Jump to content

Talk:Hash collision

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

hub?

[edit]

The article currently mentions the word "hub" 5 times, without ever explaining or linking to the meaning of this term. Very poor form for a Wikipedia article. — Preceding unsigned comment added by 1.21.121.16 (talk) 05:58, 14 April 2024 (UTC)[reply]

Wiki Education Foundation-supported course assignment

[edit]

This article was the subject of a Wiki Education Foundation-supported course assignment, between 28 January 2019 and 25 May 2019. Further details are available on the course page. Student editor(s): Mhou1.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 23:07, 16 January 2022 (UTC)[reply]

Partial revert

[edit]

I moved some sections from here back to hash table, where they belong; hash collision resolution strategies are not relevant outside of the context of hash tables. I've left a brief summary here along with a prominent link to the hash table article. I hope this is acceptable to everyone. Deco 07:48, 7 May 2005 (UTC)[reply]

"Most" -> "All"?

[edit]

The article says "Most hash functions have potential collisions".

All hash functions that have a finite output are vulnerable to collisions. This should be changed. --Tim1988 talk 18:18, 23 August 2006 (UTC)[reply]

Not quite. With a small, fixed set of keys one can construct a perfect hash function for those particular keys. HFuruseth 11:19, 30 October 2006 (UTC)[reply]

Merge Collision attack and Hash collision

[edit]

I feel that the Collision attack article should be merged into this article. The two articles describe essentially the same thing, with Collision attack merely being a specific use of a Hash collision. It doesn't seem to make much sense to keep both articles, as even if they were separate enough to warrant separate topics, a lot of the information from Hash collision would have to be copied into Collision attack for it to make much sense. --Rodzilla 05:49, 30 December 2006 (UTC)[reply]

Collision resistance

[edit]

The part about collision resistance needs a review. It looks like the sentences of weak and hard collision resistance are somewhat the same! In the part about weak resistance it is stated, that it is hard to find an x!=y with H(x)==H(y) and that the possibility of finding such a y is negligible.

This is the same as in the part about strong resistance and so it can't be right. —Preceding unsigned comment added by 81.221.221.30 (talk) 14:27, 28 September 2008 (UTC)[reply]

Collision vulnerabilities

[edit]

When hashing n strings into n buckets -- and the hash function is not a perfect hash -- why is the best possible percentage of hash collision 36.8%? See http://simonhf.wordpress.com/2010/09/26/brute-force-testing-of-hash-function-collision-vulnerabilities/ SimonHF (talk) 00:37, 26 September 2010 (UTC)[reply]

Why isn't this two pages?

[edit]

This page covers 2 very separate topics. Collisions in a hashtable and collisions on a bus network. Why isn't this two pages? Mdnahas (talk) 15:56, 7 May 2015 (UTC)[reply]

@Mdnahas: Information icon Thank you for your suggestion. When you believe an article needs improvement, please feel free to make those changes. Wikipedia is a wiki, so anyone can edit almost any article by simply following the edit this page link at the top.
The Wikipedia community encourages you to be bold in updating pages. Don't worry too much about making honest mistakes—they're likely to be found and corrected quickly. If you're not sure how editing works, check out how to edit a page, or use the sandbox to try out your editing skills. New contributors are always welcome. You don't even need to log in (although there are many reasons why you might want to). -- intgr [talk] 08:43, 8 May 2015 (UTC)[reply]

Content removal

[edit]

@Editor Guy Dude: Removing the content of the page, you turned it in to a sort of disambiguation page, but this is useless since we already have Collision (disambiguation).

The confusion about the two topics was introduced in 2015 with this edit. But the original article was about hash collision, and so it should be. For collision in telecommunications we have a distinct page. --Horcrux (talk) 08:36, 22 November 2018 (UTC)[reply]

My apologies. I read in the discussion a suggestion that information on the page was redundant. When another account provided the advice to "be bold" when making changes, I (regretfully) did so. I hope the page gets successfully modified by a more competent user. Editor Guy Dude (talk) 12:00, 22 November 2018 (UTC)[reply]
@Editor Guy Dude: You should always do be bold :-) I just wanted to inform you what the article was originally intended to talk about. --Horcrux (talk) 00:10, 24 November 2018 (UTC)[reply]

Rename article

[edit]

Apparently, following the discussion in 2015 (#Why isn't this two pages?), the article has been split along the line "Collisions in a hashtable" and "collisions on a bus network". But then, both parts should be named appropriately. That is, "Collision (computer science)" should be renamed to "Hash collision (computer science)", or simply "Hash collision". Are there any objections? - Jochen Burghardt (talk) 09:30, 7 April 2021 (UTC)[reply]

 Done - Jochen Burghardt (talk) 12:38, 20 April 2021 (UTC)[reply]

@Jochen Burghardt: FYI I've requested a technical move at WP:RMT to rename this again: Hash collision (computer science)Hash collision

I think the previous move was an improvement, but the disambiguation suffix is no longer needed, as the term "Hash collision" is not ambiguous. See WP:PRECISION, WP:QUALIFIER. -- intgr [talk] 14:05, 19 May 2021 (UTC)[reply]

Funny Math

[edit]

“Scalable Data Warehouse Architecture" apparently claims that MD5 has a 50% collision risk after 5.06 billion records. I can’t find anyone else matching that claim. The usual number is 2^64 records for a 128-bit hash.

If this is accurate, could we reference a publicly-available source that shows the math? 100.34.225.166 (talk) 15:05, 14 May 2022 (UTC)[reply]

Im also dubious of this, the maths is shown in this wikipedia article Birthday attack which shows that MD5 has a 50% collision risk for 2.2*10^19 records. SamuelBurke (talk) 13:55, 18 August 2022 (UTC)[reply]
It seemed wrong to me as well so I checked the source and it doesn't mention MD5 or hash collisions at all. That whole sections seems wrong. The other source cited doesn't mention hash collisions either. I'm removing it. Tsattolo (talk) 16:42, 9 August 2024 (UTC)[reply]