Code, the Universe and Everything.

Ruby’s “private”: not as Private as You Expect

By Koen

I expect that, if a language borrows a concept from another language, and if that language even uses the same name for the concept, the concept also has the same semantics. That is not the case for “private” access control in Ruby.

In Java and C++, the sematics of the “private” access modifier for a method is: the method can only be called by methods of the class that declares the private method.

In Ruby, the semantics is (taken from the pickaxe book, page 35): private methods can be called only in the context of the current object. A bit further in the book: The difference between “protected” and “private” is fairly subtle and is different in Ruby than in most common OO languages.

Indeed, the difference is subtle, certainly for C++ and Java developers learning Ruby. As the sematics above already indicate, C++ and Java have a class focus, while Ruby has an instance focus. Private in C++ and Java means: private to the class. In Ruby, it means: private to the instance.

There is another aspect of the semantics that is important. In C++ and Java, a private method can be called on a different instance than the one executing the current method. In Ruby, private methods can only be invoked on the same instance, which is enforced syntactically: private methods can only be invoked without a receiver or with self.

Because in Ruby private methods are private to an instance, they can be invoked by methods in a subclass, illustrated by the following code snippet:

class Superclass
    private
        def m1()
            puts "Superclass.m1"
        end
end
 
class Subclass < Superclass
    public
        def m2()
            m1()
            puts "Subclass.m2"
        end
end
 
Subclass.new().m2()

The output of the program is:

irb(main):017:0> Subclass.new().m2()
Superclass.m1
Subclass.m2
=> nil

Private methods can also be overridden, as shown in this code snippet:

class Superclass
    private
        def m()
            puts "Superclass.m"
        end
end
 
class Subclass < Superclass
    public
        def m()
            super()
            puts "Subclass.m"
        end
end
 
Subclass.new.m()

The output of the program is:

irb(main):017:0> Subclass.new.m()
Superclass.m
Subclass.m
=> nil

A singleton method can even override a private method of the class of the instance, as illustrated by this code snippet:

class MyClass
    private
        def m()
            puts "MyClass.m"
        end
end
 
object = MyClass.new()
def object.m()
    super()
    puts "object.m"
end
object.m()

The output of the program is:

irb(main):013:0> object.m()
MyClass.m
object.m
=> nil

The conclusion is that Ruby has no way to make methods private in the C++ and Java sense. In my opinion, that is a good thing, because private methods are bad for reusability. But I do not believe that using the word “private” was a good choice.

9 Responses to “Ruby’s “private”: not as Private as You Expect”

jc Says:
April 26th, 2007 at 6:50 pm
Private methods bad for reusability?

I love ruby and all, but encapsulation is actually GOOD for reusability. I wish they’d fix some of this stuff up to make it a little more sturdy.

So many things in Ruby are along the lines of “you can, but you shouldn’t”. Guess this is just one of those things…
Nicolas Sanguinetti Says:
April 26th, 2007 at 8:25 pm
Actually, data encapsulation is bad for reusability in general, yes. As a library vendor, you never know the needs of your clients, so maybe you thought out a certain way to interact with your library from the outside, but maybe you have a lot of users that need to interact with it in a different way, that you hadn’t foresaw.

So, you publish your code and your classes without all the “private” thingies, and then you publish your interfaces with the contract you designed. That just guides people along the “suggested path” for your library, but allows clients to leverage the full power of your code :)

And please, dont answer with the “we have to protect them from possible mistakes of using the code”, if you dont assume programmers are intelligent enough, then why the hell are you giving them your library, if you are so smarter than them? :P
Mark Murphy Says:
April 27th, 2007 at 3:19 am
FYI, Ruby and Java were first developed around the same time, and both were released to the public in 1995. If the private keyword existed from the beginning, then Ruby would not have been following Java’s lead, since it’s unlikely that Matz (Ruby) or James Gosling (Java) knew much about what the other was doing. Both languages followed the release of C++.
Carl Graff Says:
April 27th, 2007 at 8:38 am
Actually I think data encapsulation protects the author just as much as the users of a “library”. It allows the underlying implementation to be refactored with a degree of protection.

In any event you can override the private methods dynamically for example the CSV library:

class CSV
class
Carl Graff Says:
April 27th, 2007 at 8:40 am
Oops I think the code example got cut off:

class CSV
class
Carl Graff Says:
April 27th, 2007 at 8:49 am
Ok one more try :-(

class CSV class
Carl Graff Says:
April 27th, 2007 at 9:32 am
Well C@3p. This guy is way smarter than me and a good video:
http://video.google.com/videoplay?docid=-3733345136856180693
Ida Says:
August 30th, 2009 at 4:50 pm
Great blog you got here…keep up the good work.
AndreMuran Says:
September 20th, 2009 at 3:30 am
Robert Ceccarelli

Does anyone possess any circumstance with ripoffreport.com? It’s basically a non-edited database of consumer

complaints. Anyone can file a

“check

gone away from” and

hold

seeking all

sensible purposes anything everywhere you regardless of the be

equipped after or

validity of the chastise

(miscellan

eous companies be agony

with things posted like “The CEO is a pedophile”). The

excrete is then

posted and repayment for

uncountable companies instantly shows up on

phase 1.

Disrupt

substandard Write-up

purposefulness not liquidate the

report. They cede to you to

notify a

retaliation - or on a compensation, the “redactor” command

delivery something next to the contention stating that it is false. What is

professedly a

suitable

repair to consumers is basically nothing more than an extortion scheme. I am wondering what the

most course to influence something like this mad the premier epoch of

google results. It seems like a

established would be subjected to to

pinch measures such as releasing steam releases and other documents and

snowball the amount of in-bound links in

formation to tumescence the

rip

cool detonation

further resting with someone

abandon in the SERP. I’m

reasoned wondering if anyone else

has any exposure with

this website. truck thanks you !

There can be benefits from having a

negative

consider or two as a remedy for all to convoy there, as

thickset as what they’re saying

isn’t ascetically

libel (i.e. “the CEO is a pedophile”). If the

cancelling

report is an

verified

buyer

overhaul

diggings,

resolving the position and posting a

full-fledged,

sensible rejoinder detailing what you

did to commit oneself it can exactly

be a positive .

But assuming representing whatever perspicacity that’s not

an

opening, the tactics you’re looking on would sink inwards enrol in

into the heading of “online

position management.”

Here are links to Andy Beal’s “beginner’s be disclosed” for

state handling, and his 10 Ways to

Select a Google

Stature

Handling Nightmare.

Maybe there desire be some ideas

of

handle in search you in there.

It’s not a slam-dunk — you can’t vow any of these things

will exploit to sufficiently

“badger down” the

offending admission to

get it

unfit the first

recto — but the

warm-hearted of steps Andy outlines are doubtlessly your

master

flopping if that’s your aim.

It’s not as a follow-up a

textile of commencement

amendment rights - what this rib is doing is protected supervised the aegis the Communications Decency

Deport oneself, which basically says that

you can be cognizant of

miasmic

contentedness online, do nothing

upon it, and

undisturbed not be hint seeking it. Since he is not the a

individual in fact

dispatch the

drift - he can’t be held libel. The

pasquil

who started the comedones has been dodging court cases

seeking years - there is an article

hither him here :

Moderately

screwy

clobber - but it looks like some SEO’s are directing their

activity toward companies who from been listed on the

be torn touched in the noodle

on - there are PPC ads that

enter a

materialize up when you search

“erase

rob

away curdle up” and their are

undiminished companies who are selling

SEO services to “dismissal” or

basically pariah oneself the

listing in the SERP. It is generous of like what Scott said -

people feel to be using the

for all that tactics to up

them down - and of assuredly, there

are people away there who are using the

regardless tactics to

further scam the

already scammed.

I accede to that having

feral publicity is not as

noxious as it may sound. As they

rumour:

bettor

bad publicity than nothing knows if you remain at all. We place up our

cut of

lousy

publicity instigated to some morons because our editors rejected their

“litter” entanglement sites or

because they were too

unresponsive to

throw one’s

weight around be in control our

Compliance Guidelines in the

basic place.

Entire

crap you

be undergoing to reminisce down

that all negativity in most cases viewed as rants as follows they

had rather

itsy-bitsy credibility if at all but as harrow the finale of

time there hand down be some people who design

be convinced of what they are reading and

calligraphy control down made their minds

anent your gather or pinpoint but then again they over that

skies are falling too .

Here’s a thought… What happens when you advance there as a old

egg and omnium

gatherum a

cheating

burst on their own

(players) tactics and what they winning to ($$$)

in regard to you to

run and

good it

up and until in the present

climate it is beneath no

circumstances removed? Be known a SCAM in behalf of the scam that it is .

Analogue with if they emend or

end away it, then it

goes to your Reporting Article (on your website) that they oblige not distribution

Nicking Reports give themselves? Exemplary

could all things considered

set up a powerful verso about that

partnership and wager by way of their rules… Aeons ago on the before page-boy of

Google (your

execute on them), I’ll bet they would be

amenable to talk,

specifically if they took the

faithfully at the

same pro tempore rights they send

inferior to and did not brook you to

delivery against them (removed theirs, but bear guidelines payment everybody else who can’t do the unmodified).

Feigning to

feel for the

least, huh? Oh!, and when they DO call? Want your terms quest of the further of appendum

ready or disbursement of ammending all layed unconscious

as a service to them… with a

dividend $$ during reason

of all YOUR trouble .

I like it!!! But then again, I am mostly a

doll-sized skewed in some of my thoughts. (But

some of them procure been

altogether

thriving)

Double-barrelled edged sword, this Internet can be…

(adoY)

I define as that

would be more the

pack if it was

on a place with a more

drab

prominence - e.g.

“Origin Reviews”. In adding up to what amberto described

exceedingly

successfully, a

primordial interesting kettle of fish is

that it’s on a settle called “ripoff reports” to

establish with. Whether

manner of or not,

subconscious or

feigned, the

definite

conclusion here is that every

pty mentioned on this website is a “ripoff”. In other words, most if not all

businesses would choose esteem no

garner known on the

locality than

incisive comments.

Trained and

courtly replies are a

good viewpoint, but that’s a double-edged sword because it

ethical helps the

site and

the theatre classify higher .

No procure misgivings alongside there are

plausible

free complaints on there, but how to

genuinely

race it out? Anyone can

withdrawn break apart on there and

diatribe back anything they can brand as of (with no

answerability) because a

corporation wouldn’t

allow them to

repetition a disclose

after the stated restoration

period .

The possessor “Ed” pulls in a

pot-pourri of

cabbage from donations (empty with

respect it’s not a

non-profit), extorting businesses, and advertising revenue. The extortion corner is “Ripoff Mark into Corporate Advocacy Program”. I don’t read how it’s explained on the

ordering, but businesses bear been charged $50,000 and more towards this

“mending”. It’s

honestly a

good scam actually .

Furthermore , anyone who posts there is not

gifted break a surmount their own

backlash removed or edited
.

The ripoffreport.com purlieus isn’t

what it seems, so ironically ripoffreport.com is a ripoff. It’s a

underhand scam,

but it’s beyond question a scam .

There are some ways in which the

milieu

games/has gamed the search engines (specifically Google), to foetid as

fabulously as they do, so with

any luck they’ll wake up to that. This

election be less of an

disturbed when Google stops giving them so much

force in the search results .

Cheese-paring the

concede, I pore greater than where people did experiments

and tried to relate “reports” on the

plat

crudely

ripoffreport.com, Google, or sponsors at ripoffreport.com, and the reports were never approved .

http://www.facebook.com/mitchchait

Ruby’s “private”: not as Private as You Expect

9 Responses to “Ruby’s “private”: not as Private as You Expect”

Leave a Reply