So the disagreement is over whether clause 6 of the GPL:
6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License.
applies to the patent grant. Does "rights granted herein" refer to the rights in clause 6 ("...copy, distribute, or modify...") or to all rights mentioned in the license?
This statement actually has multiple difficulties. Firstly, only natural persons are allowed to claim invention of patents, and vision is not a natural person. Secondly, only persons and corporations are allowed to own things; vision is not. Thirdly, regardless of how "thrown out" is interpreted, it's not patentable. There are two main interprations:
* The words "thrown out" are themselves patented. One cannot use them anywhere without the permission of machine vision -- This doesn't work because patents only apply to processes. Perhaps the patent applicant would prefer to register a copyright or trademark?
* The words "thrown out" identify the process of throwing a thing out. One may not throw a thing out without the permission of machine vision -- This has two defects. On the one hand, one should use the gerund form instead of the participle form when identifying any thing. On the other hand, the "throwing out" process is usually realized by a person (e.g., "The judge threw out the case."), but, again, vision is not a person.
So clearly "they" -- regardless of that pronoun's antecedent -- cannot "do that" because the statement "Machine vision patents thrown out" can never be true.
It appears those that do not want MDI want to be able to arrange their windows around the desktop leaving little areas to peer through to other apps in the background. This group likes to use the mouse to focus windows and may enjoy being able to swap to another workspace to preserve this environment.
MDI and virtual workspaces are basically the same idea -- group together related windows. The difference: in MDI, the groupings follow vendor/application lines. In virtual workspaces, the user groups windows as he or she wishes. If you want to have immaculately arranged windows from different applications which enable you to compare just the right data (and have focus follow the mouse pointer), then fine. If you want to group a hundred windows from one application (and use keyboard navigation), then fine. If you want a mess, fine.
In this case, having to select objects by visual means is almost impossible. Instead, we prefer a single Alt+Tab or mouse click on the window list to switch away, and another to return to the graphic application environment. The MDI has it's own window list which aids in having to decide between different names and other applications in the same list. It also has its own separate Ctrl+Tab key combination to page between them.
You get the same functionality with virtual workspaces. In my GNOME/Metacity setup, I use Ctrl-Alt-{Left,Right} to move among workspaces, and Alt-Tab to move among windows within a workspace.
In some cultures it is considered improper to offer your guest an executive when sharing IT problems.
For example, if your guest is a vegetarian, he will probably just push the executive off to the side and focus on the parts of the problem which he can digest.
I love osnews little version, with all the directories in the path displayed at the top, the idea being you could click on them to go back to that directory.
I like it, too: going up a couple of directories only requires one short click. Using a combo box either requires two clicks or a dragging motion.
The problem is that it will suck for really deep directory trees. (Getting to the root will require clicking the arrow button several times.) I would like a compromise where the first button pops up a list (like the combo box) and then several buttons show the deepest directories.
For example, if the directory were
/home/me/music/artists/Crazy Fool/Greatest Hits
I would layout the path navigator like this:
[_ICON_] [artists] [Crazy Fool] [Greatest Hits]
Clicking on the icon would pop up a list (like the combo box):
This makes it easy (one click) to go up when you're browsing a small subtree in the file system, but more difficult (two clicks or dragging) to browse all over the file system. The latter case shouldn't be common, though, since "Favorites" would abbreviate navigation among distant subtrees.
File selectors should be longer horizontally since file names can be long. Having something that is taller than it is long is just dumb, there just isn't a polite way to say it.
The file list widget -- but not necessarily the file selector dialog -- should be long horizontally, and all the mockups are better than the current layout (a narrow widget for directories to the left of a narrow widget for files). Eugenia's file list widget is actually wider and contains more information than tigert's. In fact, if you give Eugenia's dialog and tigert's dialog the same dimensions, Eugenia's will handle long file names better (though it will show fewer file names).
There are the whiners, and there are the doers. What he doesn't realize is that doers just do it, and ignore these whiners.
I think that's a bit unfair. Our society embraces a lot of specialization, and there are necessarily more consumers than producers. Now, individual consumers in most markets (should) have an interest in how the market works -- but (individually) lack the power to change the market. That creates a feeling of powerlessness -- which leads to legitimate whining. A few individuals who move from consumption to production may complain less, but you and the journalist will still be overwhelmed by the remaining consumers. In short, there must be and should be whining, and meta-whining won't stop it.
Yes it would. There already is such a court, the International Criminal Court.. The problem is, the USA opposes it.
Well, there are many reasons the ICC is a non-issue:
* The current administration dislikes any and all forms of international cooperation, especially ones which can try US citizens. * Membership -- The US, Iraq, Kuwait, and Iran are not members of the ICC. * Complementarity -- ICC complements national courts. It can only try a case if the national court systems cannot or do not. * Retroactivity -- ICC jurisdiction is not retroactive. It can only try crimes after 2002, which would exclude most of the interesting Iraq-related crimes (except the US invasion...)
I work at the School of Communications for a major state University.
I work in the CS department at a smallish, private university with a liberal arts focus, and I'm responsible for the Oracle systems used in DB/DBA courses. Your question seems strange to me. Some questions/comments:
1) Why on Earth is the School of Communications teaching database concepts?
2) Why does University IT care if you use SQL or Access?
3) Perhaps University IT cares because it has to run an SQL database server for you. It wants to unload that liability. The problem is that doing so actually has academic consequences: it forces students, teachers, and TA's to spend time on the various installation and configuration details of MS Access. _I_ can't say if that's good or bad, but it must be part of your decision. You should only replace server-side SQL coding with MS Access databases if you believe that Access offers pedagogical benefits.
4) Perhaps University IT cares because they think SQL sucks; because the IT director likes GUI's; because the IT director owns Microsoft stock; etc. In any of these circumstances, you can politely nod when they share their opinions -- and then forget the conversation ever happened. If University IT gets bitchy about it, then you can tell them to piss off and let you do your job.
5) The first and most important issue is deciding what you want the students to learn. Theory-vs-practice, Access-vs-SQL, GUI-vs-code, University-vs-School, etc should fall into place after that is decided. Nobody on the Internet or in University IT can help you till this is decided.
It looks like their facing an uphill legal battle using a questionable legal argument and few legal resources. It's might be better to walk away from this fight.
They could still have a name that reflects their "Linux Gazette" history. A few names:
* lg-news (re-using their lg-* naming convention) * Likeable Gazette * Loveable Gazelle * LG Gazette ("LG" stands for "LG Gazette"...)
Debian Project Leader (DPL) Martin Michlmayr announces that Yellow Dog Linux is hiding the terrorists. x86 users are unconcerned when Debian terminates its PowerPC port and then destroys the Yellow Dog militia.
Patrick Volkerding claims that the Debian breach proves that Debian's committee style of development is cumbersome and prone to release bad software. DPL accuses Volkerding of condoning terrorism and issues demands that Slackware:
* Use.deb archives instead of.tar.gz arcvhies * Use apt-get to fetch software * Set/var/mail to mode 2775 (had been 1777) * Store MySQL database configuration files under/var/lib/mysql instead of/usr/local/mysql * Remove of all non-free, SCO code
In an impassioned speech to Linux Standards Base, the DPL proclaims, "We cannot be safe while these anti-metadata tar-mongers continue distributing SCO intellectual property!" Volkerding insists he has no SCO code but promises to admit LSB inspectors.
Fedora Core leader Michael Johnson agrees that SCO code is bad, but he wants proof that Slackware is involved. Klaus Knopper backs up Debian's claims, asserting that secret Microsoft documents show Volkerding obtained an illicit copy of UNIX sources near Seattle. Debian release manager Anthony Towns presents lkml with photos of a known Slackware user editing a document that looks suspiciously like C code. After several days of contentious arguments, lkml releases a statement condemning the use of C code to infect any distribution with SCO intellectual property.
A slashdot poll reveals that 83% of F/OSS zealots believe Slackware is trying to plant non-free code in Debian. 65% believe Slackware is trying to plant SCO code. 70% believe Debian should act even without lkml support.
Eight months after 11-20, Debian officially invades Slackware. EMACS-enabled developers quickly overrun Slackware's vi defense force.
As a user, I only expect that sending non-free off to a separate organization would cause quality regression. I don't see an advantage in it.
... put non-free somewhere else with people who care about it...
Debian has a number of individuals who maintain non-free packages. Aren't the packages already with people who care about them?
... APT will handle this very easily...
APT might, but Debian's quality doesn't come from APT. It comes from Debian -- the organization, the policies, the developers, the large umbrella.
Suppose non-free were the ugly duckling of the open source community. Then a separate Non-Free Organization should have trouble _getting_ facilities. Would a user be well served with fewer mirrors and fewer build hosts (representing fewer architectures) for the non-free portion?
Suppose non-free Debian poses different technical issues from free Debian. Why should I believe that those issues are better dealt with separately? For example: testing and unstable periodically change ABI when they adopt a new gcc or libc. The two organizations would have to coordinate the transition in order to maintain quality. Also: licensing issues and compilation issues (eg hard-coded paths, patch constraints, binary-only releases) may make non-free software harder to integrate with Debian. That difficulty is good: it is a check against open-source's flexibility. A Debian which ignores non-free issues can only become _less_ compatible. That would make it harder for _me_, a supporter of free software principles, to keep Debian at work (we use Oracle and MatLab); it would make it harder to persuade my bosses or my users that free alternatives can satisfy _their_ needs.
Finally (on another note), Debian provides a vetting process for developers and packages. That process may be overburdened or flawed, but I can't imagine that duplicating the effort at Debian and Non-Free Organization would improve it.
Now (*begins stepping off soap box*) I haven't run any large, multinational non-profit software organizations. But I think you need to make a more compelling case. What can a separate, official organization offer that an integrated, unofficial organization cannot? (*actually gets off soap box*)
GtkFileChooser: a replacement for GtkFileSelection with replaceable backends, many new API features, better user interface (UI is still a work in progress) [Owen Taylor, Federico Mena Quintero]
It's not clear from the changelog what is supposed to be replaceable, but looking at the gtk-devel mailing list (it took me 1 minute), a fine description was posted a couple weeks ago. Applications and users can both provide shortcuts. It can use GnomeVFS instead of Unix file access, so you get access to remote folders and all that...
The message from W3C indicates that Microsoft will drop infringing features from Internet Explorer. You can expect Mozilla and Konqueror to avoid this patent like the plague, causing a chain-reaction encompassing Netscape, Firebird, Galeon, Safari, and others. It won't matter if one or two obscure browsers adopt this patented technique -- it will be dead on the web, and that will be a lose-lose-lose situation for the users, developers, and patent holder.
enable users to build/bin and/sbin dynamically linked on FreeBSD
I am having difficulty parsing this, and neither the article nor the comments here help me. This is my best guess. Someone please correct me.
SITUATION: For some executable program $P in/bin or/sbin; for some executable library $L in/lib; there exists some subroutine $p in $P and some object $l in $L such that $p uses $l.
OLD BEHAVIOR: When building $P, static-linker resolves name "$l", yielding an address or the desired data.
NEW BEHAVIOR: When executing $P, dynamic-linker resolves name "$l", yielding an address or the desired data.
DETAILS OF CHANGE: The kernel enforced the old behavior by examining every request sent to the generic 'dynamic-link' facility and blocked any requests which involved programs which happened to be in/bin or/sbin. The new behavior is achieved by removing the arbitrary, stupid prohibition.
ALTERNATIVE DETAILS OF CHANGE: The old behavior was enforced by the build scripts for $P and $L; we didn't want our super-important $P to be disturbed if something as lame as the dynamic linker crapped out on us. The new behavior is achieved by changing some compiler flags. We will all die when the dynamic linker craps out.
To update, run, install and fix a service is consistent... the configuration for say, Apache, is wildly different to just about anything else, and anything else from each other.
That's an illusion. A good deployment requires a firm understanding of what is being deployed, and that requires the same amount of work for Apache or IIS. Your employee just feels more secure about configuring IIS with a GUI because it seems to require less creative input, and it allows him to deflect creative mistakes onto Microsoft rather than accept them himself.
I'll bite!...those FS are designed with efficiency in mind... end-users (non-root accounts) do not even NEED to see the FS hierarchy, they see/home/$USER...
The filesystem standards are designed to balance many different needs. The beauty of the idealized Unix filesystem is that all system objects fit into a single, logically-ordered namespace which the user and the system both understand. The problem is that there are many reasonable, overlapping ways to organize those objects: where, on the disk or on the network, is data stored? who can modify which data? what kind of data is it? who is interested in the data? when are they interested in it? Answers to these questions affect the security and efficiency of systems, networks, administrators, and users, and no single tree structure can efficiently satisfy all concerns all the time. The FHS provides a reasonable compromise, but it doesn't do a good job of satisfying any one need. (How do I support three versions of the same program? How do I satisfy/var and/etc dependencies for nfs-read-only binaries?)
In a simple Unix deployment, it's true that users may not need to know where executables and libraries and manuals are stored. Environment variables and configuration files provide a layer or three of indirection, and a potentially ugly file system becomes a nice, flat, type-specific namespace. (eg, all the man directories become one list of valid manuals.) Adventerous users can toy with these variables, but they'll have to grok the filesystem.
Linux distro's, however, aren't simple Unix deployments, and they keep adding new object types. For example: wallpapers are stored in/usr/share/pixmaps,/usr/share/wallpaper,/usr/local/share/pixmaps, and/home directories. Gnome or KDE or whoever do not support the indirection of a PATH or MANPATH. If I want to change my wallpaper, I don't get a bunch of thumbnails -- instead I get to search the filesystem.
In short, the Unix fs is a compromise which stores some of the logical and physical ordering among data. It leaves a lot of the work to be done in env vars and config files and program code, where it's easily forgotten or poorly implemented. The Unix way is not a perfect, time-proven way.
Is this sort of like the one time I ran the TI-85 emulator inside of vMac inside of VirtualPC on a Mac, or am I just missing something?
Judging by the screenshot, it's more like the time you recompiled Eclipse to run on.NET, ported.NET to run on top of POSIX/X11, but short-circuited the.NET graphics subsystem (which you hadn't written yet) so that you could directly ask GTK to make X11 primitives, so that the POSIX/X11 could directly send Win32 API calls to the Win32 environment layer on top of the Win2k microkernel, but sometimes bypass the Win32 API calls and the Win32 environment layer so that you could directly send graphics to the graphics subsystem inside the no-longer-a-microkernel Win2k microkernel.
...what the relationship between the netbeans effort and Sun is. Is it like Mozilla and AOL?
Off the top of my head... yes and no.
Mozilla and NetBeans both started out as college research projects; both grew into businesses; both were assimilated by larger businesses; and both were eventually released as open source. NetBeans is availabile under a variant of the Mozilla license.
That said, I think NetBeans is more important to Sun than Mozilla is to AOL. As far as I know, AOL doesn't get any commercial fruits from Mozilla directly, and itisn't being used in any significant commercial projects. Sun, however, does market some heavish software on top of NetBeans (Sun ONE Studio), and so do several other companies.
...one of the standard OOP pitfalls -- implementing object properties using inheritance...
No. That is not some accidental pitfall that clutzes randomly walk into. Ascribing special properties to the special class in a generalization relation is part of the plan. It's intended usage.
I might be a bit more inclined to believe the article if the example given wasn't one of the standard OOP pitfalls...
The OO literature tells us that we can leverage our vast body of thinking about objects, classes, relations, and such to solve programming problems. That's some pretty complex knowledge, and it's reasonable to expect that an OOP would support that complexity. But most OOP's don't. So we have a problem in theory.
And _lots_ of people run into this problem. They run into it when they're first learning to program; when they're starting a project; when they're maintaining a project. So we have a problem in practice.
Problems beg for solutions. The article cites the problem, and then discuses several solutions. Your post describes a solution. How, exactly, does this make it hard "to believe the article"?
Earnest: I really like you. Skeptic: What? Nooo, you don't reeeally. Do you? Earnest: I want to have your baby. Skeptic: Hahaha. I don't believe you can have babies. But you almost got me there.
Re:debian political parties vs. a national ones
on
Martin Michlmayr Wins DPL
·
· Score: 3, Informative
You clearly haven't tried reading one of the platforms chosen by the Democrats, the Republicans, the Reform Party, the Greens, the Libertarians, or any of the dozen other groups that run national candidates.
These are not trivial documents, and they're chosen at the same time as the party's official candidate (at the national convention). The candidates themselves also publish volumes of press releases and opinion papers. These papers might not be quite as pleasant the DPL platforms, but they cover more material on more difficult issues, and they're influenced by many more people. But then, given the development model of any national organization, could you really expect them to be as nice?
Why does data need to be "verified" or "researched" before we accept its validity? Whatever data they accept, I'm sure it'll pass the Slashdot Standard, and -- since I'm here -- that's evidentally good enough for me.
Slashdot Mirror
So the disagreement is over whether clause 6 of the GPL:
applies to the patent grant. Does "rights granted herein" refer to the rights in clause 6 ("...copy, distribute, or modify...") or to all rights mentioned in the license?
Nope. A blueprint is a plan. A house is an implementation of the plan. Likewise source code is an _implementation_ of a plan - not a plan.
:: blueprint:house :: blueprint:house :: blueprint:house
:: blueprint:house :: blueprint:house :: blueprint:house /* a bit tenuous */
:: blueprint:house :: blueprint:house :: blueprint:house :: blueprint:house
These things are relative:
1) flow chart:source code
2) source code:machine code
3) machine code:execution
4) building requirements spec:blueprint
5) blueprint:house
6) house:daily life
What I find neat is that the relation is transitive, i.e.
1+2) flow chart:machine code
1+2+3) flow chart:execution
2+3) source code:execution
4+5) building requirements spec:house
Well, let's examine the statement a little more closely.
"Machine vision patents 'thrown out.'"
I parsed this as:
->(sentence)
---> (subject, noun) vision
-----> (appositive) machine
---> (predicate, verb) patents
-----> (direct object) 'thrown out'
This statement actually has multiple difficulties. Firstly, only natural persons are allowed to claim invention of patents, and vision is not a natural person. Secondly, only persons and corporations are allowed to own things; vision is not. Thirdly, regardless of how "thrown out" is interpreted, it's not patentable. There are two main interprations:
* The words "thrown out" are themselves patented. One cannot use them anywhere without the permission of machine vision -- This doesn't work because patents only apply to processes. Perhaps the patent applicant would prefer to register a copyright or trademark?
* The words "thrown out" identify the process of throwing a thing out. One may not throw a thing out without the permission of machine vision -- This has two defects. On the one hand, one should use the gerund form instead of the participle form when identifying any thing. On the other hand, the "throwing out" process is usually realized by a person (e.g., "The judge threw out the case."), but, again, vision is not a person.
So clearly "they" -- regardless of that pronoun's antecedent -- cannot "do that" because the statement "Machine vision patents thrown out" can never be true.
It appears those that do not want MDI want to be able to arrange their windows around the desktop leaving little areas to peer through to other apps in the background. This group likes to use the mouse to focus windows and may enjoy being able to swap to another workspace to preserve this environment.
MDI and virtual workspaces are basically the same idea -- group together related windows. The difference: in MDI, the groupings follow vendor/application lines. In virtual workspaces, the user groups windows as he or she wishes. If you want to have immaculately arranged windows from different applications which enable you to compare just the right data (and have focus follow the mouse pointer), then fine. If you want to group a hundred windows from one application (and use keyboard navigation), then fine. If you want a mess, fine.
In this case, having to select objects by visual means is almost impossible. Instead, we prefer a single Alt+Tab or mouse click on the window list to switch away, and another to return to the graphic application environment. The MDI has it's own window list which aids in having to decide between different names and other applications in the same list. It also has its own separate Ctrl+Tab key combination to page between them.
You get the same functionality with virtual workspaces. In my GNOME/Metacity setup, I use Ctrl-Alt-{Left,Right} to move among workspaces, and Alt-Tab to move among windows within a workspace.
In some cultures it is considered improper to offer your guest an executive when sharing IT problems.
For example, if your guest is a vegetarian, he will probably just push the executive off to the side and focus on the parts of the problem which he can digest.
I love osnews little version, with all the directories in the path displayed at the top, the idea being you could click on them to go back to that directory.
/home/me/music/artists/Crazy Fool/Greatest Hits
//////Greatest Hits /////Crazy Fool ////artists ///music //me /home /
I like it, too: going up a couple of directories only requires one short click. Using a combo box either requires two clicks or a dragging motion.
The problem is that it will suck for really deep directory trees. (Getting to the root will require clicking the arrow button several times.) I would like a compromise where the first button pops up a list (like the combo box) and then several buttons show the deepest directories.
For example, if the directory were
I would layout the path navigator like this:
[_ICON_] [artists] [Crazy Fool] [Greatest Hits]
Clicking on the icon would pop up a list (like the combo box):
|
|
|
|
|
|
|
This makes it easy (one click) to go up when you're browsing a small subtree in the file system, but more difficult (two clicks or dragging) to browse all over the file system. The latter case shouldn't be common, though, since "Favorites" would abbreviate navigation among distant subtrees.
File selectors should be longer horizontally since file names can be long. Having something that is taller than it is long is just dumb, there just isn't a polite way to say it.
The file list widget -- but not necessarily the file selector dialog -- should be long horizontally, and all the mockups are better than the current layout (a narrow widget for directories to the left of a narrow widget for files). Eugenia's file list widget is actually wider and contains more information than tigert's. In fact, if you give Eugenia's dialog and tigert's dialog the same dimensions, Eugenia's will handle long file names better (though it will show fewer file names).
There are the whiners, and there are the doers. What he doesn't realize is that doers just do it, and ignore these whiners.
I think that's a bit unfair. Our society embraces a lot of specialization, and there are necessarily more consumers than producers. Now, individual consumers in most markets (should) have an interest in how the market works -- but (individually) lack the power to change the market. That creates a feeling of powerlessness -- which leads to legitimate whining. A few individuals who move from consumption to production may complain less, but you and the journalist will still be overwhelmed by the remaining consumers. In short, there must be and should be whining, and meta-whining won't stop it.
Yes it would. There already is such a court, the International Criminal Court..
The problem is, the USA opposes it.
Well, there are many reasons the ICC is a non-issue:
* The current administration dislikes any and all forms of international cooperation, especially ones which can try US citizens.
* Membership -- The US, Iraq, Kuwait, and Iran are not members of the ICC.
* Complementarity -- ICC complements national courts. It can only try a case if the national court systems cannot or do not.
* Retroactivity -- ICC jurisdiction is not retroactive. It can only try crimes after 2002, which would exclude most of the interesting Iraq-related crimes (except the US invasion...)
I work at the School of Communications for a major state University.
I work in the CS department at a smallish, private university with a liberal arts focus, and I'm responsible for the Oracle systems used in DB/DBA courses. Your question seems strange to me. Some questions/comments:
1) Why on Earth is the School of Communications teaching database concepts?
2) Why does University IT care if you use SQL or Access?
3) Perhaps University IT cares because it has to run an SQL database server for you. It wants to unload that liability. The problem is that doing so actually has academic consequences: it forces students, teachers, and TA's to spend time on the various installation and configuration details of MS Access. _I_ can't say if that's good or bad, but it must be part of your decision. You should only replace server-side SQL coding with MS Access databases if you believe that Access offers pedagogical benefits.
4) Perhaps University IT cares because they think SQL sucks; because the IT director likes GUI's; because the IT director owns Microsoft stock; etc. In any of these circumstances, you can politely nod when they share their opinions -- and then forget the conversation ever happened. If University IT gets bitchy about it, then you can tell them to piss off and let you do your job.
5) The first and most important issue is deciding what you want the students to learn. Theory-vs-practice, Access-vs-SQL, GUI-vs-code, University-vs-School, etc should fall into place after that is decided. Nobody on the Internet or in University IT can help you till this is decided.
It looks like their facing an uphill legal battle using a questionable legal argument and few legal resources. It's might be better to walk away from this fight.
They could still have a name that reflects their "Linux Gazette" history. A few names:
* lg-news (re-using their lg-* naming convention)
* Likeable Gazette
* Loveable Gazelle
* LG Gazette ("LG" stands for "LG Gazette"...)
Will we not be able to have male and female ends on our 1/4" audio cable for fear of offending the transgendered?
Well, I don't know about offending people... but I do get some weird looks whenever I describe a USB connector...
Debian Project Leader (DPL) Martin Michlmayr announces that Yellow Dog Linux is hiding the terrorists. x86 users are unconcerned when Debian terminates its PowerPC port and then destroys the Yellow Dog militia.
.deb archives instead of .tar.gz arcvhies /var/mail to mode 2775 (had been 1777) /var/lib/mysql instead of /usr/local/mysql
Patrick Volkerding claims that the Debian breach proves that Debian's committee style of development is cumbersome and prone to release bad software. DPL accuses Volkerding of condoning terrorism and issues demands that Slackware:
* Use
* Use apt-get to fetch software
* Set
* Store MySQL database configuration files under
* Remove of all non-free, SCO code
In an impassioned speech to Linux Standards Base, the DPL proclaims, "We cannot be safe while these anti-metadata tar-mongers continue distributing SCO intellectual property!" Volkerding insists he has no SCO code but promises to admit LSB inspectors.
Fedora Core leader Michael Johnson agrees that SCO code is bad, but he wants proof that Slackware is involved. Klaus Knopper backs up Debian's claims, asserting that secret Microsoft documents show Volkerding obtained an illicit copy of UNIX sources near Seattle. Debian release manager Anthony Towns presents lkml with photos of a known Slackware user editing a document that looks suspiciously like C code. After several days of contentious arguments, lkml releases a statement condemning the use of C code to infect any distribution with SCO intellectual property.
A slashdot poll reveals that 83% of F/OSS zealots believe Slackware is trying to plant non-free code in Debian. 65% believe Slackware is trying to plant SCO code. 70% believe Debian should act even without lkml support.
Eight months after 11-20, Debian officially invades Slackware. EMACS-enabled developers quickly overrun Slackware's vi defense force.
As a user, I only expect that sending non-free off to a separate organization would cause quality regression. I don't see an advantage in it.
Debian has a number of individuals who maintain non-free packages. Aren't the packages already with people who care about them?
APT might, but Debian's quality doesn't come from APT. It comes from Debian -- the organization, the policies, the developers, the large umbrella.
Suppose non-free were the ugly duckling of the open source community. Then a separate Non-Free Organization should have trouble _getting_ facilities. Would a user be well served with fewer mirrors and fewer build hosts (representing fewer architectures) for the non-free portion?
Suppose non-free Debian poses different technical issues from free Debian. Why should I believe that those issues are better dealt with separately? For example: testing and unstable periodically change ABI when they adopt a new gcc or libc. The two organizations would have to coordinate the transition in order to maintain quality. Also: licensing issues and compilation issues (eg hard-coded paths, patch constraints, binary-only releases) may make non-free software harder to integrate with Debian. That difficulty is good: it is a check against open-source's flexibility. A Debian which ignores non-free issues can only become _less_ compatible. That would make it harder for _me_, a supporter of free software principles, to keep Debian at work (we use Oracle and MatLab); it would make it harder to persuade my bosses or my users that free alternatives can satisfy _their_ needs.
Finally (on another note), Debian provides a vetting process for developers and packages. That process may be overburdened or flawed, but I can't imagine that duplicating the effort at Debian and Non-Free Organization would improve it.
Now (*begins stepping off soap box*) I haven't run any large, multinational non-profit software organizations. But I think you need to make a more compelling case. What can a separate, official organization offer that an integrated, unofficial organization cannot? (*actually gets off soap box*)
It's not clear from the changelog what is supposed to be replaceable, but looking at the gtk-devel mailing list (it took me 1 minute), a fine description was posted a couple weeks ago. Applications and users can both provide shortcuts. It can use GnomeVFS instead of Unix file access, so you get access to remote folders and all that...
The message from W3C indicates that Microsoft will drop infringing features from Internet Explorer. You can expect Mozilla and Konqueror to avoid this patent like the plague, causing a chain-reaction encompassing Netscape, Firebird, Galeon, Safari, and others. It won't matter if one or two obscure browsers adopt this patented technique -- it will be dead on the web, and that will be a lose-lose-lose situation for the users, developers, and patent holder.
enable users to build /bin and /sbin dynamically linked on FreeBSD
/bin or /sbin; for some executable library $L in /lib; there exists some subroutine $p in $P and some object $l in $L such that $p uses $l.
/bin or /sbin. The new behavior is achieved by removing the arbitrary, stupid prohibition.
I am having difficulty parsing this, and neither the article nor the comments here help me. This is my best guess. Someone please correct me.
SITUATION: For some executable program $P in
OLD BEHAVIOR: When building $P, static-linker resolves name "$l", yielding an address or the desired data.
NEW BEHAVIOR: When executing $P, dynamic-linker resolves name "$l", yielding an address or the desired data.
DETAILS OF CHANGE: The kernel enforced the old behavior by examining every request sent to the generic 'dynamic-link' facility and blocked any requests which involved programs which happened to be in
ALTERNATIVE DETAILS OF CHANGE: The old behavior was enforced by the build scripts for $P and $L; we didn't want our super-important $P to be disturbed if something as lame as the dynamic linker crapped out on us. The new behavior is achieved by changing some compiler flags. We will all die when the dynamic linker craps out.
I'll bite! :)
... the configuration for say, Apache, is wildly different to just about anything else, and anything else from each other.
To update, run, install and fix a service is consistent
That's an illusion. A good deployment requires a firm understanding of what is being deployed, and that requires the same amount of work for Apache or IIS. Your employee just feels more secure about configuring IIS with a GUI because it seems to require less creative input, and it allows him to deflect creative mistakes onto Microsoft rather than accept them himself.
I'll bite! ...those FS are designed with efficiency in mind... end-users (non-root accounts) do not even NEED to see the FS hierarchy, they see /home/$USER...
/var and /etc dependencies for nfs-read-only binaries?)
/usr/share/pixmaps, /usr/share/wallpaper, /usr/local/share/pixmaps, and /home directories. Gnome or KDE or whoever do not support the indirection of a PATH or MANPATH. If I want to change my wallpaper, I don't get a bunch of thumbnails -- instead I get to search the filesystem.
The filesystem standards are designed to balance many different needs. The beauty of the idealized Unix filesystem is that all system objects fit into a single, logically-ordered namespace which the user and the system both understand. The problem is that there are many reasonable, overlapping ways to organize those objects: where, on the disk or on the network, is data stored? who can modify which data? what kind of data is it? who is interested in the data? when are they interested in it? Answers to these questions affect the security and efficiency of systems, networks, administrators, and users, and no single tree structure can efficiently satisfy all concerns all the time. The FHS provides a reasonable compromise, but it doesn't do a good job of satisfying any one need. (How do I support three versions of the same program? How do I satisfy
In a simple Unix deployment, it's true that users may not need to know where executables and libraries and manuals are stored. Environment variables and configuration files provide a layer or three of indirection, and a potentially ugly file system becomes a nice, flat, type-specific namespace. (eg, all the man directories become one list of valid manuals.) Adventerous users can toy with these variables, but they'll have to grok the filesystem.
Linux distro's, however, aren't simple Unix deployments, and they keep adding new object types. For example: wallpapers are stored in
In short, the Unix fs is a compromise which stores some of the logical and physical ordering among data. It leaves a lot of the work to be done in env vars and config files and program code, where it's easily forgotten or poorly implemented. The Unix way is not a perfect, time-proven way.
Rant finished!
Is this sort of like the one time I ran the TI-85 emulator inside of vMac inside of VirtualPC on a Mac, or am I just missing something?
.NET, ported .NET to run on top of POSIX/X11, but short-circuited the .NET graphics subsystem (which you hadn't written yet) so that you could directly ask GTK to make X11 primitives, so that the POSIX/X11 could directly send Win32 API calls to the Win32 environment layer on top of the Win2k microkernel, but sometimes bypass the Win32 API calls and the Win32 environment layer so that you could directly send graphics to the graphics subsystem inside the no-longer-a-microkernel Win2k microkernel.
Judging by the screenshot, it's more like the time you recompiled Eclipse to run on
Duh.
...what the relationship between the netbeans effort and Sun is. Is it like Mozilla and AOL?
Off the top of my head... yes and no.
Mozilla and NetBeans both started out as college research projects; both grew into businesses; both were assimilated by larger businesses; and both were eventually released as open source. NetBeans is availabile under a variant of the Mozilla license.
That said, I think NetBeans is more important to Sun than Mozilla is to AOL. As far as I know, AOL doesn't get any commercial fruits from Mozilla directly, and itisn't being used in any significant commercial projects. Sun, however, does market some heavish software on top of NetBeans (Sun ONE Studio), and so do several other companies.
...one of the standard OOP pitfalls -- implementing object properties using inheritance...
No. That is not some accidental pitfall that clutzes randomly walk into. Ascribing special properties to the special class in a generalization relation is part of the plan. It's intended usage.
I might be a bit more inclined to believe the article if the example given wasn't one of the standard OOP pitfalls...
The OO literature tells us that we can leverage our vast body of thinking about objects, classes, relations, and such to solve programming problems. That's some pretty complex knowledge, and it's reasonable to expect that an OOP would support that complexity. But most OOP's don't. So we have a problem in theory.
And _lots_ of people run into this problem. They run into it when they're first learning to program; when they're starting a project; when they're maintaining a project. So we have a problem in practice.
Problems beg for solutions. The article cites the problem, and then discuses several solutions. Your post describes a solution. How, exactly, does this make it hard "to believe the article"?
Earnest: I really like you.
Skeptic: What? Nooo, you don't reeeally. Do you?
Earnest: I want to have your baby.
Skeptic: Hahaha. I don't believe you can have babies. But you almost got me there.
You clearly haven't tried reading one of the platforms chosen by the Democrats, the Republicans, the Reform Party, the Greens, the Libertarians, or any of the dozen other groups that run national candidates.
These are not trivial documents, and they're chosen at the same time as the party's official candidate (at the national convention). The candidates themselves also publish volumes of press releases and opinion papers. These papers might not be quite as pleasant the DPL platforms, but they cover more material on more difficult issues, and they're influenced by many more people. But then, given the development model of any national organization, could you really expect them to be as nice?
Why does data need to be "verified" or "researched" before we accept its validity? Whatever data they accept, I'm sure it'll pass the Slashdot Standard, and -- since I'm here -- that's evidentally good enough for me.