I am afraid that you still not see my point: there is nothing you can do with a registry that you cannot do with a filesystem and config files. A registry just adds complexity.
And I have seen multiple ways of storing data in the registry, so there is no convention that is followed, just like there is no convention that is followed on conf files.
It indeed is standardized; very badly. It only add complexity and solves nothing of the problems caused by conf files.
Easily accessible it is not; I would not try to tell my mother to start regedit32. And from a programmer pov: take a look at the Preferences API in Java. That API is simple, straightforward, and allows for multiple implementations.
All in one place: that is actually a very bad thing. Look at all the troubles the registry has caused, and for so many people, just because the registry is in one place. Single point of failure + Microsoft == headaches. Sometimes you need to change things in your registry for a certain application, but Microsoft warns that any changes to the registry can make the whole OS unstable. In no other modern operating system you are confronted with these complexities. Changing a conf file for an application on a *nix machine will not cause the whole OS to become unstable.
Also look at all the crapware that is able to hide in the registry; all possible because there is no rights management; either you can write in the registry, or you cannot. Application A can always read/write settings for application B. That is very, very bad design.
But that is just my whole point: the registry does not have a strict syntax, in any case no more specific than conf files. You can still write binary data in the registry. There are conventions, sure, but you can have conventions in conf files as well. And 'specifically designed for settings': you do not need a registry for that, you need an API for that. Sun solved that in Java: there is an Preferences API that can use the windows registry as backend implementation, as well as separate files as implementation for that API.
And again: I do not claim that conf files are as nice as sliced bread, I just find the registry the wrong solution for a very real problem. The registry does not solve the problems that conf files have brought us; it just added problems and complexity. A clean and well thought out API could save those problems.
Both the registry and a filesystem are hierarhical storage meganisms. Both the registry and a filesystem allow for binary data. Conceptually they are the same. A filesystem however is standard and allows one to copy single files. This has as added advantage that an application can be copied to another computer, settings and all included. Backups are also less complicated. Partial backups of files (only the files that have changed) is a breeze. Partial backups of the registry are non-standard.
A thousand.conf files spread out throughout the directory structure, each with it's own peculiar syntax. Oh yeah, that worked real well.
A million different keys spread out throughout the registry, each with it's own peculiar syntax. That works equally bad as a filesystem/conf file solution.
I am not saying that conf files are good, they have their problems. What Microsoft has done with the registry was add more complexity, while not addressing the real deficiencies of conf files. In effect, we now have the worst of 2 worlds.
My pleasure. Since my visits to the buddhist centre in Amsterdam I got more and more convinced that you can train your brain just like any other part of your body. Meditation and stuff like this sleep-exercise can do remarkable things for your brain. And the nice thing about buddhists is that even if you state you do not believe in reincarnation but like to try some meditation, they still help you for the full 100%.
Pavlov might come in handy here: every time you go to sleep try to think about the same thing, place yourself in the same situation. Make it the last thing you think about before you really doze off.
Your mind will make a link between these two and probably make it easy to fall asleap when you picture yourself in that specific situation. It might take some time, maybe even a month or two, for the link to sink in.
Be careful to not choose something you can run into in everyday life though; it looks a bit silly if everytime someone offers you a piece of apple pie, you fall asleep.
I also like the theory of the registry. All configuration held in a single, easy to access place with a consistent interface.
Well, we already had such a thing. It is called a filesystem. The registry is nothing more than a hierarchical storage meganism with a standard editor.
I still can't stand Finder, I think it is awful, and curse it every time I need to move a few files to some other folder on another drive (usually I just use "mv")
Pure curiosity: what exactly is it that makes Finder bad at moving files for you? For me it works a lot better than Windows' Explorer.
I am from '71. I wish the ?IA?'s of the world would go away.
convert the film stock to ethanol to power their SUVs
Could be, but you forgot the probable reason: companies want to sell stuff; having a lot of stuff in the public domain could lower demand on new stuff, so the old stuff should be destroyed before it enters the public domain.
If we were to build a solar farm in the sahara (550 x 550 km), this would generate enough energy to fill the current need of the whole of humanity.
If at a distance of 10 light years we could detect solar panels and confuse them for an asteroid belt, there are some aliens with quite an energy bill.
It saddens me that every time someone on/. states that capitalism is not the end-all, there are always people that seem to think communism is the only other option. And they seem for the most part to be coming from the USA.
Does having only 2 relevant political parties make people limited in their views and reasoning or something?
Ah, you also think Rockefeller was a great guy because he handed out money? billg's wealth comes from Microsoft's illegal behaviour. Please note that he only started that foundation after he had shitloads of wealth. It is very easy to give away 10% of your wealth when you are wealthy. And do not forget that the Gates foundation is used to hand out computers with (surprise) Windows on it.
Magellan Learning Suite.
Yes, how noble. Microsoft started this initiative when exactly? Oh, that's right, after they feared kids could use computers with other OSes on it, like the initial OLPC. That would cost marketshare, so they responded.
Microsoft "School of the Future" in Philadelphia...I could go on.
Yes, please go on, because these are by no mean indications of doing good. These are all initiatives to get to the goal that billg formulated years ago: Windows (and MS software) on _every_ computer. And baybe to let billg sleep better at night ("All those wasted hours I forced on humanity, all those bluescreens, maybe I can buy that off by giving some of it away).
Anytime MS does something good, the story gets tagged itsatrap.
Would you be so kind as to give examples where Microsoft did something good (as in an ethical and moral point of view that is prevalent in the western world), those deeds were reported on slashdot and tagged itsatrap?
My memory may be bad, but I cannot remember seeing one action from Microsoft that I classified as morally or ethically just. Neutral maybe, and loads and loads of immoral stuff for sure, but good behaviour...
Hmmm, I somehow have this feeling that stating the whole of the web would look significantly different because of a single law in only one of the countries that are on that web, is a bit presumptious.
So Cerf having invented something of value, years ago, makes him a reliable commentator on things political?
No, Cerf having invented something that has grown beyond prediction while still using most of the same protocols, tells me that he is a smart guy and probably knows what he is talking about when it comes to the internet. I would not simply trust his economic ideas or his gardening tips.
This cuts to the core of our problems: responsible exercise of the franchise can't be left to the uninformed. I'm not speaking of Cerf, but of those who would change their minds simply because of his -- or any other -- endorsemen
Could not agree more with you on that one. The USA political system together with the biased media and the fake-isolationist attitude are almost guaranteed to lead to abismal results when it comes to voting capable people into offices. Simply put: US citizens are actively kept stupid.
Would that mean that in the US you would finally be able to see on national television what we here in the Netherlands have been able to see since the 60s (if we want to): naked people?
From around '89 I regularly started using PCs and DOS/Windows. In 98 I started fooling around with linux-based os-es, and only the last 2 years I have been using MacOS X. I never used the classic Mac OS and I know nothing about its internals, but it comes as quite a shock that its security was, as you claim, even worse compared to what MS had to offer at the time. DOS/Win3 effectively had no security, Win95/98 was abysmal. That anything could be worse from a security standpoint (especially an OS that seemed to have quite a loyal and happy following) comes as a surprise.
This is also the first time I hear anything about it; never heard Apple pull ActiveX-like architectural stunts for example. As google seems unable to do so, do you have any decent pointers to articles that give an overview on classic MacOS security issues?
Ah, I mistakenly thought you implied 2000-XP was a much different change than Vista-Win7 shall probably be. Thanks for clearing that up.
(...) there was a substantial (if behind-the-scenes) feature cut while I was there.
It seems your being there did not result in any behavioural change within the company; feature cuts are commonplace within MS before it seems to be able to ship its OSes. Did the feature cut by any chance include that ol' slashdot favorite, winfs?
You are right. I should have written something like "also in a wrong line of reasoning".
And on MacOS: no, I did not misunderstand. Technically, classic MacOS was ditched. From a marketing/non-tech/whatever point of view however, it was a 'new and improved!!!' version which could still run old apps through a compatibility layer. MacOS9 ran Mac apps and only ran on Apple h/w, MacOSX ran those same Mac apps (and new ones) and only ran on Apple h/w. For most end-users it meant that their new Mac was more secure than their old Mac.
My brain is not functioning optimal at this moment, but I hope you get my drift.
I was trying to make clear that it is very hard to state things like [A1 will be M1 when Z1] so far outside of the realm mathematics and logic. In a complex world as ours it is very hard to predict what is going to happen. Should have been more clear there, sorry.
Oh, bull. That's why so much effort was expended on classic MacOS, right?
Well, one could argue that a lot of effort was indeed put into making MacOS more secure. Ditching its core for a *nix core for instance. Steve probably thought that that was the least effort with the best results.
Only assuming roughly equivalent effort required.
Yes, but also there is available knowledge, risks, effect, etc. Cracking ibm.com is far higher profile than cracking aunt Ann's site. Quite complex.
Following that line of reasoning, criminals do not target people with the top 1% income, because they go for the 99%.
Yes, I know, this is nonsense because I left something out. Just like you left out that criminals would most certainly go for the 1% of the 'market', if that part were using an inferior operating system, in terms of security.
If I were a criminal and I would need N machines (say for a botnet) where N
The amount of effort is never the same, just like the potentials gains are never the same. That is very hard to tackle in statistics. Also, folowing your line of reasoning, a lot more Apache servers should have been compromised that IIS servers. History should this to be false. I know, these are not OSes but webservers, but still, I would guess the same reasoning would apply.
I don't want to put you down, but...
"having worked" implies (to me at least) that you are currently not working on that project anymore. You also state that there is both a 2000-XP kind of difference and a refined-version-of-vista difference. That, together with the fact that the product will not ship for some time (if MS' release history is kept in mind one could argue this would not happen before 2010), leaves me unimpressed.
Win7 might change considerably before being released.
Slashdot Mirror
I am afraid that you still not see my point: there is nothing you can do with a registry that you cannot do with a filesystem and config files. A registry just adds complexity.
And I have seen multiple ways of storing data in the registry, so there is no convention that is followed, just like there is no convention that is followed on conf files.
It indeed is standardized; very badly. It only add complexity and solves nothing of the problems caused by conf files.
Easily accessible it is not; I would not try to tell my mother to start regedit32. And from a programmer pov: take a look at the Preferences API in Java. That API is simple, straightforward, and allows for multiple implementations.
All in one place: that is actually a very bad thing. Look at all the troubles the registry has caused, and for so many people, just because the registry is in one place. Single point of failure + Microsoft == headaches. Sometimes you need to change things in your registry for a certain application, but Microsoft warns that any changes to the registry can make the whole OS unstable. In no other modern operating system you are confronted with these complexities. Changing a conf file for an application on a *nix machine will not cause the whole OS to become unstable.
Also look at all the crapware that is able to hide in the registry; all possible because there is no rights management; either you can write in the registry, or you cannot. Application A can always read/write settings for application B. That is very, very bad design.
Again: lookup articles about it: the registry is a p.o.s.
http://www.codinghorror.com/blog/archives/000939.html
http://www.pcworld.com/article/149951/how_to_clean_your_windows_registry_and_speed_up_your_pc.html
I recon that for every registry-fan you can find at least 10 foes. Go to any programmer convention and ask around. I did.
If you really think that the registry is the answer, you did not understand the question.
But that is just my whole point: the registry does not have a strict syntax, in any case no more specific than conf files. You can still write binary data in the registry. There are conventions, sure, but you can have conventions in conf files as well. And 'specifically designed for settings': you do not need a registry for that, you need an API for that. Sun solved that in Java: there is an Preferences API that can use the windows registry as backend implementation, as well as separate files as implementation for that API.
And again: I do not claim that conf files are as nice as sliced bread, I just find the registry the wrong solution for a very real problem. The registry does not solve the problems that conf files have brought us; it just added problems and complexity. A clean and well thought out API could save those problems.
That is why I said 'current need'.
A thousand .conf files spread out throughout the directory structure, each with it's own peculiar syntax. Oh yeah, that worked real well.
A million different keys spread out throughout the registry, each with it's own peculiar syntax. That works equally bad as a filesystem/conf file solution.
I am not saying that conf files are good, they have their problems. What Microsoft has done with the registry was add more complexity, while not addressing the real deficiencies of conf files. In effect, we now have the worst of 2 worlds.
My pleasure. Since my visits to the buddhist centre in Amsterdam I got more and more convinced that you can train your brain just like any other part of your body. Meditation and stuff like this sleep-exercise can do remarkable things for your brain. And the nice thing about buddhists is that even if you state you do not believe in reincarnation but like to try some meditation, they still help you for the full 100%.
Pavlov might come in handy here: every time you go to sleep try to think about the same thing, place yourself in the same situation. Make it the last thing you think about before you really doze off.
Your mind will make a link between these two and probably make it easy to fall asleap when you picture yourself in that specific situation. It might take some time, maybe even a month or two, for the link to sink in.
Be careful to not choose something you can run into in everyday life though; it looks a bit silly if everytime someone offers you a piece of apple pie, you fall asleep.
I also like the theory of the registry. All configuration held in a single, easy to access place with a consistent interface.
Well, we already had such a thing. It is called a filesystem. The registry is nothing more than a hierarchical storage meganism with a standard editor.
It's a filesystem done badly.
I still can't stand Finder, I think it is awful, and curse it every time I need to move a few files to some other folder on another drive (usually I just use "mv")
Pure curiosity: what exactly is it that makes Finder bad at moving files for you? For me it works a lot better than Windows' Explorer.
Everyone born after 1975 hates the RIAA
I am from '71. I wish the ?IA?'s of the world would go away.
convert the film stock to ethanol to power their SUVs
Could be, but you forgot the probable reason: companies want to sell stuff; having a lot of stuff in the public domain could lower demand on new stuff, so the old stuff should be destroyed before it enters the public domain.
Sad.
If we were to build a solar farm in the sahara (550 x 550 km), this would generate enough energy to fill the current need of the whole of humanity.
If at a distance of 10 light years we could detect solar panels and confuse them for an asteroid belt, there are some aliens with quite an energy bill.
i know humanity isn't 851 million years old, not even by the 24th century.
That's right, everybody knows the Earth is only 6000 years old!
*ducks*
It saddens me that every time someone on /. states that capitalism is not the end-all, there are always people that seem to think communism is the only other option. And they seem for the most part to be coming from the USA.
Does having only 2 relevant political parties make people limited in their views and reasoning or something?
Gates Foundation.
Ah, you also think Rockefeller was a great guy because he handed out money? billg's wealth comes from Microsoft's illegal behaviour. Please note that he only started that foundation after he had shitloads of wealth. It is very easy to give away 10% of your wealth when you are wealthy. And do not forget that the Gates foundation is used to hand out computers with (surprise) Windows on it.
Magellan Learning Suite.
Yes, how noble. Microsoft started this initiative when exactly? Oh, that's right, after they feared kids could use computers with other OSes on it, like the initial OLPC. That would cost marketshare, so they responded.
Microsoft "School of the Future" in Philadelphia...I could go on.
Yes, please go on, because these are by no mean indications of doing good. These are all initiatives to get to the goal that billg formulated years ago: Windows (and MS software) on _every_ computer. And baybe to let billg sleep better at night ("All those wasted hours I forced on humanity, all those bluescreens, maybe I can buy that off by giving some of it away).
Anytime MS does something good, the story gets tagged itsatrap.
Would you be so kind as to give examples where Microsoft did something good (as in an ethical and moral point of view that is prevalent in the western world), those deeds were reported on slashdot and tagged itsatrap?
My memory may be bad, but I cannot remember seeing one action from Microsoft that I classified as morally or ethically just. Neutral maybe, and loads and loads of immoral stuff for sure, but good behaviour...
Hmmm, I somehow have this feeling that stating the whole of the web would look significantly different because of a single law in only one of the countries that are on that web, is a bit presumptious.
So Cerf having invented something of value, years ago, makes him a reliable commentator on things political?
No, Cerf having invented something that has grown beyond prediction while still using most of the same protocols, tells me that he is a smart guy and probably knows what he is talking about when it comes to the internet. I would not simply trust his economic ideas or his gardening tips.
This cuts to the core of our problems: responsible exercise of the franchise can't be left to the uninformed. I'm not speaking of Cerf, but of those who would change their minds simply because of his -- or any other -- endorsemen
Could not agree more with you on that one. The USA political system together with the biased media and the fake-isolationist attitude are almost guaranteed to lead to abismal results when it comes to voting capable people into offices. Simply put: US citizens are actively kept stupid.
"What is the Answer to Life, the Universe, and Everything?"
And to any answer that follows: "Could you explain that answer?"
Would that mean that in the US you would finally be able to see on national television what we here in the Netherlands have been able to see since the 60s (if we want to): naked people?
I am a bit shocked.
From around '89 I regularly started using PCs and DOS/Windows. In 98 I started fooling around with linux-based os-es, and only the last 2 years I have been using MacOS X. I never used the classic Mac OS and I know nothing about its internals, but it comes as quite a shock that its security was, as you claim, even worse compared to what MS had to offer at the time. DOS/Win3 effectively had no security, Win95/98 was abysmal. That anything could be worse from a security standpoint (especially an OS that seemed to have quite a loyal and happy following) comes as a surprise.
This is also the first time I hear anything about it; never heard Apple pull ActiveX-like architectural stunts for example. As google seems unable to do so, do you have any decent pointers to articles that give an overview on classic MacOS security issues?
(...) there was a substantial (if behind-the-scenes) feature cut while I was there.
It seems your being there did not result in any behavioural change within the company; feature cuts are commonplace within MS before it seems to be able to ship its OSes. Did the feature cut by any chance include that ol' slashdot favorite, winfs?
You are right. I should have written something like "also in a wrong line of reasoning".
And on MacOS: no, I did not misunderstand. Technically, classic MacOS was ditched. From a marketing/non-tech/whatever point of view however, it was a 'new and improved!!!' version which could still run old apps through a compatibility layer. MacOS9 ran Mac apps and only ran on Apple h/w, MacOSX ran those same Mac apps (and new ones) and only ran on Apple h/w. For most end-users it meant that their new Mac was more secure than their old Mac.
My brain is not functioning optimal at this moment, but I hope you get my drift.
I was trying to make clear that it is very hard to state things like [A1 will be M1 when Z1] so far outside of the realm mathematics and logic. In a complex world as ours it is very hard to predict what is going to happen. Should have been more clear there, sorry.
Oh, bull. That's why so much effort was expended on classic MacOS, right?
Well, one could argue that a lot of effort was indeed put into making MacOS more secure. Ditching its core for a *nix core for instance. Steve probably thought that that was the least effort with the best results.
Only assuming roughly equivalent effort required.
Yes, but also there is available knowledge, risks, effect, etc. Cracking ibm.com is far higher profile than cracking aunt Ann's site. Quite complex.
Following that line of reasoning, criminals do not target people with the top 1% income, because they go for the 99%.
Yes, I know, this is nonsense because I left something out. Just like you left out that criminals would most certainly go for the 1% of the 'market', if that part were using an inferior operating system, in terms of security.
If I were a criminal and I would need N machines (say for a botnet) where N
The amount of effort is never the same, just like the potentials gains are never the same. That is very hard to tackle in statistics. Also, folowing your line of reasoning, a lot more Apache servers should have been compromised that IIS servers. History should this to be false. I know, these are not OSes but webservers, but still, I would guess the same reasoning would apply.
I don't want to put you down, but...
"having worked" implies (to me at least) that you are currently not working on that project anymore. You also state that there is both a 2000-XP kind of difference and a refined-version-of-vista difference. That, together with the fact that the product will not ship for some time (if MS' release history is kept in mind one could argue this would not happen before 2010), leaves me unimpressed.
Win7 might change considerably before being released.
Do you really want to imply that 1% of a really, really huge market is not interesting enough for amoral types and criminals?