It installs, asks me for the activation thing, I don't want an account with Netcentre or anything, so i press cancel. It asks "Do you want to continue with activation", with two options, OK and Cancel (that should be Yes and No, and the question should be do you want to cancel - dirty trick). Any way, I press Cancel, and it brings me back to the activation page, I press cancel again, and it crashes.
Trying it a second time, I find out that I'm supposed to wait about 5 seconds without touching the activation page and it'll dissapear. God knows why.
Anyway, I'm less than impressed with Mozilla's ability to get Access Violations (Segmentation Faults for Unix guys), it seemed to do it alot during the Milestone releases here and there too.
Maybe I just have a talent for crashing Netscape.
:|
The rendering engine is really fast though, and so is the load time. (IE5 is still faster tho:P).
But, if IE is not in the NT kernel, how does it manage to write outside of its assigned window on my Windows NT system? It shouldn't be able to do that.
Which browser is available on nearly every platform and OS out there? Internet Explorer? wrong Netscape is.
yes, but IE works and runs on the most important platforms (Windows, MacOS, Slowlaris, HP-UX).
Re:woo hoo (Score:1) by fsck (xcp@linuxstart.com) on Sunday April 02, @10:01PM EST (#465) (User Info) http://brunching.com/toys/toy-smilinwilly.html
"Netscape are absolute idiots, a browser company which doesn't have the vision to know that within 5 years every major and minor OS in the world would ship with a webbrowser, just like they would ship with text editors."
I'm not sure what you mean by this.
Which browser is available on nearly every platform and OS out there? Internet Explorer? wrong Netscape is.
Which browser was bundled with the operating system in order to discourage the consumers of said operating system from using Netscape? Internet Explorer.
Which browser is bundled with the most Operating systems?
Bill got greedy. His greed hurt a lot of people, made a lot of enemies.
What a load of absolute bullshit. It's so easy to say something like that about a guy who, on paper is worth more then you could even imagine.
But Bill was protecting his company, he was defending windows against java by continuuing to evangelise developers to windows. And he was adding internet and www support to his os for *free*, this is really greedy i must say *rolls eyes*.
Netscape are absolute idiots, a browser company which doesn't have the vision to know that within 5 years every major and minor OS in the world would ship with a webbrowser, just like they would ship with text editors.
Piping and redirecting and filtering make Unix a breath of fresh air to any user who has been trapped in a windows only environment.
Gee and here I'm using pipes, redirecting and filtering on windows and I didn't even realise windows doesn't have them. don't i feel stupid.
The primary reason why windows is so successful is the developer support, microsoft knows this, that's why there's Visual Studio (not to mention various other offerings from borland and ibm).
So you're telling me that the MFC class hiearchry wall chart you got with VC++, the numerous and well documented class descriptions (you're expected to know Win32 cause MFC just wraps Win32), the 1200MB MSDN samples and documentation that comes with VC++ and the source code to MFC is not enough for you? Try searching for once, you'll be able to find heaps of books on MFC programming on amazon. Yeesh.
And what's this about "don't get me started on VC++", i'm interested to know what problems you have with it. It's the best C++ IDE out there, excellent compiler and debugger, intellisense, edit-and-continue etc etc etc.
Pictures, word art: These seem to be more of a problem with the terrible cut & paste functionality in X.
Um, what does cutting and pasting have to do with it? You still need an object model so that the container application is able to delagate with the COM objec t(wordart/pictures etc).
Application development: we already have widgets and things without bonobo.
Yes, but how easy is it for example..to use T widgest from C++, Java and TCL? Pretty easy cause there are wrappers for these languages right? Well yes, but what happens if in the future you want to add a new widgest, you'll have you add to the wrappers. With COM, you define interfaces for containing and serving widgest, that way ANY new widget that comes along will work with existing applications.
Like if someone goes and write a widget that displays HTML pages (yeah like IE:P), you can use it in existing pre IE applications like Word pad etc to show HTML.
It's important cause Unix programmers (or rather younger ex-windows programmers) are starting to realise that reinventing widgets and components is a waste of time perhaps?
Every geek might get a thrill out of reinventing the wheel, but it gets tiring after a while.
You don't insert excel documents inside word...oik, but do you insert pictures? word art? how about view webpages with java applets on them? what about serious application development? Surely you don't go and rewrite your own data access objects everytime...no in perl you use perl's data access objects etc
well bonobo, like COM will make all objects - no matter what language they are written in/for work together. every application will be able to 'reuse' any object using the same well defined mechanism.
This short sightedness of people who seem to think componentization is bad, monolithic design is good etc, are the same people stopping large modern desktop application development on unix.
Yes, compared to every other software company in the world, they do. Check out MSDN some time.
Stuff like SMB etc aren't what I'm talking about. I'm talking about documentation for application developers.
MSDN is a cohesive and well maintained resource for developers. It has hundreds of GB of useful articles, documentation and source code, updated basically hourly.
So much of the work done on Linux is poorly documented. Sure I can look up the man pages for "printf", but what about documentation on for example...KEdit's exposed interfaces? No I don't want to read the source.
Many OSS developers leave it up to people to read their source rather than document their work, cause the documentation isn't the fun part.
There's a hell of a lot of professional developers out there that manage to develop windows applications without windows source code - cause microsoft are the only company in the world to document ALL their work well enough (look at MSDN & Technet).
And open source maybe good for some people, but not the consumer, forking windows isn't good.
No, NT and 95 don't count, cause most homeusers are guided to use 95, and all the versions of 95 and compatible (with very few exceptions).
Even win95 software will generally run ok on NT.
Not that source code for NT would be bad, just that saying it won't harm consumers is very narrow sighted.
CE was designed to make it easier for Win32 developers to develop software for CE. Binaries should not 'run', you need to 'port. But the porting process is releatively painless, anyway since that 5% is the most least used, or inappropriate or CE devices.
You don't know anything about interfaces. Complex interface adds waste of resources and security and reliability problems? Uh uh, so what you are saying is that functions are a waste of resources and insecure and unreliable? That's all interfaces are, vtables of function pointers - like C++ virtual functions.
Have you even looked at COM interfaces? They demand more than what existing architectures provide? You haven't been listening at all. You're saying that the ability for an application to embed another application inside it and call that appliation's provided methods thru interfaces is SLOWER than your lovely way of passing commands thru a pipe or stdinput and parsing each 'text' command.
Yeah, sure you understand COM interfaces. Most COM applications use eraly binding, there is very little overhead, and certainly much less than 'conventional' or existing 'embedding' there is (pipes and the such - which isn't really embedding anyway).
I'm going to have one last go to actually make you understand things.
When I talk about interfaces, I'm talking about a COM (or CORBA) interface. Eg. A table of functions. Not user interfaces.
- COM is much faster than your command line options and piping stuff into programs. Why? Cause you COM you ARE CALLING THE FUNCTIONS directly (usually). It's not, read from a stream, parse it, process it.
- Your previous mumblings of Windows NOT supporting pipes or any other form of IPC, your suggestion that parsing input is faster than direct function calls (COM), your suggestion that parsing input from the command line is better than COM, and your suggestion that all software components should be out of process (hence, no functions, just make everything small executables with one function - main) are all signs of your unability to understand modern software design.
COM is usually inprocess (though not neccessarily). Why? Cause COM is a natural extension to C++ (and other languages). It allows you to - at runtime - create and call on any objects you might want. This is not the same as running an execuable and parsing output. That is why it is inprocess. Suggesting that windows is 'horribly' designed because component extensions (COM) or plugins (COM) are inprocess is so absolutely fucking stupid.
And only the idiocy of people like you is holding unix back from becoming a successful and modern desktop enviroment, and making it easy to reuse software. People at the GNOME and KDE projects, and modern Linux developers understand this. You do not. Part of the reason COM/CORBA/BONOBO etc was developed was that programmers got sick of reinventing wheels. We don't need 10000 button widgets, and we don't want to write another one *YET* again so you can have a button in your application. TK has some quite nice widgets, QT has some quite nice widgets, and GTK has some quite nice widgets. These widgest however have no commonality. I can't go and use the exact same METHOD/PROCESS to use QT widgets in my application, or use GTK+ widgets in my application or use TK widgests in my application, cause they all use different mechanisms (note how QT and GTK widgets are *SHOCK* inprocess). Now imagine if all Linux developers agreed that all widgets would be put into shared libraries that export certain functions that allows applications to find out the attributes and behaviours of each of these widgest and call on them - all in the exact same way - all using the exact same function calling procedure. Then you've got a primitive form of OLE/COM.
Now you can see it's a natural extension from writing C++ classes to draw your widgets, to writing C++ classes to draw your widgest, but putting these classes in a library so others can dynamically load them and use them too. Then it's another natural step to put these classes in a library, and expose them in a STANDARD way with standard rules for querying available widgest in the library and creating them, and embedding them into your application.
As you can see now, your comments on that inprocess issue is really close minded.
This is your problem. You never used Unix, so you can't imagine that things like this can be more efficient than masochism used in Windows. That's your loss.
ROFL. Whatever. You're the one who's never programmed in windows, or more than likely haven't really used windows. I have 2 windows boxes and 2 linux boxes here. I'm also a student, at an all Unix University.
[x]emacs is in lisp -- this is why it has more complex interface. However you have guessed wrong again -- while interface is different, running gnuclient with a command line isn't any different from running netscape -remote, except that gnuclient can wait until editing is finished (what makes sense considering the purpose of [x]emacs).
ROFL. That's not what I'm talking about. The interface is different fullstop. Command line? WTF? You're an idiot.
In-process embedding of completely different application is a blasphemy to efficient, secure and reliable software design. Only such a horribly designed system as Windows could require or benefit from such a thing in user-interface programs. Only idiocy of unprotected memory and cooperative multitasking in early versions of Windows could cause that monstrosity to be born.
Again you're so misinformed it makes me cry. Think of COM objects as more abstract C++ classes, C++ classes are OBVIOUISLY inprocess. Like I said, when you do any programming (you've tried that haven't you?) do you write thousands of little executatbles with command line options, or do you write functions and classes?
Just like COM classes, but they can dynamically be loaded at runtime. COM offers no benefits? Would you listen to yourself. I've tried to avoid this, but I just have to say it. You're an absolute total idiot. COM is about REUSE.
You obviously have not developed any modern large scale applications. Cause what you say is so depressingly STUPID.
That's a horrible way to do it. That's not what I want at all. I want to be able to directly access elements on an existing HTML page without resorting to hackery. I want to be able to say access an object that represents an image on the HTML page, and be able to pass that object to another application to manipulate it.
Eseentially COM standardizes everything. Standardises name mangeling (or lack of), and all sorts of mechanisms for accessing data or sharing data of all types.
The kind of thing COM does really well is make OO programming style great. The most important parts of COM is abstraction.
Writing applications as a collection of widgets or objects which expose interfaces with 'callable' methods is just so much better than using some kind of third party protoocl and hack - all different from each other ofcourse.
Like the way I'd be able to control netscape would be different form the way I'd be able to control xemacs. Ofcourse, I was talking about inprocess embedding which is something you can't do with netscape at all.
Slashdot Mirror
If you're talking about Living Witness, Voayger was not destroyed, but simply involved in a war.
If you're talking about Timeless, voyager was destroyed trying to slipstream home. However, Harry and Chakotay used time travel to change that.
He's one of transmeta's primary investors.
It installs, asks me for the activation thing, I don't want an account with Netcentre or anything, so i press cancel. It asks "Do you want to continue with activation", with two options, OK and Cancel (that should be Yes and No, and the question should be do you want to cancel - dirty trick). Any way, I press Cancel, and it brings me back to the activation page, I press cancel again, and it crashes.
:P).
Trying it a second time, I find out that I'm supposed to wait about 5 seconds without touching the activation page and it'll dissapear. God knows why.
Anyway, I'm less than impressed with Mozilla's ability to get Access Violations (Segmentation Faults for Unix guys), it seemed to do it alot during the Milestone releases here and there too.
Maybe I just have a talent for crashing Netscape.
:|
The rendering engine is really fast though, and so is the load time. (IE5 is still faster tho
It uses the MainWin libraries for Unix (HP-UX version does this too). I don't see how that invalidates the fact that IE5 running on Slowlaris.
Still faster than Netscape on solaris.
Um, IE5 works on Windows, Slowlaris, HP-UX and MacOS.
But, if IE is not in the NT kernel, how does it manage to write outside of its assigned window on my Windows NT system? It shouldn't be able to do that.
WTF are you talking about?
Since when was IE in the kernel? It's an activex control for god sakes.
And why is HTML controls different from a Rich text control? Or picturebox control or button control?
IT's different cause some idiot company decided they could make money off giving browsers away for free.
Which browser is available on nearly every platform and OS out there?
Internet Explorer? wrong
Netscape is.
yes, but IE works and runs on the most important platforms (Windows, MacOS, Slowlaris, HP-UX).
Re:woo hoo (Score:1)
by fsck (xcp@linuxstart.com) on Sunday April 02, @10:01PM EST (#465)
(User Info) http://brunching.com/toys/toy-smilinwilly.html
"Netscape are absolute idiots, a browser company which doesn't have the vision to know that within 5 years every major and minor OS in the world would ship with a webbrowser, just like they would ship with text editors."
I'm not sure what you mean by this.
Which browser is available on nearly every platform and OS out there?
Internet Explorer? wrong
Netscape is.
Which browser was bundled with the operating system in order to discourage the consumers of said operating system from using Netscape?
Internet Explorer.
Which browser is bundled with the most Operating systems?
Netscape.
And what does that have to do with Gates?
Bill got greedy. His greed hurt a lot of people, made a lot of enemies.
What a load of absolute bullshit. It's so easy to say something like that about a guy who, on paper is worth more then you could even imagine.
But Bill was protecting his company, he was defending windows against java by continuuing to evangelise developers to windows. And he was adding internet and www support to his os for *free*, this is really greedy i must say *rolls eyes*.
Netscape are absolute idiots, a browser company which doesn't have the vision to know that within 5 years every major and minor OS in the world would ship with a webbrowser, just like they would ship with text editors.
I disagree... but then why would that stop anyone from using cygwin to get access to your precious gnu/unix tools.
or even 4NT/4DOS.
Piping and redirecting and filtering make Unix a breath of fresh air to any user who has been trapped in a windows only environment.
Gee and here I'm using pipes, redirecting and filtering on windows and I didn't even realise windows doesn't have them. don't i feel stupid.
The primary reason why windows is so successful is the developer support, microsoft knows this, that's why there's Visual Studio (not to mention various other offerings from borland and ibm).
So you're telling me that the MFC class hiearchry wall chart you got with VC++, the numerous and well documented class descriptions (you're expected to know Win32 cause MFC just wraps Win32), the 1200MB MSDN samples and documentation that comes with VC++ and the source code to MFC is not enough for you?
Try searching for once, you'll be able to find heaps of books on MFC programming on amazon.
Yeesh.
And what's this about "don't get me started on VC++", i'm interested to know what problems you have with it. It's the best C++ IDE out there, excellent compiler and debugger, intellisense, edit-and-continue etc etc etc.
Pictures, word art: These seem to be more of a problem with the terrible cut & paste functionality in X.
Um, what does cutting and pasting have to do with it? You still need an object model so that the container application is able to delagate with the COM objec t(wordart/pictures etc).
Application development: we already have widgets and things without bonobo.
Yes, but how easy is it for example
Like if someone goes and write a widget that displays HTML pages (yeah like IE
It's important cause Unix programmers (or rather younger ex-windows programmers) are starting to realise that reinventing widgets and components is a waste of time perhaps?
Every geek might get a thrill out of reinventing the wheel, but it gets tiring after a while.
You don't insert excel documents inside word...oik, but do you insert pictures? word art? how about view webpages with java applets on them? what about serious application development? Surely you don't go and rewrite your own data access objects everytime...no in perl you use perl's data access objects etc
well bonobo, like COM will make all objects - no matter what language they are written in/for work together. every application will be able to 'reuse' any object using the same well defined mechanism.
This short sightedness of people who seem to think componentization is bad, monolithic design is good etc, are the same people stopping large modern desktop application development on unix.
Yes, compared to every other software company in the world, they do. Check out MSDN some time.
Stuff like SMB etc aren't what I'm talking about. I'm talking about documentation for application developers.
MSDN is a cohesive and well maintained resource for developers. It has hundreds of GB of useful articles, documentation and source code, updated basically hourly.
So much of the work done on Linux is poorly documented. Sure I can look up the man pages for "printf", but what about documentation on for example...KEdit's exposed interfaces? No I don't want to read the source.
Many OSS developers leave it up to people to read their source rather than document their work, cause the documentation isn't the fun part.
There's a hell of a lot of professional developers out there that manage to develop windows applications without windows source code - cause microsoft are the only company in the world to document ALL their work well enough (look at MSDN & Technet).
And open source maybe good for some people, but not the consumer, forking windows isn't good.
No, NT and 95 don't count, cause most homeusers are guided to use 95, and all the versions of 95 and compatible (with very few exceptions).
Even win95 software will generally run ok on NT.
Not that source code for NT would be bad, just that saying it won't harm consumers is very narrow sighted.
You on drugs? Cisco stock goves above Microsoft a wee bit and suddenly Mcnealy is correct?
"The network is the computer" ha!
Yeah, and without a network there's no computer right?
CE was designed to make it easier for Win32 developers to develop software for CE.
Binaries should not 'run', you need to 'port. But the porting process is releatively painless, anyway since that 5% is the most least used, or inappropriate or CE devices.
Yes I do.
/..
But why would that matter anyway? One finds his place in a community such as
Windows CE and Windows NT don't share much except a start button, when it comes down to brass tacks.
you mean apart from the fact that CE's kernel is based on NT's, and that Windows CE basically supports 95% of the Win32 APIs?
You are a dumbass 24 hours a day.
You don't know anything about interfaces. Complex interface adds waste of resources and security and reliability problems? Uh uh, so what you are saying is that functions are a waste of resources and insecure and unreliable? That's all interfaces are, vtables of function pointers - like C++ virtual functions.
Have you even looked at COM interfaces? They demand more than what existing architectures provide? You haven't been listening at all. You're saying that the ability for an application to embed another application inside it and call that appliation's provided methods thru interfaces is SLOWER than your lovely way of passing commands thru a pipe or stdinput and parsing each 'text' command.
Yeah, sure you understand COM interfaces. Most COM applications use eraly binding, there is very little overhead, and certainly much less than 'conventional' or existing 'embedding' there is (pipes and the such - which isn't really embedding anyway).
I'm going to have one last go to actually make you understand things.
When I talk about interfaces, I'm talking about a COM (or CORBA) interface. Eg. A table of functions. Not user interfaces.
- COM is much faster than your command line options and piping stuff into programs. Why? Cause you COM you ARE CALLING THE FUNCTIONS directly (usually). It's not, read from a stream, parse it, process it.
- Your previous mumblings of Windows NOT supporting pipes or any other form of IPC, your suggestion that parsing input is faster than direct function calls (COM), your suggestion that parsing input from the command line is better than COM, and your suggestion that all software components should be out of process (hence, no functions, just make everything small executables with one function - main) are all signs of your unability to understand modern software design.
COM is usually inprocess (though not neccessarily). Why? Cause COM is a natural extension to C++ (and other languages). It allows you to - at runtime - create and call on any objects you might want. This is not the same as running an execuable and parsing output. That is why it is inprocess. Suggesting that windows is 'horribly' designed because component extensions (COM) or plugins (COM) are inprocess is so absolutely fucking stupid.
And only the idiocy of people like you is holding unix back from becoming a successful and modern desktop enviroment, and making it easy to reuse software. People at the GNOME and KDE projects, and modern Linux developers understand this. You do not. Part of the reason COM/CORBA/BONOBO etc was developed was that programmers got sick of reinventing wheels. We don't need 10000 button widgets, and we don't want to write another one *YET* again so you can have a button in your application. TK has some quite nice widgets, QT has some quite nice widgets, and GTK has some quite nice widgets. These widgest however have no commonality. I can't go and use the exact same METHOD/PROCESS to use QT widgets in my application, or use GTK+ widgets in my application or use TK widgests in my application, cause they all use different mechanisms (note how QT and GTK widgets are *SHOCK* inprocess).
Now imagine if all Linux developers agreed that all widgets would be put into shared libraries that export certain functions that allows applications to find out the attributes and behaviours of each of these widgest and call on them - all in the exact same way - all using the exact same function calling procedure.
Then you've got a primitive form of OLE/COM.
Now you can see it's a natural extension from writing C++ classes to draw your widgets, to writing C++ classes to draw your widgest, but putting these classes in a library so others can dynamically load them and use them too.
Then it's another natural step to put these classes in a library, and expose them in a STANDARD way with standard rules for querying available widgest in the library and creating them, and embedding them into your application.
As you can see now, your comments on that inprocess issue is really close minded.
This is your problem. You never used Unix, so you can't imagine that things like this can be more efficient than masochism used in Windows. That's your loss.
ROFL. Whatever. You're the one who's never programmed in windows, or more than likely haven't really used windows. I have 2 windows boxes and 2 linux boxes here. I'm also a student, at an all Unix University.
[x]emacs is in lisp -- this is why it has more complex interface. However you have guessed wrong again -- while interface is different, running gnuclient with a command line isn't any different from running netscape -remote, except that gnuclient can wait until editing is finished (what makes sense considering the purpose of [x]emacs).
ROFL. That's not what I'm talking about. The interface is different fullstop. Command line? WTF? You're an idiot.
In-process embedding of completely different application is a blasphemy to efficient, secure and reliable software design. Only such a horribly designed system as Windows could require or benefit from such a thing in user-interface programs. Only idiocy of unprotected memory and cooperative multitasking in early versions of Windows could cause that monstrosity to be born.
Again you're so misinformed it makes me cry. Think of COM objects as more abstract C++ classes, C++ classes are OBVIOUISLY inprocess. Like I said, when you do any programming (you've tried that haven't you?) do you write thousands of little executatbles with command line options, or do you write functions and classes?
Just like COM classes, but they can dynamically be loaded at runtime.
COM offers no benefits? Would you listen to yourself. I've tried to avoid this, but I just have to say it. You're an absolute total idiot. COM is about REUSE.
You obviously have not developed any modern large scale applications. Cause what you say is so depressingly STUPID.
I stress again, YOU ARE STUPID.
That's a horrible way to do it. That's not what I want at all.
I want to be able to directly access elements on an existing HTML page without resorting to hackery. I want to be able to say access an object that represents an image on the HTML page, and be able to pass that object to another application to manipulate it.
Eseentially COM standardizes everything. Standardises name mangeling (or lack of), and all sorts of mechanisms for accessing data or sharing data of all types.
The kind of thing COM does really well is make OO programming style great. The most important parts of COM is abstraction.
Writing applications as a collection of widgets or objects which expose interfaces with 'callable' methods is just so much better than using some kind of third party protoocl and hack - all different from each other ofcourse.
Like the way I'd be able to control netscape would be different form the way I'd be able to control xemacs. Ofcourse, I was talking about inprocess embedding which is something you can't do with netscape at all.