http://tonybain.disqus.com/Innovations in Information ManagementenThu, 03 Sep 2009 15:59:28 -0000http://blog.tonybain.com/tony_bain/2009/07/the-nosqls-need-to-engage-the-dbas.html#comment-15897843<p>I created a "No SQL" t-shirt and mug on CafePress with 0% markup.<br>You can find them here: <br><a href="http://www.cafepress.com/nosql" rel="nofollow noopener" target="_blank" title="www.cafepress.com/nosql">www.cafepress.com/nosql</a></p><p>Ryan<br></p>disqus_qCQ0mSyQ8kThu, 03 Sep 2009 15:59:28 -0000http://blog.tonybain.com/tony_bain/2009/07/the-nosqls-need-to-engage-the-dbas.html#comment-13661874<p>Time was, it wasn't going to get adoption until Microsoft productised it.</p><p>Perhaps we're finally leaving that sad era. But if non-R DBMS catch more mainstream attention, I am sure they'll wake up and do a Zune. (Note they already jumped on the functional bandwagon with F#, LINQ.)</p>qu1j0t3Thu, 30 Jul 2009 11:30:07 -0000http://blog.tonybain.com/tony_bain/2009/07/the-nosqls-need-to-engage-the-dbas.html#comment-13648900<p>Well, we (i.e. <a href="http://neo4j.org" rel="nofollow noopener" target="_blank" title="http://neo4j.org">http://neo4j.org</a>) have certainly tried. But it's come to the point when for example in VC evaluations, I normally say "If you talk to any programmer in our target demographic, they're going to love us but if you talk to any DBA or RDBMS implementor, they're going to hate us." And I'd say that's 95% true.</p><p>I don't think it has to do with lack of enterprise features. We certainly have everything you list: transactions (*real* transactions, i.e. transaction recovery battle tested over 5 years in 24/7 production), JTA/JTS/XA support (yes, including 2PC), online backups, etc. We do lack in reporting and in some aspects monitoring (working on that tho!).</p><p>But I don't think that's it. I think a large part has to do with the fact that half of our offering is programmer convenience. It's just so much *easier* to develop with a graph database, even if we completely ignore the runtime benefits (performance, scalability) we have over RDBMS for a class of applications. And that part is of little value to a DBA, who sees no or modest benefit in our model over an O/R mapping layer.</p><p>I agree with the sentiment of your post (as I read it): ops people are important. I've just been unsuccessful so far at having a productive dialogue with them.</p><p>--<br>Emil Eifrem<br><a href="http://neo4j.org" rel="nofollow noopener" target="_blank" title="http://neo4j.org">http://neo4j.org</a><br><a href="http://twitter.com/emileifrem" rel="nofollow noopener" target="_blank" title="http://twitter.com/emileifrem">http://twitter.com/emileifrem</a></p>Emil EifremThu, 30 Jul 2009 02:09:38 -0000http://blog.tonybain.com/tony_bain/2009/07/the-nosqls-need-to-engage-the-dbas.html#comment-13646961<p>so true!</p>VisuThu, 30 Jul 2009 01:24:34 -0000