I don't think that DLL Hell was commonly caused by device drivers. Moreover, what does driver signing have to do with DLL Hell?
TBH, It seems that most people who invoke the spectre of "DLL Hell" don't know what the hell they're talking about. You may not be one of those people. : D
MSFT does get some credit for resolving "DLL Hell" by following in libtool's footsteps with Side-by-Side Assemblies.;)
Heh. I bet that you could install Windows 98 on that same RAID array.
True story: I have a system with a Silicon Image 31(something or other) SATA controller. The WinXP installer didn't know what to make of it and required a driver floppy. The Win98 installer (and DOS for that matter) installed just fine. (Albeit very *slowly*. :/)
Maybe the community should just step up and write them? I mean they do it for Linux, why not Microsoft?
'Cause it's Microsoft. Really, there's no other reason than that. Why should we reward their reprehensible behaviour by adding valuable functionality to their systems?
If they don't have developers, their operating systems are useless. : D
This is the reality of the information carrying infrastructure in the United States. Things are not going so well for us here right now. Hopefully this will turn around sometime in the next decade; I am prepared to be mightily disappointed.
Yeah. I know. You live on an unpopulated island continent and must get *everything* over undersea fiber. That really sucks. I'm glad that I don't live there. (Or New Zealand. Ugh!)
Also, I know that most "anti-cap" examples thrown out there are hysterical, but here me out. (Bear in mind that I haven't eaten in a while, so this might ramble a little.:( )
We here in America (and elsewhere) are worrying about the future. Will these caps strangle things like IPTV? HD IPTV? VoIP? Home HD broadcasting? What as-of-yet unimagined things might these caps strangle? I know of at least one FOSS developer at RedHat whose work has been impeded due to Comcast's 250GB caps.
250GB is ~6 HD movies. Think about that for a minute. If I run my line at full throttle for four days, I've hit my 250GB cap for the month. [1] Have I ever hit the cap? No. Is it a hard cap like ATT has? Comcast says no. This is a marginally good thing.
But, why are American telcos imposing caps in the first place? How can Japan, the ex-Soviet Bloc, and India deliver uncapped 100Mbps FTTH at a reasonable price when we can't do the same in *ANY* of our major cities? Moreover, what happened to the shitpot of money that the telcos were given during the 90's? They were supposed to use it to build fiber, but from what I can see, this never happened.
Finally, we were (and probably still are) told that these services we were signing up for were "Always On" and "Unlimited". Being lied to really stings.
So, to summarize, and in conclusion: I've really gotta go eat something. Thanks for your kind attention. : D
[1] My average download speed is ~750KB/s: 768000 bytes in a KB ~87658 seconds in a day 268435456000 bytes in 250GB (768000*87658)/268435456000 = ~0.25079 Please correct my math if it's off.
Athlon XP 2600, 2GB RAM? Pentium III 850, ~780MB RAM?
Am I supposed to do this test with Windows, or xorg? If with xorg, what window manager? Finally, how many windows is many? 10? 100? 1000? Can they all be xterms?
My desktop machine has been dual-booting Windows and Linux since 2002, BTW. xorg and Windows "feel" equally fast to me, whether we're talking about 3D or 2D accelerated, or with the framebuffer.
I've seen Remote Desktop in action. In 2002, it was *marginally* faster than TightVNC. See the discussion upthread [1] for more details about VNC vs. X forwarding.
Putting a field into a struct is different from packing it. For instance, if you write an X server you need to support both endian-ness for all the numbers (clients can use either endian)....Furthermore, not all systems are so nice as x86...
So, if Windows or OS X supplied builds that supported more than just x86, they would have to write this duplicate code?
X11 *can* be more efficient than VNC, and VNC *can* be more efficient than X11... it really depends on the kind of app and what it's doing.
Eh, that is true. I remember X forwarding to be the faster way to run mplayer, but I'll test that out later tonight.
Also, VNC does not need to have a large delay. It does currently because the implementations are mostly focused on bandwidth and/or low memory use.
I probably misspoke. I don't think that VNC adds a large amount of latency. It is, however, either bandwidth hungry as all hell, or looks terrible ugly. (in general, YMMV, corner cases, etc, etc)
Ok I have to give you this one, because it's cool; however, the same comparison re VNC vs X11 above applies to 3d graphics. Sometimes it'll be faster, sometimes slower. It depends on the app.
So we can now say this with confidence: If an app needs *lots* of round trips [1] to the X server, VNC will be generally faster. The converse is also true. [2]
I don't use email.;-P
Then why are you using the internet?;)
[1] Not that I know what operations need lots of round trips. : ( [2] So, for what I do (remote web browsing and xterms over my cable modem) X Forwarding is faster than VNC. (nyah nyah nyah! : D )
Slashdot Mirror
I don't think that DLL Hell was commonly caused by device drivers. Moreover, what does driver signing have to do with DLL Hell?
TBH, It seems that most people who invoke the spectre of "DLL Hell" don't know what the hell they're talking about. You may not be one of those people. : D
MSFT does get some credit for resolving "DLL Hell" by following in libtool's footsteps with Side-by-Side Assemblies. ;)
Must shit on a WHAT?!?
I don't think that you know what you're talking about.
Cute! What're you using to generate that?
Heh. I bet that you could install Windows 98 on that same RAID array.
True story: I have a system with a Silicon Image 31(something or other) SATA controller. The WinXP installer didn't know what to make of it and required a driver floppy. The Win98 installer (and DOS for that matter) installed just fine. (Albeit very *slowly*. : /)
Maybe the community should just step up and write them? I mean they do it for Linux, why not Microsoft?
'Cause it's Microsoft. Really, there's no other reason than that. Why should we reward their reprehensible behaviour by adding valuable functionality to their systems?
If they don't have developers, their operating systems are useless. : D
*seconding the "antiquated?! wtf?!" sentiment*
Unless -of course- you're worried about not having gigabit LAN capability, or 802.11n capability... then I can understand the comment.
*points to my $60 WRT54G*
It has been running OpenWRT in my apartment for the past three or four years. I couldn't be happier.
What nice things do I get for 3X the cost of this setup?
Fuggin sweet! Thanks for this!
But in that case, you'd need to include the size of the source files that created the static libs you're linking in!
Odd. My install of Win 2k3 works in 2D with my old Voodoo 3 PCI.
poke me harder!!! i wanna see that v.h. fly!
This is the reality of the information carrying infrastructure in the United States. Things are not going so well for us here right now.
Hopefully this will turn around sometime in the next decade; I am prepared to be mightily disappointed.
Yeah. I know. You live on an unpopulated island continent and must get *everything* over undersea fiber. That really sucks. I'm glad that I don't live there. (Or New Zealand. Ugh!)
Also, I know that most "anti-cap" examples thrown out there are hysterical, but here me out. (Bear in mind that I haven't eaten in a while, so this might ramble a little. :( )
We here in America (and elsewhere) are worrying about the future. Will these caps strangle things like IPTV? HD IPTV? VoIP? Home HD broadcasting? What as-of-yet unimagined things might these caps strangle? I know of at least one FOSS developer at RedHat whose work has been impeded due to Comcast's 250GB caps.
250GB is ~6 HD movies. Think about that for a minute. If I run my line at full throttle for four days, I've hit my 250GB cap for the month. [1] Have I ever hit the cap? No. Is it a hard cap like ATT has? Comcast says no. This is a marginally good thing.
But, why are American telcos imposing caps in the first place? How can Japan, the ex-Soviet Bloc, and India deliver uncapped 100Mbps FTTH at a reasonable price when we can't do the same in *ANY* of our major cities? Moreover, what happened to the shitpot of money that the telcos were given during the 90's? They were supposed to use it to build fiber, but from what I can see, this never happened.
Finally, we were (and probably still are) told that these services we were signing up for were "Always On" and "Unlimited". Being lied to really stings.
So, to summarize, and in conclusion: I've really gotta go eat something. Thanks for your kind attention. : D
[1] My average download speed is ~750KB/s:
768000 bytes in a KB
~87658 seconds in a day
268435456000 bytes in 250GB
(768000*87658)/268435456000 = ~0.25079
Please correct my math if it's off.
What does this have to do with the man's limo driver!?
...unless you're of the worse-is-better school.
Well, given that I use Linux at home...
Is it an inkjet printer? Did it cost less than $150? ;)
If so, it's likely crap.
Meh. The Adobe-supplied Firefox Flash plugin works almost all of the time, but is randomly crashy. /me weeps in frustration.
What's your wifi device?
What's a low-end machine?
Athlon XP 2600, 2GB RAM?
Pentium III 850, ~780MB RAM?
Am I supposed to do this test with Windows, or xorg? If with xorg, what window manager? Finally, how many windows is many? 10? 100? 1000? Can they all be xterms?
Cheers!
Simon
Do you have a good benchmark to back this up?
My desktop machine has been dual-booting Windows and Linux since 2002, BTW. xorg and Windows "feel" equally fast to me, whether we're talking about 3D or 2D accelerated, or with the framebuffer.
I've seen Remote Desktop in action. In 2002, it was *marginally* faster than TightVNC. See the discussion upthread [1] for more details about VNC vs. X forwarding.
[1] http://slashdot.org/comments.pl?sid=1017147&cid=25630607
Anyone with an early WinMo phone can attest to the suckyness of the Windows UI on that mofo. : <
Putting a field into a struct is different from packing it. For instance, if you write an X server you need to support both endian-ness for all the numbers (clients can use either endian). ...Furthermore, not all systems are so nice as x86...
So, if Windows or OS X supplied builds that supported more than just x86, they would have to write this duplicate code?
X11 *can* be more efficient than VNC, and VNC *can* be more efficient than X11... it really depends on the kind of app and what it's doing.
Eh, that is true. I remember X forwarding to be the faster way to run mplayer, but I'll test that out later tonight.
Also, VNC does not need to have a large delay. It does currently because the implementations are mostly focused on bandwidth and/or low memory use.
I probably misspoke. I don't think that VNC adds a large amount of latency. It is, however, either bandwidth hungry as all hell, or looks terrible ugly. (in general, YMMV, corner cases, etc, etc)
Ok I have to give you this one, because it's cool; however, the same comparison re VNC vs X11 above applies to 3d graphics. Sometimes it'll be faster, sometimes slower. It depends on the app.
So we can now say this with confidence:
If an app needs *lots* of round trips [1] to the X server, VNC will be generally faster. The converse is also true. [2]
I don't use email. ;-P
Then why are you using the internet? ;)
[1] Not that I know what operations need lots of round trips. : (
[2] So, for what I do (remote web browsing and xterms over my cable modem) X Forwarding is faster than VNC. (nyah nyah nyah! : D )
Heh. Copy and paste seems to work just fine between GTK apps. Text styles and hyperlinks seem to be preserved.
Regardless, do you know how difficult it is to create a new window manager for MS Windows or OS X? I don't, that's why I'm asking! : D