[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: fighting spam || avoiding spam

On Tue, Jun 08, 2004 at 03:34:28PM +0200, Mathieu Roy wrote:
> Wouter Verhelst <wouter@grep.be> wrote:
> > Your suggestion is "I know you don't think it's possible, but I do, so
> > please pursue my idea for me".
> I doubt it is impossible. 

Then prove it. By providing working code. It doesn't have to do
/everything/, but it has to do the core bit of the job.

> It is already possible to obfscute all email addresses in a text --
> will you challenge that assertion.

No, because it is indeed possible; the point isn't that it's not
possible to obfuscate email addresses, the point is that it'll reduce
the usefullness of the archives, which is a trade most people are not
willing to make.

The alternative is to make it more intelligent. Since that has to be
done through mathematical rule sets (which all programs are), there's a
limit to how intelligent you can make it. Since nobody (except you)
thinks it can be done with current technology, it won't happen -- unless
you prove all of us wrong.

> What could be interesting is to list the traps to avoid and handle
> them. 

Be my guest.

> > Either provide a working solution (and prove it works), 
> If it was essential in my life, I would do so. It is not, I will find
> workaround less time consuming on my side. 

If you're not willing to do that, because it's not essential in your
life, then why are you wasting people's time with this?

Do you think someone else, who doesn't even believe your idea is valid,
would see this as 'essential in their life'?

We are all volunteers. If you pay me, I'll pursue your fantasie. If you
want me to do it on my free time, then prove me that it's not just a
waste of time, or shut up.

> The sad thing is the fact I'm not even surprised to read people
> suggesting that proposal should not be made before the work is done. 

Nobody says you have to do _all_ the work before they'll listen; But
it's fair to require some prove that the idea is valid before wasting
your time on it.

> > or shut up. 
> You cannot even be polite and you think you deserve attention?

Don't quote me out of context.

> > It's been told you now, what, three times(?) that doing such a thing is
> > indeed impossible, by people who have tried.
> >
> > If you insist that they're wrong, prove it. Don't blatter on endlessly
> > that you think it's possible. 
> Funny to hear that for most debian developers, find a way to modify a
> specific string in a text with the least false-positive is something  
> impossible. 

The thing is that it's not a "specific" string.

"Find all occurances of the string 'Mathieu Roy' and replace them by
'Someone wasting everybody's time with unproving theories'" talks about
a specific string. 

"Find all occurances of email addresses (and nothing else, and not in
cases where it would do harm to the rest of the text) and replace them
by an obfuscated version of that same string so that legitimate people
can still read it if required, but web spiders can't", doesn't.

You're suggesting the second thing, which people with experience in the
area agree is impossible. If you think it isn't, prove it.

> So I think everything is said, change nothing, everything is perfect
> as it is,

Nobody said that everything is perfect as it is (unless I missed
something, in which case I'd like to see a message ID). What people did
say, however, is that your suggesting would make matters worse rather
than better; and that is not usually a trade people should make.

If you disagree, prove your point.

> it is definitely not worth trying anything else.

Sure it is. But it has to be valid, not just a waste of time.

If you think it is, pr...

Hey, wait. I *think* I said that before.

     smog  |   bricks
 AIR  --  mud  -- FIRE
soda water |   tequila
 -- with thanks to fortune

Attachment: signature.asc
Description: Digital signature

Reply to: