That’s what I used to read in some newsgroups during the 90’s. Back then, the number of usenet postings was already growing exponentially. The bad “signal” to “noise” ratio, i.e. rubbish versus useful postings, led to predictions of the imminent collapse of the net.
People at universities, who then formed the bulk of internet users, started to FTP large data files; 500KB was large and it took forever! At the same time, newbies were struggling with things like usenet netiquette and making their first HTML pages for the XMosaic browser. The first overly long ASCII based e-mail signatures appeared, and usenet legends like James “Kibo” Parry were drawing attention by replying to thousands of usenet articles. Nowadays, we are continuously chatting and twittering away like there’s no tomorrow. MSN is currently more used than e-mail and snail-mail together. I think I can safely say that the signal to noise ratio is worse than ever, but the net is still alive.
Next up was the universal worry that the IP-address space would be completely used up. It was obvious that every PC, microwave oven, and fridge would eventually need its own fixed IP-address. Moreover, with the increasing number of networks, the internet routing tables were getting too big, causing serious delays. Again, the death of the net seemed imminent. For this, the next IP protocol, IPv6 a.k.a. IPng, was being developed that — with its huge address space — would once and for all solve the problem. But before IPv6 was ready, technologies like CIDR, DHCP and Virtual Networks, much sooner than expected, averted IP-number shortage. Only now, IPv6 is finally implemented in TCP/IP stacks of all common OS’s, but it is mostly unused. Eventually we will probably switch to IPv6, but there is no hurry to do so, we can introduce it gradually. IPv6 can accomodate IP addresses for every interface on the planet and beyond.
From the year 2002 on, we started collectively worrying about the increase of spam: it would render e-mail eventually completely unusable. Spam would kill the web. But unexpectedly, DNS blacklists, heuristic server-side spam filters, and intelligent e-mail clients again saved the day. Although more than 90% of all e-mail traffic today is some form of spam, we are still happily e-mailing.
The current internet doomsday hype is the upcoming bandwidth shortage because of increasing demand by P2P and webvideo traffic. Recently, Nemertes Research revealed that by 2012 bandwith demand will (mostly locally) surpass the technical capacity of the infrastructure. In other words, the internet would come to a grinding halt. Been there, done that, deja vu! I’m not worried at all, because the shape of things to come is visible. Just to mention a few: Google is building data centers around the world that will serve as a huge local (low hop count) cache for your video data, there are many advances in wire opticals, and new WiFi devices will be able utilize local bandwidth in the air. Moreover, Web 2.0 applications with locally stored data — harddisk space is cheap and plenty — will decrease the amount of unneccessary requests, or … just pre-download it ALL when you sleep.
To summarize, I have the wonderful suspicion that internet technology is evolving faster than we can think of new problems. The death of the net is not imminent: the internet is more alive and resilient than ever!
Entries (RSS)