Absolutely, I wish I could reply in as many words as you do. But "Coward Anonymous" is just being a full-time arrogant, I don't know why but perhaps whatever gene enables social intelligence has malfunctioned. But I'm always fascinated by how brains work, so that's my usual theory.
Over at XDA, people are constantly trying out new things and posting the results. Some of them even develop complete ROMs you can install packed with customisations, having only little development experience (which is sometimes worrying but the good ones always stay around longer). The things that can be accomplished is just amazing. Besides rooting and overclocking, there's TouchWiz on HTC and Sense on Samsung, mtd repartitioning, alternate cli shells, native SIP over UMTS/HSPA, backtrack for ARM, various webservers, qemu and bochs... too much to mention. If your phone is on the list of forums, the community developments are just wonderful. And the people of Stack Overflow are great, too. The site has become an invaluable resource of answers to common development problems (started out as mainly.net orientated but covers a wide range of topics).
But I like my home screen easy and simple, though. So I just adb pull Launcher2.apk from an AVD for the version I'm currently running and adb install them onto my phone. Switching home screens can optionally be done with EasyHome. OTOH it could definitely be fun to customise it a bit more then that!
So your solution applies to the Compaqs and Dells but not custom-built PCs, I think the big vendors will just say something along the lines of "we include a free version of $shitty_antivirus, so we don't need this". And then continue to charge extra for support plans, so the "stupid tax" works (clueless customers pay extra).
A simple hardware solution without any software support would be far superior. Sometimes, the more low-tech the solution is the greater the chance that it'll actually work.
For what CPU architecture will the install/live/rescue OS be compiled? How is it going to download the security software? Will it automatically set up 802.1x/PPPoE with your own chap-secrets/a USB UMTS modem or whatever Internet connection the customer might have?
I'd rather see a hardware failsafe with a manual override switch which resets the CPU whenever the SATA controller detects a write to a block below, say, 8. It should be done without using an interrupt. This way, an infection is prevented rather then worked around after the fact. Also you get to use your existing OS installation media to fix whatever is left to fix.
Thanks for that! How can one accomplish anything at all without something as fundamental and basic as CreateWindow(Ex)? They can't say this even remotely resembles "Win32 and COM" with a straight face. This is about as useless as a Matchbox car on a Formula 1 track and makes Windows CE look like a wonderful paradise all of the sudden.
LoadLibrary was replaced with LoadPackagedLibrary (it was never deprecated) and CreateProcess is gone, so how are they going to... you know... run a frickin' program?
As I understand it they merely said no BC for x86 win32 programs on ARM (that is, no "WOWARM" will be made, or Windows on Windows for ARM like there is WOW64 for x86 programs on x86_64). But it would be pretty difficult indeed, to right-click a touch panel. The accessibility options can partially solve that, but it's a bit awkward to work with.
Though I, for one, would love to see Gnome 3.4 on a tablet when it's out. Paired with a bluetooth keyboard and placed on an old fashioned book stand it could replace a laptop and be much more useful and portable as well. Add in wireless HDTV and a setup like this could mark the end of the laptop era.
I see no reason why the win32 API (kernel32, gdi32, user32, comdlg32, etc32, etc32) couldn't be recompiled to run on ARM. It has been on PowerPC, MIPS, Alpha, Alpha64 and is now on x86, x86_64 and IA64. Applications should be recompilable if they use the right data types (e.g. INT_PTR instead of int or void* where msdn says INT_PTR) and don't make assumptions about implementation details (like the fact that the 2 least significant bits of a handle are always 0).
If they don't do this, that would be quite a radical shift away from what they're selling today. It would also drive more attention towards Wine, as large businesses may have high stakes in keeping win32 software around. This is in itself reason enough for me to believe they won't say goodbye to win32 just yet. But it would still be cool to have some kind of dynamic recompilation going on in wine and run old x86 programs unmodified anyway albeit at a major performance hit.
Complaining about MS not using standards is like complaining a car isn't using rocket fuel.
The main problem I've encountered with Visual Studio is that it crashes. A lot. I don't always write.net applications, but when I do I use SharpDevelop. It is everything you need to write for.net, isn't a memory hog and it's stable. Without.net it's useless though but Eclipse is fine. Just enable "show heap status" and click the "Run Garbage Collect" button every so often.
I don't know if LA Noire is good as a game, but I love the albums! CheckItOut!. Amazon hasthesethree, Spotify has these and one more (sorry, no linky).
Ive never run into that situation except (I think) in an addon, where it did. Im not sure I would be boasting that your browser's userland was asking a user if they wanted to terminate some of the code that the browser uses; that seems like that should get caught in qa. That seems a pretty good time to halt the browser and relaunch it.
I ran into this situation on an 8 year old office PC. The QA dept probably doesn't have any of those. Restarting wasn't an option because it would've taken ages.
No argument. They also allow massive memory leaks, hence the complaints for the last umpteen versions about firefox using gigs of memory (HINT-- its your addons). And honestly, as we now have before-the-resource-loads adblock now (yes, it works the SAME as it does in firefox), Im not sure what Im missing with Chromes extension API.
Don't install crappy add-ons then, same as with Chrome. Not loading most javascript (NoScript) and unnecessary images (ABP, RequestPolicy) helps to reduce ram usage.
Thats NOT a safe assumption, as you can install NPAPI plugins without restarting the browser, too (worst case, tab reload is needed, sometimes not even that).
Because it was a design goal to allow run-time loading of netscape plugins, they can be created and destroyed at any time and only represent a single DOM element. Plugins solely communicate through a defined API, they don't have the same access add-ons do.
Not being a webdev, im not the person to argue this in detail, but about 1 month ago I was doing some site troubleshooting, and in order to do in Firebug what I did in Chrome's web dev tools, I did need to download an additional piece-- I believe it was JS related. Anyways, I gave you my reasons why I prefer Chrome's tools (much easier live editing, much better performance), and your retort is "youre wrong"? No reasons, or explanation? Care to expound on that?
I was writing from meat memory. It does seem to have been improved in Webkit, but I haven't attempted any serious debugging with it lately. The profiler is nice, though.
Compared to a C++ implementation, yes, they do, and youd be foolish to argue that interpreted code could possibly outperform compiled code.
Chrome extensions are written in HTML, Javascript and CSS. FF Add-ons can be written in any language, but what are you trying to argue here? That browsers should implement all the functionality in all add-ons/extensions available? Or that from now on all add-ons should be written in C++? Rewriting a program in another language doesn't automatically make it faster and most add-ons/extensions aren't programs of their own but depend heavily on the browser. Thus, they may spend more time in the rest of the browser rather then in their own scripts. A C++ rewrite won't achieve anything except to make it more difficult to maintain and increase the download size. SpiderMonkey already has a JIT compiler anyway.
When I just want to do some quick research or whatever, having a responsive browser that wont pause for 30+seconds checking if my plugins work with a new version, or waiting for the tabs to load is worth a whole lot more.
By default, as of FF 5 IIRC, tabs from a previous session aren't loaded until you select them.
Not sure we're talking about the same thing. Im talking about addon-induced leaks, which the devs have LONG talked about (I remember an Asa Dotzler post from v3.0 about plugins being the cause of leaks). This isnt a hypothetical thing; its a real problem. Its also why when troubleshooting they ask for a clean profile with no addons.
Plugins and add-ons aren't the same thing. Also, so far I've found an issue in FF and
What about when some idiot JS dev had an infinite loop thats killing performance? Def a good idea to A) have Chrome catch it and ask if you want to halt it, and B) have the option to kill the page yourself if you dont want to wait that long.
FF catches this too, and allows you to stop the script even if it's internal to FF itself. Does chrome allow you to stop an internal Javascript? Also, infinite loops in add-ons are unheard of. Are Chrome extensions that much more crappy in general?
Updates can only happen before the browser has fully loaded. Once the browser is fully loaded, any change requires a restart. With chrome, I can install and remove extensions instantly. Trying to argue that Chrome isnt better in this regard is ridiculous; why should it be a GOOD thing that I need to restart constantly if eg Im testing changes to an Extension's javascript? In chrome, I make my changes, save the javascript, and install the updated version, and can see changes immediately.
FF add-ons are more flexible then Chrome add-ons, you can do pretty much anything in an FF add-on. If Chrome extensions could tap into the internals, perhaps it would require regular restarting as well. But it has nothing on XUL, so it's apples and oranges. Thunderbird and Firefox can both run on the same framework (XULRunner). They have their own copy but it's not strictly necessary.
Thats great, but you need more than just firebug to equal Chrome's dev tools. Pretty sure you need a few extra pieces to cover all the javascript, css, etc (dont remember which pieces exactly). There goes some of your hypothetical memory gains.
Chrome is inferior here. Sure, it comes a long way but it's not as good as Firebug yet. No extra pieces involved.
I want to be clear here. Youre worried about memory and inter-process communication, but your solution to deficiencies in firefox is "we can install an extension for that". Thats true, and is one of Firefox's strengths, but if all of your needs are native in Chrome, you will end up sacrificing a ton of performance trying to get them built in as CSS / JS / HTML based extensions.
I have 15 add-ons and the performance with all of them enabled is *better*. They don't slow down anything at all, but ABP, NoScript and RequestPolicy are blocking things I don't need. There's no RequestPolicy for Chrome.
You also introduce the chance of massive memory leaks, which is probably (has LONG been believed to be) the cause of most people's complaints about Firefox.
It doesn't leak memory, but the heap can get fragmented. It's not the same thing. If you go into about:memory and press "Minimize memory usage", not a lot of ram will be freed. Chrome bypasses this problem with their multi-process model but the extra ram cost of this model is higher then the fragmented heap of Firefox.
Thats wonderful, and this isnt a "I hate firefox forever" thread. Its a "chrome is currently superior IMHO" thread. And I am hoping it also addresses tab tearing and creation speeds. I do want firefox to improve, because sometimes one site gives one browser a problem, so options are nice.
Creation and removal imply a switch unless you do it in the background, but yes FF does it instantaneously just like Chrome does.
Another piece of memory and screen real estate gone, for that once every 6 months need.
You can use the Google toolbar without having it shown, for instance you can have it replace the default search box. The extra ram it uses is insignificant, unless the add-on you install is something like picasa or flickr with lots of images it won't cost much at all.
And you missed the point, which was that I dont need to worry about installing it when such a 1-off situation arises.
Sure, but the situation is a bit different. Google has a translation servi
Slashdot Mirror
Absolutely, I wish I could reply in as many words as you do. But "Coward Anonymous" is just being a full-time arrogant, I don't know why but perhaps whatever gene enables social intelligence has malfunctioned. But I'm always fascinated by how brains work, so that's my usual theory.
Over at XDA, people are constantly trying out new things and posting the results. Some of them even develop complete ROMs you can install packed with customisations, having only little development experience (which is sometimes worrying but the good ones always stay around longer). The things that can be accomplished is just amazing. Besides rooting and overclocking, there's TouchWiz on HTC and Sense on Samsung, mtd repartitioning, alternate cli shells, native SIP over UMTS/HSPA, backtrack for ARM, various webservers, qemu and bochs... too much to mention. If your phone is on the list of forums, the community developments are just wonderful. And the people of Stack Overflow are great, too. The site has become an invaluable resource of answers to common development problems (started out as mainly .net orientated but covers a wide range of topics).
But I like my home screen easy and simple, though. So I just adb pull Launcher2.apk from an AVD for the version I'm currently running and adb install them onto my phone. Switching home screens can optionally be done with EasyHome. OTOH it could definitely be fun to customise it a bit more then that!
So your solution applies to the Compaqs and Dells but not custom-built PCs, I think the big vendors will just say something along the lines of "we include a free version of $shitty_antivirus, so we don't need this". And then continue to charge extra for support plans, so the "stupid tax" works (clueless customers pay extra).
A simple hardware solution without any software support would be far superior. Sometimes, the more low-tech the solution is the greater the chance that it'll actually work.
I don't get it. Is OP FunnyBot?
That's just not the Stingray that I used to know :(
I'd rather see a hardware failsafe with a manual override switch which resets the CPU whenever the SATA controller detects a write to a block below, say, 8. It should be done without using an interrupt. This way, an infection is prevented rather then worked around after the fact. Also you get to use your existing OS installation media to fix whatever is left to fix.
That's what Logical Volume Management its friends are for.
Thanks for that! How can one accomplish anything at all without something as fundamental and basic as CreateWindow(Ex)? They can't say this even remotely resembles "Win32 and COM" with a straight face. This is about as useless as a Matchbox car on a Formula 1 track and makes Windows CE look like a wonderful paradise all of the sudden.
LoadLibrary was replaced with LoadPackagedLibrary (it was never deprecated) and CreateProcess is gone, so how are they going to... you know... run a frickin' program?
As I understand it they merely said no BC for x86 win32 programs on ARM (that is, no "WOWARM" will be made, or Windows on Windows for ARM like there is WOW64 for x86 programs on x86_64). But it would be pretty difficult indeed, to right-click a touch panel. The accessibility options can partially solve that, but it's a bit awkward to work with.
Though I, for one, would love to see Gnome 3.4 on a tablet when it's out. Paired with a bluetooth keyboard and placed on an old fashioned book stand it could replace a laptop and be much more useful and portable as well. Add in wireless HDTV and a setup like this could mark the end of the laptop era.
Almost...
Yes. Every time I go there I lose a couple of IQ points. Oh shii~
I see no reason why the win32 API (kernel32, gdi32, user32, comdlg32, etc32, etc32) couldn't be recompiled to run on ARM. It has been on PowerPC, MIPS, Alpha, Alpha64 and is now on x86, x86_64 and IA64. Applications should be recompilable if they use the right data types (e.g. INT_PTR instead of int or void* where msdn says INT_PTR) and don't make assumptions about implementation details (like the fact that the 2 least significant bits of a handle are always 0).
If they don't do this, that would be quite a radical shift away from what they're selling today. It would also drive more attention towards Wine, as large businesses may have high stakes in keeping win32 software around. This is in itself reason enough for me to believe they won't say goodbye to win32 just yet. But it would still be cool to have some kind of dynamic recompilation going on in wine and run old x86 programs unmodified anyway albeit at a major performance hit.
Complaining about MS not using standards is like complaining a car isn't using rocket fuel.
The main problem I've encountered with Visual Studio is that it crashes. A lot. I don't always write .net applications, but when I do I use SharpDevelop. It is everything you need to write for .net, isn't a memory hog and it's stable. Without .net it's useless though but Eclipse is fine. Just enable "show heap status" and click the "Run Garbage Collect" button every so often.
Hurrrrrrrrr
THIS
You mean Jell-O. Jello is a german band with hits such as The race and Oh Yeah.
You can still set a password, pin code or pattern to lock the screen. 3rd Party apps also allow for other unlocking methods such as a gesture.
Shit like this is why we can't have nice things.
Would you install your vendors driver CD on a fresh windows installation?
I don't know if LA Noire is good as a game, but I love the albums! Check It Out!. Amazon has these three, Spotify has these and one more (sorry, no linky).
(and here's how to d/l yt vidyas)
It says "HTML5 Video is a subscriber-only feature that works in H.264-compliant browsers."
The video requires Flash 9.0, I have neither (what's a "H.264-compliant browser" anyway?).
Ive never run into that situation except (I think) in an addon, where it did. Im not sure I would be boasting that your browser's userland was asking a user if they wanted to terminate some of the code that the browser uses; that seems like that should get caught in qa. That seems a pretty good time to halt the browser and relaunch it.
I ran into this situation on an 8 year old office PC. The QA dept probably doesn't have any of those. Restarting wasn't an option because it would've taken ages.
No argument. They also allow massive memory leaks, hence the complaints for the last umpteen versions about firefox using gigs of memory (HINT-- its your addons). And honestly, as we now have before-the-resource-loads adblock now (yes, it works the SAME as it does in firefox), Im not sure what Im missing with Chromes extension API.
Don't install crappy add-ons then, same as with Chrome. Not loading most javascript (NoScript) and unnecessary images (ABP, RequestPolicy) helps to reduce ram usage.
Thats NOT a safe assumption, as you can install NPAPI plugins without restarting the browser, too (worst case, tab reload is needed, sometimes not even that).
Because it was a design goal to allow run-time loading of netscape plugins, they can be created and destroyed at any time and only represent a single DOM element. Plugins solely communicate through a defined API, they don't have the same access add-ons do.
Not being a webdev, im not the person to argue this in detail, but about 1 month ago I was doing some site troubleshooting, and in order to do in Firebug what I did in Chrome's web dev tools, I did need to download an additional piece-- I believe it was JS related. Anyways, I gave you my reasons why I prefer Chrome's tools (much easier live editing, much better performance), and your retort is "youre wrong"? No reasons, or explanation? Care to expound on that?
I was writing from meat memory. It does seem to have been improved in Webkit, but I haven't attempted any serious debugging with it lately. The profiler is nice, though.
Compared to a C++ implementation, yes, they do, and youd be foolish to argue that interpreted code could possibly outperform compiled code.
Chrome extensions are written in HTML, Javascript and CSS. FF Add-ons can be written in any language, but what are you trying to argue here? That browsers should implement all the functionality in all add-ons/extensions available? Or that from now on all add-ons should be written in C++? Rewriting a program in another language doesn't automatically make it faster and most add-ons/extensions aren't programs of their own but depend heavily on the browser. Thus, they may spend more time in the rest of the browser rather then in their own scripts. A C++ rewrite won't achieve anything except to make it more difficult to maintain and increase the download size. SpiderMonkey already has a JIT compiler anyway.
When I just want to do some quick research or whatever, having a responsive browser that wont pause for 30+seconds checking if my plugins work with a new version, or waiting for the tabs to load is worth a whole lot more.
By default, as of FF 5 IIRC, tabs from a previous session aren't loaded until you select them.
Not sure we're talking about the same thing. Im talking about addon-induced leaks, which the devs have LONG talked about (I remember an Asa Dotzler post from v3.0 about plugins being the cause of leaks). This isnt a hypothetical thing; its a real problem. Its also why when troubleshooting they ask for a clean profile with no addons.
Plugins and add-ons aren't the same thing. Also, so far I've found an issue in FF and
OpenVZ was never mainlined, but LXC (Linux Containers) is.
I hereby pledge that we absolutely must have multi-OS action rule 34 on the intertubes.
But you can only get to pay with Google Checkout. Which sits right next to useless.
What about when some idiot JS dev had an infinite loop thats killing performance? Def a good idea to A) have Chrome catch it and ask if you want to halt it, and B) have the option to kill the page yourself if you dont want to wait that long.
FF catches this too, and allows you to stop the script even if it's internal to FF itself. Does chrome allow you to stop an internal Javascript? Also, infinite loops in add-ons are unheard of. Are Chrome extensions that much more crappy in general?
Updates can only happen before the browser has fully loaded. Once the browser is fully loaded, any change requires a restart. With chrome, I can install and remove extensions instantly. Trying to argue that Chrome isnt better in this regard is ridiculous; why should it be a GOOD thing that I need to restart constantly if eg Im testing changes to an Extension's javascript? In chrome, I make my changes, save the javascript, and install the updated version, and can see changes immediately.
FF add-ons are more flexible then Chrome add-ons, you can do pretty much anything in an FF add-on. If Chrome extensions could tap into the internals, perhaps it would require regular restarting as well. But it has nothing on XUL, so it's apples and oranges. Thunderbird and Firefox can both run on the same framework (XULRunner). They have their own copy but it's not strictly necessary.
Thats great, but you need more than just firebug to equal Chrome's dev tools. Pretty sure you need a few extra pieces to cover all the javascript, css, etc (dont remember which pieces exactly). There goes some of your hypothetical memory gains.
Chrome is inferior here. Sure, it comes a long way but it's not as good as Firebug yet. No extra pieces involved.
I want to be clear here. Youre worried about memory and inter-process communication, but your solution to deficiencies in firefox is "we can install an extension for that". Thats true, and is one of Firefox's strengths, but if all of your needs are native in Chrome, you will end up sacrificing a ton of performance trying to get them built in as CSS / JS / HTML based extensions.
I have 15 add-ons and the performance with all of them enabled is *better*. They don't slow down anything at all, but ABP, NoScript and RequestPolicy are blocking things I don't need. There's no RequestPolicy for Chrome.
You also introduce the chance of massive memory leaks, which is probably (has LONG been believed to be) the cause of most people's complaints about Firefox.
It doesn't leak memory, but the heap can get fragmented. It's not the same thing. If you go into about:memory and press "Minimize memory usage", not a lot of ram will be freed. Chrome bypasses this problem with their multi-process model but the extra ram cost of this model is higher then the fragmented heap of Firefox.
Thats wonderful, and this isnt a "I hate firefox forever" thread. Its a "chrome is currently superior IMHO" thread. And I am hoping it also addresses tab tearing and creation speeds. I do want firefox to improve, because sometimes one site gives one browser a problem, so options are nice.
Creation and removal imply a switch unless you do it in the background, but yes FF does it instantaneously just like Chrome does.
Another piece of memory and screen real estate gone, for that once every 6 months need.
You can use the Google toolbar without having it shown, for instance you can have it replace the default search box. The extra ram it uses is insignificant, unless the add-on you install is something like picasa or flickr with lots of images it won't cost much at all.
And you missed the point, which was that I dont need to worry about installing it when such a 1-off situation arises.
Sure, but the situation is a bit different. Google has a translation servi