[ruby-talk:19233] Re: Some remarks from a nembie in Ruby

ts wrote:
> 
> >>>>> "R" == Renaud HEBERT <[e-mail]> writes:
> R> I wouldn't call this "changing completely the language":
> R> - the octal notation is merely syntactical sugar for "consistency
> R> purpose"
> 
>  You break some scripts

Not until you change the meaning of 010, just adding the 0o notation
won't break scripts.
And the change of the "default meaning" shouldn't occur until a change
in the "major" number.

Python is trying to evolve, Java is trying to evolve, evolution means
breaking some scripts.
It has to be done as rarely as possible and minor changes such as the
octal notation are definitely not enough important to "break scripts"
all by themselves, but they can accompany important major changes which
will be needed anyway ie Ruby will evolve and it will break some
scripts.

 
> R> - the addition of "getlines without end-of-line" methods is a suggestion
> R> to avoid portability nightmares: hardly a complete change.
> 
>  In [ruby-talk:19192] you want change #gets, or at least this is why I've
> understood.
> 
>  #chomp is portable.

And how many people will use chop, str[0..-1] ?


> > IMHO, it would have been much better if by default:
> > gets and getline return without the end-of-line and there are other
> > methods which returns the line with the end-of-line: get_full_line for
> > example.
> 
>  For me, your proposition is to add #get_full_line, not to add "getlines
>  without end-of-line"

It would definitely be better: the shell is much better than Perl here..
Emultating the good points of Perl is good, emulating its bad point is
just a pity.

But I do realise that it may be too late to overcome this mistake, but
adding getps and getpline and writing in the documentation in big red
letters, be carefull about portability when using gets, and getline,
chop may help.

OTOH it may not help if we take as reference all the security problems
which happens in C/C++ everyday because coders use function which are
labelled as dangeurous in every books.. 

 
> R> While it may be too pretentious, I noted what bothered me on the fist
> R> reading of the book, because afterwards you get "too used" to the
> R> 'little quircks' of the language and you can't see what is bothering the
> R> newbies..
> 
>  Well, perhaps this is the difference with me. I've not read the book (it's
>  written in a strange language :-)) but writing scripts.
> 
> Guy Decoux

And what do you prefer?
1.step(12,3) { ... } or (1..12).step(3) { ... } ?

-- 
Renaud Hebert
phone: [phone]
mailto:[e-mail]