Google Speed

[anon]

Let's make the web faster

What would be possible if browsing the web was as fast as turning the pages of a magazine? We invite you to join us in exploring and innovating across the entire spectrum of performance - from Internet protocols to the browser to website development. Together, let's make the web faster!


IWWBnJEsUtU

Link (http://code.google.com/intl/en-US/speed/) :thumbsup:

[splicer]

If Google really wanted to make "the web faster", then they should start by laying telecommunication cables to improve the availability of bandwidth to all countries, then improve the internal infrastructure of each country, and ensure that due to supply and demand, that the average internal price of an internet connection drops so that everyone can afford one by creating their own ISPs.

They can spend all they want trying to make programs use a connection more efficiently but ultimately data is being transfered and the rates available to people is what should be improved, especially in 3rd world countries, as their countries are less likely to have the capital to invest in such technology. This increased access will mean jobs are created; from laying the cables on land (labour) to such things as call centres and allow for training from other countries (a Linux administrator can get paid upto $80000).

Obviously this will all cost money. And probably lots of it; but that's something Google has - just look at their total assets and equity Google - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Google). Laying the cables may not be as expensive as you think; as technology progresses so does the cost of cables: Undersea Cable Costs (http://www-dsg.stanford.edu/holbrook/CableCosts.html) . On the basis that the Earth has an equator of 40,075.02 km, it would cost Google just over $ 1 billion dollars to lay a cable around the equator, if they used the same wire that AT&T used for Columbus III in 1999, meaning 14 countries would see improvements to communications available to them, not including off shooting cable to near by countries.

We must bear in mind that the transfer rates of the cable will also effect the availability of internet; in the example above, Columbus III had a bandwidth of 10Gbits/s. Ofcourse as mentioned as time goes on, technology improves; the Europe-India Gateway, being built currently, has a bandwidth of 3.84Tbits/s (yes, terabits!).

[slikrapid]

google is reinventing hot water here - the trend towards increasing speed of internet connections has been ongoing through decades, so nothing new here - imo they are more concerned about their web applications demanding more and more bandwidth compared to a slower rate of user connection speed increase


If Google really wanted to make "the web faster", then they should start by laying telecommunication cables to improve the availability of bandwidth to all countries, then improve the internal infrastructure of each country

that's an idealistic/utopian view - google doesn't have nearly enough money needed for such infrastructure changes and remember - they are a company interested in profit so don't expect many efforts towards investments that are risky in providing a positive money return - there is also a possible problem of control/privacy if they are present in all ISP's on a hardware/software level - no private company should have such a large access


They can spend all they want trying to make programs use a connection more efficiently but ultimately data is being transfered and the rates available to people is what should be improved

well, apart from raw connection speed & managing a balanced load on servers, there is a few percent improvement possible through better protocols & software programming with appropriate caching & compression


especially in 3rd world countries, as their countries are less likely to have the capital to invest in such technology.

investing in such countries is risky and the countries themselves should be careful not to give up majority of their top companies to foreign capital - as we all know investments are done with a reason, usually being to enter a rather 'successful' company and acquire control over its decisions and money flow


On the basis that the Earth has an equator of 40,075.02 km, it would cost Google just over $ 1 billion dollars to lay a cable around the equator, if they used the same wire that AT&T used for Columbus III in 1999, meaning 14 countries would see improvements to communications available to them, not including off shooting cable to near by cables.

even if you put such a mega-cable through the world, the questions are:
- will it be sufficient for everyone to use,
- who controls the data flow & contents,
- who can guarantee for privacy protection,
- who owns it and how much will the leasing cost be,...

imo it is better to use the p2p non-centralized approach and builed & interconnect many networks with multiple intercontinental backbones that are owned by different companies or maybe countries so that abuse can be reduced as opposed to a fewer big players (Tier1) who nowadays control the internet inter-connectivity and charge massive amounts to new/lower level ISPs for getting access to the internet as a whole



Columbus III had a bandwidth of 10Gbits/s. Ofcourse as mentioned as time goes on, technology improves; the Europe-India Gateway, being built currently, has a bandwidth of 3.84Tbits/s (yes, terabits!).

afaik even this 10Gbps connection can be increased by applying DWDM for example without the need to lay new fiber

[splicer]

Agreed it's utopian and agreed they are also a corporation interested in profit, but such a cable would be 100% theirs, they could charge ISP's to rent bandwidths segments, or even sell the cable to the ISPs themselvesm obviously with profit in mind. Many large corporations often worry about brand image; many want to put off an image of an image of a global company, one which would effect us daily, but also have the ability to control humans ( if you've seen Resident Evil the type of thing the Umbrella Corporation would be); they instead want to replace this with a caring image, Virgin does, so that people aren't worried and think of them positively. Many pay for expensive internet connections, something which has now been dubbed (by the UK government atleast) as essential - like having piped water in the home, by Google investing in such a wire, it would lower the costs of internet in many countries, hence taking a worry of many minds.

Modifying protocols may mean that many software would become incompatable, or may not meet new standards; would it not be a possibility that worldwide a significant amount of money would be given to programmers to update programs?

Expansion can lead to profit; by investing in such a cable to benefit 3rd world countries would mean Google would be the main communications provider in that county, perhaps a monopoly; once again something that would be in Google's business interest.

The problems with multiple ISPs of various countries expanding internationally to eachothers networks vary. Many ISPs, especially small one simply do not have the capital; it would be the large ISPs just getting larger; with Google laying a wire, each could buy/rent it's own share, and again Google seeing profit. Also many ISPs charge ISPs in other countries a lot for rent, due to cultural differences (not to bring up an argument, such an example would be Argentinian ISPs chargings Falklands Islands ISPs alot for rent out of cultural spite).

Ultimately these international cables will always be built, by who is the only variable and Google, expanding into everything internet, this may be the next target.