OpenSourceInLargeCommercialCompanies
From SPA Wiki
OpenSourceInLargeCommercialCompanies was run by PaulSimmons and TomAyerst.
DresdnerKleinwortWasserstein (DrKW) have open sourced mission critical software via collab net. The product is open adapter 1, a 100% Java/XML-based software platform which allows for rapid business system integration with little or no custom programming.
This lively bof examined why companies go open source, what's in it for them.
Contents |
Why?
- Cost
- IT retention/attraction/motivation (programmers like to contribute back to the community).
- Attract business partners (it is easier for partners to see evidence that the s/w is clean if it is open source. They are more likely to run it on their site).
- Legal implications for laboratories - ie. open source calculations which would normally require all sorts of legal checks, this would save labs £££$$$. (this last topic was not discussed openly in the bof).
What to open source?
- Only infrastructure/horizontal?
- Most people are not sure about whether vertical apps can be open sourced? Too much commercial competition to make this viable.
- e.g: Databases? What business model would Oracle use if they open sourced their product?
- Can only one of each application survive in the open source world? (nb: currently their are 2 open source db offerings).
3 Levels of Use for Open Source Software (OSS)
- Use OSS(normally comes under the corporate radar)
- Contribute to OSS(e.g: safeways).
- Found OSS(e.g: DrKW) BUT the organisation must support and enhance it until a critical mass of developers in the Community forms.
Do we naturally progress from use->contribute->found?
3 Negative reasons to produce OSS
- Dumping - companies finish some software and no longer want to support it.
- Get on the "band-wagon"
- Give it for free to ruin your competitors.
Note that the last one may well be an effective strategy but it will not work if the Community doesn't feel that the product is of sufficient quality or the commitment to it is weak.
3 Grades of OSS support
- OSS with no-one to support it.
- OSS with initial critical mass but no corporate support ("no one to pay")
- "Red Hat" model of support
Support issues: Are the support companies big enough to support your (large commercial) company? Do you just want someone to sue? (General bof opinion was that its not about someone to blame or sue).
I'd be interested to know what level of exposure both the participants and their employers had to Open Sourced software:
- Heard about Open Source
- Have evaluated Open Source
- Actively employ Open Source
- Have contributed to Open Source
- Actively contribute to Open Source
From above: Found OSS(e.g: DrKW) BUT the organisation must support and enhance it until a critical mass of developers in the Community forms.
Even then the product will still need direction and a central point of management (forgive me if I'm wrong, but I don't think OSS just 'happens' - someone (or some group) has to plan and manage releases, act as architect, and so on. Companies supporting OSS in this way must realise they're in it for the long run.)
Also - there's another model of OSS within an organisation, and that is to build an internal community around internally produced (typically infrastructure) technology which is evolved by projects as they use the software and require enhancements. I've thought about introducing this once or twice - has anyone actually done this?
