Transactional Memory: History and Development | Hacker News Reader

Transactional Memory: History and Development

kukuruku.co

50 pointsnkurz14 comments

Weird. This article was originally written by me on Russian IT-resource habrahabr, here: http://habrahabr.ru/post/221667/. I don't see any link to it. Guys from kukuruku leaves only url on github, which shows it is my work. I'm not a fan of copyright or something, no blame on them. I would not mind if they attach a url to original article though. I was going to translate the article but now I leave it.

I study STM/HTM for my university work now, and this article is some kind of overview of technology.

Looks like this one is as well: https://news.ycombinator.com/item?id=7753166

(copy of http://habrahabr.ru/post/220841/)

And this one that showed up recently: http://kukuruku.co/hub/cpp/lock-free-data-structures-basics-...

(copy of http://habrahabr.ru/company/ifree/blog/195948/)

So that appears to be kukuruku's business model: Take articles from habrahabr, translate them (poorly) into English and publish them.

It's great that we get to see content that was previously only available in Russian, but I wish they would show more openly what they were doing. (I did actually wonder why their articles didn't have an author)

The second weird thing is that I compile an article from the English sources and translated them on Russian. Kukuruku folks translated them back to English, probably losing a bunch of meaning.

The Russian is terrible too, reads like Google translate. They left in the "silver bullet" idiom, which makes zero sense in Russian.

Russian translation is mine. I hear "серебряная пуля" (silver bullet) many times on Russian, so I don't think it makes zero sense. Many people have read Mythical Man-Month or at least knows what this idiom means.

A similar thing was done to another article from Habrahabr.ru.

Compare:

http://habrahabr.ru/company/ifree/blog/219201/

With:

http://kukuruku.co/hub/cpp/lock-free-data-structures-introdu...

The favicon also looks a bit familiar.

'made for hn'.

I went to the site and left a comment complaining about how they stole your article. They deleted the comment, and you still aren't credited. Does HN have a blacklist of sites that steal content or publish only blogspam? I think they should consider it.

Not cool at all if this is true. They gave you no obvious credit.

Joe Duffy has a really interesting retrospective on his team's attempt to build a STM for .NET:

http://joeduffyblog.com/2010/01/03/a-brief-retrospective-on-...

There is work being done to get STM into python, but it is currently underfunded. See the complete proposal at http://pypy.org/tmdonate2.html

There is also a good talk from the FOSDEM 2014 about STM in PyPy: http://www.youtube.com/watch?v=u0evRytGE0Y

An interesting read for someone interested in the performance characteristics of STM.

Atomic Quake: Using Transactional Memory in an Interactive Multiplayer Game Server

(pdf warning) http://www.bscmsrc.eu/sites/default/files/atomicquake-ppopp0...

There's a naive implementation of an STM in the concurrent-ruby gem that is a good overview of how an STM can work.

https://github.com/jdantonio/concurrent-ruby/wiki/TVar-(STM)

<shameless plug> If anyone's interested in a full-featured transactional memory implementation for .NET, have a look: https://github.com/jbakic/Shielded ... It's more or less complete, feature-wise, and should be (ahem) bug free. </shameless plug>

Reply on news.ycombinator.com