF is very incomplete. To actually implement it, and have a chance of your documents interoperating with others, you have to base your implementation on reverse engineering OpenOffice. You cannot base it just on the spec and have any hope of interoperability.
Second, there are no binary blobs in OOXML that aren't also in ODF. One wonders if you ever actually have downloaded and looked at the ODF spec?
Anti-OOXML arguments would be a lot more persuasive if they didn't rely on declaring things as bad that work exactly the same way in ODF!
Wow, a site run by a paid anti-Microsoft activist doesn't like it when the ODF editor says something good about OOXML? Shocking!
With digital products that are not tied to a physical medium, one major new issue they bring to the table is that a used copy is as good as a new copy.
Think about that a minute. If I buy a copy of a normal book, and read it, and sell it to a used bookstore, that copy is going to be a little worn. And if you buy that used copy, read it, and sell it, it is a little more worn, and so on.
Since people like new books better than used books, this puts a limit on the secondary market. Not so with a digital book.
Let's do some numbers, to get a feel for this. Say a new digital book sells for $10. Someone opens a used digital book site, which buys the used books for $8, and sells them for $9. I have no idea if $8/$9 would be what the market will settle on, but the exact numbers don't matter.
As a reader, I have no reason to buy my books from the publisher for $10, if I can find them for $9 at the used site.
Furthermore, if I think I'm going to read a book 9 or less times, I have no reason to keep it. Buy it for $9 the first time, and sell it back for $8 when I'm done. If I want to read it again, I can buy it again, and sell it back when done again. It effectively costs me $1 to rent the book for a single read, versus $9 to buy it.
When a new book is released, you'll have the first few readers buy it from the publisher. As the faster readers finish it, and sell to the used site, some new readers will pick it up there. So what you have is a rapidly growing used market, which, unlike with physical books, is a perfect substitute for the new books from the publisher.
Essentially, the original publisher's market size is the maximum number of people that want to be reading the book simultaneously.
And note that because the book is purely a digital file, the storage costs for the used store are insignificant, as are the transaction costs for buying and selling. In fact, the used bookstore site could be set up to pretty much run itself. It could make a profit with the spread between their buy price and sell price being much lower than the $1 I hypothesized in my example. And there will be competition between different used sites, which will drive the spread down. I think the spread will be pennies, not dollars.
Consider what that means. If you can sell back a used digital book for $0.10 less than what you paid for it, it now might take a hundred reads to make it worth keeping the book, instead of buying it temporarily whenever you want to read it. This ensures that most people will immediately sell back any book they read, which makes sure the used market always has a good supply of every book.
The good news for the reader is that this greatly lowers the cost of reading. The bad news for the publishers is that this means that when they publish some new book, a lot more of us readers will be willing to wait--we have a bunch of other material from the used stores to read. We can wait for the new book to show up used. The number of people who will want to simultaneously read a new book will be much lower than it is now.
So, what digital text brings to the table is a complete and fundamental change in how the book market works
Do you actually know anything at all about OOXML?
I'm shocked to find out that Pages, Numbers, and TextEdit on my Mac, not to mention my iPhone, are Microsoft products.
And when did Microsoft acquire Thinkfree Office? Dataviz? Intergen? NeoOffice? Zoho Writer?
That hasn't been a problem for any prior standard. Off the top of your head, can you name any other comparable standards (or any ISO standards at all?) for which that has been done?
Can you give your definition of "open"?