My favorite Fool writer, Brian Stoffel, wrote a piece on Mongo last week: https://www.fool.com/investing/2018/10/08/why-im-buying-shar…
It’s got some great information for anyone not familiar with Mongo, a little detail for those of us who are, and a nice vote of confidence for anyone who owns it.
This month has been really volatile, but we’ve discussed lots of businesses that are doing great. I think they’ll have a bright future as well.
Bear
Stoffel’s article introduced me to MongoDb Stitch which is a very exciting development.
MongoDB Stitch
https://docs.mongodb.com/stitch/
They call it “MongoDB Stitch serverless platform.” I find the “serverless” part misleading, the server is still there, what changes is how the server is accessed, instead of directly by the app, access goes through Stitch. This is state of the art architecture, using various APIs as middlemen or “transaction brokers” as they might be called.
I’m liking MongoDB better every day.
Denny Schlesinger
I talked a bit about serverless at the bottom of this post: https://discussion.fool.com/it39s-really-not-that-hard-34020522…
What Stitch does, in my view, is turn the Cloud into a giant Object-Oriented paradigm. See this article: https://www.zdnet.com/article/mongodb-stitch-serverless-comp… , for instance.
The main difference from most other serverless interfaces like AWSD Lambda is that through Stitch a developer can maintain “state.” Typically with Lambda, the developer passes data to the function being called. With state, the data stays in the cloud. This makes using Stitch sort of like programming in an Object-Oriented language (pretty much state of the art for years now for application languages like C++, etc.). Makes sense given it’s a database, after all.
Mongo Stitch is more than just a Mongo DB on, say AWS Lambda. It’s actually its own JavaScript runtime. That lets me them make some things easy, but it also means some compatibility issues with commonly used tools like Node.js
This article takes you step by step through an example: https://thecodebarbarian.com/building-a-serverless-app-with-… . Even if you’re not technical, you see that it’s a mouse-click style interface, not a Command-Line Interface. Don’t even have to be a programmer in this example.
BTW, from the first article: An upcoming update to MongoDB will add new in-database analytic capabilities, with built-in visualisations. Is MongoDB looking to take some of the analytics and visualization revenue away from companies like Alteryz and Tableau? It’s an interesting development, and shows how hot and competitive this whole space is.
Smorgasbord1, you are right, technically it’s more than I wanted to know! LOL
Investment-wise the simpler it is to get your app up and running on a paid service the better.
Thanks.
Denny Schlesinger
technically it’s more than I wanted to know! LOL
Yeah, but my point is that this is a real advantage, not just some marketing thing. Sometimes companies announce products or services that are nothing more than marketing hype. It might sound good, but in practice it’s not compelling.
In this case, however, it’s the real deal.
Rather than just saying that, I thought I’d provide something to back that statement up. All good.
In this case, however, it’s the real deal.
I realized that when I read the documentation. At NPI:
I went in search of “MongoDB Stitch.” They call it a “serverless platform.” I find the “serverless” part misleading, the server is still there, what changes is how the server is accessed, instead of directly by the app, access goes through Stitch. This is state of the art architecture, using various APIs as middlemen or “transaction brokers” as they might be called. Curiously, it has a similar goal as Pivotal, to make implementation simpler, Stitch connects all the various clients to the database on the server hiding all the ugliness of the implementation. The concept has been in development for a long time, with libraries that would connect a generic interface to any of a number of databases. Stitch does the same but in a more modern fashion with implementations in iOS, Android, and Javascript (browsers).
https://discussion.fool.com/tinker-i39m-sold-on-mongodb-34038885…
Denny Schlesinger
MongoDB is the modern general purpose database for enterprises. Mongo not only has the networking effect in a winner take most market (we are all familiar with how in enabling software the larger get larger even if they are not the best) but they also are the best.
Cassandra, as an example, may do one function better than Mongo or other, but no NoSQL database has the useful functionality Mongo has.
As an example, not touted (as ACID gets all the P.R.), but is the CIO’s favorite feature is aggregation pipeline. This will be a much more used functionality than ACID, although ACID may sell more business. No other NoSQL database has ACID, and no other NoSQL database (to my knowledge) has aggregation pipeline.
You can dig further, but here is one article in regard: https://www.datanami.com/2018/06/27/a-mongodb-secret-weapon-…
Note how this is the CTO’s baby, and this CTO reminds me very much of Arista’s CTO. Both are founders, both are already rich, and both thoroughly love their ‘babies’.
Twitch is yet another exclusive feature. There is a reason why Mongo is seeing little to no competition with many of its sales to enterprises. The real competition is either little to none in greenfield uses, and otherwise against SQL databases for applications that SQL databases are becoming non-optimal.
Overtime, since Mongo can now do things that heretofore only SQLs could do, plus they can do all the other stuff and more enterprises will just start to put SQL functionality on their already working MongoDB just because it is there and will work.
I cannot see how we are wrong on this analysis. May have some choppy market waters from here to there.
Tinker
MongoDB is the modern general purpose database for enterprises.
While I appreciate Mongo’s position, one has to be careful about statements like this because of the temptation to gloss over the word “modern”. It it still the case that, by far, the “general purpose database for enterprises” is a SQL database … and likely to continue to be so for some time. SQL database sales are not yet declining, as far as I know.