DLL Hell is when Foo DLL 1.0 and Foo DLL 6.0 both stored in the file foo.dll (unlike libfoo.so.1.0 and libfoo.so.6.0) and brain damaged installer programs blindly replace foo.dll version 6.0 with foo.dll version 1.0, thus breaking every single program that depends on the newer version of foo.dll
... or blindly replacing ver. 1.0 with ver. 6.0 leaving apps dependent on features and bugs of 1.0 unusable.
Before MSI, every application had to use its own installer, and each installer might or might not be so smart about checking for versioning / existance of DLLs before forcibly overwriting newer versions with older versions of the same library or blowing away a more recent registry key.
But you'll still have the problem of new versions not being backward compatible. MFC42.DLL exists as a 4.2 as well as 6.0 version. MSVCRT.DLL ver 6.0 tigthened up memory management, but that uncovered bugs in existing programs, rendering them unusable. This problem wasn't fixed until SP3 of Visual Studio. Note that all of these problems are related to new versions of a DLL without changing it's name. Being backward compatible also means keeping the old bugs. At which point it just becomes riduculous.
If there is no money in music, then a lot of the best musicians will simply cease to exist. Yes, we will always have amateur musicians, but a lot of the best musicians will never happen unless they are able to practice all day, every day, and you can't do that unless you do it professionally. There is a reason why professional athletes, for example, will kick almost any amateur's ass.
And why all commercial software kicks all free software. NOT!
I still think the Borland Community has the stupidest disclaimer I've seen:
The materials on this Site are copyrighted and protected by
worldwide copyright laws and treaty provisions. You may download one copy of
the information ("Materials") found on this Site on a single computer for your personal, non- commercial internal use only unless specifically licensed to do otherwise by Borland in writing.
Really? Ever understood what's really important when building software? it's not the IMPLEMENTATION. It's the DESIGN, the algorithms behind the implementation. And these are not stored in GPL-ed code, but in research documents, patents, whatever.
I didn't really understand this answer. The question was how BSD would be better than GPL. I can't see how your reply has any bearing on wheather said code is BSD or GPL or any other license. As long as you can read the source, you can extract the design. Or at least, the license on the code is not what stops you from understanding the design.
If MS wants to get some designs implemented, they get the research documents, bring that to their huge staff of university graduated programmers and let them implement these designs, these algorithms. They don't need BSD licensed code, nor will they use it.
This is plain wrong. The TCP/IP implementation in all Windows versions are based on the original BSD TCP/IP implementation.
Also, a lot of places, the code is actually a lot better to copy than the design. For instance with anything security related it is better to copy the entire source since a lot of breaches are caused by bad implementation. More generally, it's better to copy the entire source if it's a complete 'piece', and has a good interface (this would be called a component today, but a library 10 years ago).
When people really investigate it, they will find the truth. GPL has it's pitfalls. You may want to modify code, but can't cause you don't want to release the changes.
There's nothing in the GPL that prevents you from modifying the source. The only thing the GPL says is that you have to give the source to whomever you give the binary. And that the recipient then can do whatever they like with that source. There's nothing preventing you from charging a million dollars for that binary.
The Apache-example you gave seems a bit weird. If you make a custom modification for custom need, that sounds like a single web-site (or at least custom-built ones ("site" her might be embedded)). And there's nothing in the GPL preventing you from doing that either. You just have to give the customer the source.
Now, nobody that answered me was in any way coerced - they were all free to refuse to cooperate. And those who did answer got something in return.
Think of it this way: I allow you to use some information of me anyway you like as long as you don't give it to anybody else. That has one price. A t-shirt might be enough. Or maybe I jus like you. If you want to share my information with others, you need to pay me more. Since I cannot limit the number of copies you make of the information. The only possibility is to make laws that enforce that limit. As such, it is very similar to other copy-limitation-agreements for other easy-to-copy items.
I program mostly in Borland Delphi, so dropping to Assembler is as simple as
asm
end;
So, if I need to, I do. My main reason is when Borland's handcoded assembly code sucks... For instance, string searching uses REPNE SCASB. Which was the best way of doing searching on 8086 and possibly 80286 processors. But on todays Pentiums, a tight loop is faster. So, if I have a lot of string searching to do....
*sigh* ASP, like JSP, PHP and Perl CGI works by generating HTML on the server, which is then sent to the browser. It's not like some weird ASP code is sent to Netscape, and Netscape gets confused and doesn't know what to do
I *think* the original poster meant the ASP code gets a bit complicated if you have to check for the clients' browser all around the place. Which is what you'll have to do if you want to make use of new (or rather, something that *was* new when that particular version of a browser shipped...)
I wrote: When she later spoke with someone from InterBase R & D, the "fix" he described was merely a change to a string - no fix at all.
MadAhab wrote: That seems to indicate that they DID just switch passwords.
Ok, I was a bit unclear on who said what. But it was the InterBase folks at Borland ('he') who did the 'change passwords' trick, and the Interbase2000 people ('she', Open Source people) discovered it and called it a no-fix.
This is, of course, no proof that the patch doesn't just change the password...
Maybe they didn't fix it at all but just changed the password.
Hopefully not...
There's another line in there that indicates they didn't just switch passwords: When she later spoke with someone from InterBase R & D, the "fix" he described was merely a change to a string - no fix at all.
Looking at the Kylix FAQ there are som interesting points, seemingly contradictory...
Under Is Kylix an Open Source project?, Borland states "The details of the Kylix open source project are currently under development and will be announced at a later date."
But under How will Kylix be priced?, there's the usual "ranges between $99 and $799 for the standard to professional editions and up to $2,500 for the Enterprise editions"
From the press release: Trolltech, creator of the Qt cross-platform graphical user interface (GUI) toolkit, today announced that the current version of Qt/Embedded will be licensed under both a commercial license and an open source license; the GNU General Public License (GPL).
If they delay longer, and release something harder to crack (even for the sake of argument, impossible to crack)
Nah, unless they put the decoder in the speakers, it will be really easy to work around. Don't even need to crack it. Just resample the analog output, or write a sound card driver that copies to disk instead of to the actual sound hardware.
Are you saying they would make a Windows emulator that doesn't run on Linux?
Guessing: non-free Wine. Meaning only Corel can sell, meaning (possibly) no other distro can run it.
Then what would it run on? Windows? A Windows emulator on Windows? What's that? A whino?
Nope, WOW (Windows On Windows). Check an NT box for "wowexec" some time.
Slashdot Mirror
... or blindly replacing ver. 1.0 with ver. 6.0 leaving apps dependent on features and bugs of 1.0 unusable.
M.
But you'll still have the problem of new versions not being backward compatible. MFC42.DLL exists as a 4.2 as well as 6.0 version. MSVCRT.DLL ver 6.0 tigthened up memory management, but that uncovered bugs in existing programs, rendering them unusable. This problem wasn't fixed until SP3 of Visual Studio.
Note that all of these problems are related to new versions of a DLL without changing it's name.
Being backward compatible also means keeping the old bugs. At which point it just becomes riduculous.
M.
And why all commercial software kicks all free software. NOT!
:*)
Now, that's what I call community building.
M.
I didn't really understand this answer. The question was how BSD would be better than GPL. I can't see how your reply has any bearing on wheather said code is BSD or GPL or any other license. As long as you can read the source, you can extract the design. Or at least, the license on the code is not what stops you from understanding the design.
If MS wants to get some designs implemented, they get the research documents, bring that to their huge staff of university graduated programmers and let them implement these designs, these algorithms. They don't need BSD licensed code, nor will they use it.
This is plain wrong. The TCP/IP implementation in all Windows versions are based on the original BSD TCP/IP implementation.
Also, a lot of places, the code is actually a lot better to copy than the design. For instance with anything security related it is better to copy the entire source since a lot of breaches are caused by bad implementation.
More generally, it's better to copy the entire source if it's a complete 'piece', and has a good interface (this would be called a component today, but a library 10 years ago).
M.
There's nothing in the GPL that prevents you from modifying the source. The only thing the GPL says is that you have to give the source to whomever you give the binary. And that the recipient then can do whatever they like with that source. There's nothing preventing you from charging a million dollars for that binary.
The Apache-example you gave seems a bit weird. If you make a custom modification for custom need, that sounds like a single web-site (or at least custom-built ones ("site" her might be embedded)). And there's nothing in the GPL preventing you from doing that either. You just have to give the customer the source.
M.
Think of it this way: I allow you to use some information of me anyway you like as long as you don't give it to anybody else. That has one price. A t-shirt might be enough. Or maybe I jus like you.
If you want to share my information with others, you need to pay me more. Since I cannot limit the number of copies you make of the information. The only possibility is to make laws that enforce that limit.
As such, it is very similar to other copy-limitation-agreements for other easy-to-copy items.
M.
M.
asm
end;
So, if I need to, I do. My main reason is when Borland's handcoded assembly code sucks... For instance, string searching uses REPNE SCASB. Which was the best way of doing searching on 8086 and possibly 80286 processors. But on todays Pentiums, a tight loop is faster. So, if I have a lot of string searching to do....
I *think* the original poster meant the ASP code gets a bit complicated if you have to check for the clients' browser all around the place. Which is what you'll have to do if you want to make use of new (or rather, something that *was* new when that particular version of a browser shipped...)
M.
Isn't this called taglibs?
M.
When she later spoke with someone from InterBase R & D, the "fix" he described was merely a change to a string - no fix at all.
MadAhab wrote:
That seems to indicate that they DID just switch passwords.
Ok, I was a bit unclear on who said what. But it was the InterBase folks at Borland ('he') who did the 'change passwords' trick, and the Interbase2000 people ('she', Open Source people) discovered it and called it a no-fix.
This is, of course, no proof that the patch doesn't just change the password...
M.
Hopefully not...
There's another line in there that indicates they didn't just switch passwords:
When she later spoke with someone from InterBase R & D, the "fix" he described was merely a change to a string - no fix at all.
M.
For security reasons, the patch is available only as a binary and you will be required to register for this download.
Nice, eh?
M.
to the GNOME Foundation.
Meaning, IBM will see it, you and I won't, maybe?.
M.
M.
Short version: Kylix is Delphi for Linux.
Borland already has Interbase on Linux, Open Source and the works.
M.
It will work on both.
Or, at least, that's what they say...
M.
Under Is Kylix an Open Source project?, Borland states "The details of the Kylix open source project are currently under development and will be announced at a later date."
But under How will Kylix be priced?, there's the usual "ranges between $99 and $799 for the standard to professional editions and up to $2,500 for the Enterprise editions"
So this means, yes, no, maybe, and perhaps later?
Believe it when I see it, perhaps...
M.
From the press release:
Trolltech, creator of the Qt cross-platform graphical user interface (GUI) toolkit, today announced that the current version of Qt/Embedded will be licensed under both a commercial license and an open source license; the GNU General Public License (GPL).
M.
Right?
Nah, unless they put the decoder in the speakers, it will be really easy to work around. Don't even need to crack it. Just resample the analog output, or write a sound card driver that copies to disk instead of to the actual sound hardware.
M.
This is a rather shortsighted view. Humans dominate this world because they cooperate. Not because we go on whatever ego trip we like.
And, besides, everything you do interferes with somebody else. That also goes for non-doing.
M.
Guessing: non-free Wine. Meaning only Corel can sell, meaning (possibly) no other distro can run it.
Then what would it run on? Windows? A Windows emulator on Windows? What's that? A whino?
Nope, WOW (Windows On Windows). Check an NT box for "wowexec" some time.
If they can replicate it, why can't anyone else?