But what the HELL does that have to do with packages, installers,...
Generally packages/installers handle menu items. Hence why I see it relevant point to bring up when it comes to comparing software management systems.
let alone whether Apple is open?
I have no idea at this point, kind of got side-tracked addressing other comments.
Or are you mistaking me for an Apple fanboi and think you can troll me just by being critical of Apple?
No, I have just disagreed with some of your comments.
I would not of developed on OS X had I found it entirely worthless. I am just generally critical of anyone saying something is a certain 'case'. I have had way too many people 'hype' Windows, Linux, OS X, BSDs and their organizations. Ended up believing it until I started using them and interacting in the communities.
Even now, I am not particularly satisfied with any system or their respective organizations.
The dock isn't a menu. And not all UNIX window managers even HAVE an application menu, let alone automatically add them.
I am not aware of currently popular window managers/desktop environments that do that. I am aware of unpopular and ones which are like that however.
And that has nothing to do with an installer or the lack of an installer, it has to do with Apple's design decisions in the Dock and Finder.
That is obvious, but in my opinion, that is a flawed design. For a operating system that concentrates so much on being user friendly and 'intuitive' to the user's wishes, it doesn't seem to fulfill that role there.
So they bundle all the libraries they depend on? Or they don't depend on libraries?
Depends, might be dependent on LSB, or they come with all the necessary libraries, static builds etc.
And what do you do if you have an application that has a hook in to Open Office and to Gimp?
Your question is vague and I'm not sure I understood you correctly. I have done some hooks into Open Office, I have done it with the extensions support. Gimp on the other hand.. I can only think of the Gimp Toolkit that can is useful outside of Gimp (unless you mean a script-fu script?). In which case the LSB certainly provides that - although static builds of the gimp toolkit is a definite possibility too. One can also install runtimes for certain things such as Java application,.net applications (with runtimes, the site usually tells you to do so) etc.
On Windows, that kind of thing is handled by god knows what.
Active X, plugin support, extensions, runtimes?
Under the UNIX package systems I'm used to, that's handled by dependencies. Or do you simply not do that?
I generally just follow the LSB for most things.
Try it with two versions of Firefox 2, or two versions of Firefox 3.
That would require a redirect install (using dpkg), but entirely possible. Although if ran under the same user, it would modify the same ~/.configuration path, as it would under OS X - not very desirable.
I don't remember whether it was called Netscape or Firefox at the time, but I do remember having to rebuild it with a different $PREFIX to get it to work.
In the past I had to do this with Wine, but not anymore.
I also had to go through insane hoops to install two different JVMs and two different versions of Apache libraries to get two versions of Tomcat working using RPMs under RHEL 4.
Redhat, Fedora, CentOS - the only popular distributions I have never used and have no interest in doing so. What I can say is that I have run JVM 1.2.x, 1.5.x at the same time with Tomcat under SuSE Linux Enterprise Server. It did require some fiddling (I think I spent 30 minutes total?), but it would of been a lot more manual if I had to do it on OS X server or Windows 2003 Server.
I'd have set up separate chroot environments, but I didn't have the disk space for two sets of all the support THOSE beggers would have needed in a chroot environment.
Why wouldn't "mount -o bind/orig-path/new-path" of worked?
Even if that's all *you* can see in it... what the hell is wrong with that?
I have brought up the 'issues' I see with it in my previous posts already. Using that system over a centrally software managed system takes away too many advantages for something that doesn't really provide that many benefits.
I don't think I've ever seen that and I've been using FreeBSD since it was a set of patchkits for 386BSD (I did patchkit 23, an
Yes, friend, I know that. One of their licenses is open source. I'm not referring to that one as a "shared source" license, the whole "shared source" debacle is about Microsoft trying to confuse the issue, and I do not go along with their shenanigans.
I have seen a lot of crap from Microsoft, but I have seen decent stuff come out of them too. The same is with Apple.
I have already brought up a number of protocols that Apple came up with that aren't "Samba, CUPS, etc". You have even acknowledged that one of them is useful, is it beyond conception that the rest might be?
Apple did not come up with multicasting DNS or zero configuration. They did a implementation which was considered useful by some Linux distributions (although considering the numerous issues it caused in Mandriva, doesn't make it a winner in my book).
WHat the **** does that have to do with anything? If they want it in the task bar in Windows they have to do that too. Sheesh.
Difference is that in Windows, Linux, BSDs and pretty much most other desktop operating systems, new software automatically are added to menus.
And then find all the rpm or deb files that it depends on
Uhm, no? Cross distro RPMs don't have dependencies. Checkout popular commercial RPMs, Staroffice's RPMs, Crossover RPMs etc.
All those repositories have to be in the config file.
You are confusing distribution specific packages with cross-distribution packages.
I've been in "RPM Hell" looking for dependencies often enough.
Which tells me you've been trying to install packages made for another operating system/distribution.
you don't get any of the management advantages you're talking about for that application.
Conflict management, dependency management (if it's a distro specific package), once the software is available in a repository, it will also become automatically upgradable.
Most people only use a small part of the ecosystem. But it's a different part for everyone. And it only exists because the system encourages it. Linux, both in culture and in design, doesn't.
To be honest, I don't see it. I don't see how Apple's pkg/application in directory encourage it anymore than Linux's package management options.
Then why the **** are they shipping GPLed and other non-BSD code, *by default*, as part of FreeBSD
Last time I poked at FreeBSD, the installer threw up a message asking me if I wanted to install GPL utilities, considering them less free. The default option was 'no'. These days I work more with OpenBSD (it really needs a lot of improvement for friendliness) than any other BSD, and it still seems to be the same, non-BSD-like code tends to get shunned.
And it lets you install multiple versions of applications, just by putting them in different appdirs.
And? I can install firefox-2, firefox-3. Having both browsers on my system at the same time, it's the same thing with different libraries.
I've been a system admin for 20 years and I've frequently had to work around installers for users
I haven't lived long enough to do *nix that long. I think a lot of your conceptions are from older times in issues. For one, you're bringing up DLL hell, rpm dependency hell etc. as reasons why one shouldn't use the windows or Linux methods of managing software. But those issues have been long resolved.
having a singe appdir that, in almost all cases, "just works"... is a huge advantage.
I don't disagree that it has it's merits, but I don't really see this an issue anymore. I can't even come up with a modern day issue related to package management, windows installers anymore.
The only 'benefit' I can see, is the application is easily 'installable' on other computers by just copying it.
Microsoft came into open source kicking and screaming, and they still have source out there under restrictive "shared source" licenses. The absolutely worst behaviors you have pointed to from Apple are petty compared to the whole "shared source" debacle.
Believe it or not, one of Microsoft's shared source licenses is fully compatible with BSD and GPL licensing.
And Microsoft's open source doesn't include the things you most need to interoperate with Windows for open source developers, like file systems, kernel and network APIs and protocols. To get that kind of thing you need to sign NDAs. Which makes them completely irrelevant to how open they are.
Since Apple uses Samba, CUPS etc. Not really much from Apple either.
OK, you're Joe Average, and you want to install Mikes Cool App. On Windows you download MikesCoolApp.exe and run it, and it installs.
Pretty much.
On Mac OS X you download MikesCoolApp.dmg and run the.pkg or dragg the.app to Applications (or, in my case, I put it in/Local/Applications to reduce the amount of 3rd party stuff I have to copy out of/Applications when I upgrade).
And don't forget that if they want it in the dock, they have to add it there too (or, do what I do, which is add the applications folder to the dock and categorize the applications like KDE's application menu)
On Linux... you edit a config file and add the URL for a new repository and run the installer.
Wrong, to install a specification application, you would normally download and open the.rpm or.deb file. Or, if the organization/developer provides a repository, they usually give you a single command to copy paste into a run dialog box (which I think is stupid since they could essentially just make a deb/rpm to do the same thing).
And then when you want to install a new box, you hope that all those repositories are still there.
I don't hope, they are still there. Unless I reinstall from scratch, in which case it's just a matter of copying the/etc/apt/sources.list.d/ folder (contains all the custom repositories) over - that's it.
No "installers", no central database, just a bunch of property lists that the system keeps track of.
I don't see what is wrong with the system keeping track of installed applications on the system, updates and software management. Unless you are manually modifying this stuff in the first place, manually forcing things that are not supposed to be installed - there shouldn't ever be an issue and there shouldn't be a reason to either.
No, I'm saying that the biggest advantage of Mac OS over Linux is the fact that there's lots of commercial software out there.
Beyond Apple, Adobe, Microsoft and a few games, I honestly haven't seen that much better commercial software out there (unless you want to count the commercial IRC, bittorent clients, text editors which are lacking in features compared to their opensource equivalents) for Mac. I've seen a lot more proprietary stuff on Linux, from virtualization technologies to Microsoft Exchange alternatives - But that stuff is mostly server based. Certain commercial software like Crossover games runs a lot better on Linux due to not having as many issues with broken opengl interfaces and buggy drivers (for more info, see BSDtalk #148).
I sure wouldn't be bothering with OS X if it was just UNIX with a white color scheme.
I actually don't run Kubuntu Linux on my main laptop because it's opensource, free - I have bought a lot of software on my system, including VMware, staroffice, crossover, crossover games, Microsoft Office 200
Most of this thread still makes no sense to me, sorry. I'm having a good deal of trouble figuring out how any of it relates to the original point about Apple being more proprietary than Microsoft.
I brought this up earlier in the thread already. I just don't really see Apple being more open than Microsoft is. Microsoft has made numerous opensource contributions as has Apple. Faith on both counts can be questionable with certain behaviour.
That's great, but a central repository seems pretty much incompatible with a thriving commercial software ecosystem to me.
You've made a assumption about me which is incorrect. While a central repositories are good for the software that comes with the OS, commercial software/specialist software should provide their own repositories, which would fulfill the needs for both scenarios. Thankfully, that is beginning to be the case on Linux systems now.
For example, on my Kubuntu system, I have additional sources for certain specific applications:
deb http://wine.budgetdedicated.com/apt hardy main deb http://dl.google.com/linux/deb/ stable non-free deb http://download.skype.com/linux/repos/debian/ stable non-free deb http://getswiftfox.com/builds/debian unstable non-free deb http://www.getdropbox.com/static/ubuntu hardy main
However, it is unfortunate that these companies/organizations have not yet standardized a way to add these. a simple.deb file (for debian based distributions), adding the sources to a file in/etc/apt/sources.list.d/[product/company name] would be sufficient, instead of the copy-paste command they give (not difficult, but still - stupid in my opinion). The same could be done for RPM based distributions (fortunately there is only two major package management systems on Linux).
If you don't care about that, fine, but then why are you programming on the Mac?
Are you trying to imply all development on the Mac should be for commercial purposes (that has been the reason why I have developed in the Mac in the past - I just find this question confusing)?
I think you misunderstand the point of the BSD license. Really, I do.
You are not going to see FreeBSD, OpenBSD or NetBSD developers using code that is more restrictive than their current BSD license - yes, you can take their code and make it proprietary, but those developers have a goal to actually make code that is not closed/restricted in the first place. They are not including closed source/proprietary/very restrictive software with their software for a reason. This is why Apple's licenses do not work with the BSD opensource community either.
By massive brute force that isn't practical for anyone that doesn't have as tight control over the OS as Microsoft.
Not quite. Microsoft changed the way shared DLLs worked by requiring developers use Microsoft redistributable packs for DLL updates if they needed to - which would only update the system should those DLLs be out of date. Additionally, Windows XP has some mechanisms to present older software with the correct version of the DLL they are looking for.
For new software and new dramatically different DLLs, Microsoft simply names their new DLLs new names, such as riched20.dll, riched32.dll, Msftedit.dll
Linux libraries work the same way, where you can find libgtk-x11.so libgtk-x11-2.0.so.. etc. Preventing such conflicts.
That's precisely what I'm talking about. OS X packages using the NeXT Framework and appdir mechanisms don't have those problems. Those problems only come when you drop back to the traditional UNIX model.
I have done plenty of OS X development, I could package 'mc' into a application directory with all it's required files, like any other OS X applications and trust me - it would segfault on newer and older versions of OS X. These issues are not just because the binaries happened to be arranged in the traditional UNIX model.
That's not "hilarious", that's "how it should work".
Wait? You advocating people pirating software? Because that hilarious comment was only about me seeing someone do piracy.
You've been brainwashed into thinking that software distribution NEEDS to be complex and require complex packaging schemes.
I personally prefer to have all the software on my system centrally managed by a software management systems, that automatically update, install, uninstall all the software with just one click - the end result is not complex.
I don't want to goto websites, downloading.dmg images, drag and dropping the application into a folder, then creating dock menus (I'm not sure if my parents could even handle that), having to update the software myself... - This end result just ends up being a lot of work, for flexibility in moving applications around which is... Not very useful at the end of the day.
That is technically possible, if you move an application by copying it and deleting the original, then don't run it even once.
Just a note, don't do this with Tinkertool, you will be sorry.
You can, I suppose, but I've been downright abusive and I haven't managed to break anything yet.
You have to work REALLY hard to make it impossible to use your software with BSD code. Even with the old BSD license about the only license I can think of that wasn't compatible with it was, well, the GPL.
If you want your BSD license code to remain BSD, as in, only the same restrictions as BSD, it is also incompatible for developers who develop only under the BSD license.
Um, the top level link to Apple's open source releases has always been www.opensource.apple.com and the link to Darwin is on the front page. That *is* posting modern releases of Darwin.
Seriously, unless you knew that link or guessed it to begin with. You wouldn't find it on Apple's site. A link from a site that is unlinked in Apple's main site is not useful.
Trolltech's QT documentation is focussed on QT, not Linux, too.
Trolltech does not have a manual section on Linux to begin with. Apple has a section on Darwin and talks about OS X. Darwin is also Apple's creation. The link I linked to is what Apple claimed to be documentation for Darwin.
You don't go to XOrg to get your UNIX documentation either.
Again, this has not related to what I initially brought up. You're just making up irrelevant stories now.
The basic UNIX documentation, on any UNIX system, is in the reference manual.
Uhuh...
And there Apple does a better job than pretty much any Linux distro I've ever used.
I don't know what you're talking about here. Apple's OS X reference manual? Apple's unix (Note: XNU is not Unix) reference manual? Apple's general manuals?
See, Linux documentation itself tends to be a bit scattershot.
What the hell do I care about Linux? I was comparing the openness of OS X to Windows. Linux is just a kernel, period. The documentation for the kernel is available on kernel.org.
Of course, you're probably using some "shrinkwrap" term to describe common Linux desktop distributions... In which case, the documentation is provided with the distribution themselves. Since the major distributions all have regulations that require each and every executable binary, each library in their main repository to have documentation (including the kernel), standardized in a fashion that each module/application has it's own documentation.
You call this worse than OS X's splishsplash of documentation? Where documentation on frameworks and such are often muddled which have numerous bits that refer to different bits of the system and there is often no QA involved in ensuring each piece has accessible documentation in a specific location?
You think? It's missing some subsystems (UNIX tape APIs, for example) but it's a pretty straight UNIX implementation.
And broken threading, signaling, which Apple's own documentation fails to bring up.
Well, first of all, I was responding to your claim that the open source code was "just the kernel and some old BSD tools". That's complete balderdash, and you know it.
Alright, kernel, BSD tools and some pieces of OS X that are not quite useful.
As to the benefit of things like the NeXT code... the whole NeXT application framework model is the best scheme I've seen anywhere for distributing software that has to interoperate with each other, because it bring "late binding" to libraries and shared files. It avoids Microsoft's "DLL Hell" and Linux's "RPM Hell".
Microsoft resolved the DLL hell issues long ago with Windows XP. "RPM hell" -- I think you mean RPM dependency hell, was caused by users installing RPMs for the wrong distribution/version. This isn't much different from forcing a a install of a.pkg file that was made for a specific version of OS X.
Fortunately users aren't doing that anymore and they are using cross-distribution LSB RPMs/universal binaries with loki installers etc. for the works in all distros software.
Oh, and by the way, I have numerous experiences with finf, fink, macports etc. That software runs into numerous problems on OS X than anything else. Requiring you to install very OS X
Publicly documenting what you're doing, so that other people who aren't you, or your partners, can work with you.
There is substantial documentation on MSDN. You cannot say Microsoft has not done this at all.
Have you actually looked at what they're distributing?
Infact I have, I have spent substantial amount of time digging through Darwin's code base too. I have also looked at the 'holy grail' of documentation. I have found that Apple's documentation is practically useless on Darwin, because it is all focused on OS X, rather than Darwin. Opendarwin died due to lack of interest, the documentation from OpenDarwin is also practically useless.
Apple have been starting to lose touch with the FOSS community when it comes to Darwin, they don't even post modern releases of Darwin anymore, and one has to figure out the URL http://www.opensource.apple.com/darwinsource/ which I might add is not simply posted anywhere on Apple's site telling you, you can find the latest sources there.
On the other hand, thanks to the insane amounts of deliberation with KHTML developers (you should read the older KHTML mailing list archives - it was quite ridicules), the developers moved over to a Apple controlled project "webkit", just so they didn't have to continuously battle with the forks, trying to sync up from both sides. In this case, Apple is doing a lot better than what they used to do.
That includes most of the non-GUI code from NeXT, most of the new tools
Most of which are not really useful.
(like launchd)
Perhaps there is merit to launchd? But how does having access to the sourcecode to a already extremely modifiable "system initialization" bring any more "interoperability"?
The SysVinit system was already open, but this did not bring interoperability to the Mac platform with SysVinit systems. Upstart was created because systems like launchd were incompatible with SysVinit. I am not buying your interoperability arguments.
the whole framework behind their Framework library model.
Which is useful in "interoperability", how?
I can think of it being useful for perhaps porting things to OS X that use that framework... This code has been essentially useless for projects like Afterstep, which have been reimplementing the entire framework that is used in OS X.
Reading technical books on Windows you come up against situations over and over again where the documentation says one thing, Windows does something else, and the author has to throw their hands up and say things like "it appears that the FooObject returns a BarObject under all situations, even though it's supposed to return the object requested in the Baz method..."
The BSD subsystem in OS X does not perform to specifications either. There have been also known quirks in various APIs in OS X that are similar in behavior to what you describe. I don't see how this is any different.
But only by abdicating from the open source and open systems communities. That's a hell of a string.
Just because Apple isn't as open as Red Hat, that doesn't mean they're less open than Microsoft.
Well, I can get all the source code to products from Microsoft provided I sign certain NDAs with certain assurances and give a reasonable explanation as to why.
With Apple... Well, beyond the stuff that was already GPL before they adopted it (thus, preventing them from close sourcing it - ie: webkit). I can get access to... The kernel and some old BSD utilities. Which, to be perfectly honest... Isn't useful, since the useful/interesting things aren't the kernel or the outdated BSD utilities that we can get from else where.
Sure, it's nice that they give out the sourcecode to XNU/darwin, but nobody has any practical use for it. It's not much different from Microsoft opensourcing some of their stuff (MechCommander 2 etc.)
One one hand, Apple is slightly more open in the way that they give the source code because they have to, or because the source code isn't really useful to begin with. On the other hand, I can get useful sourcecode from Microsoft via some restrictive terms. To be honest, I don't really see either company more open than the other.
They both seem pretty closed to me, the only advantage is that with Microsoft, there is a chance to get into everything.
The PC market changed and started demanding audio, yet open source linux is still caught up in the change by trying to standardize on one of 6389382 audio APIs.
Wait? What distribution isn't using ALSA?
In the further quest for desktop multi-media, the markets have changed and demanded Flash capability for web browsing. Uhhh, nope, Linux hasn't made it there yet either.
One click install on Linux distributions compared to "your version of Flash is out of date, click here to upgrade" on a new windows installation? Not really seeing what you're complaining about?
Making the game only playable by interacting with their servers in a online mode, where some of the content is only available through their servers. By holding some of the logic (for example 'quests'). Additionally forcing a online component into the mix where the user has to authenticate with a valid key or such - no way around it.
We've demonstrated that we dislike invasive DRM, and that DRM increases piracy.
Before copy protection became standard in software, software piracy was so rampant that many companies just closed up shop.
If they keep pushing they'll eventually run out of customers, and more civilised companies will pick up the slack.
I don't think that will ever happen to companies like EA. EA generally hold all the cards, they maybe forced to pursue different revenue making games, such as the one mentioned above. Since these larger companies are also the driving force of the market, many (but not all) companies would follow in their foot steps.
If they didn't think piracy cost them sales, they would just ignore it.
Before any kind of copy protection, average joe user could do casual copying. This stops that casual copying.
Either way, one way I get to play the game, the other way I don't. I'll go for the way that benefits me.
Is it any good? I haven't heard anything great about it yet.
No posts on the account, everyone on Slashdot is going to consider it brand new. Except perhaps for some possible next slur of posts by you.
You've embarrassed yourself on so many levels. It's brilliant.
And yet you tell me this, as if, you're trying to make this happen, but hasn't happened.
Not that your kind has any dignity.
Here you go again with your character assassination. Rather than debating the subject, you would rather go rampage on about how I apparently "embarrassed" myself by getting modded up to +5 Karma on two posts.
I'm just going to go now and leave you to making yourself look foolish on your own merits.
You are the one who is doing character assassination instead of coming up with a logical argument that cannot be refuted. I think you need to look in the mirror.
Perhaps you should of done some more research into believable argument - something that didn't rely on a low percentage of web users that use Firefox (only browser I'm aware of that uses Google's phishing lists) and the fact that Yahoo (doesn't use Google's phishing list) and Google are pretty much on par with each other for user usage.
Additionally not taking into account that the company in question could of reported a false positive and the fact that Google does not filter websites or match websites based on IP addresses.
I would like to point out that, since you insist there was IP matching, I have to wonder what kind of business this was, since obviously HTTPS is specific to IP addresses and a specific domain, they couldn't of had a HTTPS site since you suggest it was the same IP address and there were other sites on the same machine.
You've learned the term "IP address" and you're trying to apply it to everything. You're right, Google does not filter by IP address. This is not what I or anybody else has claimed. Go back, read through properly and fill in the gaps. There are now yet more individuals posting to tell you that you don't understand this.
I don't take any credence from a brand new Slashdot account. I know it's you, Mr. AC.
Don't draw a naked wolf giving a blowjob to a horse please, furry.
http://d.furaffinity.net/art/pinkuh/1190102989.pinkuh_cerisepony01.jpg - Oops, too late:(
That's what I thought at first, but the problem is probably that with American drinking regulations, if you do that while you're under 21 (college applicants usually are), it's illegal. I regularly forget that about the US. This whole can-die-for-your-country-but-can't-drink-legally thing is weird.
Ah, I'm sorry. I didn't realize this. That said, being such a common practice, I'm still surprised they even bother to discriminate with that in mind.
Slashdot Mirror
Generally packages/installers handle menu items. Hence why I see it relevant point to bring up when it comes to comparing software management systems.
I have no idea at this point, kind of got side-tracked addressing other comments.
No, I have just disagreed with some of your comments.
I would not of developed on OS X had I found it entirely worthless. I am just generally critical of anyone saying something is a certain 'case'. I have had way too many people 'hype' Windows, Linux, OS X, BSDs and their organizations. Ended up believing it until I started using them and interacting in the communities.
Even now, I am not particularly satisfied with any system or their respective organizations.
I am not aware of currently popular window managers/desktop environments that do that. I am aware of unpopular and ones which are like that however.
That is obvious, but in my opinion, that is a flawed design. For a operating system that concentrates so much on being user friendly and 'intuitive' to the user's wishes, it doesn't seem to fulfill that role there.
Depends, might be dependent on LSB, or they come with all the necessary libraries, static builds etc.
Your question is vague and I'm not sure I understood you correctly. I have done some hooks into Open Office, I have done it with the extensions support. Gimp on the other hand.. I can only think of the Gimp Toolkit that can is useful outside of Gimp (unless you mean a script-fu script?). In which case the LSB certainly provides that - although static builds of the gimp toolkit is a definite possibility too. One can also install runtimes for certain things such as Java application, .net applications (with runtimes, the site usually tells you to do so) etc.
Active X, plugin support, extensions, runtimes?
I generally just follow the LSB for most things.
That would require a redirect install (using dpkg), but entirely possible. Although if ran under the same user, it would modify the same ~/.configuration path, as it would under OS X - not very desirable.
In the past I had to do this with Wine, but not anymore.
Redhat, Fedora, CentOS - the only popular distributions I have never used and have no interest in doing so. What I can say is that I have run JVM 1.2.x, 1.5.x at the same time with Tomcat under SuSE Linux Enterprise Server. It did require some fiddling (I think I spent 30 minutes total?), but it would of been a lot more manual if I had to do it on OS X server or Windows 2003 Server.
Why wouldn't "mount -o bind /orig-path /new-path" of worked?
I have brought up the 'issues' I see with it in my previous posts already. Using that system over a centrally software managed system takes away too many advantages for something that doesn't really provide that many benefits.
I have seen a lot of crap from Microsoft, but I have seen decent stuff come out of them too. The same is with Apple.
Apple did not come up with multicasting DNS or zero configuration. They did a implementation which was considered useful by some Linux distributions (although considering the numerous issues it caused in Mandriva, doesn't make it a winner in my book).
Difference is that in Windows, Linux, BSDs and pretty much most other desktop operating systems, new software automatically are added to menus.
Uhm, no? Cross distro RPMs don't have dependencies. Checkout popular commercial RPMs, Staroffice's RPMs, Crossover RPMs etc.
You are confusing distribution specific packages with cross-distribution packages.
Which tells me you've been trying to install packages made for another operating system/distribution.
Conflict management, dependency management (if it's a distro specific package), once the software is available in a repository, it will also become automatically upgradable.
To be honest, I don't see it. I don't see how Apple's pkg/application in directory encourage it anymore than Linux's package management options.
Last time I poked at FreeBSD, the installer threw up a message asking me if I wanted to install GPL utilities, considering them less free. The default option was 'no'. These days I work more with OpenBSD (it really needs a lot of improvement for friendliness) than any other BSD, and it still seems to be the same, non-BSD-like code tends to get shunned.
And? I can install firefox-2, firefox-3. Having both browsers on my system at the same time, it's the same thing with different libraries.
I haven't lived long enough to do *nix that long. I think a lot of your conceptions are from older times in issues. For one, you're bringing up DLL hell, rpm dependency hell etc. as reasons why one shouldn't use the windows or Linux methods of managing software. But those issues have been long resolved.
I don't disagree that it has it's merits, but I don't really see this an issue anymore. I can't even come up with a modern day issue related to package management, windows installers anymore.
The only 'benefit' I can see, is the application is easily 'installable' on other computers by just copying it.
Believe it or not, one of Microsoft's shared source licenses is fully compatible with BSD and GPL licensing.
Since Apple uses Samba, CUPS etc. Not really much from Apple either.
Pretty much.
And don't forget that if they want it in the dock, they have to add it there too (or, do what I do, which is add the applications folder to the dock and categorize the applications like KDE's application menu)
Wrong, to install a specification application, you would normally download and open the .rpm or .deb file. Or, if the organization/developer provides a repository, they usually give you a single command to copy paste into a run dialog box (which I think is stupid since they could essentially just make a deb/rpm to do the same thing).
I don't hope, they are still there. Unless I reinstall from scratch, in which case it's just a matter of copying the /etc/apt/sources.list.d/ folder (contains all the custom repositories) over - that's it.
I don't see what is wrong with the system keeping track of installed applications on the system, updates and software management. Unless you are manually modifying this stuff in the first place, manually forcing things that are not supposed to be installed - there shouldn't ever be an issue and there shouldn't be a reason to either.
Beyond Apple, Adobe, Microsoft and a few games, I honestly haven't seen that much better commercial software out there (unless you want to count the commercial IRC, bittorent clients, text editors which are lacking in features compared to their opensource equivalents) for Mac. I've seen a lot more proprietary stuff on Linux, from virtualization technologies to Microsoft Exchange alternatives - But that stuff is mostly server based. Certain commercial software like Crossover games runs a lot better on Linux due to not having as many issues with broken opengl interfaces and buggy drivers (for more info, see BSDtalk #148).
I actually don't run Kubuntu Linux on my main laptop because it's opensource, free - I have bought a lot of software on my system, including VMware, staroffice, crossover, crossover games, Microsoft Office 200
I brought this up earlier in the thread already. I just don't really see Apple being more open than Microsoft is. Microsoft has made numerous opensource contributions as has Apple. Faith on both counts can be questionable with certain behaviour.
You've made a assumption about me which is incorrect. While a central repositories are good for the software that comes with the OS, commercial software/specialist software should provide their own repositories, which would fulfill the needs for both scenarios. Thankfully, that is beginning to be the case on Linux systems now.
For example, on my Kubuntu system, I have additional sources for certain specific applications:
However, it is unfortunate that these companies/organizations have not yet standardized a way to add these. a simple .deb file (for debian based distributions), adding the sources to a file in /etc/apt/sources.list.d/[product/company name] would be sufficient, instead of the copy-paste command they give (not difficult, but still - stupid in my opinion). The same could be done for RPM based distributions (fortunately there is only two major package management systems on Linux).
Are you trying to imply all development on the Mac should be for commercial purposes (that has been the reason why I have developed in the Mac in the past - I just find this question confusing)?
You are not going to see FreeBSD, OpenBSD or NetBSD developers using code that is more restrictive than their current BSD license - yes, you can take their code and make it proprietary, but those developers have a goal to actually make code that is not closed/restricted in the first place. They are not including closed source/proprietary/very restrictive software with their software for a reason. This is why Apple's licenses do not work with the BSD opensource community either.
Not quite. Microsoft changed the way shared DLLs worked by requiring developers use Microsoft redistributable packs for DLL updates if they needed to - which would only update the system should those DLLs be out of date. Additionally, Windows XP has some mechanisms to present older software with the correct version of the DLL they are looking for.
For new software and new dramatically different DLLs, Microsoft simply names their new DLLs new names, such as riched20.dll, riched32.dll, Msftedit.dll
Linux libraries work the same way, where you can find libgtk-x11.so libgtk-x11-2.0.so.. etc. Preventing such conflicts.
I have done plenty of OS X development, I could package 'mc' into a application directory with all it's required files, like any other OS X applications and trust me - it would segfault on newer and older versions of OS X. These issues are not just because the binaries happened to be arranged in the traditional UNIX model.
Wait? You advocating people pirating software? Because that hilarious comment was only about me seeing someone do piracy.
I personally prefer to have all the software on my system centrally managed by a software management systems, that automatically update, install, uninstall all the software with just one click - the end result is not complex.
I don't want to goto websites, downloading .dmg images, drag and dropping the application into a folder, then creating dock menus (I'm not sure if my parents could even handle that), having to update the software myself... - This end result just ends up being a lot of work, for flexibility in moving applications around which is... Not very useful at the end of the day.
Just a note, don't do this with Tinkertool, you will be sorry.
I have, plenty of times.
GNU has some notes.
If you want your BSD license code to remain BSD, as in, only the same restrictions as BSD, it is also incompatible for developers who develop only under the BSD license.
Seriously, unless you knew that link or guessed it to begin with. You wouldn't find it on Apple's site. A link from a site that is unlinked in Apple's main site is not useful.
Trolltech does not have a manual section on Linux to begin with. Apple has a section on Darwin and talks about OS X. Darwin is also Apple's creation. The link I linked to is what Apple claimed to be documentation for Darwin.
Again, this has not related to what I initially brought up. You're just making up irrelevant stories now.
Uhuh...
I don't know what you're talking about here. Apple's OS X reference manual? Apple's unix (Note: XNU is not Unix) reference manual? Apple's general manuals?
What the hell do I care about Linux? I was comparing the openness of OS X to Windows. Linux is just a kernel, period. The documentation for the kernel is available on kernel.org.
Of course, you're probably using some "shrinkwrap" term to describe common Linux desktop distributions... In which case, the documentation is provided with the distribution themselves. Since the major distributions all have regulations that require each and every executable binary, each library in their main repository to have documentation (including the kernel), standardized in a fashion that each module/application has it's own documentation.
You call this worse than OS X's splishsplash of documentation? Where documentation on frameworks and such are often muddled which have numerous bits that refer to different bits of the system and there is often no QA involved in ensuring each piece has accessible documentation in a specific location?
And broken threading, signaling, which Apple's own documentation fails to bring up.
Alright, kernel, BSD tools and some pieces of OS X that are not quite useful.
Microsoft resolved the DLL hell issues long ago with Windows XP. "RPM hell" -- I think you mean RPM dependency hell, was caused by users installing RPMs for the wrong distribution/version. This isn't much different from forcing a a install of a .pkg file that was made for a specific version of OS X.
Fortunately users aren't doing that anymore and they are using cross-distribution LSB RPMs/universal binaries with loki installers etc. for the works in all distros software.
Oh, and by the way, I have numerous experiences with finf, fink, macports etc. That software runs into numerous problems on OS X than anything else. Requiring you to install very OS X
There is substantial documentation on MSDN. You cannot say Microsoft has not done this at all.
Infact I have, I have spent substantial amount of time digging through Darwin's code base too. I have also looked at the 'holy grail' of documentation. I have found that Apple's documentation is practically useless on Darwin, because it is all focused on OS X, rather than Darwin. Opendarwin died due to lack of interest, the documentation from OpenDarwin is also practically useless.
Apple have been starting to lose touch with the FOSS community when it comes to Darwin, they don't even post modern releases of Darwin anymore, and one has to figure out the URL http://www.opensource.apple.com/darwinsource/ which I might add is not simply posted anywhere on Apple's site telling you, you can find the latest sources there.
On the other hand, thanks to the insane amounts of deliberation with KHTML developers (you should read the older KHTML mailing list archives - it was quite ridicules), the developers moved over to a Apple controlled project "webkit", just so they didn't have to continuously battle with the forks, trying to sync up from both sides. In this case, Apple is doing a lot better than what they used to do.
Most of which are not really useful.
Perhaps there is merit to launchd? But how does having access to the sourcecode to a already extremely modifiable "system initialization" bring any more "interoperability"?
The SysVinit system was already open, but this did not bring interoperability to the Mac platform with SysVinit systems. Upstart was created because systems like launchd were incompatible with SysVinit. I am not buying your interoperability arguments.
Which is useful in "interoperability", how?
I can think of it being useful for perhaps porting things to OS X that use that framework... This code has been essentially useless for projects like Afterstep, which have been reimplementing the entire framework that is used in OS X.
The BSD subsystem in OS X does not perform to specifications either. There have been also known quirks in various APIs in OS X that are similar in behavior to what you describe. I don't see how this is any different.
Point.
Well, I can get all the source code to products from Microsoft provided I sign certain NDAs with certain assurances and give a reasonable explanation as to why.
With Apple... Well, beyond the stuff that was already GPL before they adopted it (thus, preventing them from close sourcing it - ie: webkit). I can get access to... The kernel and some old BSD utilities. Which, to be perfectly honest... Isn't useful, since the useful/interesting things aren't the kernel or the outdated BSD utilities that we can get from else where.
Sure, it's nice that they give out the sourcecode to XNU/darwin, but nobody has any practical use for it. It's not much different from Microsoft opensourcing some of their stuff (MechCommander 2 etc.)
One one hand, Apple is slightly more open in the way that they give the source code because they have to, or because the source code isn't really useful to begin with. On the other hand, I can get useful sourcecode from Microsoft via some restrictive terms. To be honest, I don't really see either company more open than the other.
They both seem pretty closed to me, the only advantage is that with Microsoft, there is a chance to get into everything.
You only questioned the availability of the source code, not it's "openness".
Stolen NT sourcecode has been floating around the web for ages.
Microsoft has numerous agreements to let one view the sourcecode of Windows too. Although it requires signing some NDAs.
I find it very rare to find a average Joe who will not obtain a illegal copy of software if he could unfortunately.
Considering how popular VLC, Media player classic, Pidgin, Linux are... I would have to disagree.
Wait? What distribution isn't using ALSA?
One click install on Linux distributions compared to "your version of Flash is out of date, click here to upgrade" on a new windows installation? Not really seeing what you're complaining about?
Standard is ALSA in LSB if I am not mistaken.
Google says: No results found for "Improve your jihad: nuclear weapons".
There have been recent papers by Stanford on how to effectively use nuclear bombs against civilians? Where?
What happened to all open source software is crap arguments?
Surely companies likes Microsoft were not jesting!
Personally, I prefer it when people are uninformed about the furry fandom. Makes it so much funnier having people call me a "animal fucker".
Making the game only playable by interacting with their servers in a online mode, where some of the content is only available through their servers. By holding some of the logic (for example 'quests'). Additionally forcing a online component into the mix where the user has to authenticate with a valid key or such - no way around it.
Before copy protection became standard in software, software piracy was so rampant that many companies just closed up shop.
I don't think that will ever happen to companies like EA. EA generally hold all the cards, they maybe forced to pursue different revenue making games, such as the one mentioned above. Since these larger companies are also the driving force of the market, many (but not all) companies would follow in their foot steps.
Before any kind of copy protection, average joe user could do casual copying. This stops that casual copying.
Is it any good? I haven't heard anything great about it yet.
After careful thought, I must agree. I apologize for the manor I disagreed with.
Ah, I like this trick. By posting earlier in the thread it looks like I'm leading on the person. Nice character assassination.
Moderators, I should caution you to look at the time of each post.
No posts on the account, everyone on Slashdot is going to consider it brand new. Except perhaps for some possible next slur of posts by you.
And yet you tell me this, as if, you're trying to make this happen, but hasn't happened.
Here you go again with your character assassination. Rather than debating the subject, you would rather go rampage on about how I apparently "embarrassed" myself by getting modded up to +5 Karma on two posts.
You are the one who is doing character assassination instead of coming up with a logical argument that cannot be refuted. I think you need to look in the mirror.
Perhaps you should of done some more research into believable argument - something that didn't rely on a low percentage of web users that use Firefox (only browser I'm aware of that uses Google's phishing lists) and the fact that Yahoo (doesn't use Google's phishing list) and Google are pretty much on par with each other for user usage.
Additionally not taking into account that the company in question could of reported a false positive and the fact that Google does not filter websites or match websites based on IP addresses.
I would like to point out that, since you insist there was IP matching, I have to wonder what kind of business this was, since obviously HTTPS is specific to IP addresses and a specific domain, they couldn't of had a HTTPS site since you suggest it was the same IP address and there were other sites on the same machine.
This "company" sounds really shady.
I don't take any credence from a brand new Slashdot account. I know it's you, Mr. AC.
http://d.furaffinity.net/art/pinkuh/1190102989.pinkuh_cerisepony01.jpg - Oops, too late :(
I would assume the original AC is lying because Google's practices on filtering bad sites were disclosed long ago on http://www.stopbadware.org/
Ah, I'm sorry. I didn't realize this. That said, being such a common practice, I'm still surprised they even bother to discriminate with that in mind.