User talk:Royfriedman123

CAP theorem
Hi,

I saw you reverted my changes on the CAP Theorem. I'm not sure I understand why (is it because I removed your external link? I didn't see its relevance in the context of this article. Maybe it could fit better as a reference rather than an external link?), but I would be happy to discuss those. 7804j (talk) 14:31, 29 September 2016 (UTC)

Answer: Hi,

I do not feel very strongly if these items should be listed as references or external links, but do have a slight preference toward external link for the following reason: in my mind, a reference is the original result, or the sources based on which the Wikipedia page was defined. But if you prefer references, I am fine with those - but then I do not understand why you left the "CAP 12 Years Later" under external links. At any events, the links you removed are definitely useful for people interested in CAP theorem as they help understand the concept and its practical implications. The last link - the paper by Birman and Friedman - is in fact the same as CAP, but more rigorous and formulated 4 years earlier. Historic academic justice mandates that it be mentioned in this page. By the way, similarly, the PACELC theorem (that I think you added) rediscovers the famous result by Attiya and Welch from 1991 stated in "Sequential Consistency vs. Linearizability" that both the paper by Abadi and the PACELC fail to cite. In summary, these removed links are helpful, and so they should be restored, and if you insist on references rather than external links, this is fine, although I personally think that external links is more appropriate, as I explained above. Royfriedman123 (talk) 17:47, 30 September 2016 (UTC)


 * Hi Royfriedman123,


 * The reason why I kept the 2012 article is because it is an important statement from the original author of the theorem, which was even discussed in the core of the article.


 * I removed the PACELC link, because it belongs more to the PACELC page (which I just updated; it was already existing) rather than to the CAP theorem. As for the three other links, I felt at the time that they weren't of particular notability (maybe I was wrong) and did not add anything that could not be included directly in the core of the article (see external links policy: "If the website or page to which you want to link includes information that is not yet a part of the article, consider using it as a source for the article, and citing it"). I agree that the paper of Birman and Friedman for example is important if it discussed of the subject before the CAP theorem was first stated, but then maybe we should put that directly in the "history" part of the article and cite it as a reference?


 * But I don't feel especially strongly either about external links. If you feel like they should still be brought back, please proceed; just don't override my other changes ;)


 * Also if you feel like PACELC is "stealing" the work of previous academic research, please feel free to mention it on the PACELC page. I am not an expert in the field, I just tried to improve the description of the theorem based on my understanding 7804j (talk) 13:44, 1 October 2016 (UTC)


 * Hi 7804j,


 * Good, so I will re-add the links without canceling your other changes. As for the PACELC page, I'll think of a way to add the Attiya&Welch result without implying any intentional wrongdoing on behalf of Abadi. Perhaps saying something like that "the PACELC can be viewed as unifying PAC with earlier results about the latency vs. consistency results such as Attiya&Welch". Royfriedman123 (talk) 05:17, 2 October 2016 (UTC)