Slashdot Mirror
Vista's 'Next Gen' TCP/IP Stack
boyko.at.netqos writes "Microsoft's new Vista TCP/IP stack might be beneficial to businesses looking to increase use of their IT infrastructure... if they did it right. Ted Romer at Network Performance Daily writes: '[Vista] now allows us to throttle outbound traffic at a client or server. For example, you can throttle the bandwidth of a particular subnet to a particular server, giving some departments more access to the servers that they need. You can even restrict outgoing bandwidth for certain peer-to-peer applications like bit torrent. This shaping can also be handy when applied to servers, allowing less bandwidth for certain users/departments, and more for others. While consumers may debate whether Vista is a worthwhile upgrade, I believe it to be important for enterprise customers who will best be able to put Vista's capabilities to their fullest potential. Of course, I'm getting it for DirectX 10 games, but that's just me.'"
Why not just put a linuz router with wondershaper in between?
'Once scientists, even the dim-witted social scientists, get muzzled, the Western Civilization is finished.' - oldhack
"redesigned from the ground up"
668: Neighbour of the Beast
Using TCP/IP stack fingerprinting...
...let me choose how much bandwidth to allocate to each app, and their relative priority? I want my browser to go first, then Google talk, then any updates (OS, virus checker, firewall) and finally P2P. It's quite annoying that I can't do that on XP. Perhaps it's a tricky problem though.
because it sure reads like one
Microsoft is desperate to get business interested in their Vista product so will trot about all manner of reasons to buy it, but business are not biting, unless this Vista can make workers type faster or calc spreadsheets quicker or email faster than there is NO productivity gains unless wowing the coworker with a 3D AIGLX/Beryl like desktop counts as productive
if an Enterprise is worried about client bandwidth they would already be using a tool dedicated for the job like, say a Router
So that you can't download the latest Linux iso's on bittorrent. And hopefull it could be an open standard and not another ms only product. Then the whole interweb has it and we're screwed.
OK - it is nice, but it certainly is not new.
Of course, I'm getting it for DirectX 10 games, but that's just me
Just you? Wow, those will be some disappointing sales figures.
If that guy in accounting is spending all his time downloading movies off from bittorent, wouldn't it be better to fire him instead of shaping his packets??
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
The innovation! What will MS think of next?<sarcasm>
GRC | Security Now! Transcript of Episode #51 "Vista's Virgin Stack" http://www.grc.com/sn/SN-051.htm
"A gun is a tool, Marian. No better, no worse than any other tool. An axe, a shovel, or anything." Shane (1953)
Big deal. ______ has had this in the kernel since ______.
The difference between stupidity and genius is that genius has its limits.
If you have Windoze Vista and want to surf the web or download email, the server must be running and Windoze server software.
Throttle network traffic is great ... if either your network or your server capacity suck.
I hadn't the slightest objection to his spending his time planning massacres for the bourgeoisie... (P.G. Wodehouse)
Yay! Now people will hopefully fix their firewalls so I can turn those on again in my Linux boxes.
Finally! A year of moderation! Ready for 2019?
I'm having problems with Vista's DNS Suffixing:
s tID=884630&SiteID=1
- My entire network has the dns suffix : work.intranet and I have a BIND dns server that resolves a.work.intranet and *.a.work.intranet to 192.168.0.2 so that if I ping bbb.a.work.intranet or ccc.a.work.intranet they all resolve to 192.168.0.2 (at least up until Vista)
- If I ping a.work.intranet it correctly resolves to 192.168.0.2;
- If I ping a it correctly resolves to 192.168.0.2;
- If I ping bbb.a.work.intranet it correctly resolves to 192.168.0.2;
- BUT IF I ping bbb.a it no longer resolves. (could not find host)
- If I do a nslookup bbb.a it correcly resolves to 192.168.0.2
So what appears to be happening is that it isn't adding the dns suffix when the domain has more than two parts (xxx.yyy).
Any ideas how to solve this?
http://forums.microsoft.com/MSDN/ShowPost.aspx?Po
http://blog.ebserver.org
What does Vista TCP/IP do that http://en.wikipedia.org/wiki/Ipv6">IPV6 cannot and I don't mean such feetures that are welded to the Vista API.
davecb5620@gmail.com
Rather interesting that the quote in the summary here on slashdot skipped this (emphasis mine):
... it really doesn't do much. It may be slightly more convenient to configure QoS on your routers based on the tags rather than port numbers ... but that's about it.
FTFA: "Vista's ability to use centrally configured group-policies to push out policies to specific users or servers, and allows tagging of packets with the Diffserv code point values, so that our network infrastructure can see the marking and react to it in different ways - whether it's VoIP traffic, or TCP/IP business critical traffic, or web-surfing traffic. (Granted, this QoS doesn't guarantee anything, it just marks the packet in Windows and it is up to your network infrastructure to honor those tags.)"
So
- Roach
...of iptables or netfilter did they steal from Linux to make up their new traffic-shaping capabilities that has been in Linux and other *nix variants for years?
Woot, MS just got a bit more professional in their offering. How nice.
Bandwidth management _must_ not rely on the host's cooperation. All will work beautifully until a virus totally rapes the network because QoS responsibility had been shifted from the network to the hosts. Damn, this isn't just stupid, it's freaking pathetic. What next Microsoft, pull in layer 2 into your stack as well?
QoS requires support from your network hardware.
The Internet doesn't have that.
Note also QoS doesn't actually solve all problems. For example, if you have two network applications running, and you want one of them to have priority such that it can take bandwidth from the other when it needs it - well, you're out of luck. QoS doesn't handle that situation.
...throttle users not traffic.
Regards, Phil
We really should not view any improvement in the Borg IP stack as a bad thing. They have already assimilated the world, the least they could do is provide a quasi-robust core set of features w/ some improvement in stability.
There are large amounts of small companies that probably do not have campus grade layer 2/3 devices that can accomodate QoS or traffic shaping, so adding the functionality at an end-point is honestly a good idea. However I will be the first to admit that it is a little late in the game, but they can now check another box from the list when comparing their product to other offerings.
When the only tool you have is a hammer, every problem looks like a nail
The Compound TCP talked about in TFA is disabled in Vista by default. If you want to turn it on, you can open a console with admin privs (right click Command Prompt -> Run as Administrator) and enter:
This was one of the first commands I ran after Vista installed, and the difference is noticable.
The network has different characteristics depending upon what point you are at on it.
... and then put a shaping router on the WAN links.
The WAN routers see the low bandwidth, higher latency serial links and such.
The servers/workstations see the high bandwidth, low latency ethernet links.
Do you really want your server(s) calculating its(their) window(s) based upon whether the request is originating across the WAN or next to it on the LAN?
This sounds like a good idea when you're talking about a single workstation, at home, connected to a cable connection or xDSL or whatever. But it sounds like soooooo many problems in the corporate environment.
Right now it is easy to find the server/workstation that is flooding the network. It's going to be very difficult when you have hundreds(thousands?) of machines that are ALL trying to maximize their bandwidth usage.
Personally, I'd prefer the ability to set the LAN parameters for the machines
Many people on Slashdot have been screaming for over a year that Vista doesn't offer anything new, nothing worth upgrading for, etc. Well, it seems to me it does. I think anyone who is intellectually honest would agree... I might be on the wrong site?
This article points out 1 cool thing, a new networking stack, but it isn't the only thing. And actually he doesn't even talk about IPv6. For example, my lab at home I has 3 Vista installs, and the communicate out of the box over IPv6. In a couple of years IPv6 will be main-stream because of MS, and we all know the benefits from using the upgraded protocol.
-I think it's cool that when you browse the network people can see a picture of the person instead of the Computer Icon.
-I also do photography, and you use to be able to open an image file on an OSX machine and XP and it would look better on the OSX box. Not with vista.
-For Remote Access: PNRP. Again, really cool... do a search if you don't know what it is.
-Even the average business user will benefit from little things like the snippet tool (prety cool by the way, it's in the accessories folder if you haven't tried it yet).
-I have clients that are going to love the way the Windows clock works now. They can jump around by month, year, or decade in seconds. Those little things are pretty cool.
These are just random features that popped into my head, but it seems that Vista has LOTS of things other than Aero to encourage upgrading on all fronts (Security, tools, toys, looks, games, etc.).
Seriously, apple announces multiple desktops and have this site has a heart attack.... then praises Steve Jobs for being an inventor, a genius, etc. Meanwhile Windows has had those features for years, hell, Unix has for decades!
MS may not have invented the notion of every new feature in Vista, but it's a good product, and way better than XP. A worthy upgrade. It's not one feature that makes it a good product, but the cumulative of many features. I think you anti-ms people lose a lot of credibility when you blindly bash MS and say Vista sucks and it offers no reasons to upgrade for anyone. For all users it has some pretty enticing plusses.
No, but you can do that in XP and (presumably) Vista using the excellent Netlimiter program.
http://www.netlimiter.com/
Back when I worked at an ISP with a shared bandwidth broadband solution, we would politely suggest that to the college jackasses downloading Bittorrent without setting it up to be network friendly (like we asked/demanded/etc). It really does work quite well.
So windows now has as part of the default configuration a tool that allows you to control the performace of different networking applications....
How long until malware takes advantage of this to slow down your connection (litterally) until you pay for their new "tool" to fix your system....
I hope vista security on this feature is well designed, otherwise this is an easy target for malware to trick more naive users into parting with their cash.
Though to be fair, with this feature and others in vista MS is actually approaching having a real OS - rather than just a GUI + disk manager... I wonder which BSD licensed code they used....
$_="Slashdotter";$syn="OTT";s;..;;;sub _{print shift||$_};s!ash!Perl !;s=$syn=ack=i;tr+LLEd+BLAH+;_"Just Another ";_
I've been using L7-Filter, on my Linksys router of all things, for a couple years now and it performs very nicely. Can have gnutella/G2/ED2K/BitTorrent transfers happening and still get 20ms pings in Counter-Strike with a game as smooth as it would be without any of the P2P stuff going on. It is absolutely great.
It's a big deal because now, viruses and malware can slow your network access automagically, so that it'll take weeks for you to download those security patches and antivirus signatures that you should've already downloaded. :-)
I wonder if it will have a checkbox to allow you to throttle the traffic of the next zero day worm.
Another great reason to reboot your workstation with a Knoppix DVD. Problem solved.
All of you are saying that linux has this and ahrdware does this. What they dont get bashed but microsoft adds the feature and you right away bash them. Doesnt linux have it too? I dont see why people dont bash linux either. What people dont want a microsoft os to get better? why all the hypocritical bashing?
Is it just me or is this article describing abilities Linux + IPTables has had for years upon years?
Hell my Linksys WRT54G with modded firmware at home does application-based traffic shaping (no way I would be able to use Vonage reliably while downloading huge ISOS if it didn't).
How is this next generation?
"I think you're confused here. This is just a specific implementation of TCP/IP, not a different protocol or anything"
..
Then please go ahead and enlighten me. What does Vista TCP/IP do that requires a specific implementation
was Re:IPV6
davecb5620@gmail.com
Microsoft astroturf in action.
Ahh, more Microsoft bashing.
To paraphrase a quote on bash.org, one day I'm going to invent a device to stab people in the face over the Internet and get very rich. Oh, and you'll be my first target.
The bandwidth throttling may not be a big deal to you, but on high bandwidth high latency links you can get huge performance improvements (i.e. 10-100x) with proper use of TCP window scaling. In the original TCP spec the window size could be no more than 64 KB, but this behavior was later amended and a TCP option was added to allow you to increase this value.
The optimal window size is (Round Trip Time)*(Bandwidth). For my internet connection (600 KBps) that means that a 64KB window is only adequate for sites whose ping time is no greater than 110 ms. For sites with a higher latency, the amount of bandwidth I can get in a TCP connection between me and this host is artificially limited by my TCP window size.
Right now it generally isn't possible to get a reliable connection after increasing the window size past 64 KB because some older/cheapo routers will not work with TCP windows greater than 64K. But if this gets into Vista and TCP window scaling options started getting heavy use, there would be a lot of pressure on sites with broken routers to get them fixed, and then those of us with high bandwidth connections would reap the benefits.
#include ".signature"
Interesting. I have tried to setup two subnets on the same physical interface (WLAN) in Windows XP Home... (wanted to have localnetwork + internet): no way - only one network at a time! A friend of mine (network professional) told me that its possible but very hard to do. Is this going to be solved in Windows Vista? So, what is this talk really about? Why Windows users even need to know about subnets let alone shaping traffic?
Well
On this week in tech someone commented that this new stack had a lot of identical bugs bugs that had been fixed years ago in the older stacks.
You will never get to heaven with an Ak 47... But A Zu 30 is good for Low Flying Cherubim
If you consider those gimmicks "cool" then good for you. But I don't see any bottom line productivity gain.
Looking through the first Halloween document over at http://www.catb.org/~esr/halloween/halloween1.html #_Toc427495768 I noticed this familiar tactic.
(quoted from the document)
"Blunting OSS attacks
Generally, Microsoft wins by attacking the core weaknesses of OSS projects.
De-commoditize protocols & applications
OSS projects have been able to gain a foothold in many server applications because of the wide utility of highly commoditized, simple protocols. By extending these protocols and developing new protocols, we can deny OSS projects entry into the market."
No thanks Microsoft, I'll stick with standard protocols.
Why is this called "next-gen"? There is nothing "next-gen" about this. If anything, Microsoft is finally catching up with the rest of the world in this department.
Such stuff was possible with Linux (and, I'm sure, BSD) servers for years. I know for sure because I used to have such a setup (to do traffic shaping on our -then- relatively slow internet connection shared by too many people) on a Linux server, more than 5 years ago!
Please stop this silly use of marketingspeak of calling something "next-gen" when in fact the company under consideration is just finally catching up with what the rest of the world has been doing for ages.
Every expression is true, for a given value of 'true'
"easier to use" means "requires less knowledge", then Linux might not be "easier to use". But if "easier to use" means "consistently behaves the way a knowledgable person expects", then Linux is much "easier to use".
s x_adaptor_en.php>googled web page]:
... [Yeah, good luck downloading ./config ./make ./make install, by now my girlfriend was asking me when were we going to play the darn thing ]
Just read at what you just wrote, "consistently behaves the way a knowledgable person expects", WTF is a "knowedgable person"? one that knows how to use the system? if that is, then both systems are "easier to use" to the people that already know how to use it DOH!!
On the other side, just to backup what the other poster said, and as a proof of the "easiesterst" of use of Linux, see the difference in "easiability of use" between Linux (Ubunut 6.10) and Windows (XP SP2) when wanting to connect a PSX Dance Mat using a PSX 2 Parallel port adapter [I *tried* to do this some weeks ago and gave up and installed it in my girlfriends Windows XP machine):
Windows XP:
- Download PSXPAD program
- Run program
- Click NEXT button until it changes to "Finish" button
- Click Finish button.
- Go into the Control Panel/Joysticks/PSXPAD/Properties window and select the option to treat axis as buttons
Ubuntu Linux[via http://www.raphnet.net/electronique/psx_adaptor/p
- Open console (woah! 80% of the users would have installed Windows by then)
If the gamecon driver is compliled in your kernel (head explodes):
# modprobe gamecon gc=0,7,0,0,0,0 [ Note that, after several head explosions and hours of google search you realize that you should write 8 instead of 7 to enable "dance mat" compatilibity there]
If you have a rescent kernel, try this instead:
# modprobe gamecon map=0,7,0,0,0,0
If the module does not exist, you will have to compile it yourself. When you configure your kernel, select the
Then of course you connect the dance mat [wireless] and it does not work, it just sits there, i tried mod probing enabling disabling and what not without sucess.
In summary, I use Linux for everything [I use Fedora Core at work, and Ubuntu at home and ssh -Y quite often to a RHEL server) but for the love of god leave those blatant lies to comp.linux.advocacy fanbois
Ubuntu is an African word meaning 'I can't configure Debian'
And just to emphasize my point, we are not talking about an obscure hyper closed source driver here, it is a driver that is already *supported* in the kernel. WHY the FUCK is it SOOOOOO cumbersome to connect and use a dance mat in Linux while in Windows XP [which does NOT support the hardware and you have to download a THIRD PARTY DRIVER] is as easy as [lots of things] double click, next next next, finish.
Soo? any comments? Linux zealots please display your fury ahaha.
p.s. As I said I work and develop software [Mainly Java but I do some C++/wxWidgets/OpenGL code which I always try to make 100%] portable in Linux. My everyday platform is Linux, I recently bought Lemmings Revolution game and plan to spend a weekend making it work with wine, but for the love of god, IT IS NOT AS EASY TO USE AS WINDOWS.
Ubuntu is an African word meaning 'I can't configure Debian'
Why doesn't the uevent consumer (e.g., hotplug or udev) load the appropriate kernel module automatically when the dance mat is plugged in?
I think the main reason for the negative reaction is the title:
If MS really was claiming it was 'Next Gen', we'd have a problem. Of course, they haven't, because everyone else has been supporting this for ages. So, as usual, MS is the last to implement it, while somehow pretending they invented it.
Only this time, it wasn't MS pretending that, it was whatever moron thought up the Slashdot title.
Don't thank God, thank a doctor!
Lack of interest.
Someone interested in SELLING you a dancemat has done the work for you in the WinDOS environment.
Networking, OTOH, is not that sort of thing. It's been well supported in Linux before there was any TCP/IP libraries even included in Windows.
A Pirate and a Puritan look the same on a balance sheet.
Not for Vista, but for net neutrality.
Think about it: You no longer need special software, you no longer need to mess with your router, you just get a "Vista-ready" router, or an ISP that cooperates, and you have the QoS built in to the OS, probably in an easy enough way that there's no longer a question of your ISP having to enforce some draconian policy over your own Internet usage.
Don't thank God, thank a doctor!
Just read at what you just wrote, "consistently behaves the way a knowledgable person expects", WTF is a "knowedgable person"? one that knows how to use the system? if that is, then both systems are "easier to use" to the people that already know how to use it DOH!!
If you actually read what he said instead of just making fun of it, you'll see that he was discussing how "easier to use" was not necessarily limited to your notion of "requires less training". His point was not that linux requires less training, or that someone with know prior knowledge would be able to run linux more easily than windows. What he said was that "behaves consistently" might be a more important part of "easier to use" than "requires less training" for someone who already has knowledge of the system.
You also whine about having to google to find the proper command to enable the driver for your mat. It's a little unfair to ignore the time you spent googling to find PSXPAD. Or to pretend that copying and pasting a search result to run a single command from a prompt is more difficult than downloading and installing a program. They are different tasks, but I have trouble believe that search->copy->terminal->paste is "harder" than search->download->execute->follow prompts.
As far as a feature goes, it's really a complete non-story. Aside from the inefficiency of shaping traffic on every client computer as opposed to the router, it's also the sort of thing that you can do with ease on linux.
For a home router I recommend taking a look at the Fairnat script http://www.metamorpher.de/fairnat/ although it doesn't include rules for firewalling - you have to do that yourself.
To take full advantage of traffic shaping a connection on linux for users who use p2p it's a good idea to get a version of iptables (or compile your own) with the ipp2p module. It can match against pretty much all p2p protocols that people use (bittorrent, edonkey, soulseek etc).
That'll allow you to prioritise your interactive traffic (ssh, IRC, instant messengers), let your mid-priority traffic (HTTP, IMAP, POP) come in and then put your bulk transfers and p2p lower down in terms of priority.
If you've just got the one machine connected to a broadband modem then take a look at the wondershaper http://lartc.org/wondershaper/ script which will shape traffic on your machine.
Writing a new stack from the ground up is a VERY BAD IDEA from a security perspective. Why? Because you have no idea what weaknesses, exploits, buffer vulnerabilites, etc are written in your brand new stack. The stack in WinXP was really good for only one reason - Microsoft simply copied a UNIX stack that had been around for ages and had already had a ton of bugs fixed.
Now, all that is scrapped and written from the start. Who knows how many mistakes that were already made and fixed in the past were re-made in Vista.
"Trying is only the first step towards failure." - Homer
I'm using two network subnets on same ethernet segment without much trouble (except ICS and related DHCP problem...) in XP. Masquerading connection to DSL link (DSL modem is awful, freezes under heavy load when in NAT mode).
Isn't this what routers already can do for the past ohhh 7+ years? Why is Micro$oft spending time & resources in implementing something on a server that is actually the job of something at the network layer? Also "shaping entire subnets" is not doable from a server, the only thing that can traffic shape an entire subnet is the router that actually controls all packets going in and out of the subnet... a server at the edge can't control other IP device's traffic in the same subnet.
Microsoft should first fix their internal OS issues before spreading their ambitions into #1) Areas they are not experts in #2) Areas that are much lower on the OSI layer.
The function of an OS should be to provide stability & functionality to the applications that depend upon it. Trying to control all network traffic is going outside of that scope and I have doubts it can ever do a better job than a router.
Ultimately I would like to see MS & Network vendors work together. MS can allow users to MARK (QoS) the packets for specific applications, but then let the routers do the bandwidth shaping. This to me makes a heck of a lot more sense.
Adeptus.
No trees were killed in the making of this post; however, many trillions of electrons were horribly inconvenienced.
I don't have any experience with PSXPAD in particular, but in situations like this, I'm generally quite paranoid about downloading a random .exe and running it on my Windows machine to get some random functionality. Usually, it's only available as a compiled binary, and the site that hosts it is a template php site that could be put together in an hour. If one is to put in the due diligence to make sure such a file is legitimate and/or to repair their newly created WinZombie(R) when it's not, the difficulty of installing measured in effort and expertise required is much higher for the Windows machine.
Chances are that any IT group that would manage to use this feature w/o f*cking up, already has tools in place to do it. As useless as one-way firewalls.
Well, that depends. You can set the ToS bits in the IP header. What actually happens depends on the router. If you have one application set to "Maximize bandwidth" while the other one is set to "normal service", AND the router looks at that and actually does what you asked for, you can get it to work that way.
Or you can use a Linux box and set up traffic shaping, in which case you can divide bandwidth in any way you want. HTB makes what you say easy to do. Say: 20% for server A, 50% for server B, 30% for server C, servers A and B can use the full pipe if nothing else wants it, and C is always limited to 30%.
ipfw pipe 1 config bw 300Kbit/s
mod me funny
"They can certainly communicate with Vista happily, with no problems whatsoever (one of the requirements of Compound TCP is that it doesn't break the TCP standard, unlike some other ways of getting past this problem). If they wanted to implement the same features, I think at least some of the features might be patent-protected. Which is a bit rubbish"
Now come off it, you know specifically what was asked - can a non Vista OS provide full functionality without a license. Not whether it could communicate or if Compound TCP would break TCP/IP standard. Which features are not patent protected and does it make a difference to a developer if he can't impliment his own full version without violating some patent. Why patent a specification to TCP/IP at all.
You call me confused yet when I ask you to give some details backing up your statements you respond with 'not entirely sure' and answer the question I didn't ask. Some earlier not answering the question samples
Q. "Is such functionality not in IPV6", RS
A. "I think you're still a bit confused here", zootm
was Re:specific implementation
davecb5620@gmail.com
Wow! If your best means of proving that Windows is easier to use than Linux, is some corner-case about a Dance Dance Lemmings gaming peripheral, then Linux is even closer to WorldDomination(tm) than I thought.
Censorship is telling a man he can't have a steak just because a baby can't chew it. --Mark Twain
The way I see this working is that in your enterprise environment you come up with your default workstation image. The image contains various applications. ERP, CRM, web browser, IM client, mail client. You configure QoS tagging via group policy so that the ERP package gets 1st priority, CRM package gets 2nd, mail client 3rd, IM client 4th, web browser 5th (yes I know that you don't tag packets like that). The workstation connects to the switch, the switch recognizes the QoS tags without having to inspect the packets and traffic gets around the network better.
What is there to bitch about here? QoS is the future that is here now. Everyone who cares about getting the most out of their network is looking for the best way to impliment QoS. The fact that Microsoft is letting administrators do it at the workstation, based on specific applications, controlled via an enterprise wide mechanism look Group Policy is a GREAT implimentation of QoS. Just because it isn't some ubah-l337 phresh no-day implimentation of QoS doesn't mean it isn't a good idea.
It seems to me like the majority of people who have replied to this thread are looking at the world from their home PC. They don't see how QoS is going to help them get more out of their DSL line. You know what, it won't. As I've said in other posts, Microsoft doesn't really give two shits about your home machine. They are focused on helping enterprises make the most of their 1000+ workstation deployments. As a network administrator, I'd much rather that the workstation tell the network what applications it is running and tag those packets, instead of having to configure QoS policies based on subnets, VLANs, and other arbitrary delimiters.
"Networking, OTOH, is not that sort of thing. It's been well supported in Linux before there was any TCP/IP libraries even included in Windows."
Hmmm. Windows NT shipped (and existed for a year prior) at the same time as the very first Slackware distribution. Only the most hardcore would claim that Linux at that time supported anything "well". No one was rolling out their IT infrastructure using Linux at that time.
Well, what's here?
Fooling around with TCP slow start and congestion window management. That's nothing new. We were arguing over that back in the 1980s. (I'm the inventor of the Nagle algorithm and fair queuing, and was the first to describe congestion collapse and the tragedy of the commons problem in networking. See my old RFCs.) Back then, we were more worried about causing transient network congestion; if you didn't have slow start, you could lose packets at the gateway from the LAN to the WAN because there wasn't enough memory in the router. And we were really worried about congestion in the middle of the network backbone. So the early versions of those algorithms were on the conservative side. Nobody worries about protecting the network in desktop clients any more, and the Internet now has more backbone capacity than edge capacity. Strategies in those areas have been more aggressive for years.
Using the quality of service field. That's not new. The main issue is how to express priorities between the applications and the operating system. It's probably good enough to give video and audio higher quality of service. (Gamers who are trying to play streaming audio while fragging might complain about lag, so there will probably be some obscure registry key to mess with this.)
Multithreading the network stack so it will use multiple CPUs. Reasonable enough, but not new. QNX has had that for years. Since its network stack runs in user space, put QNX on a multiprocessor and the network stack speeds up. Cisco uses QNX in their big routers. On the consumer desktop, it's not going to be noticed; your DSL line or cable modem is nowhere near fast enough to need this.
Given Microsoft's usual poor code quality, we should all be cowering in fear. The IP stack is something that needs to be battle-tested for years before we get comfortable with it. Uncle Bill and his minions have chosen to inflict an unproven stack on us for the sake of a few bells and whistles.
This is another fine reason to delay your Vista "upgrade" until at least the second service pack -- assuming you upgrade at all.
I'm taking bets on how many critical patches will be on the ip stack this year (2007).
It's Linux, damnit! Pay no attention to renaming attempts by self-aggrandizing blowhards.
You allow server admins to throttle their servers, you're allowing them to hose up the switches which in turn causes delays all around the network and then network has to join confrence calls when the server people are saying its a network problem when it's their own damn fault. no thanks
well thats an obscure device, the thing is, you spend much less time maintaining linux, on my windows partition i have to reformat and reinstall every couple of months just to keep things working, simply because its easier than tracking down the problem. in linux the only time i have to actually do anything to make things work is when i install a newer version. installing binaries is incredibly easy. i have a whole directory full of updated packages that i keep in case i want to install a different version or what not, after a fresh install i just cd to the dir, type installpkg *.tgz, and presto everythings ready to go. ease of maintenance, takes about 30 minutes for the whole process, in windows the same thing would take god knows forever. switch to slack youll learn how to use linux, and you will never look back.
> Just read at what you just wrote, "consistently behaves the way a knowledgable person expects", WTF is a "knowedgable person"? one that knows how to use the system? if that is, then both systems are "easier to use" to the people that already know how to use it DOH!!
That's not true at all. Windows consistently behaves in a rather flaky manner for me. I have programs made by Microsoft that hang for no reason, give no indication of what's wrong, and there's nothing I can do except restart the program and hope that the problem doesn't occur too often.
Once you get something working in Linux, it tends to *stay* working, not randomly flake out on you for no apparent reason. Yes, it might be harder to get things working initially in Linux, but you pay more in the long run in using Windows when you have to do all sorts of crap to work around the general flakiness.
(see subject)
http://outcampaign.org/
Cool, now ms can degrade network acceess to applications they don't favor, oh, say Itunes, for example. I'm sure they would never try to leverage some kind of control like this against their competition.
Fortunately none of my clients is even remotely interested in upgrading to Vista, so this might not be an immediate problem. They use XP because they need to, and offering them a "new" product, that was considerably behind schedule (hence problematic), from the same company doesn't really appeal to them.
Usually, I hate this sort of comment, but I can't help it. Since decent OSs have had this sort of feature for at least a generation (or more, in some cases), shouldn't this be a "Three Generations Old" IP stack?
steve
Oh, you're not stuck, you're just unable to let go of the onion rings.
This would be like bringing up Windows 1.0 in a "who's gui came first discussion".
Slackware's support for serious networking was actually very repectable in those days. I was a Slackware user then and even used ip masquerading to share a serial connection in those days.
A Pirate and a Puritan look the same on a balance sheet.
Which will do for most apps.
If you need text styles to communicate then you don't have a message.
Instead of trying to get each app to use different IPs, when most will just use the IP on the same subnet as your default gateway, use users. Replace "firefox" with a shell script that does "sudo -u firefox_user firefox", and write your firewall rules for traffic shaping based on the users. This is also a good idea in general so that if you get hit with a malicious site exploiting firefox, it can't delete/change any of your data.
I can count all the people affected my that little "problem" on one finger.
Guess which one.
So, vista now has traffic shaping...
Linux has had this for years, as i`m sure have many other OS's, i've been using the linux traffic shaping facilities to deprioritise bittorrent traffic and prioritise voip traffic for quite some time.
As for the stack having been rewritten from the ground up, it's amusing to see them make lots of the same mistakes that have been made in other tcp stacks when they were newly written too.
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
..that, AND
your ISP can throttle those precious VOIP and bittorrent packets down to near zero (since the traffic is signed) in an effort to encourage use of their own telephone and video services.
I suppose this software package would also be worth having a look at. Supports OS/2, Linux, and Windows OS's.
Administration
Superior Protection
Access Management
Traffic Shaping
IPSec VPN Support
NAT Gateway
http://www.fx.dk/firewall/shaping.html
Bandwidth throttling over particular subnets, stateful packet inspection, the ability to inspect every part of a packet and filter/route based on however many rules you wish to apply to it.... has been in Linux since 1999... only 7 years now. Oh, and the Linux distro is free and stable and tested over years of service. But thanks for asking...
The problem is how to setup it thru control panel. One network uses DHCP, another one static IP...
'Your wording is consistently confusing to me. You're not asking proper questions. Define "full functionality"'
..
You define your understanding of the meaning of "full functionality" in the context of IPV6 and Vista TCP/IP.
Rules of non-debating tactics:
# 47: Pretend to not understand what the other fella said and accuse him of being confusing.
# 48: Pretend to misunderstand the meaning of commonly used phrases, for instance 'full functionality'.
# 49: Move the discussion to the meaning of particular words, for instance 'full functionality'.
# 50: Personally insult the other fella.
re: Mr. 'objective third-party': You're not on Usenet now so do try and be polite.
was Re:full functionality != can communicate
davecb5620@gmail.com
Color management? Expensive calibrated monitors? A broken JPEG decoder on the Windows box?
There are legit possibilities, but it's impossible to know if he meant anything like that or was just grasping his Mac in a religious fashion.
Laws do not persuade just because they threaten. --Seneca
"This would be like bringing up Windows 1.0 in a "who's gui came first discussion"."
It's nothing like that. It's an undisputable fact that MS offered Windows products with integral TCP/IP software at a time when Linux was still in its infancy. The claim that Linux offered networking for years prior to MS's support of TCP/IP is just plain wrong.
"Slackware's support for serious networking was actually very repectable in those days. I was a Slackware user then and even used ip masquerading to share a serial connection in those days."
I would venture to guess that Windows NT 3.1 was a more robust server platform than Slackware 1.0. Nevertheless, they came into existence within a month of one another. Considering that Slackware was the first real distro and NT was available for a year prior to release in developer kits suggests that the statement was utterly false (regardless of how useful you personally found Slackware in those days).
I didn't start using Linux until Red Hat 5.x and even at that mature (for linux) date I found it to be a joke. I immediately dumped it for FreeBSD which was far superior. I wouldn't consider even by RH5 for networking to be "well supported in Linux" in my experience.
I've been asking you for clarification! ..
.."What does Vista TCP/IP do that IPV6 cannot and I don't mean such feetures that are welded to the Vista API", RS
.. (Score:2)
Rules of non-debating tactics:
# 51: Ignore repeated requests for clear answers.
# 52: Accuse the other fella of not being clear.
# 53: Move attention off the original question.
# 54: Incorrectly re-state what the other fella said.
You know bloody well what I mean
I'm sorry, your being very confusing and I'm going to have to ask you for clarification. I will of course mistate what you said and proceed to dissect the meaning of specific words while not actually addressing what you asked. I will of course neglect to actually give a straight answer.
was Re:classic non-debating tactic
davecb5620@gmail.com