Seriously, do you think a run-time profiling algorithm can do better than a proficient human with a profiling tool?
Ok, Mr. Proficient Human With Profiling Tool, quickly state how you'll optimize your program for each and every input it will ever see so that those optimizations will be coded intop the program.
You see, runtime profiling has the ability to examine the state of the program as its running and respond to changing conditions with chaning optimizations.
Now, perhaps you mean a C++ programmer could write code into his program that monitors various values in the program and then starts swapping function pointers around to use function implementations that are more optimal for the state. If that is your point, then go have fun with that!
If you'rebenchmarking "Java" against "C++", you need to define what each of those are. YOu say HotSpot profiling is unfair unless the C++ compiler cna also be allowed to use profiling data. But if "Java" means "A Java program running on Sun's HotSpot VM" (after all, a Java program is useless without a VM so it is reasonable to assume "Java" includes a VM), then that is part of "Java". Is it part of "C++"? It's not Java's fault that "C++" doesn't include runtime profiling.
That's simply not true if you expect benchmarks to reflect real world applications. Sure, it might be true for trivial utilities but for applications that stay alive for long periods of time chugging along, which is probably the majority of the worlds useful applications, it makes the results invalid. Or, at a minimum, sets greatly unrealistic expectations.
No, it is true. THink about. What application do you know of that uses 100% of the CPU? Consider a GUI which spends most of its time waiting for user input. BY deferring memory cleanup until the garbage collector feels like doing it, you can make the GUI respond quicker because it's not watsing valuable repaint time cleaning up Objects as temporary objects outlive their usefulness. Instead, the GC can run while the human is reading the results just painted to the screen.
A GUI is an extreme example. But consider anytime the processor can't continute, like when it's blocked on I/O or something. These are the same principles that allow multi-process, multi-thread systems to work better.
Debunked? I don't know. Certainly shows that with hand-tuning, C++ can be faster. BUt again, as I brought up in another post in this topic, what is C++? Is it the language? Is it the "standard" libraries? I mean, what the benchmark showed for String hashign is that Java is faster. But that seems to only be because the standard hash routine chosen (and forgive me if it was standard and correct me) for C++ is slow. But if one were to write C++ programs, should they expect to have to start throwing out parts of the standar dlibraries in favor of alternativees because the standard sucks so bad? I suppose one could throw out Java's implementation for one linked through JNI to an ASM-implemented String hashing routine. Or what if Java standard were updated to include a VM instruction for hashing Strings so that that instruction could be optimized in the VMs dpeloyed to various platforms?
The benchmark is junk because the benchmarkere didn't systematically state -- or even understand -- hwta it is they were testing. In most if not all case, it was apples ot oranges.
And FireFox is guaranteed free of unknown exploits? The very nature of unknown exploits is that they're unknown. And just because its opens ource doesn't mean they won't happen. They may be less likely, but even Apache HTTPD has had exploits in it.
Then why are you using mouse at all? All of the cool kids are using hotkeys
Actually, I use both. And you know Microsoft is decent with hot keys. Not every third party app for Windows is, but Microsoft makes sure they are. I would say that 80% of the daily functions I do I kow both keyboard shortcuts and mouse clicks for and I use whichever my hands are already doing at the time. BUt, when my hand is on the mouse, I use aas few clicks as possible. That's why the "Start Menu" is evil in Windows. But I use that as a sort of master listing. All of my frequently used programs are assigned hotkeys or kept in taskbar pop-outs/quicklaunch menus for access with far fewer clicks.
But still, to me, convenience of syncing a single file across the network is alone worth the trouble.
I still don't see it. Even when I "backup" my favorites across the network, I right-click the Favorites folder, go to a network share, right-click, paste (note, in honor of our previous comments, I could just as easily use CTRL+C and CTRL+V hotkeys:)). Now when I copy that folder, everything underneath it comes.
Now, I will give you this. When you copy a file, it replaces the destination. When you copy a folder, it typically "merges" the destination -- otherwise, you'd have to delete the destination first.
As for file management to manage the bookmarks...
Yes, you're right, it does require shell integrationt o be able to fully edit bookmarks. And that is where the OS integration cam ein handy for MS, I suppose. But I also have shell integrationf or editing MP3 ID3 tags. And as far as more common editing tasks -- renaming, reorganizing, etc. -- I can do that in the naked shellwithout any integrations. As someone else pointed out, in Unix, everything is a file... except bookmarks!
Incidentally, are there any "contact management" systems in *nix that store individual contacts in idividual files (probably not vCards since they're designed for portability, not speed)?
But to me, Firefox is good for browsing the Web
Honestly, I think IE too is fine for browsing around the web. It ain't slow (people kept saying FireFox is fast, but I didn't notice when I tried it). It renders everys ite I visit just fine.
There are good tools for every task. I could do all three from a web browser
I think this is your best point. Honestly, though, in the Microsoft Widows world, most "good solutions" end up getting commercialized. Since I hate programming in Visual Dot Crap ++ # anything (I don't have the flipping IDE, how can I do it?), I like the fact that using Windows Scripting Host (a brilliant idea), or HTML pages with JavaScript, I can extend my shell. YEs, it would be nice to have a dedicated dictionary text field tool on myh taskbar, but I ain't registering $9.95 for it wehen I can code it myself to hot dictionary.com from HTML embedded in my taskbar. Perhaps it's my only bit of freedom in MS-land.
It's not that I'm outsmarting a hacker, it's that I'm not inconveniencing myself. Look, let's say I enable a firewall. Now, I open port 80 so my Apache HTTPD server can be reached by the outside world. Then, someone hacker finds a hole in Apache HTTPD, hits my box on port 80 right through my firewall,a nd takes over my machine. Now, what did that firewall do for me? The very services that might be "hacked" on my machine -- HTTP, Telnet, etc. -- could be hacked through a firewall because I'd have them open! The other services Microsoft turns on my default I turn off for precisely security reasons.
Apples? Oranges? How *should* we comapre them?
on
Java Faster Than C++?
·
· Score: 1
I think everyone has pointed out in many ways why these tests and comparisons were flawed. Much of it boils down to deciding what exactly "C++" is and what exactly "Java" is. IN fact, I don't think either of those terms is even specific enough to build benchmarks around.
With C++, there is the language syntax, the standard libraries, which compiler you use, which optimizaitons you chose, etc. Within Java, there is the version, the libaries, the VM vendor, the VM optimizations, etc. And then you have to consider that some tests might perform better in one of the languages on one platform but the reverse on another -- where the platform might be OS, hardware, or both.
And then how you measure "performance"? The first execution? Subsequent executions? If you ask a business manager, they might include time-to-develop as a performance measurement. OR howw about time spent mediating the religious war started in an IT department over which language to use???
Java's implementation of hashCode() for String could vary, as I understand it, by language implementor (i.e. Sun vs. IBM). Regardless, I looked in Sun's JDK 1.4.2 source code. Here's the basic algorithm:
1. If the stored hashcode is not zero, return it 2. get local vairable copies of the member variables (I'm assuming these are trivial stack allocations) 3. Looping index from 0 to length, do...
4. hashcode = 31*hashcode + character at index 5. Store hashcode 6. return hashcode
As I read it, steps 3 and 4 are the meat of this algorithm, and looks equivalent to the C++ version the grandparent post had. I'm nore sure how much the checking/storage of the hashcode would impact things -- unless your hashcode happened to be the unlucky zero in which case you'd be screwed! But to have 0 hashcode, you're string would have to be empty (in which case the loop would terminate quickly) or every character would have to be '\0' which is probably not likely.
Show me your white hat first! When I run portscans, I do so with the perceived trust that the one scanning my ports is trustworthy. Iknow nothing about you. Besides, if you were good, you coudl already find my IP.
You're seriously naive if you think that IE is in any way secure by default, or secure when patched up
Why is it naive? Is IE listening on some ports for people to hijack? Did someone hijack slashdot such that click ing on a link in it will do something malicious? Am I going to naively install some ActiveX control?
You win. I found that same thing in Opera. BUt I had to play devil's advocate! The truth is, though, I did go tabbed and I did go back. The tabs alone just weren't enough the other suffering I'm getting flamed for elsewhere in this discussion:)
As I replied in another thread, I couldn't do what I wanted.
I can cleverly pick necessary stuff so that there's no need for anything to be hidden behind pop-out anything...
Like many others, I've found that every click counts. I have several top-level things one-click away, and collapse some secondary items under pop-outs so they're effectively two clicks away. And I can easilyr e-size or re-arrange toolbars to my liking if my needs change. It's one of those things I would've laughed at as too trivial to care abotu until I realized how much easier it made things.
That was pure pain on the FAT16 days, please don't remind me. Millions of files eating away the precious diskspace.
First, if you're using FAT16, you've got other problems. Second, do your eally have millions of bookmarks? Even thousands? And how much disk space is that wasting? More or less than your pr0n collection? Seriously, are you suggesting that the disk-space efficiency of tiny little bookmarks is a concern for anyone? Not that disk space is something trivial, because it is the scale that matters -- but even this will never scale to a worrisome level.
Gimmicks. All gimmicks
AHh, now you sound like a talk radio host who waves his hand at something he doesn't understand or would rather not touch. If you actually sat down and tried this without bias (which is probably an impossibility for you from here on out), you might like it. BUt you've already written it off as a gimmick. Know what else were gimmicks? Mouse gestures. Toolbars. WIndow managers. GUIs in general. High-level languages. The abacus. How far back are you?
You don't need to melt everything into an unholy combination
You are exactly right. You don't have to. Even I reserve the right not to use any unholy union I don't want to. I could switch to FireFox if I wanted to. But, you see, then I would have two browsers on my system because IE would still be integrated into the OS for the things I use it for. Now, if MS would allow me to plug in any old browser for the OS-integrated part (which I still think would be a grand idea), I might do it in a heartbeat. BUt the question is, would the Mozilla/FireFox zealots ever write an adapter to such an API... or would that still be another unholy combination?
#1. I did try re-arranging them. I couldn't re-arrange them as I wanted. Two on same horiztonal row? Shrinking one so the overflow is turned into a popout button? I have the same complaint for ThunderBird, btu it's not as big of a problemf or me personally there as it is in my web browser.
#2. See #1.
#3. I know. I read them. They didn't have everything I use in teh real Google Toolbar.
#4. Bookmarks can't be used as files. Thus, everywhere I've taken advantage of my favorites being files and folders I wouldn't be able to use bookmarks. Hard-linking subfolders to other locations. Using pop-out style quick menus out of folder son taskbar. And so on.
So if I enable, say, Microsoft's XP firewall, am I ok? It is, after all, another XP firewall. Or how about some of the NetGear hardware firewalls. Hasn't NetGear also had security problems as of late (I only have a dum NetGear hub, so it's not applicable there).
I opccasioanlly scan my ports. I check my processes. I run cirus scans, spyware scans, etc. I've never found anything. My wife's machine is another story, but those spyware's didn't come in through security holes -- just unsecure humans (kids) clicking 'yes' to anything that got themto the cool website their friends told them to check out. And even then, a firewall wouldn't have helped. Yeah, I could run FireFox on it to stop ActiveX crap from getting in, but we were talking purely about firewalls here.
Agreed! This is the one thing where Microsoft didn't crap stuff into the registry, a proprietary XML format, or anything else. Just simple, individual files.
Probably not as space-efficient as keeping them in one HTML file, but I doubt anyone is complaining that their bookmarks are their primary consumer of diskspace.
Seeing as how simple and straight forward this is, hwo hard would it be to build an extension for Mozilla/FireFox to support MS's *.url files and folders, or fi not, some new, open URL format (I prefer the first because it would make it easier for people to migrate from IE to Mozilla/FireFox).
And that is precisely the reason somebody can hack your windows machine using IE.
How? Seriously. It's not been done yet. I don't have a firewall, so it should be easy. PUt your white hat on, hack my machine, and show me the proof when it's all done. Maybe that will be my wakeup call to finally secure my system with FireFox.
I started using Opera couple of years ago. The mouse geastures were nice. The tabs were nice. The little HTML panels you could drop in were nice too. In the end, it had problems rendering many of the sites I used. I also started to take advantage of the IE-XP integration (you know, the vil OS integration Microsoft got in hot water for).
And even when I did use Opera with tabs, I often would start separate instances of Opera so all of my tabs wouldn't be clumped together. Sometimes I have the weather open in one window to watch the coming storms while I have another window open to an API reference. There is no reason for those to be crammed into the same tabbed window.
And that is the mindset that has gotten me to appreciate NON-tabbed windows. In IE, I can have browser windows fully distinguishable in the taskbar (so I can pick the window I want with one click instead of one to get the browser window and another to get the right tab). If I wanted them to collapse, I could use XP's generic task collapsing. But, as I stated above, I like to be able to see one taskbar button for one task. To be, the fact that two pieces of content are delivered via the same protocol and markup language does not mean I necessarily want them in the same window.
So, are you gonna keep preaching about Microsoft and them not being able to buy their way out, or are you going to try and "get it". I know Open Source *could* be better if we all stopped thinking like theorists and mixed in a little practicality all the way. We're blinding ourselves with our own bullshit!
I installed FireFox just a few days ago, wne to SlashDot and comapred it to IE. A little different. Odd, I thought. The greeen header displaying my username was contacting the greenpart of the top-most article. Then I increased my font-size one notch with the scrool-wheel and it was now separated by the gap I'm accustomed to. Then I reduced my font-size back down one notch to try and reproduce it. But the gap was still there. Did I scroll too far? I closed FireFox and started again from scratch. Again, no-gap then the gap and then the gap stays.
I just recently decided to give FireFox a try on my XP box after all the type about it here on SlashDot. I had high hopes after switching to Mozilla Thunderbird recently.
However, I was severely disappointed. It didn't do what I can do with IE. Look, you may remember that MS got in some hot water for "integrating" the browser into the OS. We all said how evil that was. But, you know what? I've taken advantage of it!
Here's a breif list of the things I like about IE over FireFox (and if I'm wrong about being able to do these in FireFox, please correct me):
1. I can re-arrange the toolbars to my satisfaction.
2. I can cleverly size my toolbars such that extra items are hidden behind a pop-out button so I can effectively make quick 1-click menues.
3. The unbeatable *real* Google Toolbar
4. The Favorites are arranged as files & folders so I can manipulate them easily (i.e. but them as pop-outs on my taskbar, make hard-links to subfolders in them in other logical locations)
5. I can embed HTML in my TaskBar to accomplish all sorts of useful things (wallet-size photoalbum, dictuinary, phone number lookup, etc.)
My browser is not just a browser. It bleeds into my operating system and vice-versa. I'm not blindly pro-Microsoft, I just happen to take advantage of the integration Microsoft chose to thrust upon us.
Honestly, I wish the "browser-integration" API were documented so Mozilla couled wholly replace the IE integration in XP. But to not have it at all is a big hole -- at least to someone who has become accustomed to the convenience it offers.
The first problem I would come up with is: the MS Word......By the way, openoffice 1.1.1 with python (pyUNO) looks promising for the above problem
This is the exact same problem that keeps resurfacing in these postings about Windows bashing. Microsoft Word has nothing to do with Windows in this context. I don't use Word. Apparently, neither do you. However, I do use Windows, but you don't.
I run OpenOffice.org on Windows. And you can run Python on Windows too. So, what is it about the example you cited that is intended to be an example of why you don't use Windows?
I stopped using MS Word because I found a very decent alternative in OpenOffice.org (same going for spreadsheet and presentations). I hope to keep finding more open source alternatives for other prgrams as I go on (i.e. I switched to Palm Desktop + Mozilla Thunderbird, away from Outlook, a few months ago and have marginally survived). I'm also running Apache + TomCat instead of IIS + ASP. BUt I still use all of that on Windows.
sent via a broadband IP address, I refer to it in my spam complaints to ISP's
So you're the kind of punk that blindly condemns every broadband address to be SPAM making it increasingly difficult for me to run my own, home SMTP server?
Did you also know that most spyaware programs are capable of "phoning home" when the PC connects so that even dial-up can be used? Sure, dial-up is slow, but e-mail is relatively lightweight and even if one sends only a couple of messages, multiply that by thousands of zombie dial-up PCs and you've got something.
TO this day,I can't e-mail my father on his RR account from my SMTP server because RR blocks mail it. Their failure messages direct me top a web page that enumerate severla possibilities. Attempts to contact them for exact details have been unanswered. So, I have to still keep my crappy ISP's SMTP server handy so I can use it occasionally.
The probelm you've just described isn't about the possibility of control, it's about exercising it. If you willy-nilly install crap on Linux and don't keep up with it, I'm sure your box could get out of control too (fortunately, no evil marketer has decided to write a spyware program for Linux that I know of yet, but if Linux on the desktoop such as Lindows ever does catch on -- they will!).
What you described however is exercising your control. You clearly demonstrated in your post that you exercised your control over Windows on behalf of the novice Windows user innundated with spyware and trojans. That same knowledge and mindset you have to take, keep and use that control is what allows your to also administer a Linux box.
Let it be clear, you *can* control Windows if you want to.
"Oh well, can't do anything about it". (and you can't you know.)
Oh please! Can't do anything about it? Have you ever looked at the registry? We're you too intimidated? It offers nice things like the fact that it's explicitly structured as opposed to an open-ended text file. HEll, I once used a Windows program to manually fix the NTFS partition table that was corrupted by a bad driver (From a 3rd party, not Microsoft).
Just because you couldn't do anything about it doesn't mean something can't be done abotu it.
If anything, the advantage of Linux is that it forces you to understand how the system operates so you can fix things whereas in Windows you can get pretty far knowing nothing unitl something breaks -- then you're screwed. But if you are not afraid to try and fix things, it can be done.
Slashdot Mirror
Seriously, do you think a run-time profiling algorithm can do better than a proficient human with a profiling tool?
Ok, Mr. Proficient Human With Profiling Tool, quickly state how you'll optimize your program for each and every input it will ever see so that those optimizations will be coded intop the program.
You see, runtime profiling has the ability to examine the state of the program as its running and respond to changing conditions with chaning optimizations.
Now, perhaps you mean a C++ programmer could write code into his program that monitors various values in the program and then starts swapping function pointers around to use function implementations that are more optimal for the state. If that is your point, then go have fun with that!
If you'rebenchmarking "Java" against "C++", you need to define what each of those are. YOu say HotSpot profiling is unfair unless the C++ compiler cna also be allowed to use profiling data. But if "Java" means "A Java program running on Sun's HotSpot VM" (after all, a Java program is useless without a VM so it is reasonable to assume "Java" includes a VM), then that is part of "Java". Is it part of "C++"? It's not Java's fault that "C++" doesn't include runtime profiling.
I'd write pooling code to handle the objects instead
Actually, object pooling may hurt!
That's simply not true if you expect benchmarks to reflect real world applications. Sure, it might be true for trivial utilities but for applications that stay alive for long periods of time chugging along, which is probably the majority of the worlds useful applications, it makes the results invalid. Or, at a minimum, sets greatly unrealistic expectations.
No, it is true. THink about. What application do you know of that uses 100% of the CPU? Consider a GUI which spends most of its time waiting for user input. BY deferring memory cleanup until the garbage collector feels like doing it, you can make the GUI respond quicker because it's not watsing valuable repaint time cleaning up Objects as temporary objects outlive their usefulness. Instead, the GC can run while the human is reading the results just painted to the screen.
A GUI is an extreme example. But consider anytime the processor can't continute, like when it's blocked on I/O or something. These are the same principles that allow multi-process, multi-thread systems to work better.
Debunked? I don't know. Certainly shows that with hand-tuning, C++ can be faster. BUt again, as I brought up in another post in this topic, what is C++? Is it the language? Is it the "standard" libraries? I mean, what the benchmark showed for String hashign is that Java is faster. But that seems to only be because the standard hash routine chosen (and forgive me if it was standard and correct me) for C++ is slow. But if one were to write C++ programs, should they expect to have to start throwing out parts of the standar dlibraries in favor of alternativees because the standard sucks so bad? I suppose one could throw out Java's implementation for one linked through JNI to an ASM-implemented String hashing routine. Or what if Java standard were updated to include a VM instruction for hashing Strings so that that instruction could be optimized in the VMs dpeloyed to various platforms?
The benchmark is junk because the benchmarkere didn't systematically state -- or even understand -- hwta it is they were testing. In most if not all case, it was apples ot oranges.
And FireFox is guaranteed free of unknown exploits? The very nature of unknown exploits is that they're unknown. And just because its opens ource doesn't mean they won't happen. They may be less likely, but even Apache HTTPD has had exploits in it.
Then why are you using mouse at all? All of the cool kids are using hotkeys
Actually, I use both. And you know Microsoft is decent with hot keys. Not every third party app for Windows is, but Microsoft makes sure they are. I would say that 80% of the daily functions I do I kow both keyboard shortcuts and mouse clicks for and I use whichever my hands are already doing at the time. BUt, when my hand is on the mouse, I use aas few clicks as possible. That's why the "Start Menu" is evil in Windows. But I use that as a sort of master listing. All of my frequently used programs are assigned hotkeys or kept in taskbar pop-outs/quicklaunch menus for access with far fewer clicks.
But still, to me, convenience of syncing a single file across the network is alone worth the trouble.
:)). Now when I copy that folder, everything underneath it comes.
I still don't see it. Even when I "backup" my favorites across the network, I right-click the Favorites folder, go to a network share, right-click, paste (note, in honor of our previous comments, I could just as easily use CTRL+C and CTRL+V hotkeys
Now, I will give you this. When you copy a file, it replaces the destination. When you copy a folder, it typically "merges" the destination -- otherwise, you'd have to delete the destination first.
As for file management to manage the bookmarks...
Yes, you're right, it does require shell integrationt o be able to fully edit bookmarks. And that is where the OS integration cam ein handy for MS, I suppose. But I also have shell integrationf or editing MP3 ID3 tags. And as far as more common editing tasks -- renaming, reorganizing, etc. -- I can do that in the naked shellwithout any integrations. As someone else pointed out, in Unix, everything is a file... except bookmarks!
Incidentally, are there any "contact management" systems in *nix that store individual contacts in idividual files (probably not vCards since they're designed for portability, not speed)?
But to me, Firefox is good for browsing the Web
Honestly, I think IE too is fine for browsing around the web. It ain't slow (people kept saying FireFox is fast, but I didn't notice when I tried it). It renders everys ite I visit just fine.
There are good tools for every task. I could do all three from a web browser
I think this is your best point. Honestly, though, in the Microsoft Widows world, most "good solutions" end up getting commercialized. Since I hate programming in Visual Dot Crap ++ # anything (I don't have the flipping IDE, how can I do it?), I like the fact that using Windows Scripting Host (a brilliant idea), or HTML pages with JavaScript, I can extend my shell. YEs, it would be nice to have a dedicated dictionary text field tool on myh taskbar, but I ain't registering $9.95 for it wehen I can code it myself to hot dictionary.com from HTML embedded in my taskbar. Perhaps it's my only bit of freedom in MS-land.
It's not that I'm outsmarting a hacker, it's that I'm not inconveniencing myself. Look, let's say I enable a firewall. Now, I open port 80 so my Apache HTTPD server can be reached by the outside world. Then, someone hacker finds a hole in Apache HTTPD, hits my box on port 80 right through my firewall,a nd takes over my machine. Now, what did that firewall do for me? The very services that might be "hacked" on my machine -- HTTP, Telnet, etc. -- could be hacked through a firewall because I'd have them open! The other services Microsoft turns on my default I turn off for precisely security reasons.
I think everyone has pointed out in many ways why these tests and comparisons were flawed. Much of it boils down to deciding what exactly "C++" is and what exactly "Java" is. IN fact, I don't think either of those terms is even specific enough to build benchmarks around.
With C++, there is the language syntax, the standard libraries, which compiler you use, which optimizaitons you chose, etc. Within Java, there is the version, the libaries, the VM vendor, the VM optimizations, etc. And then you have to consider that some tests might perform better in one of the languages on one platform but the reverse on another -- where the platform might be OS, hardware, or both.
And then how you measure "performance"? The first execution? Subsequent executions? If you ask a business manager, they might include time-to-develop as a performance measurement. OR howw about time spent mediating the religious war started in an IT department over which language to use???
Java's implementation of hashCode() for String could vary, as I understand it, by language implementor (i.e. Sun vs. IBM). Regardless, I looked in Sun's JDK 1.4.2 source code. Here's the basic algorithm:
1. If the stored hashcode is not zero, return it
2. get local vairable copies of the member variables (I'm assuming these are trivial stack allocations)
3. Looping index from 0 to length, do...
4. hashcode = 31*hashcode + character at index
5. Store hashcode
6. return hashcode
As I read it, steps 3 and 4 are the meat of this algorithm, and looks equivalent to the C++ version the grandparent post had. I'm nore sure how much the checking/storage of the hashcode would impact things -- unless your hashcode happened to be the unlucky zero in which case you'd be screwed! But to have 0 hashcode, you're string would have to be empty (in which case the loop would terminate quickly) or every character would have to be '\0' which is probably not likely.
What's your IP?
Show me your white hat first! When I run portscans, I do so with the perceived trust that the one scanning my ports is trustworthy. Iknow nothing about you. Besides, if you were good, you coudl already find my IP.
You're seriously naive if you think that IE is in any way secure by default, or secure when patched up
Why is it naive? Is IE listening on some ports for people to hijack? Did someone hijack slashdot such that click ing on a link in it will do something malicious? Am I going to naively install some ActiveX control?
You win. I found that same thing in Opera. BUt I had to play devil's advocate! The truth is, though, I did go tabbed and I did go back. The tabs alone just weren't enough the other suffering I'm getting flamed for elsewhere in this discussion :)
Can do. View - Toolbars - Customize
As I replied in another thread, I couldn't do what I wanted.
I can cleverly pick necessary stuff so that there's no need for anything to be hidden behind pop-out anything...
Like many others, I've found that every click counts. I have several top-level things one-click away, and collapse some secondary items under pop-outs so they're effectively two clicks away. And I can easilyr e-size or re-arrange toolbars to my liking if my needs change. It's one of those things I would've laughed at as too trivial to care abotu until I realized how much easier it made things.
That was pure pain on the FAT16 days, please don't remind me. Millions of files eating away the precious diskspace.
First, if you're using FAT16, you've got other problems. Second, do your eally have millions of bookmarks? Even thousands? And how much disk space is that wasting? More or less than your pr0n collection? Seriously, are you suggesting that the disk-space efficiency of tiny little bookmarks is a concern for anyone? Not that disk space is something trivial, because it is the scale that matters -- but even this will never scale to a worrisome level.
Gimmicks. All gimmicks
AHh, now you sound like a talk radio host who waves his hand at something he doesn't understand or would rather not touch. If you actually sat down and tried this without bias (which is probably an impossibility for you from here on out), you might like it. BUt you've already written it off as a gimmick. Know what else were gimmicks? Mouse gestures. Toolbars. WIndow managers. GUIs in general. High-level languages. The abacus. How far back are you?
You don't need to melt everything into an unholy combination
You are exactly right. You don't have to. Even I reserve the right not to use any unholy union I don't want to. I could switch to FireFox if I wanted to. But, you see, then I would have two browsers on my system because IE would still be integrated into the OS for the things I use it for. Now, if MS would allow me to plug in any old browser for the OS-integrated part (which I still think would be a grand idea), I might do it in a heartbeat. BUt the question is, would the Mozilla/FireFox zealots ever write an adapter to such an API... or would that still be another unholy combination?
#1. I did try re-arranging them. I couldn't re-arrange them as I wanted. Two on same horiztonal row? Shrinking one so the overflow is turned into a popout button? I have the same complaint for ThunderBird, btu it's not as big of a problemf or me personally there as it is in my web browser.
#2. See #1.
#3. I know. I read them. They didn't have everything I use in teh real Google Toolbar.
#4. Bookmarks can't be used as files. Thus, everywhere I've taken advantage of my favorites being files and folders I wouldn't be able to use bookmarks. Hard-linking subfolders to other locations. Using pop-out style quick menus out of folder son taskbar. And so on.
#5. My taskbar is here:
http://www.trinition.org/taskbar.PNG
So if I enable, say, Microsoft's XP firewall, am I ok? It is, after all, another XP firewall. Or how about some of the NetGear hardware firewalls. Hasn't NetGear also had security problems as of late (I only have a dum NetGear hub, so it's not applicable there).
I opccasioanlly scan my ports. I check my processes. I run cirus scans, spyware scans, etc. I've never found anything. My wife's machine is another story, but those spyware's didn't come in through security holes -- just unsecure humans (kids) clicking 'yes' to anything that got themto the cool website their friends told them to check out. And even then, a firewall wouldn't have helped. Yeah, I could run FireFox on it to stop ActiveX crap from getting in, but we were talking purely about firewalls here.
Agreed! This is the one thing where Microsoft didn't crap stuff into the registry, a proprietary XML format, or anything else. Just simple, individual files.
Probably not as space-efficient as keeping them in one HTML file, but I doubt anyone is complaining that their bookmarks are their primary consumer of diskspace.
Seeing as how simple and straight forward this is, hwo hard would it be to build an extension for Mozilla/FireFox to support MS's *.url files and folders, or fi not, some new, open URL format (I prefer the first because it would make it easier for people to migrate from IE to Mozilla/FireFox).
And that is precisely the reason somebody can hack your windows machine using IE.
How? Seriously. It's not been done yet. I don't have a firewall, so it should be easy. PUt your white hat on, hack my machine, and show me the proof when it's all done. Maybe that will be my wakeup call to finally secure my system with FireFox.
Once you go tabbed you'll never go back.
I went tabbed. I went back.
I started using Opera couple of years ago. The mouse geastures were nice. The tabs were nice. The little HTML panels you could drop in were nice too. In the end, it had problems rendering many of the sites I used. I also started to take advantage of the IE-XP integration (you know, the vil OS integration Microsoft got in hot water for).
And even when I did use Opera with tabs, I often would start separate instances of Opera so all of my tabs wouldn't be clumped together. Sometimes I have the weather open in one window to watch the coming storms while I have another window open to an API reference. There is no reason for those to be crammed into the same tabbed window.
And that is the mindset that has gotten me to appreciate NON-tabbed windows. In IE, I can have browser windows fully distinguishable in the taskbar (so I can pick the window I want with one click instead of one to get the browser window and another to get the right tab). If I wanted them to collapse, I could use XP's generic task collapsing. But, as I stated above, I like to be able to see one taskbar button for one task. To be, the fact that two pieces of content are delivered via the same protocol and markup language does not mean I necessarily want them in the same window.
So, are you gonna keep preaching about Microsoft and them not being able to buy their way out, or are you going to try and "get it". I know Open Source *could* be better if we all stopped thinking like theorists and mixed in a little practicality all the way. We're blinding ourselves with our own bullshit!
PLEASE!
I installed FireFox just a few days ago, wne to SlashDot and comapred it to IE. A little different. Odd, I thought. The greeen header displaying my username was contacting the greenpart of the top-most article. Then I increased my font-size one notch with the scrool-wheel and it was now separated by the gap I'm accustomed to. Then I reduced my font-size back down one notch to try and reproduce it. But the gap was still there. Did I scroll too far? I closed FireFox and started again from scratch. Again, no-gap then the gap and then the gap stays.
Is *that* part of the standard???
I just recently decided to give FireFox a try on my XP box after all the type about it here on SlashDot. I had high hopes after switching to Mozilla Thunderbird recently.
However, I was severely disappointed. It didn't do what I can do with IE. Look, you may remember that MS got in some hot water for "integrating" the browser into the OS. We all said how evil that was. But, you know what? I've taken advantage of it!
Here's a breif list of the things I like about IE over FireFox (and if I'm wrong about being able to do these in FireFox, please correct me):
1. I can re-arrange the toolbars to my satisfaction.
2. I can cleverly size my toolbars such that extra items are hidden behind a pop-out button so I can effectively make quick 1-click menues.
3. The unbeatable *real* Google Toolbar
4. The Favorites are arranged as files & folders so I can manipulate them easily (i.e. but them as pop-outs on my taskbar, make hard-links to subfolders in them in other logical locations)
5. I can embed HTML in my TaskBar to accomplish all sorts of useful things (wallet-size photoalbum, dictuinary, phone number lookup, etc.)
My browser is not just a browser. It bleeds into my operating system and vice-versa. I'm not blindly pro-Microsoft, I just happen to take advantage of the integration Microsoft chose to thrust upon us.
Honestly, I wish the "browser-integration" API were documented so Mozilla couled wholly replace the IE integration in XP. But to not have it at all is a big hole -- at least to someone who has become accustomed to the convenience it offers.
The first problem I would come up with is: the MS Word... ...By the way, openoffice 1.1.1 with python (pyUNO) looks promising for the above problem
This is the exact same problem that keeps resurfacing in these postings about Windows bashing. Microsoft Word has nothing to do with Windows in this context. I don't use Word. Apparently, neither do you. However, I do use Windows, but you don't.
I run OpenOffice.org on Windows. And you can run Python on Windows too. So, what is it about the example you cited that is intended to be an example of why you don't use Windows?
I stopped using MS Word because I found a very decent alternative in OpenOffice.org (same going for spreadsheet and presentations). I hope to keep finding more open source alternatives for other prgrams as I go on (i.e. I switched to Palm Desktop + Mozilla Thunderbird, away from Outlook, a few months ago and have marginally survived). I'm also running Apache + TomCat instead of IIS + ASP. BUt I still use all of that on Windows.
sent via a broadband IP address, I refer to it in my spam complaints to ISP's
So you're the kind of punk that blindly condemns every broadband address to be SPAM making it increasingly difficult for me to run my own, home SMTP server?
Did you also know that most spyaware programs are capable of "phoning home" when the PC connects so that even dial-up can be used? Sure, dial-up is slow, but e-mail is relatively lightweight and even if one sends only a couple of messages, multiply that by thousands of zombie dial-up PCs and you've got something.
TO this day,I can't e-mail my father on his RR account from my SMTP server because RR blocks mail it. Their failure messages direct me top a web page that enumerate severla possibilities. Attempts to contact them for exact details have been unanswered. So, I have to still keep my crappy ISP's SMTP server handy so I can use it occasionally.
What would be nice is a console where I could enter WSH instructions at a command-like prompt.
The probelm you've just described isn't about the possibility of control, it's about exercising it. If you willy-nilly install crap on Linux and don't keep up with it, I'm sure your box could get out of control too (fortunately, no evil marketer has decided to write a spyware program for Linux that I know of yet, but if Linux on the desktoop such as Lindows ever does catch on -- they will!).
What you described however is exercising your control. You clearly demonstrated in your post that you exercised your control over Windows on behalf of the novice Windows user innundated with spyware and trojans. That same knowledge and mindset you have to take, keep and use that control is what allows your to also administer a Linux box.
Let it be clear, you *can* control Windows if you want to.
"Oh well, can't do anything about it". (and you can't you know.)
Oh please! Can't do anything about it? Have you ever looked at the registry? We're you too intimidated? It offers nice things like the fact that it's explicitly structured as opposed to an open-ended text file. HEll, I once used a Windows program to manually fix the NTFS partition table that was corrupted by a bad driver (From a 3rd party, not Microsoft).
Just because you couldn't do anything about it doesn't mean something can't be done abotu it.
If anything, the advantage of Linux is that it forces you to understand how the system operates so you can fix things whereas in Windows you can get pretty far knowing nothing unitl something breaks -- then you're screwed. But if you are not afraid to try and fix things, it can be done.