Talk:Pearson hashing

Removal of Python snippet

I've removed the Python code snippet for the following reasons:

• One code snippet is sufficient - the pseudocode is enough to clearly demonstrate the principle of the algorithm
• It's redundant - the Python snippet implements the algorithm in precisely the same way as the pseudocode
• Precedent on most algorithm articles is to remove duplicate code examples in different languages if there's already one canonical example. Sometimes they've been moved to Wikibooks, but I hardly think that's worth it for 5 lines of code.

Oli Filth^{(talk|contribs)} 17:36, 11 September 2009 (UTC)

1) It's not sufficient. Pseudo-code is good for showing the concept, since it's not wedded to an implementation. Sample code is good for actually being able to step through a working implementation, as well as using it as a reference for your own.

2) It's not redundant. Look at them for yourself and you'll see there are optimizations in the Perl that were not present in the pseudo-code and which make it less than obvious that the two have identical behavior.

3) This precendent is disputed.

208.80.104.2 (talk) 18:20, 11 September 2009 (UTC)

Please stop this; you've already made it explicitly clear that you're willing to edit war until someone gets blocked, if you keep this up I'll report you for tendentious editing.

I've pointed you at examples where consensus has been to remove pointless extra code snippets, and your claim that the Python example is "optimised" is nonsense (there's no such optimisation); even if it were true it would not be relevant, as Wikipedia is not a how-to guide. Furthermore, the pseudocode is from a referenced source (apparently), whereas the Python isn't. Oli Filth^{(talk|contribs)} 18:26, 11 September 2009 (UTC)

Please don't be uncivil my misrepresenting what I've said. I said repeatedly that I would prefer to avoid conflict. I also said that if you try to get me blocked on false charges of incivility, you will expose yourself to a block on true charges of the same. In the interest of avoiding conflict, I will wait until you can manage a civil response. Until then, I will edit as I see fit, within the rules of Wikipedia. 208.80.104.2 (talk) 18:53, 11 September 2009 (UTC)

So you're opting not to reply to the counterarguments I made?

For reference, I was referring to this edit of yours. Oli Filth^{(talk|contribs)} 18:56, 11 September 2009 (UTC)

And conveniently ignoring the [1] next edit. Very selective of you. 208.80.104.2 (talk) 18:58, 11 September 2009 (UTC)

To be clear, so long as you are uncivil, I will not pretend that you have any arguments to refute. 208.80.104.2 (talk) 18:59, 11 September 2009 (UTC)

That approach will get you nowhere fast. If you aren't going to participate in discussion, then don't make edits to the article! Oli Filth^{(talk|contribs)} 19:02, 11 September 2009 (UTC)

Practical impact

I have a bit of a conflict here, as I know that Pearson's Hash benchmarks well for many uses, from my own testing. So the practical impact is quite significant, and worth noting. However, as I understand, Wikipedia "no original research" policy frowns on linking back to your own articles, so I have NOT linked back to my own writings. Which also seems to make writing a weblog article to document my own benchmarks as not-quite-suitable. So how to make this all "legitimate"? pbannister (talk) 21:50, 17 January 2010 (UTC)

Write a good article/blog post on it and wait for somebody else to consider it to be a good source for this article. That's all you can do. 82.33.61.156 (talk) 16:00, 18 January 2010 (UTC)