BTW because there is no backwards compatibility IPV6 will never become mainstream. The cost to switch would be $500 billion PLUS (and probs htat it just for the states). If IPV6 was backwards compatible we could just progress ot it.. if only :(
Thomas Morton ++ No problem should ever have to be solved twice ++ 2008/11/27 Thomas Morton <[EMAIL PROTECTED]> > I never said IPV6 was new - just that it's not really ready. > > TBH it's a load of crap - it makes the network stack a lot more > complicated. And because it has taken so long we have shore up methods that > mean large companies aren't worried about adopting it (because it means > screwing around with stuff which costs..) and new implementations are not > necessairily any cheaper or easier in IPV6 (so we stick with IPV4 to stay > backwards compatible). > > Routing is NOT better (in fact probably it is worse). where did you get > that from? > > As a practical implementation IPV6 is pants. It might work on an extremely > large network (say 100,000 units) but for sub 10,000 units you will find > network suppliers will without fail build IPV4. And TBH I have not yet seen > one with a standard IPV6 offering.... (we work closely with a lot of network > suppliers across the world) > > IPV6 doesnt work for the internet either - on a super-scale it throws > around too much of the clouds resources and bandwidth to implement. Plus > there is no backwards compatibility so adoption will be patchy and > complicated. > > All in all a total mess. > > What we need is a simple, clean and lean network protocol. IPV6 should have > been a light but extensible stack offering NOTHING but a huge adress pool > and translation layers. Then we can build more efficient routing (NAT's > equivalant - and I agree with you, it sucks) and other bits and bobs > (security) on top of that. > > The major problem is we no know IPV6 wont work - even in the relatively > short term. But becuase of it's complexity there is no way to "fix" it. If > we had gone with the light approach we could now pull off and modify or > recreate all the wrong extensions... then the base IPV6 would have lasted > forever and been constantly refreshed. > > As a programmer who has worked on network protocols at a kernel level (we > run our own in house, from scratch, OS) I can tell you IPV6 is a PAIN to get > working. :( OS builders hate it, programmers hate it, network techs hate > it..... sucky.... :P > > I draw your attention to one of the better appraisals of where IPV6 is > "wrong": > http://www.renesys.com/blog/2008/11/its-the-end-of-the-internet-as.shtml > > Sorry for the major OT ;) > > Thomas Morton > > ++ No problem should ever have to be solved twice ++ > > > 2008/11/27 Japje | Jasper Wonnink <[EMAIL PROTECTED]> > > >> Thomas Morton wrote: >> > Implementing IPV6 is probably more of a pain than they can be bothered >> to >> > work at (I would imagine). >> > >> > Seeing as it *still* is not ready, complete or accepted as standard >> (another >> > 3 or 4 years do we reckon?) it seems fairly pointless. >> >> It is ready, and has been for many years. Major hardware company's like >> juniper and cisco allready support it on all their core routers. Ipv6 >> is not something new, its been around as early as 1995. >> >> >> > Plus of course you have the IPV6 jack of all trades useless for anything >> > else problem - I can see it ending up as a stop gap. >> >> You do know what a network stack is right? >> >> > Until IPV4 drops off as the standard protocol I doubt anyone much will >> be >> > implementing it - and by that time I fully expect we will see IPv10 at a >> > decent (better) standard... >> >> Ipv4 will never be dropped within the next decade because there is >> legacy equippment that will never support ipv6 (products not being made >> anymore but still in use) >> >> And dont expect an ipv10 within the next half a decade. >> >> Ipv6 brings a lot of good things like performance and stability and easy >> configuration. Better routes and we are finally get rid of NAT. which is >> _NOT_ a security feature! >> >> >> >> > >> > Just my 2p :) >> > >> > But IPV6 support WOULD be awesome nevertheless :D >> > >> > Thomas Morton >> > >> > ++ No problem should ever have to be solved twice ++ >> > >> > >> > 2008/11/27 Saint K. <[EMAIL PROTECTED]> >> > >> >> Agreed, we're closing into a country wide IPv6 rollout as well here, so >> it >> >> would be more then advisable to get IPv6 support up as soon as >> possible! >> >> >> >> Cheers, >> >> >> >> -----Original Message----- >> >> From: [EMAIL PROTECTED] >> >> [mailto:[EMAIL PROTECTED] On Behalf Of Japje >> | >> >> Jasper Wonnink >> >> Sent: Thursday, November 27, 2008 12:20 PM >> >> To: Half-Life dedicated Linux server mailing list >> >> Subject: [hlds_linux] hlds/srcds ipv6 support >> >> >> >> Hey guys, >> >> >> >> After a long time i decided to put up some servers again, and lately im >> >> working on a lot of ipv6 business cases so i wondered if hlds supported >> >> it and i tried to bind my hlds to an ipv6 address, which it failed >> (came >> >> as little surpise). >> >> >> >> And i wondered if you are planning to support the ipv6 stack (which >> >> include ipv4 as a bonus). >> >> >> >> And to the fellow admins if they have ipv6 and would like to see it >> >> supported aswell. >> >> >> >> -- >> >> Regards, >> >> Japje >> >> >> >> >> >> _______________________________________________ >> >> To unsubscribe, edit your list preferences, or view the list archives, >> >> please visit: >> >> http://list.valvesoftware.com/mailman/listinfo/hlds_linux >> >> >> >> >> >> _______________________________________________ >> >> To unsubscribe, edit your list preferences, or view the list archives, >> >> please visit: >> >> http://list.valvesoftware.com/mailman/listinfo/hlds_linux >> >> >> > _______________________________________________ >> > To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> > http://list.valvesoftware.com/mailman/listinfo/hlds_linux >> >> _______________________________________________ >> To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> http://list.valvesoftware.com/mailman/listinfo/hlds_linux >> > > _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
