Dr. Mark Humphrys

School of Computing. Dublin City University.

Home      Blog      Teaching      Research      Contact

My big idea: Ancient Brain


CA114      CA170

CA668      CA669      Projects


The 1980s - "A PC on every desktop"

Before the 1980s, computers were scarce and users would share large mainframe computers.

The PC revolution of the 1980s was about giving a computer to every user. This was a good idea because:

  1. The machine did not get overloaded by other users.
  2. You could take the machines home, on travel, and other places where you could not access your mainframe.
  3. You got your own dedicated machine to run a user-friendly graphical user interface (which is CPU and memory intensive).
  4. It was usually easier to install new software on PCs because you had complete ownership of the machine. On mainframes new software would have to install in the user space only. To install in the Operating System space would require the co-operation of the system manager. As a result, people often find it easier to customise their PCs than their user spaces on mainframes.

The future? - "The network is the computer"

But the PC era is seen now as defined by the fact that networks were nonexistent or at least slow and unreliable.

As the network got faster and "always on", problems with PCs (or any user device) emerged:

The Network Computer

Many PC users have grown tired of maintaining software, incompatible upgrades, missing libraries. They have other, real, work to do.

An attractive alternative is a login at some central operation with a paid professional at the other end of the line keeping it running for me, making backups, installing upgrades, etc. I want to be able to access my account no matter where I am in the world.

One issue of concern is privacy: PCs are more private. Would the NC company be able to read your files? This can be solved by an encryption-decryption layer between home and NC company, so the NC company cannot actually read your files.

Long-term storage - a possible example of the NC model

As a historian I am interested in the long-term, something technologists often pay little attention to.

Consider where most human data is kept in electronic form. How does a business keep its long-term records (e.g. land ownership records) viable for 50, 70 or 100 years? How does a family store its digital photo album for 50 or 100 years?

You can't store it yourself on removable media, since these media get lost, stolen, decay over time, and anyway the formats will be incompatible and unreadable in a few decades time.

The only way to store long-term data is through an NC-type operation, a company that you pay to manage your data, copy it into new formats and new media, etc.

But what happens if you stop paying your subscription to the NC company? Perhaps the service is combined with libraries (who keep long-term data of all sorts) and banks (who do long-term storage of valuables for a fee). Perhaps if you stop paying your subscription, the data is not destroyed, it keeps on being re-copied and saved, but you have to pay the arrears to actually see it.

Privacy on the Internet

Packet readable by every machine it passes through, like a postcard.

Eavesdropping harder since messages travel different routes. Also messages broken into packets on different routes. Also (for non-government eavesdropping) where to eavesdrop? Most of route is along links you will not have access to.

Also packets mixed with other packets, binary data, etc. Also sheer volume. Manpower to eavesdrop expensive. Software not very powerful.

Still, only real privacy/security is encryption.


Email is all very well, but many people are drowning in email. It is simply too easy to send someone a message.

In an attempt to control the flood coming into their Inbox every morning, many people now make it difficult to email them. e.g. See Jakob Nielsen. Or indeed me:

It is often not recognised that email is a very inefficient method of communication (compared to telephone or meeting in person) because email is one-way. Someone asks you a question, but their message is garbled and semi-literate. You reply asking what do they mean. Sometime later, they reply clarifying their question. You reply to that. Then they reply and finally you come to a conclusion. The process takes days, whereas it could have been solved in minutes if you met face to face.


"Spam should be illegal" - else everyone will get 10,000 messages a day. No reason not to send them.

"No need to make it illegal" - Sysadmins can deal with this problem themselves, by forming networks, sharing info in automatic filters. No need for the law to get involved.

The future of spam

Already over 90 percent of email is spam. Despite this, spam is still legal in the USA.

Until spam is made illegal in the advanced countries, and countries that allow spam are treated to international sanctions, this is the long-term future:

  1. Spam in your email every day.
  2. Spam in your children's email every day.
  3. Spam in your employees' email every day.
  4. Porn in your email every day.
  5. Porn in your children's email every day.
  6. Porn in your employees' email every day.
  7. Spam forged to come from you.
  8. Porn spam forged to come from you.

Spam sent by viruses from compromised PCs

Spam may be technically legal, but most spam is now sent through illegal means - with fake From lines, fake headers, and from compromised PCs infected with viruses.
(Sending spam normally from a spammer's own site is too easy to block.)

So, since most spammers are breaking the law already, making the actual spam illegal may change little. It will not stop the flood of spam generated by viruses.

ancientbrain.com      w2mind.org      humphrysfamilytree.com

On the Internet since 1987.

Wikipedia: Sometimes I link to Wikipedia. I have written something In defence of Wikipedia. It is often a useful starting point but you cannot trust it. Linking to it is like linking to a Google search. A starting point, not a destination. I automatically highlight in red all links to Wikipedia and Google search and other possibly-unreliable user-generated content.