Grace Hopper and UNIVAC

Yearly subscriptions start from £38 (digital), which we think is pretty good considering how much we pack into each issue (take a look here for the contents list of issue 3: http://www.linuxvoice.com/issue-3-is-out/).

Print subs are obviously more expensive, mostly due to the cost of postage.

I had a quick flick through to the third issue to get a rough estimate of how many pages of ads there were and...erm, I couldn't find any except an EFF ad on the inside cover. I don't know if that means they're struggling to find advertisers until they're more established but frankly I'm happy to pay a bit more to have more content pages, given the choice.

In the life cycle of a significant program, which is more important how easily it was initially written, or how easily it can be read and understood over the years?

Because no one can write write a self-documenting Regex, Regex's are only easier to write because any need to document programs is ignored. The extension of this practice is an argument is that code should not contain a lot of comments because comments tend not to be maintained. As Larry Wall noted, Laziness, Hubris, and Impatience all play their role - He being the person responsible for a language so obfuscating as to garner a reputation for author's being unable to understand their own programs {heavy on Regex a coincidence or correlation?}

As Perlis said "It is easier to write an incorrect program than understand a correct one." Particularly if the language makes it harder to read the correct program and speeds up the process of writing the incorrect one. Flomatic and Cobol are verbose for exactly the same reason that some programming languages have static type systems and nobody really takes claims about their verbosity seriously because we recognize the tradeoff of keystrokes for clarity.

Hopper was a Naval officer. She took programming seriously and she understood human nature. They're both reflected in her approach to programming languages. She realized that people can die if a program on the fire control computer is incorrect and a 16" artillery shell falls short onto friendlies or long onto a maternity ward. Nothing promotes the importance of clear communication for robustness like combat - as in civilian life you never know who might be dead or retired or reassigned when the code needs to change.

BTW, I was being a bit tongue in cheek. On the other hand, programmers have been known to use obfuscation as a means for protecting their jobs, and bosses being able to read code is the primary argument for using technical people as managers rather than those with general experience as managers. Programmers are no less immune from creating fiefdoms than anyone else.

There is a temporal nexus circa 1980 between the vilification of Cobol; the rise of Ada and with it a myth of the woman programmer as subordinate to the brilliant genius; and the beginnings of the decline in women computer science majors in US universities [from 36% to 7% today].

There is a trade off between a language that is easy to read and that is easy to write.

There is no factual support for this assertion yet it stands as the unquestioned argument against languages like Hopper's[et al] Cobol. It is used to silence an alternative way of writing, to suppress an alternative means of expression and explicitly the only one with a claim to a feminine origin.

Certainly it's not the case that English-like programming languages have disappeared. There isn't a single widely used language that doesn't use English keywords and one of the major arguments against Lisp is it's non-English like syntax...all those parenthesis and verbs at the front. No, the argument against Grace Hopper's languages is that they are not the missionary position.