You didn't cite my last sentence, and it was one of the most important.
That is one of the very reasons for the creation of the EU, to eliminate economic barriers between the countries. Before the states became united they were effectively seperate countries. One of the reasons the US has been so successful was because the states are forbidden to meddle in interstate commerce.
The US lowers barriers in interstate commerce within the US. It raises them to outside state. NAFTA enlarges that protected area, but doesn't change the principle. The EU does the same, for their member states. I cannot see the difference. Both the EU and the US give a shit about free trade when it comes to protecting their own companies. That's simply a fact of life, and I won't quarrel with it.
You're half right that it's unworkable. I don't think there will be any consequences for deals with private customers, government authorities are not able to track deliveries of digital goods to them anyhow.
There might be a difference for deals with companies or incorporated non-profit organizations (gov, NGOs, etc.), though. Foreign companies are usually interested to make such deals - they are by far larger than sells to end customers. They cannot hide such deals since they appear in the books of the buyer. And for EU sellers, it levels the fields: They have to pay VAT already, and now their overseas competitors have to do the same.
If that's a protection of European companies, then I don't care. I run one, so that's fine with me.;-) If it comes to protection of companies, neither the US nor the EU may throw the first stone, they're sitting both in a glasshouse.
Within the EU when a company collects this tax the money goes to the government of that company. If an American buys something from an EU company the tax goes to an EU government.
No and no.
On our VAT forms, we need to declare sells to other EU countries. These declarations are used for money transfer calculations between EU member states.
American buyers don't pay VAT. If they did, they can go to custom when leaving the EU and reclaim their money. Many shops in touristic areas have appropriate forms available already (look for signs that say "tax free sales"). Of course, they also have to declare their goods on import into the US, where they have to pay custom and local taxes (which go the the state government, usually).
Btw, I'm a CEO of a European country, and have to know about these things. Don't tell BS like that if you have obviously no idea about tax systems and international treaties.
fvwm2 and Destop Environments
on
fvwm Turns Ten
·
· Score: 1
As a fvwm user since more than 15 years, and one who has tried both KDE and Gnome and has gone back to fvwm -- to those, who complain that fvwm is no substitute for a desktop environment:
Boys, I don't need no friggin' desktop environment. I got XEmacs. Go home to your kindergarten and keep on playing in your sandpit.
Now, where is my Usenet-proofed asbestos underwear...
No, DEK doesn't even write Email by himself. You know you got an email from DEK if the subject says "message from Knuth" and the sender is his secretary.
(For context: I'm a member of the TeX developer community since 20 years and had the opportunity to receive these letters and have DEK as a guest a few times.)
Yes, this special case is outrageous; just like many US laws or US court decisions. We have a different law system than you, and that doesn't necessarily mean it's better or worse -- first of all, it's different. (I'm German, as you will guess.:-)
But this case is different. Here, no 3rd party law firms are involved. Linux companies and organizations use legal means to stop unlawful behaviour from SCO Germany that harm them. That's OK in my book.
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Hmpff, what did you write? It's code copied from SCO to Linux, not the other way round? Well, who knows...:-):-)
From my viewpoint, it was the LaTeX people, especially Frank, who worked very hard. Looking at the discussions, it's hard to name Debian legal folks that "worked very hard". Care to name one?
At our last CTAN meeting, two weeks ago in Bremen, license issues with Debian was named as the biggest hassle that the volunteers drive away from their work they do for the community. Please, don't tell me about who's working hard -- I'm among them and these people are my best friends.
Since I am part of the "LaTeX people", I know about the discussions by heart. And I can tell you, it's more than wearing. Alone to get the Debian people to state explicitely and officially what they think is the "non-free" aspect of LPPL, needed several months.
And that after we had already months of discussions with RMS to draft LPPL in the first place. You can see it in the LaTeX bug database, ticket 1600.
The Debian license zealots at work, again. At least, they start to get consistent.
First, they started to throw out LaTeX, because the LPPL has a clause that says "you are allowed to take our code and change it, but then you must rename the package, since in LaTeX documents package names are part of the API and consistency is needed for document exchange." The invariance clause of the GFDL is very similar in nature, both accept that documents have other aspects than software packages.
Now, they must only understand that they have to throw out TeX as well. After all, the same restriction is on TeX the program (the code is factually frozen and may only be changed under an other name), Metafont, and the associated CM fonts. But suddenly, the license zealonts find lots of obscure reasons why these programs and fonts are supposed to be in the public domain.
Debian, be more consistent: Throw out LaTeX, throw out GNU project documents, and throw out TeX -- one of the first free software packages that was created as a collaborative effort! Go, forward!
He might present the current discussion about the Total Information Awareness project to the kids. Appropriate Information is available from EPIC and from other activists sites. (Warning: Turn off your popups before visiting the second link.)
First, they learn something about the threat to their own privacy. Second, one can present the dilemma to them: If you're asked to work on it, especially in the current job market, would you do it?
The teacher can also add that this is a concrete, current issue that is reality and not some fake problem. (I know this since I am exactly in this position.)
I like the idea of double jeopardy. This and other reasons are why I live in the United States. I believe that the idea of government lends itself to the abuse of power and I want every protection from them.
If one looks at the sentences and the jail praxis in the US, you really need double jeopardy. After all, your government seems to abuse its power and you need every bit of protection for them.
As you wrote,
However, there are people and societies that believe removing criminals from the streets is better for the greater good of the community and outweighs the dangers of innocent people being affected by their methods.
which is a very good description of the state of affairs in the US, isn't it? The amount of death sentences for innocents is only rivaled by countries with dictators. (Other civilized democratic countries don't have the death sentence any more, as you'll know.) US citizens with arabian heritage detained, without the right to speak to a lawyer. The possibility to ruin innocent people by dragging them to court and burying them under lawyer costs.
There are many reasons why one wants to do things different than the US justice system. After all, the image created by Hollywood court movies doesn't portray reality.
Most of my customers are multi-national companies, financial institutions. Definitively more than 10,000 employees, alone the IT staff is around this number if one includes external contractors. We're currently working on COBOL code from the 70s, and there 1000s of person years in that code.
Several of our customers try to replace their legacy systems with modern ones. All these projects are years overdue and have cost already eight-to-nine-digit numbers. None of them has gone live successfully. Most of them cannot handle the sheer number of transactions. (There is something to say for COBOL on mainframes...:-)
Legacy systems are here to stay, and here to provide work for those who know to handle the old hosties.
Then start by lobbying vendors that they ship sendmail aliases and Exchange with all RFC-proposed mboxes.
As long as Sun et.al. won't have them as default, domains won't make them available. Most sysadmins struggle to keep the MTA alive, and don't know shit about proper configuration.
This is Germany, not US. If you disagree, let it handle your lawyer. If the c&d letter (called "Abmahnung" here) is not founded on truth, they will have to pay all your legal expenses. If there is really a case, drag them to court. Again, you will have to pay no cent for your lawyers. (OTOH, you have to invest your time. That's often more worth.)
Oh yes, and this was obviously no real c&d letter. They must be on paper. Fax may work as well, there our law cases differ.
German law allows for oral contracts. It's just hard to enforce them in court. In particular, courts almost always decide in favour of the customer and against companies if the only existing contract is oral. Since the loser has to pay all legal fees, it makes much sense to demand some kind of statement that the request ain't no hoax.
Signature is usually only demanded in highly formalized situations, though. E.g., contracts between companies, which is a completely different matter. But then, you should see the legal brouhou US companies make for simple company contracts. There one learns about real bureaucracy.
Every developer who designs dialog boxes with only a single "OK" gadget on it should be sentenced to 7x24 hours Celine Dion, or something similar horrible.
For a user, these boxes often bring unpleasent messages. I don't want to confirm them with "OK" and no chance to stop. Either it is a warning, then one needs a "Cancel". Or it's a message, then a simple "Close" suffices.
Then use your Pentium 3 notebook for MP3 en- and decoding and for MPEG-4. You don't need IA64 for that.
But other people need computers for real work. E.g., for car crash simulations. And there IA64 clusters beat the hell out of IA32 clusters. (This is not an estimation, these are real jobs computing real models.) There IA64 clusters play almost in the Power4 league. Actually, IA32 clusters won't even be able to handle the larger models; but that's an other topic. Not to speak of the non-existing double-precision performance of IA32 clusters.
You're right that you won't need IA64 for the desktop tasks you cited. But to use this as a reason for "IA64 performance is very bad in the real world" is not professional. Because in the real world Itanium2-clusters finally show how good they are - but sorry, I won't count computer science classes as the real world here.
I doubt that you had anything to do with HPC. And where else do you really need performance gains of 64bit CPUs?
First, five nines are not needed. There are very seldomly needed, actually. And that is almost never when one needs CPU performance. (Therefore so many five nines installations use mainframe sysplexes.)
Second, three "two nines" servers won't give you five nines for almost all applications where you need five nines. ERP, bank systems -- those won't run fail-safe on your home made 3-node cluster.
Third, three servers when one needs performance? What are you smoking? That never suffices. E.g., we typically run clusters around 128-256 CPUs for crash simulations. If one buys very cheap cluster nodes, one will pay for that later in terms of operating costs. (This is not an estimation, this is from experience learned the hard way.)
Btw, for such demands, Itanium clusters cost similar amounts as Power4 clusters. XEON clusters don't cut it, they cannot compute complex models, double precision performance is absimal bad, and interconnection switches are under par.
Re:Immortal code - which do you know?
on
Immortal Code
·
· Score: 1
I didn't mean complete programs that are used again and again and again. I meant code that is incorporated in many programs and thus get's immortal.
The TCP/IP-stack, or zlib are very good examples; I should have thought of them. Once there was getdate.y - but I don't know if that's still used nowadays.
Immortal code - which do you know?
on
Immortal Code
·
· Score: 5, Interesting
There are code pieces I think might get immortal. Henry Spencer's regexp routines come to mind.
What other basic code pieces, used in hundreds or thousands of programs, do you know?
Sir, you don't understand. Sorry, if I was too concise for you.
I do even agree with you that one doesn't want to write LISP-style markup for text oneself. But I do also not want to write XML-markup literally. Then, it's a matter of internal markup (but still human readable and thus better debuggable). And for that there are better ways to represent marked up text than XML. More power to them.
XML was not created to be good for humans, but easy to implement for parsers. Therefore, it was not meant as a human-prefered input format. Therefore, other - better parsable - representations would have been even better technically. XML took over the world purely for political reasons, the cabal was able to present XML as the "better HTML" or the "successor to HTML" and made it.
Believe me, I know most of the cabal personally, being active in the SGML community since 1984; around 1990 I've maintained the world-wide largest repository of SGML software (the SGML archive at TU Darmstadt, Germany). Search in Google Newsgroups for Joachim Schrod, that's me - SGMLer, TeXie since 1981, and proud of it. In the last five years, I have been involved in strategic XML deployment at many multi-national companies. Since I do this stuff (markup languages) both for interest since more than 20 years, and for my living, I know by heart how bad it is in real-life.
If you really think XML is the best thing since sliced bread in textual markup representation, let's talk again about that topic when you've got 20 years of experience in that area.
'Til then, hope you can use your enthusiasm in your job. Best wishes,
First I was entitled to answer your sarcasm - because yes, data representation in LISP is able to do all that. I started to explain that LISP has been shown to be one of the best SGML/XML manipulation tools, due to the semantic similarity of s-expressions and XML documents. Almost a decade ago I released the first open-source tool (STIL - SGML Transformations In Lisp) for this task - at a time when XML didn't even exist.
But then I saw you calling XML concise. And realized the troll you are. Well done, Sir. You got me.
I'm in Germany, and had Oracle engineers flown in from the US, when the local SEs couldn't fix it in 8 hours.
As always, you get what you pay for. If it costs you 6- or 7-digit numbers per hour when your RDBMS is down, then spent a lot of money and use Oracle. That's where Oracle is exceptional.
If it's not as important (i.e., when you loose only a few thousand dollars per hour), ask in some Oracle newsgroup, or use DB2. If your database isn't mission-critical at all, use PostgreSQL. If you actually don't need a database but want SQL queries, use MySQL.
Slashdot Mirror
There might be a difference for deals with companies or incorporated non-profit organizations (gov, NGOs, etc.), though. Foreign companies are usually interested to make such deals - they are by far larger than sells to end customers. They cannot hide such deals since they appear in the books of the buyer. And for EU sellers, it levels the fields: They have to pay VAT already, and now their overseas competitors have to do the same.
If that's a protection of European companies, then I don't care. I run one, so that's fine with me. ;-) If it comes to protection of companies, neither the US nor the EU may throw the first stone, they're sitting both in a glasshouse.
s/country/company/
On our VAT forms, we need to declare sells to other EU countries. These declarations are used for money transfer calculations between EU member states.
American buyers don't pay VAT. If they did, they can go to custom when leaving the EU and reclaim their money. Many shops in touristic areas have appropriate forms available already (look for signs that say "tax free sales"). Of course, they also have to declare their goods on import into the US, where they have to pay custom and local taxes (which go the the state government, usually).
Btw, I'm a CEO of a European country, and have to know about these things. Don't tell BS like that if you have obviously no idea about tax systems and international treaties.
Now, where is my Usenet-proofed asbestos underwear...
(For context: I'm a member of the TeX developer community since 20 years and had the opportunity to receive these letters and have DEK as a guest a few times.)
But this case is different. Here, no 3rd party law firms are involved. Linux companies and organizations use legal means to stop unlawful behaviour from SCO Germany that harm them. That's OK in my book.
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Hmpff, what did you write? It's code copied from SCO to Linux, not the other way round? Well, who knows... :-) :-)
At our last CTAN meeting, two weeks ago in Bremen, license issues with Debian was named as the biggest hassle that the volunteers drive away from their work they do for the community. Please, don't tell me about who's working hard -- I'm among them and these people are my best friends.
And that after we had already months of discussions with RMS to draft LPPL in the first place. You can see it in the LaTeX bug database, ticket 1600.
First, they started to throw out LaTeX, because the LPPL has a clause that says "you are allowed to take our code and change it, but then you must rename the package, since in LaTeX documents package names are part of the API and consistency is needed for document exchange." The invariance clause of the GFDL is very similar in nature, both accept that documents have other aspects than software packages.
Now, they must only understand that they have to throw out TeX as well. After all, the same restriction is on TeX the program (the code is factually frozen and may only be changed under an other name), Metafont, and the associated CM fonts. But suddenly, the license zealonts find lots of obscure reasons why these programs and fonts are supposed to be in the public domain.
Debian, be more consistent: Throw out LaTeX, throw out GNU project documents, and throw out TeX -- one of the first free software packages that was created as a collaborative effort! Go, forward!
First, they learn something about the threat to their own privacy. Second, one can present the dilemma to them: If you're asked to work on it, especially in the current job market, would you do it?
The teacher can also add that this is a concrete, current issue that is reality and not some fake problem. (I know this since I am exactly in this position.)
If one looks at the sentences and the jail praxis in the US, you really need double jeopardy. After all, your government seems to abuse its power and you need every bit of protection for them.
As you wrote,
which is a very good description of the state of affairs in the US, isn't it? The amount of death sentences for innocents is only rivaled by countries with dictators. (Other civilized democratic countries don't have the death sentence any more, as you'll know.) US citizens with arabian heritage detained, without the right to speak to a lawyer. The possibility to ruin innocent people by dragging them to court and burying them under lawyer costs.
There are many reasons why one wants to do things different than the US justice system. After all, the image created by Hollywood court movies doesn't portray reality.
Most of my customers are multi-national companies, financial institutions. Definitively more than 10,000 employees, alone the IT staff is around this number if one includes external contractors. We're currently working on COBOL code from the 70s, and there 1000s of person years in that code.
Several of our customers try to replace their legacy systems with modern ones. All these projects are years overdue and have cost already eight-to-nine-digit numbers. None of them has gone live successfully. Most of them cannot handle the sheer number of transactions. (There is something to say for COBOL on mainframes... :-)
Legacy systems are here to stay, and here to provide work for those who know to handle the old hosties.
As long as Sun et.al. won't have them as default, domains won't make them available. Most sysadmins struggle to keep the MTA alive, and don't know shit about proper configuration.
This is Germany, not US. If you disagree, let it handle your lawyer. If the c&d letter (called "Abmahnung" here) is not founded on truth, they will have to pay all your legal expenses. If there is really a case, drag them to court. Again, you will have to pay no cent for your lawyers. (OTOH, you have to invest your time. That's often more worth.)
Oh yes, and this was obviously no real c&d letter. They must be on paper. Fax may work as well, there our law cases differ.
Signature is usually only demanded in highly formalized situations, though. E.g., contracts between companies, which is a completely different matter. But then, you should see the legal brouhou US companies make for simple company contracts. There one learns about real bureaucracy.
For a user, these boxes often bring unpleasent messages. I don't want to confirm them with "OK" and no chance to stop. Either it is a warning, then one needs a "Cancel". Or it's a message, then a simple "Close" suffices.
But other people need computers for real work. E.g., for car crash simulations. And there IA64 clusters beat the hell out of IA32 clusters. (This is not an estimation, these are real jobs computing real models.) There IA64 clusters play almost in the Power4 league. Actually, IA32 clusters won't even be able to handle the larger models; but that's an other topic. Not to speak of the non-existing double-precision performance of IA32 clusters.
You're right that you won't need IA64 for the desktop tasks you cited. But to use this as a reason for "IA64 performance is very bad in the real world" is not professional. Because in the real world Itanium2-clusters finally show how good they are - but sorry, I won't count computer science classes as the real world here.
First, five nines are not needed. There are very seldomly needed, actually. And that is almost never when one needs CPU performance. (Therefore so many five nines installations use mainframe sysplexes.)
Second, three "two nines" servers won't give you five nines for almost all applications where you need five nines. ERP, bank systems -- those won't run fail-safe on your home made 3-node cluster.
Third, three servers when one needs performance? What are you smoking? That never suffices. E.g., we typically run clusters around 128-256 CPUs for crash simulations. If one buys very cheap cluster nodes, one will pay for that later in terms of operating costs. (This is not an estimation, this is from experience learned the hard way.)
Btw, for such demands, Itanium clusters cost similar amounts as Power4 clusters. XEON clusters don't cut it, they cannot compute complex models, double precision performance is absimal bad, and interconnection switches are under par.
The TCP/IP-stack, or zlib are very good examples; I should have thought of them. Once there was getdate.y - but I don't know if that's still used nowadays.
What other basic code pieces, used in hundreds or thousands of programs, do you know?
I do even agree with you that one doesn't want to write LISP-style markup for text oneself. But I do also not want to write XML-markup literally. Then, it's a matter of internal markup (but still human readable and thus better debuggable). And for that there are better ways to represent marked up text than XML. More power to them.
XML was not created to be good for humans, but easy to implement for parsers. Therefore, it was not meant as a human-prefered input format. Therefore, other - better parsable - representations would have been even better technically. XML took over the world purely for political reasons, the cabal was able to present XML as the "better HTML" or the "successor to HTML" and made it.
Believe me, I know most of the cabal personally, being active in the SGML community since 1984; around 1990 I've maintained the world-wide largest repository of SGML software (the SGML archive at TU Darmstadt, Germany). Search in Google Newsgroups for Joachim Schrod, that's me - SGMLer, TeXie since 1981, and proud of it. In the last five years, I have been involved in strategic XML deployment at many multi-national companies. Since I do this stuff (markup languages) both for interest since more than 20 years, and for my living, I know by heart how bad it is in real-life.
If you really think XML is the best thing since sliced bread in textual markup representation, let's talk again about that topic when you've got 20 years of experience in that area. 'Til then, hope you can use your enthusiasm in your job. Best wishes,
But then I saw you calling XML concise. And realized the troll you are. Well done, Sir. You got me.
I'm in Germany, and had Oracle engineers flown in from the US, when the local SEs couldn't fix it in 8 hours.
As always, you get what you pay for. If it costs you 6- or 7-digit numbers per hour when your RDBMS is down, then spent a lot of money and use Oracle. That's where Oracle is exceptional.
If it's not as important (i.e., when you loose only a few thousand dollars per hour), ask in some Oracle newsgroup, or use DB2. If your database isn't mission-critical at all, use PostgreSQL. If you actually don't need a database but want SQL queries, use MySQL.