A modest proposal

A modest proposal for reducing World Wide Web traffick, and for preventing it from being a burden on the Inter-net's bandwidth:


It is a melancholy object to those who travel through this great Inter-net, to see the laggardly pace at which many of its World Wide Web sites send their data across the 'net; and the extreme volume of it which slows the average user's browsing of the Web to the pace of a snail. These users must wait up to 20 seconds to be able to click a cow or view an amusing image of a cat -- so burdensome is the weight of this wait that many find no time to care for their children, who are doomed to death by starvation without the attention of their parents; or, for those able to scrounge food for themselves, they are inevitably led to a life of crime and prostitution, with-out the guidance of their parents to more respectable, legitimate career paths.

I think it is agreed by all parties, that this wasteful use of the Inter-net resources must be reduced and, eventually, eliminated. To aid in this, I propose some modest changes which will improve the efficiency of Web traffick.

The Hyper-Text Transfer Protocol is a grossly verbose mechanism for so simple a task as the transfer of data; as a result I believe it alone is responsible for the long page loads which are the cause of our society's problems. For this reason, I suggest its improvement with the following changes, to increase its hastiness at delivering information across the Inter-net:

Request methods are not necessary, as experience has shown that they are adequately replaced by inclusion of method information in the URL, therefore that information should be removed from the HTTP protocol itself; as should the version, which adds no useful information and yet contributes a wasteful 9 bytes to the volume of each request. There is also no need to include any header fields in each request other than the one called "Host"; it has been the experience of many users of the Web that further information only invites invasions of the Web user's privacy, revealing to the remote server, information that users have not consented to provide.

The HTTP version and status information are likewise to be excised from the response data, as they only waste bandwidth and add no useful information to each request. Indeed, the status information is often meaningless anyway: many pages indicating the nonexistence of a requested resource are erroneously served with the "200" status code; therefore, in the interests of "paving the cowpaths" of the Inter-net, we ought to do away with distinction between them. Likewise all response header fields, almost all of which are now obsolete -- such as the one called "Content-Type", which is adequately replaced by user agents' "content sniffing" features; and "Location", which is adequately replaced by the redirect capabilities of the Hyper-Text Markup Language; and "Date", which only repeats information available thru' the user's own system.

Some would propose more advanced schemes for the reduction of traffick and making more hasty the workings of the Inter-Net's protocols. I am not so violently bent upon my own opinion, as to reject any offer, proposed by wise men, which shall be found equally effectual. But before something of that kind shall be advanced in contradiction to my scheme, and offering a better, I desire the author or authors will be pleased maturely to consider that any more complex scheme may well be significantly more difficult for the programmers of Web software to implement; I fear such difficulty would put them off, and that therefore they would not make use of such a system, and our problems would continue.

To those who would reject such a scheme, I would invite the consideration of the ill effects of the systems currently in place; the resulting increased rates of crime and whorish behaviour owing to the lack attention payed by parents to their young -- because of time wasted waiting for Inter-net pages to load.

Received on Monday, 1 April 2013 13:42:26 UTC