On the contrary. It's a statistical fallacy that leads to a lot of
wasted effort - if you are under the presumption that speeding CAUSES
accidents, you will go overboard on spending resources on reducing
speeding, instead of plucking low hanging fruit that would fight
reckless driving. It can also INCREASE accidents. Case in point: In
germany a few years ago, a massive car pileup was caused by a driver
who spotted a speed-trap (on a piece of highway that had very low
accident rates, no less - it was there because it earned the local
municipality quite a bit of cash in that location). The irony? That
driver wasn't speeding, he just knee-jerked when he spotted the speed
trap.

There's quite a big difference between positional independence, which
reduces the intrinsic complexity of the language (no need for rules
about which side 'wins', and eases ability to reason about code),
whilst the cost seems reasonable. I contrast this to the current
decision to both have no operator overloading of any kind and no
central body that picks known cases with no issues and adds them to
the language spec (e.g. BigInteger and BigDecimal) - I feel this is
overly restrictive and has quite a high cost in code readability.

My road safety analogy was merely an attempt to steer such discussions
towards a weighing of the pros and cons and away from mantras. We
should fight intrinsic complexity and we should advocate for those
measures that find a good value proposition between code that is
readable-at-a-glance (good at the WHY) and code that is easy to follow
mechanically (good at the HOW). Those two things have a nasty tendency
to be mutually exclusive. That's a much better idea than just naming a
few baddies, such as 'operator overloading', without any room for
discussion.


In regards to an eclipse plugin for replacing < with double-angle-
brackets to aid in readability - A standing offer of 50 bucks is
effective at making sure people who are thinking of it get
disillusioned. 50 bucks covers not even an hours worth and I'm
guessing that even if things go well you're looking at about a manweek
to get everything sorted out. So, 50 bucks could only be an insult. If
you want to write it yourself, I can give you quite a few pointers.
For starters, I think a javaagent that patches a few choice snippets
in the JDT internals would be easier than an eclipse plugin, and I'm
virtually positive that the extension points available for eclipse
plugins are not sufficient to do this without relying on internal
eclipse API.


On Aug 14, 1:27 am, Ben Schulz <[email protected]> wrote:
> > There's one safeguard on roads that actually saves lives: Driving
> > lessons. If you put somebody behind the wheel of a car, they can kill
> > people at the drop of a hat, and no amount of laws are going to stop
> > this from being true. You get that license because the state
> > recognizes that they are trusting the driver to not be a reckless
> > jackass.
>
> It's not that they trust you not to be a jackass, it's that you don't
> have to demonstrate that fact. It's called the presumption of
> innocence. It's also where the speed limit comes in; a facility of
> proving that you are, in fact, a jackass, and thus can not be trusted
> to drive a vehicle. In 2002 the chinese station CCTV reported 106,000
> road fatalities, that's almost 300 a day (no driving license required,
> no speed limits, etc.).
>
> > There's loads of scientific proof out there that all speeding and
> > safety belt laws are wrong; they have their causation reversed.
> > Speeders are in more accidents not
> > because speeding causes accidents; it's because reckless driving
> > causes both speeding and accidents.
>
> Yes, and x + 0 is not x, it's just the results that are equal.
>
> > This idea transfers easily to programming: an idiot programmer is
> > going to create stupid code, no matter how many rules and restrictions
> > you create. If operator overloading is there, it's just a particularly
> > convenient method for blowing your foot off. Removing that hand cannon
> > doesn't really help - you can't make things idiot proof, it doesn't
> > work.
>
> Absolutely, but wouldn't you agree that we should minimize the risk
> wherever it's reasonable? I know you do, because you're not a jackass.
> Also you would not be striving for positional independence.. :D
>
> With kind regards
> Ben
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "The 
Java Posse" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/javaposse?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to