Use of Open Source in Commercial ELN products

On the (rather hard to get into) LinkedIn ELN group, there was a question about the use of Open Source components in commercial ELN products.

We use a lot of Open Source components in our products, and I know we’re not alone.

There are vendors who are very committed to a specific platform – Windows (and associated libraries, APIs etc.), Oracle, and so on. Those will almost certainly have some Open Source components in them, but not much.

There are other vendors – Amphora (my company) and Rescentris are two that I know of – who have built on top of an Open Source stack. We do have some proprietary components but where there’s an Open Source alternative we use that.


  • The Open Source stuff just works better (and you can fix it if it isn’t).
  • Support is better.
  • Licensing issues go away (of course you have to abide by the Open Source license, but that’s not a problem as long as you check everything before the developers start using it).
  • It is dramatically cheaper for our customers to deploy. No expensive additional Windows Server or Oracle licenses (unless you want to use those of course – we can support them if you prefer).
  • We have much more latitude in deployment options. We can bundle our product in a variety of ways and on different platforms, which we wouldn’t be able to do if we were locked to a specific commercial platform. You can get PatentSafe as everything from SaaS, to an embedded device, to a traditional “Install this on your own server” software product.

A few years ago we got raised eyebrows about our platform choices (“We will only consider applications written in .Net”) but that’s not been an issue for a long time. Everyone assumes that the components we’ve assembled into the solution will work and we’re responsible for the overall performance of that – what bits we’ve chosen seldom get discussed.

From what I can see, Open Source starts at the bottom of the stack – the OS, generally – and is gradually moving up (Database, Application Server, some applications). Every commercial vendor needs to keep an eye on what value they are bringing compared to what’s provided by the community.

As an aside, we don’t consider ourselves to be a “Software Vendor”. We solve a business problem and it just so happens we deliver our expertise as some software which implements a “best in class” process. But we don’t consider we’re charging for “software” – we get paid for our expertise and how we deploy that to help customers solve their problem.

What that means is that as the software environment changes (and quite probably Open Source gets further up the stack) that’s not something that threatens our identity. I know some vendors (particularly those locked to proprietary platforms) aren’t so lucky, and I wonder how they will fare as the Open Source community begins to provide more and more of the “ELN” system.

3 Replies to “Use of Open Source in Commercial ELN products”

  1. To answer Egon’s question we don’t really do Cheminformatics – there are plenty of solutions out there and we don’t think there’s a lot more we can add.

    What we do is give you the ability to use any “front end” system as part of your ELN system – and our PatentSafe product stitches the whole thing together for you into a single repository. So chemists can do what they want from a Cheminformatics perspective, Biologists will generally use a whole variety of stuff, etc. etc. – and it can all get collated into a single system. Which is really the only approach you can take in diverse populations of users – which is to say, the situation most research organisations are in.

    So it is very typical for our customers to have one or more Chemistry-centric systems, a couple of home grown systems to support specific activity, plus a bunch of ad-hoc stuff in Word and Excel. Rather than have to throw all that out (or integrated it painfully) to get a cohesive ELN system they can just use PatentSafe to stitch it all together, and keep the lawyers happy while they are at it.

    (in case this causes concerns about the amount of integration required – rarely is any needed. The “buy it all from one vendor and you’ll save a fortune in integration costs” argument is nonsense).

    I don’t know where the “ELN” is in that scenario which is why I have some difficultly with the term. Is it the many applications that are used by the users to do their daily work, or the application that actually replaces the paper notebook? Hence why I view it as a matter of creating an “ELN System” which is the sum of the constituent parts.

    In terms of the Open Source software we use, PatentSafe is Java based and therefore use Java, Spring, Hibernate and a load of other libraries. If the customer has no firm views on Databases we’ll use Postgres although we have some where the policy is “There is no other SQL database other than Oracle” (or SQLServer) which we’re happy to work with – although Postgres is by far the most trouble free of all the databases we support! All our SaaS systems run on Postgres.

    We’re finding that combination works well, for all sizes of customer – from a handful of users to many thousands of users on one server.

    We do have a reporting/business intelligence product (“The Custodian’s Console”) written in Ruby on Rails – where we don’t need the same scalability as we have with PatentSafe. It is intended for just a few users in any given company, but the customer’s needs can be quite diverse so the flexibility of Rails really helps – and it means the customer can adjust things to taste themselves, if they so wish.

Leave a Reply

Your email address will not be published. Required fields are marked *