Dennis Howlett penned his thoughts on open sourcing SAP Hana http://diginomica.com/2015/06/17/should-sap-open-source-hana/ that led to a fairly good debate on Facebook , and I thought I will share some thoughts on the topic here on my blog
Hana is a great piece of software which is a full fledged database , and have some lightweight app server capabilities . While it could very well be a general purpose database – historically it’s been used mostly under SAP business suite and business warehouse , and some data mart type use cases . It’s a well crafted piece of software and of course I am a bit partial to it . So what about open sourcing it ?
I was a big proponent of open sourcing Hana when it first came out . I no longer think it is a great idea . Here are six random reasons why
1. By now , there are way too many open source databases optimized for many different things . No first mover advantage remains for sap
2. These open source databases all have large community following . SAP has a very large (larger than most open source databases) and loyal community who need to be nurtured on to Hana . That is a much more pragmatic approach than getting say Hadoop developers to switch to Hana
3. Hana is rather limited on drivers . If widespread adoption by community is needed , SAP will need to support drivers for many languages on Hana . I don’t see the need for that investment given the heavy focus on S4Hana for near future . River was not the right approach in my opinion – that is not how real life developers build apps . That is an academic view of the world .
4. Developers need software they can play with quickly and decide if their use case is a good fit . The trial needs to be quick to install , learn and tweak . People are happy to pay for support and enterprise grade features . Don’t confuse between open source licensing and open source business model . SAP can keep Hana commercial and just focus on making Hana extremely developer friendly for unlimited trials to get to the same results
5. What is really different between an open source database company that employs all (or most) of its commiters and SAP ? In both cases the company controls product direction – with of course some input from company . This is not a valid reason for SAP to open source Hana in my opinion . Of course not all open source companies employ all their commiters – but many successful ones do exactly that . Essentially that negates the argument that no one has all answers to a problem . The better solution is in vendors working together to make interoperability work better – consistentcy in driver support , security , provisioning , HA/DR etc
6. Hana is not the only game for SAP. For SAP business model – till cloud can pick up significantly in net new business , it makes more sense to have Hana as high ASP , lower volume as sales model . That is the opposite of typical open source subscription model by other database vendors . SAP already sells Hana as a subscription I think – but doing it at scale , at a comparable price point to say MongoDB , Cassandra etc is just too disruptive in my mind . They might get there at some point – and to do a subscription business , software doesn’t need to be open source licensed necessarily .