What went wrong with Linux on the Desktop

[wraith808]

Those who are using Visual studio knows Why I am saying this, because microsoft's own programs make the system unusable.

-mahesh2k (September 05, 2012, 04:58 AM)

I've been running every single version (not every edition, of course!) and service pack of Visual Studio from 6 to 2010 SP1, and will soon be installing 2012. I've even been using VS on Win9x. While the first VS.NET version was pretty crappy and unstable, I've never had VS affect my system stability, and never heard of stories like that from friends or co-workers.

-f0dder (September 06, 2012, 06:54 AM)

Same here all the way back to 9x. Hell we've even got one guy that (does legacy stuff) is still doing VB code with VS6, on Windows 8.

-Stoic Joker (September 06, 2012, 07:20 AM)

Same here... including unfortunately the VB code.   Anything other than anecdotal evidence of this VS making the system unusable bit?

Sure, there's some nice things about OSX, and the build quality on some macbooks is better than a lot of non-OSX PCs. But there's plenty of problems as well, and plenty of funny security holes as well :-)

-f0dder (September 06, 2012, 06:54 AM)

 +1

They're all just tools, and all of them have their pros and cons.

[TaoPhoenix]

... Your assumptions on its advantages are plain wrong, that's all.
Oh, right: Call it a flamebait when someone points out your mistakes. Well played.

-Tuxman (September 05, 2012, 05:21 AM)

Hi Tux.

Since you asked elsewhere why people perceived you as a "troll", I picked this comment to look at. I have done some little mini-studies on quantifying semantics to put some math behind "qualitative" words such as "troll". (I am not the first, I even saw some news story that some big corps are patenting this stuff too, this is just my little variant.)

The short easy explanation borrows a phrase from somewhere called the "Golden Rule vs the Silver Rule". The Golden Rule is often stated as "treat others as you want to be treated". The problem is that people often end up in the Silver Rule, stated as "treat others as you *have been* treated". Aka, someone misfires a comment and then the series rolls along.

So with that as a backdrop, let's do some number crunching. An early starting point is to assign a number scale of "sharpness" to a comment that the person makes. Your first line is slightly sharp. The number scale can be anything that appeals to you. As a simple 0-10 scale, (with the zero being important!) 0 means perfectly balanced, 10 is boiling over with sharpness. I'd rate your first line about a 2. Nothing too bad, a little staccato, but okay.

But look at the second line! It's flanked by two (sharp / sarcastic / your choice of word) elements. I'd rate it at least a 4 as a pair.

Then it's just an algorithm. The average sharpness of comments eventually crosses a (somewhat fuzzy) line to what people begin calling "troll". Per person, that line moves up and down, but finally like a bell curve it converges. Does that help?

[Tuxman]

A bit, thanks.

(OTOH this seems to be rather subjective. I mean, "4" is "medium trollness" (for you) where "trollness" is a synonym for sarcasm which is not actually a good idea. I might be sarcastic at times, but the difference between trolling and being sarcastic is that trolling is only destructive and not a good way to transport suggestions.)

[TaoPhoenix]

A bit, thanks.

(OTOH this seems to be rather subjective. I mean, "4" is "medium trollness" (for you) where "trollness" is a synonym for sarcasm which is not actually a good idea. I might be sarcastic at times, but the difference between trolling and being sarcastic is that trolling is only destructive and not a good way to transport suggestions.)

-Tuxman (September 06, 2012, 08:29 AM)

Aye, I have a lot more finesses behind my theory and I'm happy to start a thread maybe in the basement to thrash them out! You're quite right, there are elements of both each person's individual scales, and intention of the original person. And more. My early aim was to begin to use what I believe is your talent for analysis to begin to study why you are a little confused at the reception you are receiving here.

I believe that the theory can reduce to a series of mathematical equations that will yield a numerical quantitative answer to these kinds of questions.  Like all science, we can fiddle with the structure of the equations gleefully, but already at a rough level it works for me. I used it this morning when a recruiter for a life insurance company called me.  : )

[mahesh2k]

Honestly, you don't need crappy windows. I used OSX and Linux and they don't show performance issues like windows. Not even windows 7 solved performance issue problem. More you use windows, less usable it becomes, after you add more data, reg entries and stuff.

-mahesh2k (September 05, 2012, 04:58 AM)

I personally haven't seen this since I moved away from Win9x. The registry is pretty efficient, definitely a lot more efficient than re-parsing text files all the time - the only problems I've seen have been caused by really badly written 3rd party software, and the cause hasn't been "too much data in the registry", it's been "really broken data" (which just coincidentally happened to be located in the registry).

-f0dder (September 06, 2012, 06:54 AM)

I did found problems, especially on Win XP and Vista. I had plenty of things broken e.g. TCP/IP reg fix that is available for both XP and vista when net crashes to reg corruption of the entries, I am sure you are aware of it. I can dig out few more reg edit fixes out there, in fact I had one bunch saved as file when I used XP Pro SP2 to use them after fresh installation.  I don't have this on Windows 7 because those problems are pretty much solved on 7 but with windows XP SP 2 that problem is still possible to replicate. I don't know how you guys are saying that after Windows 9x, system is much stable in terms of reg performance because it's not for me.

Those who are using Visual studio knows Why I am saying this, because microsoft's own programs make the system unusable.I've been running every single version (not every edition, of course!) and service pack of Visual Studio from 6 to 2010 SP1, and will soon be installing 2012. I've even been using VS on Win9x. While the first VS.NET version was pretty crappy and unstable, I've never had VS affect my system stability, and never heard of stories like that from friends or co-workers.

-mahesh2k (September 05, 2012, 04:58 AM)

I used the Pro edition upto 2010 and after 2010 I used selective VS for web development. Prior to that almost every VS edition used to interfere with the ,NET framework which I updated and patched beyond default VS bundle. e.g. ,NET 3 and onwards. I had debugger closing my programs (this includes treedb, cintanotes and few other programs) bugging every single application and opening debugger for them. Never had this problem on your side? I wonder how you turned off debugger after VS installation to keep it from interfering with other applications and default .NET framework.

On the other hand, OSX based on unix is perfectly fine. Doesn't break or gets crashed with official softwares and upgrades.Most of the developers at my current job are on OSX laptops. A bunch of them started cursing some months ago after installing whatever-cat-named-update because their systems got bogged down (disk paging, beach ball icon, and sometime systems so unresponsive they had to hardboot them) - seems like Apple messed up the memory manager, majorly. Not something you'll see if you're just drinking caffè latte and not using your shiny laptop for facebook and hipstagram - but definitely if you're actually using the machine. And should I mention the funny instances where the battery expands somewhat (natural thing to do because of heat), messing up the touchpad? Or the various data-loss incidents there's been in Finder?Sure, there's some nice things about OSX, and the build quality on some macbooks is better than a lot of non-OSX PCs. But there's plenty of problems as well, and plenty of funny security holes as well :-)

-mahesh2k (September 05, 2012, 04:58 AM)

I don't know which part of my post people are picking up to imply "no security holes on osx". I am just comparing it with windows and on that comparison I had very few complaints on OS X in terms of patching (explained above) and the official updates, where you know how good windows performs.

[f0dder]

I did found problems, especially on Win XP and Vista. I had plenty of things broken e.g. TCP/IP reg fix that is available for both XP and vista when net crashes to reg corruption of the entries, I am sure you are aware of it. I can dig out few more reg edit fixes out there, in fact I had one bunch saved as file when I used XP Pro SP2 to use them after fresh installation.  I don't have this on Windows 7 because those problems are pretty much solved on 7 but with windows XP SP 2 that problem is still possible to replicate. I don't know how you guys are saying that after Windows 9x, system is much stable in terms of reg performance because it's not for me.

-mahesh2k (September 06, 2012, 09:41 AM)

I'm not sure I understand what you're saying here. Are you saying the registry tweaks (i.e. registry settings that are perceived to improve system performance, while often being snake-oil) that worked on XP no longer work on Vista and Win7? Or does your "net crashes to reg corruption" mean a BSOD causing corruption of the hive-files, and thus a completely broken Windows install?

Since installing Win2k (the first NT I've really used), 99% of the BSODs I've seen have been caused by flaky hardware, bad 3rd-party drivers, or myself messing around with kernel-mode debuggers or driver development. I can't recall seeing a BSOD that was caused by MS code, but I'm pretty sure there's been a few. In those 10+ years. While a BSOD does mean losing whatever unsaved open files, I've only seen filesystem corruption in a very few instances - that was with ATI video drivers. Those caused extremely nasty FS corruption, though, bad enough that I had to run filesystem recovery software. Had 3-4 of those before I realized the drivers were insanely lame and couldn't handle LargeSystemCache=1. And as late as June 2012, AMD/ATI video drivers prevented system-wide ASLR.

But please do elaborate on the problems you mention above, as I'm genuinely confused as to what you mean.

I used the Pro edition upto 2010 and after 2010 I used selective VS for web development. Prior to that almost every VS edition used to interfere with the ,NET framework which I updated and patched beyond default VS bundle. e.g. ,NET 3 and onwards. I had debugger closing my programs (this includes treedb, cintanotes and few other programs) bugging every single application and opening debugger for them. Never had this problem on your side? I wonder how you turned off debugger after VS installation to keep it from interfering with other applications and default .NET framework.

-mahesh2k (September 06, 2012, 09:41 AM)

I've never had the debugger "interfering with other applications". If a program crashes, yes, I'll get the option to attach the debugger and do post-mortem. But that's not the debugger interfering, that's the 3rd-party program crashing, for whatever reason. Rather than blaming VS, perhaps it's the 3rd-party program that's stupidly programmed and making unguaranteed assumptions, and then crashing when implementation details change in a later .NET version? (not that I'm saying .NET is bugfree, though, I've filed a couple of bug reports on it myself. Quite esoteric edge-cases, though.)

I don't know which part of my post people are picking up to imply "no security holes on osx". I am just comparing it with windows and on that comparison I had very few complaints on OS X in terms of patching (explained above) and the official updates, where you know how good windows performs.

-mahesh2k (September 06, 2012, 09:41 AM)

*shrug* - Windows update has (almost) always worked pretty well for me - it really doesn't like running out of disk space, though, but I'm not sure I can really blame it for that. I've had coworkers with bricked systems after OSX updates (to be fair, that was upgrading to a new OSX version rather than just a regular update, but still.)

[mahesh2k]

I did found problems, especially on Win XP and Vista. I had plenty of things broken e.g. TCP/IP reg fix that is available for both XP and vista when net crashes to reg corruption of the entries, I am sure you are aware of it. I can dig out few more reg edit fixes out there, in fact I had one bunch saved as file when I used XP Pro SP2 to use them after fresh installation.  I don't have this on Windows 7 because those problems are pretty much solved on 7 but with windows XP SP 2 that problem is still possible to replicate. I don't know how you guys are saying that after Windows 9x, system is much stable in terms of reg performance because it's not for me.

-mahesh2k (September 06, 2012, 09:41 AM)

I'm not sure I understand what you're saying here. Are you saying the registry tweaks (i.e. registry settings that are perceived to improve system performance, while often being snake-oil) that worked on XP no longer work on Vista and Win7? Or does your "net crashes to reg corruption" mean a BSOD causing corruption of the hive-files, and thus a completely broken Windows install.Since installing Win2k (the first NT I've really used), 99% of the BSODs I've seen have been caused by flaky hardware, bad 3rd-party drivers, or myself messing around with kernel-mode debuggers or driver development. I can't recall seeing a BSOD that was caused by MS code, but I'm pretty sure there's been a few. In those 10+ years. While a BSOD does mean losing whatever unsaved open files, I've only seen filesystem corruption in a very few instances - that was with ATI video drivers. Those caused extremely nasty FS corruption, though, bad enough that I had to run filesystem recovery software. Had 3-4 of those before I realized the drivers were insanely lame and couldn't handle LargeSystemCache=1. And as late as June 2012, AMD/ATI video drivers prevented system-wide ASLRBut please do elaborate on the problems you mention above, as I'm genuinely confused as to what you mean.

-f0dder (September 06, 2012, 09:56 AM)

I mean reg corruption,not tweaks. Remember WINSOCK issues? Yes. I am talking about some of these reg level corruption of the entries which makes XP or vista unusable. Upto Vista, it's easy to find that problem. As for BSOD, had that problem after patching uxtheme.dll which we need while changing the themes for XP. It has nothing to do with reg but performance issue you wanted to see is there when you do modify shell stuff. To be honest, never had video driver issues, unless ofcourse it was from the hardware of HP (no complaints to MS for that).

I used the Pro edition upto 2010 and after 2010 I used selective VS for web development. Prior to that almost every VS edition used to interfere with the ,NET framework which I updated and patched beyond default VS bundle. e.g. ,NET 3 and onwards. I had debugger closing my programs (this includes treedb, cintanotes and few other programs) bugging every single application and opening debugger for them. Never had this problem on your side? I wonder how you turned off debugger after VS installation to keep it from interfering with other applications and default .NET framework.I've never had the debugger "interfering with other applications". If a program crashes, yes, I'll get the option to attach the debugger and do post-mortem. But that's not the debugger interfering, that's the 3rd-party program crashing, for whatever reason. Rather than blaming VS, perhaps it's the 3rd-party program that's stupidly programmed and making unguaranteed assumptions, and then crashing when implementation details change in a later .NET version? (not that I'm saying .NET is bugfree, though, I've filed a couple of bug reports on it myself. Quite esoteric edge-cases, though.)

-mahesh2k (September 06, 2012, 09:41 AM)

Here's how I found out debugger issue. I installed the fresh XP and vista on different machines. Then used the typical programs without any problems. And after installation of VS, observed the problem with debugger. It has mostly due to .NET and the debugger issues IMO. So it's definitely not third party issue. Yet to find out how to get over that. But As I am not doing much VC/Sharp development, it's not needed for VS 2010 onwards. Web dev VS doesn't interfere with other programs. Also mind telling me how do you get rid of all reg entries when you install VS PRO? I have yet to figure out complete removal of VS without leaving some traces behind.

I don't know which part of my post people are picking up to imply "no security holes on osx". I am just comparing it with windows and on that comparison I had very few complaints on OS X in terms of patching (explained above) and the official updates, where you know how good windows performs.*shrug* - Windows update has (almost) always worked pretty well for me - it really doesn't like running out of disk space, though, but I'm not sure I can really blame it for that. I've had coworkers with bricked systems after OSX updates (to be fair, that was upgrading to a new OSX version rather than just a regular update, but still.)

-mahesh2k (September 06, 2012, 09:41 AM)

It didn't worked for me upto vista. Patch list and security fixes keeps on increasing once we have VS installed. Windows 7 on the other hand is never showed the upgrade and patching problems.

[Stoic Joker]

For some reason installers for .NET applications have a tendency to (occasionally) replace key .NET framework files as part of the install routine. There is an XML patch that I've had to reinstall over 30 times in the past year ... Simply because the EMR software that several of our clients use insists on "updating" (e.g. borking) this perticular patch every time they do an update.

[f0dder]

I mean reg corruption,not tweaks. Remember WINSOCK issues? Yes. I am talking about some of these reg level corruption of the entries which makes XP or vista unusable. Upto Vista, it's easy to find that problem.

-mahesh2k (September 06, 2012, 10:16 AM)

Never ever heard about it. I had BSODs on my nforce4 motherboard when trying to use nvidia's "hardware" firewall (which was a retarted piece of crap, requiring a full install of an apache httpd in order to manage settings locally... :rolleyes: ), and I've had BSODs from third-party firewall, VPN and antivirus products. Some of these seem to show winsock as the problem, but that's only if you look just at the top-level driver in the BSOD, not if you study the actual minidump with WinDbg. Also, I've never had filesystme or registry corruption from any of these "normal" BSODs. Got more information on the issue?

As for BSOD, had that problem after patching uxtheme.dll which we need while changing the themes for XP. It has nothing to do with reg but performance issue you wanted to see is there when you do modify shell stuff. To be honest, never had video driver issues, unless ofcourse it was from the hardware of HP (no complaints to MS for that).

-mahesh2k (September 06, 2012, 10:16 AM)

Sure you didn't use some dodgy software to patch uxtheme, which installed malware on your machine? I've had it patched on XP, XP64, Vista64 and Win7-64 without trouble. It would be a very weird cause of BSODs, since it's a usermode DLL.

Here's how I found out debugger issue. I installed the fresh XP and vista on different machines. Then used the typical programs without any problems. And after installation of VS, observed the problem with debugger. It has mostly due to .NET and the debugger issues IMO. So it's definitely not third party issue. Yet to find out how to get over that. But As I am not doing much VC/Sharp development, it's not needed for VS 2010 onwards. Web dev VS doesn't interfere with other programs.

-mahesh2k (September 06, 2012, 09:41 AM)

I still kinda doubt that has anything to do with "the debugger" (unless the 3rd party software software refuses to run on a machine with developer/debugging tools installed) - but it could very well be that the software simply doesn't work with some versions of .NET runtime libraries... whether the particular VS versions upgrade or accidentally downgrade them. If there's one thing I've learned in my years as a developer, it's that you first blame yourself, then 3rd party vendors, and only then start suspecting Microsoft. Often saves your some embarassment :-)

Also mind telling me how do you get rid of all reg entries when you install VS PRO? I have yet to figure out complete removal of VS without leaving some traces behind.

-mahesh2k (September 06, 2012, 09:41 AM)

Haven't looked at it, and honestly don't care - as long as the leftovers don't cause any trouble (which they haven't). Sure, it's not aesthetically pleasing that junk gets left behind, but even some megabytes worth of data wouldn't really affect the system.

It didn't worked for me upto vista. Patch list and security fixes keeps on increasing once we have VS installed. Windows 7 on the other hand is never showed the upgrade and patching problems.

-mahesh2k (September 06, 2012, 09:41 AM)

Are you saying that updates failed to install, or that there were too many updates for your liking?

Haven't had much trouble with Windows Update except when running on heavily modified Windows versions. Sure, every once in a while I have to install updates a few at a time for whatever reason, as installing everything fails. Not a big deal, sometimes I have to compile programs from source on Linux, because default packages sometimes have conflicting dependencies, which luckily always turn out to be default settings I don't need. Fortunately it's been long enough that I can't remember any particular details, just like it's been long enough since I've had problems with Windows Update :-)

For some reason installers for .NET applications have a tendency to (occasionally) replace key .NET framework files as part of the install routine. There is an XML patch that I've had to reinstall over 30 times in the past year ... Simply because the EMR software that several of our clients use insists on "updating" (e.g. borking) this perticular patch every time they do an update.

-Stoic Joker (September 06, 2012, 11:47 AM)

Annoying. "xcopy" installs, or custom installers? MSI based installers are at least supposed to make sure that doesn't happen (but I still loathe their insanely slow speed).

[mahesh2k]

I had BSODs on my nforce4 motherboard when trying to use nvidia's "hardware" firewall (which was a retarted piece of crap, requiring a full install of an apache httpd in order to manage settings locally... :rolleyes: ), and I've had BSODs from third-party firewall, VPN and antivirus products. Some of these seem to show winsock as the problem, but that's only if you look just at the top-level driver in the BSOD, not if you study the actual minidump with WinDbg. Also, I've never had filesystme or registry corruption from any of these "normal" BSODs. Got more information on the issue?

-f0dder (September 06, 2012, 12:10 PM)

Not BSOD due to reg but was because of the theme patching and shell related changes. In case of reg issues, most of them were related to the WINSOCK and other network related issues. I can't specifically point to the issue because I used to patch that using the reg fix for Winsock. Without that fix, It was hard to connect to the net. I am not sure if it was because of driver issue or say some malware? For some reason system was not usable without those reg fixes (most of them were released by MVPs), I hope you remember XP days with regular fixing of broken stuff.

Sure you didn't use some dodgy software to patch uxtheme, which installed malware on your machine? I've had it patched on XP, XP64, Vista64 and Win7-64 without trouble. It would be a very weird cause of BSODs, since it's a usermode DLL.

-mahesh2k (September 06, 2012, 10:16 AM)

No. I used to play with TGTSoft's uxtheme.dll which was released by them for community. StyleXP isn't alive anymore but I am sure  their community themexp.org still has that uxtheme.dll for patching the shell. Without this it was not possible to change themes freely on XP at that time. Had BSOD during these changes.

I still kinda doubt that has anything to do with "the debugger" (unless the 3rd party software software refuses to run on a machine with developer/debugging tools installed) - but it could very well be that the software simply doesn't work with some versions of .NET runtime libraries... whether the particular VS versions upgrade or accidentally downgrade them. If there's one thing I've learned in my years as a developer, it's that you first blame yourself, then 3rd party vendors, and only then start suspecting Microsoft. Often saves your some embarassment :-)

-mahesh2k (September 06, 2012, 09:41 AM)

Well it worked fine without VS, so I can't take the blame. It worked Okay for sometime with VS installed so 3rd party software can't be blamed until .NET had clashes so surely MS has to be blamed here for this. lol

Are you saying that updates failed to install, or that there were too many updates for your liking?

-mahesh2k (September 06, 2012, 09:41 AM)

Some updates with KB fixes for .NET and other VS components failed to install and sometimes had startup issues with them as well. Especially on vista, but never had startup related issue on XP with those KB Fix updates.

[Stoic Joker]

I had BSODs on my nforce4 motherboard when trying to use nvidia's "hardware" firewall (which was a retarted piece of crap, requiring a full install of an apache httpd in order to manage settings locally... :rolleyes: ), and I've had BSODs from third-party firewall, VPN and antivirus products. Some of these seem to show winsock as the problem, but that's only if you look just at the top-level driver in the BSOD, not if you study the actual minidump with WinDbg. Also, I've never had filesystme or registry corruption from any of these "normal" BSODs. Got more information on the issue?

-f0dder (September 06, 2012, 12:10 PM)

Not BSOD due to reg but was because of the theme patching and shell related changes. In case of reg issues, most of them were related to the WINSOCK and other network related issues. I can't specifically point to the issue because I used to patch that using the reg fix for Winsock. Without that fix, It was hard to connect to the net. I am not sure if it was because of driver issue or say some malware? For some reason system was not usable without those reg fixes (most of them were released by MVPs), I hope you remember XP days with regular fixing of broken stuff.

-mahesh2k (September 06, 2012, 01:41 PM)

Only time I ever saw WINSOCK get borked was usually due to an orphaned LSP from either a virus or babysitter security suite (mostly the 2nd one). Either way the fix has always been the same netsh reset winsock ... I don't ever recall using a reg patch for it.

[f0dder]

Not BSOD due to reg but was because of the theme patching and shell related changes. In case of reg issues, most of them were related to the WINSOCK and other network related issues. I can't specifically point to the issue because I used to patch that using the reg fix for Winsock. Without that fix, It was hard to connect to the net. I am not sure if it was because of driver issue or say some malware? For some reason system was not usable without those reg fixes (most of them were released by MVPs),

-mahesh2k (September 06, 2012, 01:41 PM)

Are you talking about registry changes, or the TCPIP.SYS patch to increase the maximum half-open connections (which was recommended for p2p uses)? There's a big difference. I've never heard about winsock registry fixes (apart from malware related crap), but I did use TCPIP.SYS patching back in the XP days. Stupid blind patching (that didn't check for correction version) could nuke your driver, which could definitely end up causing BSODs eventually.

I hope you remember XP days with regular fixing of broken stuff.

-mahesh2k (September 06, 2012, 01:41 PM)

Not really, no. Didn't have a lot of problems that weren't caused by 3rd party software, and while there have been a fair amount of security holes in Windows, I'm not complaining that Microsoft have actually been patching them :-)

Sure you didn't use some dodgy software to patch uxtheme, which installed malware on your machine? I've had it patched on XP, XP64, Vista64 and Win7-64 without trouble. It would be a very weird cause of BSODs, since it's a usermode DLL.

-f0dder

No. I used to play with TGTSoft's uxtheme.dll which was released by them for community. StyleXP isn't alive anymore but I am sure  their community themexp.org still has that uxtheme.dll for patching the shell. Without this it was not possible to change themes freely on XP at that time. Had BSOD during these changes.

-mahesh2k (September 06, 2012, 01:41 PM)

Can't remember which uxtheme patch I specifically used, nor whether it was a patcher or a full .dll download (but I'd definitely suspicious about a full .dll). Again, it would be highly unusual for usermode software to cause BSODs.

I'm still not convinced that anything you've put forth stems from problems with the registry or Windows in general - it does sound more and more like there could be some dodgy 3rd-party software involved, though (and by that, I include not just various OS patches, but potentially also drivers, antivirus programs, firewalls, et cetera).

[Stoic Joker]

For some reason installers for .NET applications have a tendency to (occasionally) replace key .NET framework files as part of the install routine. There is an XML patch that I've had to reinstall over 30 times in the past year ... Simply because the EMR software that several of our clients use insists on "updating" (e.g. borking) this particular patch every time they do an update.

-Stoic Joker (September 06, 2012, 11:47 AM)

Annoying. "xcopy" installs, or custom installers? MSI based installers are at least supposed to make sure that doesn't happen (but I still loathe their insanely slow speed).

-f0dder (September 06, 2012, 12:10 PM)

Hm... Maybe I'm off in the weeds on the .NET aspect, but I do seem to recall reading about the behavior while researching this (apparently other) issue with KB954430/ AmazingCharts causing the XML patch to fall off constantly.

[TaoPhoenix]

Going way up to a sky high view of the topic, I'll start a new subthread about Linux on the Desktop based on China. You know how "everything" (materially) is made in China? (I know, there's new countries in the game now, but Made in China has the ring you know what I mean.) By now they're past 50 cent party favors, they basically make everything - but with a curious flaw.

We don't yet recognize them as "tech innovators". Why not? (Elsewhere I've seen some articles that their culture somehow discourages it, but that's another day.) Trying not to go stereotypical, China was the original "Cheap Copy". So since Linux is ... wait for it ... Cheaper Than Cheap, with an open license to copy it, why haven't the Chinese pulverized the tech world with The Good Enough Linux Distro that makes everyone else react? (A medium-range problem is that China has gotten hooked with building sneaky back doors to everything, yuck.) Just do it Old School China style. Make a $200 computer with a medium-carefully tweaked Linux Distro, and just flood the market. (And that was what made old China style adorable, they "only medium tweaked" stuff, leading to amusingly-flawed-but-I-still-bought-it-because-it-was-cheap stuff.)

Since we're in daydreaming, they'd make a hybrid new OS, and hit a sweet spot when what was in fact cutting corners on design turns out to be the "Simple OS" that "Mainstream America/Europe" likes. (Maybe only desktop + 2 level deep folders, a simple System Settings set, their own copy of something like LibreOffice and Firefox but having actually fixed all known bug reports, and say like five more amazing darkhorse killer apps no one even knew were possible. They'd contact each of say 500 software vendors and provide a software port free of charge. (Good ol' China-Copying!)

It has just enough extensibility for the techies with plugins, "easy" for the masses, runs everything because they did the ports, and allowed free OS copies for OEM builders.

THAT might get Linux on the desktop.

[40hz]

^What you're describing sounds an awful lot like what Android was originally (supposedly?) intended to do - except you're suggesting it come with a better keyboard and bigger screen.

And yeah...if anybody does build it, it's gonna be China I'm guessing. Probably take it to Mars with them too at the rate they're going.

[TaoPhoenix]

^What you're describing sounds an awful lot like what Android was originally (supposedly?) intended to do - except you're suggesting it come with a better keyboard and bigger screen.

And yeah...if anybody does build it, it's gonna be China I'm guessing. Probably take it to Mars with them too at the rate they're going.

-40hz (September 06, 2012, 05:12 PM)

Heh well to get that post going, I used China, but if I had to actually make bets, I might put them on India First. There's no meme "Made in India", but because it's tech and not materials, I think I respect India's culture more right now. (Red Herring shout out to Anand! Let's hear the Audience Roar from DC's Indian contingent!        )

[40hz]

^yeah. But we're not looking for good here. All we need is good enough.
Yup  I'm still betting on China.   (Although India may still end up being who designs it.)

[TaoPhoenix]

^yeah. But we're not looking for good here. All we need is good enough.
Yup  I'm still bet on China.   (Although India may still end up being who designs it.)

-40hz (September 06, 2012, 06:33 PM)

Well, at the risk of going off topic, "Creating Axis of ____ alliances is the way to transcend Religious Differences" (finally!)
Elsewhere Muslim Iran and Communist North Korea have decided that forming an alliance is more important than the religious differences between Islam and Communism. (!!?) Too bad that doesn't work so well for Islam-Christian-Hodgepodge USA.

So I'm all for a Chinese-Indian Alliance! (I have NO idea what that means GeoPolitically, it sounds kinda neat.)

[Paul Keith]

I think if good enough was the right criteria then everyone would be trying to make a Haiku PC as soon as possible.

Desktop is plagued by the fact that until Tablet OS like Android, most were blind towards what "wows" casuals and little was being done to fix the gaming hole found in Linux. I don't mean high requirement games either. No one was simply stamping their foot on Linux on gaming unless it was cross platform. Even the "great" Linux games never got sequels. Just constant updates of which it is more obvious for casuals to see on a software made for the Tablet than they are to spot the changelog being told to them via an update manager.

Then there's still this myth that casuals like LibreOffice. Just cause casuals don't need all the features of MS Office doesn't mean they don't want MS Word. LibreOffice is not Firefox and MS Office is closer to Photoshop than Internet Explorer. You can't just flood it.

Also: Windows XP Black tried darkhorse killer apps and it did get a ton more notice than Linux as a desktop but it kept being shut down because users suddenly found out that their helpful tech forum guy/mod does not allow someone to help fix warez compromised OS issues...even if it's still just Windows XP and so no one online barely helps anyone out...which is way worse than Linux forums/chatrooms.

[40hz]

So I'm all for a Chinese-Indian Alliance! (I have NO idea what that means GeoPolitically, it sounds kinda neat.)

-TaoPhoenix (September 06, 2012, 06:40 PM)

I'm guessing it would mean major changes in the world's economic patterns and spheres of influence. And that's just for openers.
                                                           

[Edvard]

I dunno, Red Flag Linux never really took off, even in China, and even after Nanchang internet cafés were forced to install it.

[mahesh2k]

Are you talking about registry changes, or the TCPIP.SYS patch to increase the maximum half-open connections (which was recommended for p2p uses)? There's a big difference. I've never heard about winsock registry fixes (apart from malware related crap), but I did use TCPIP.SYS patching back in the XP days. Stupid blind patching (that didn't check for correction version) could nuke your driver, which could definitely end up causing BSODs eventually.

-f0dder (September 06, 2012, 02:20 PM)

No. I am talking about this. I used to get IPMON related dll error. That page was not even there when protonic.com guys helped me solve the issue with regfix. That and some of the other fixes that I downloaded from MVP kelly (from site kelly's corner or something). Like this there were plenty of problems related to reg which bugged me and required fixes.

@Stoic joker, AFAIK, netsh fix was not told to me and I never found any written stuff on MS KB for that. I take that it was from SP 2 onwards? I guess that breaking of WINSOCK could be because of the AV (I am not sure what you mean by orphaned entries).

I'm still not convinced that anything you've put forth stems from problems with the registry or Windows in general - it does sound more and more like there could be some dodgy 3rd-party software involved, though (and by that, I include not just various OS patches, but potentially also drivers, antivirus programs, firewalls, et cetera).

-f0dder

Point is not to convince you for something. Point is that I had these problems with windows versions and I suffered, patched or fixed and finally I moved onto other OS which rarely had these issues. Again, if you guys wish to nitpick words and go on about it as if it's religious or political debate,  I used "rarely" word here to imply personal experience. Windows has issues, much more than OSX because windows is hardware agnostic and it breaks with every random component not tested during the deployment or testing phase, like modem driver killing something else on the system, reg entries getting corrupted, file copy issue(vista), VS causing issues with other programs and list goes on. These problems makes system less usable compared to other OS, that was my point. Ofcourse, this experience is personal, some people had similar to me and that's why MS KB exists at first place. And some people never had it, so they tend to  disagree.

I don't know which part of my post says - "OS X has no security issues or Linux is bug free". Even if I didn't wrote about it and people implying it and then hitting on that point makes discussion rather primitive where things are like - X OS is better than Y. That is not my point and If you're following this thread you know that I never made that point rather it was forced on me and was demanded explanation for, where in reality neither facts for any of the OS points to that domination which I am supposed to hypothetically defend.

[40hz]

I dunno, Red Flag Linux never really took off, even in China, and even after Nanchang internet cafés were forced to install it.

-Edvard (September 07, 2012, 12:36 AM)

I think that's because Red Flag (could they have picked a worse name?) was a little too heavy handed and obvious about what it was all about. China won't make that mistake a second time. The next Chinese state controlled Linux distro will be created in a university by a "club" of "unaffiliated students." There will be some minor concessions to internet freedoms. And that will allow the government to get what it wants while still keeping it generally palatable to the intended audience. The Chinese apparently have far fewer issues with authority than most people from my experience. A token concession or two towards personal freedom goes a long way in China.

And besides, India will probably design it anyway. It will be India's design (based on a US authored codebase), built in China, and selling like hotcakes in most of the world.

Maybe it won't be that popular in the USA and some parts of Europe. But why should that be a concern? The US economy is in trouble, and Europe's isn't far behind. So it's not like the US/Eur is going to continue to have infinite amounts of money to spend like they did in the old days.

In the meantime there are literally billions of potential first time computer buyers in the rest of the world. All that's needed is something "good enough" and "cheap enough" to become the new reference platform for personal computing.

"Necessity furthers. It is advisable to have a goal in view. No blame." as the I Ching says.
 
 

[wraith808]

Again, if you guys wish to nitpick words and go on about it as if it's religious or political debate,  I used "rarely" word here to imply personal experience.

-mahesh2k (September 07, 2012, 02:19 AM)

Not nitpicking words, but as everyone has said, is genuinely interested.  No one is attempting to attack- it's just that much of what you said is only anecdotal, so the only way to deal with it is to ask you to see how this could have occurred.  If you've moved past this and don't really have any way to substantiate it by anything that might be other than anecdotal and unwilling to/ don't want to discuss it just say so.  Everyone here was just trying to help it seems and see things from your perspective- not pick you apart.

OSes are just like the hardware they run on- tools- and each has its uses and problems.  And each person's experience with them is going to be individual.  But if you throw out phrases like "Those who are using Visual studio knows Why I am saying this, because microsoft's own programs make the system unusable," or "I hope you remember XP days with regular fixing of broken stuff," that doesn't sound either anecdotal or personal experience, so of course people are going to chime in.  Then if you change course midstream, you can't expect everyone to change course with you so unexpectedly.

[Stoic Joker]

@Stoic joker, AFAIK, netsh fix was not told to me and I never found any written stuff on MS KB for that. I take that it was from SP 2 onwards? I guess that breaking of WINSOCK could be because of the AV (I am not sure what you mean by orphaned entries).

-mahesh2k (September 07, 2012, 02:19 AM)

Actually it's mentioned in the Vista section of the article you posted above. I've been using it for so long I forget when I started. But I do believe it was back before XP's first SP. AV apps (and viruses) tend to inject an LSP (Layer Service Provider) shim (for lack of a better term) into the stack so they can analyze what is coming over the wire. However during the uninstall these LSPs tend to get forgotten and therefore are "orphand" leaving a big hole in communications. Simplest way I found to rip them loose is to do the netsh winsock reset (and sometimes a tcp/ip reset) to get things back to normal quickly.