Re^7: Looking for old Perl CGI code

It doesn't pass the sniff test because it stinks. By that definition, phonecalls are a "computer language" because it's both a communication standard and is regularly used to control devices^*. Yet, not a single person in the world would agree that the protocol used by telephones is a "computer language".

<rant>There seems to be some fundamental misunderstanding of how computers work in the mind of the general public. What we do is not magic nor is it some completely new idiom in the history of mankind. We define tasks, then hand them off to a servant (slave, really) to do. The only difference between a computer programmer and a slaveowner in Rome is that our slave will do exactly what we say, has perfect recall, and will do it for as long as was asked for. We're not wizards in a tower stirring a bubbling cauldron containing Eye of Newt and Bat's Wings. Gah! We do nothing more than give directions from A to B.</rant>

* Remote detonation of a bomb is just the first example that comes to mind. Another is the standard automated thing that you get when you call your bank looking for a human.

My criteria for good software:

Does it work?
Can someone else come in, make a change, and be reasonably certain no bugs were introduced?

Comment on Re^7: Looking for old Perl CGI code

Replies are listed 'Best First'.
Re^8: Looking for old Perl CGI code by sauoq (Abbot) on Oct 13, 2005 at 19:58 UTC
I do understand the irritation. And I had pretty much the same reaction in that debate with my professor a decade or so ago. By that definition, phonecalls are a "computer language" because it's both a communication standard and is regularly used to control devices. How is a phonecall a communication standard? Regardless of how you answer that, though, I just see it degrading into a discussion about the definition of "language" in this context and I'm not sure how useful that is. If we can agree that HTML is a language and that it is used to control devices, that's sufficient. The only difference between a computer programmer and a slaveowner in Rome is that our slave will do exactly what we say, has perfect recall, and will do it for as long as was asked for. Provided you don't ask it to do anything too complicated (like feeding you grapes), you are very careful in the way you ask it, you don't require it to recall... say... an irrational number, and you have an infinite supply of power and hardware that will never wear out. :-) -sauoq "My two cents aren't worth a dime.";	[reply]
Re^9: Looking for old Perl CGI code by dragonchild (Archbishop) on Oct 13, 2005 at 20:24 UTC
There is a communication standard that controls how phonecalls are sent and received. It's harder to see in the analog world, but very easy to see in the digital world. In the cellphone industry, there are three protocols - GSM, CDMA (and its variants), and the protocol that only Nextel used and Motorola supported. CDMA (which is the one I'm familiar with) uses code multiplexing in order to transmit digitized voice across the air using a 2.5MHz wide spectrum. It uses messages, including messages with containers, to allow the base station and the phone to exchange information and control each other. VOIP is even more explicit in the SIP protocol. So, to claim that a phonecall doesn't follow an established communication protocol doesn't hold up. But, you're right in that a phonecall is more akin to the HTTP communication protocol vs. HTML. For that, I can point to the IVR protocol that bank's use to prevent customers from talking to real people. That's not a computer language, is it? My criteria for good software: Does it work? Can someone else come in, make a change, and be reasonably certain no bugs were introduced?	[reply]