tag:blogger.com,1999:blog-768233104244702633.post5451116418644268796..comments2018-06-05T00:05:26.577-07:00Comments on The Scale-Out Blog: Disproving the CAP TheoremRobert Hodgeshttp://www.blogger.com/profile/05379726998057344092noreply@blogger.comBlogger11125tag:blogger.com,1999:blog-768233104244702633.post-61106493352443082292012-06-26T23:07:00.068-07:002012-06-26T23:07:00.068-07:00@Andre, I think you are perhaps misreading the not...@Andre, I think you are perhaps misreading the notation. Equations 1-3 are axioms as they are stated without proof. They have an "AND" relationship to each other as CAP asserts they are true for all time. <br /><br />As @Piccolapatria pointed out there are more parsimonious formulations but this part of the "proof" correct as far as I can tell.Robert Hodgeshttps://www.blogger.com/profile/05379726998057344092noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-91836006003297276082012-06-26T05:44:12.091-07:002012-06-26T05:44:12.091-07:00With
(1) A and P => not C
(2) P and C => not...With<br />(1) A and P => not C<br />(2) P and C => not A<br />(3) C and A => not P<br /><br />you translate the CAP theorem into <br />(1) and (2) and (3),<br /><br />while it should be translated into<br />(1) or (2) or (3).Andrenoreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-80557746027934545552012-04-20T12:31:21.821-07:002012-04-20T12:31:21.821-07:00@All, Anonymous has discovered an easter egg withi...@All, Anonymous has discovered an easter egg within the easter egg. Based on this new insight we can adjust the proof as follows. <br /><br />(5) C => (A <=> P)<br /><br />In other words if C is true A and P are equivalent. The rest of the proof then proceeds as follows: <br /><br />(6) C => (A and C => not A)<br />(7) C => (C => (A=> not A))<br />(8) C => (A => not A) <br /><br />So the conclusion is the same (whew!) and we still have soggy nachos. I invite other great minds of the age to submit further thoughts on the matter.<br /><br />(arg, had to repost to remove typos!)Robert Hodgeshttps://www.blogger.com/profile/05379726998057344092noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-30928338615977479772012-04-20T09:00:13.715-07:002012-04-20T09:00:13.715-07:00@anonymous:
the author doesn't enforce 4), ...@anonymous:<br /><br /><br /><br />the author doesn't enforce 4), but is just stating that the blog post he quotes does.piccolapatriahttps://www.blogger.com/profile/09568421446391495100noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-73790485967467881112012-04-20T07:03:07.015-07:002012-04-20T07:03:07.015-07:00nice bullshit ;)
You can not do that:
(4) C and A...nice bullshit ;)<br /><br />You can not do that:<br />(4) C and A <=> C and P<br />which further reduces to<br />(5) A <=> P<br /><br />becasue when C is False then A can be true and P can be False so (4) statement is true and (5) is notAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-91142404672739786362012-04-05T08:13:35.800-07:002012-04-05T08:13:35.800-07:00I was just adding some entropy to the post...didn&...I was just adding some entropy to the post...didn't mean to be boring ;)piccolapatriahttps://www.blogger.com/profile/09568421446391495100noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-13038181605488247452012-04-05T07:46:35.667-07:002012-04-05T07:46:35.667-07:00@All, thanks for reading! I hope this proof is us...@All, thanks for reading! I hope this proof is useful to all of you in your next start-up. <br /><br />@piccolopatria, Very parsimonious! I didn't see that reduction first time through. When I submit for publication I'll be sure to include it. Reviewers like to see solid mathematics as I'm sure you are aware.Robert Hodgeshttps://www.blogger.com/profile/05379726998057344092noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-21470066700844180912012-04-05T06:11:37.610-07:002012-04-05T06:11:37.610-07:00Nice easter egg ;)
Anyway having:
(C ^ P=>!A)...Nice easter egg ;)<br /><br />Anyway having:<br /><br />(C ^ P=>!A) ^ (A ^ P=>!C) <br /><br />Leads to :<br /><br />(A=> !C v !P ) ^ (C => !A v !P)<br /><br />and thus:<br /><br />A ^ C => (!C v !P) ^ (!A v !P) => !P<br /><br />So you just need two CAP statements ;)piccolapatriahttps://www.blogger.com/profile/09568421446391495100noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-14348658012740096122012-04-01T12:26:14.932-07:002012-04-01T12:26:14.932-07:00In the immortal words a slot machine spoke to me a...In the immortal words a slot machine spoke to me and my wife after we fed it the first dollar we ever spent gambling, and stabbed randomly at buttons whose meanings were unclear to us, and thus won $91 and change in a few seconds, "you're brilliant!"<br /><br />In the less-immortal words of the computer scientist with whom we had dinner a few months later, and then discovered that he'd programmed the very slot game we played, "yeah, that's a hard one. How much money did you lose on it?"Baronhttps://www.blogger.com/profile/01621441847303652718noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-17817745264206948852012-04-01T08:17:40.184-07:002012-04-01T08:17:40.184-07:00(Ah another Blogger/Firefox user who can't get...(Ah another Blogger/Firefox user who can't get in line comments to post.)<br /><br />One is tempted to take this as an April Fool's joke. OTOH, industrial strength RDBMS with 2 Phase Commit have been SuperCAP over many "partitions" for a couple of decades.Robert Younghttps://www.blogger.com/profile/09056808374481236610noreply@blogger.comtag:blogger.com,1999:blog-768233104244702633.post-90077397669509533702012-04-01T05:39:22.174-07:002012-04-01T05:39:22.174-07:00A database system which is offline can be perfectl...A database system which is offline can be perfectly consistent and unchanging.Anonymousnoreply@blogger.com