Beware - Architect inside

Recently while having a discussion with an "architect" working with large telco company, he batted for "orientdb" as graph database. He wants to build a large, secure and stable application. My argument was that I have been following up on this database for a long time and there are not enough takers in the community for this technology. They used to encounter large number of issues related to stability. So if you really have to use an open source technology find out competitors using google and compare. Make your choice based on 

a. Momentum/Adoption by some large companies

b. Number of stars on github

c. Issues being reported in stackoverflow

d. Nature of these issue 

e. Number of committers, their profile

and last but not the least identify within your network who has worked on these open source technologies.

My problem with these so called architects is that they make decisions based out of their limited reading and emotions and endanger organisation's interest and can incur huge losses to company in longer run.

If you want to use a multi model database then why will you not use "ArangoDB" for example which fares better on all the above parameters including almost double stars on github. Not that ArangoDb or any other piece of technology will be without issues but you have a solid reasons to choose one over other.

From same Architect it was argued that node.js should be default tech for enterprises as it is backed by google. Can only come from a rookie definitely not from an experienced person

Node.js no doubt is a great piece of technology but can an enterprise rely on all the packages open sourced by the community for reasons from security to stability, I am not so sure. Lot of packages in node.js community are not even 1.0. At times you need to try 10 packages to find out one that works. on top of that if you encounter issues in packages debugging is exactly not easy. 

I have seen developers making foolish choices only to bite in the end.

Any company where software architects are so shallow in their thinking process that they just look at arguments like "if google is supporting it then must be great", its future is doomed. 

What they do not know is that google also supported something called "GWT". Where it is today?

No doubt their is weight in this argument but building large scale software is far more complex compared to which language to choose. How mature is ecosystem around that must be looked at

As of today for an enterprise there exists no open source tech which can provide the scalability, security and stability of JVM if you are looking at open source. I have written a very elaborated blog titled "Why Java is my default Choice" precisely for same reasons.

Fact of the matter is that I have seen these so called architects running to the woods when production issues surface. They simply shift the blame on shoulders of developers and devops teams. CIOs obviously do not have a clue why their environments are not getting stabilised despite spending huge amount of money. I can only sympathise with these companies.