Re:Why is printing “B” dramatically slower than printing “#”?
by logicchains on Tuesday February 25, 2014 @01:39AM
(#7296048)
There are words in other languages, French for instance, that with an accent changed or removed have completely different meaning.
Re:Why is printing “B” dramatically slower than printing “#”?
by bausson on Tuesday February 25, 2014 @03:36AM
(#7296367)
That is sadly true.
It's even worse than that. 'e' is the most used letter in french, but 'é' and 'è' are quite used too. 'ê' and 'ë' being rarer, but still existing.
So, I am exaggerating it a bit, but you can compare it to having to search for words with the consonants, because every vowel is considered the same by your search algorithm.
Yes, I'm pushing it, but you see the problem it can easily become in huge pages.
Re:Why is printing “B” dramatically slower than printing “#”?
by JetSpiegel on Tuesday February 25, 2014 @05:44AM
(#7296799)
Why sadly?
Re:Why is printing “B” dramatically slower than printing “#”?
by bausson on Tuesday February 25, 2014 @08:35AM
(#7297470)
What is sad for me is chrome new way of handling thing, not french accents (which can be a bother, but they're cute, so that's ok).
Most pages nowadays are small enough that it won't be a bother, but their still are some cases (newspaper posts, blog analysis, etc...) where accent sensitiveness will be an issue.
Is that loose of a search not actually desirable in "find in page" instances?
In other words, are there many words where 'e' matching 'é' would actually generate a false positive? This is the only case I could see that would be problematic.
And even then "find in page" is usually used as quick navigation to the minimum matching search (high recall) and not about excluding all possible mismatches (high precision).