web analytics

Don Christie - Brandon Keepers

GitHub head of open source Brandon Keepers says: “In an ideal world everything would be open source, but that’s not always the case”.

Open source has many advantages, but it isn’t always the right approach. At the Open Source Open Society 2025 conference in Wellington delegates discussed when projects should be closed and when open is best.

…it all depends on the circumstances

When Open Source is not the best choice

He says at GitHub there are three cases when it is better to stay closed:

  • If it makes money. Remember that money can be used to finance open projects.
  • When it is specific to internal business processes.
  • When you can’t expect to maintain the project in the long-term.

Eventually many of these cases will be brought into the commons.

As an example of this, he says GitHub’s billion code has not been released as an open project. The thinking here is that making it open wouldn’t make the code any better.

Coding not difficult

Catalyst IT founder Don Christie says one argument in favour of open is that coding isn’t difficult.

Most of the time that means others can quickly replicate closed software. He says: “They are going to replicate it anyway. It can be better to make it open source and get the benefits of better code.”

Another argument for keeping projects open is that there is less money in keeping them closed. Christie says: “80 percent of the value in information technology is in services. About 90 percent of New Zealand’s IT exports are in services — that’s despite all the attention given to products.”

Christie says open source also acts as a hiring strategy.

2 thoughts on “Github: when to choose closed or open source

  1. The only reason to keep something closed as it “makes money” (except the mentioned internal software) is because you’re worried someone will take it and do it better. There is nothing inherent with open source making it free. Businesses are generally just afraid it makes it easier to copy but history has shown closed source software isn’t that hard to get a hold of. (Illegally) I could install Windows just as easy as Linux right now.

  2. In the late 1990s and early 2000s I worked for Australian Linux Today. At that time open source was a vibrant, yet immature, market.
    Visiting the Open Source Open Society 2015 conference in Wellington this week was like watching someone else’s child who you knew as a smart but awkward teenager, grown into a smart, professional adult.
    Open source grew up.
    Linux didn’t dominate
    When I was writing daily about Linux, the operating system and apps were already hard at work in data centres, on servers and on high-end workstations.
    The IT market was still moving away from a model where servers came with an expensive to buy and expensive to support operating system linked to the hardware maker.
    Some of those OSes were fully proprietary. Others were versions of Unix although they often had proprietary branding and non-open components.
    Much of my writing centred on asking how people and companies could make a living from the new model. We all knew that money could be made from selling services, but that didn’t look like enough revenue to sustain technology companies and jobs.[1]
    So the open source crowd spent a lot of time trying to create related products, usually apps, that they could charge money for.
    Desktop Linux proved a diversion
    There was something else going on. The open source movement looked to displace Microsoft on the PC desktop. Linux set out to challenge Windows and open source apps jockeyed to capture share away from the likes of Microsoft Office, Adobe Photoshop and MYOB.
    In one sense the PC open source explosion never happened. In part that was because Microsoft changed tack to see off the threat. Among other things the price of desktop apps fell. The commercial software developers bundled more value into their products and attempted to shorten the time between releases.
    Most of all, they had to sharpen their act. That meant creating cleaner, more robust applications. It was about this time the overall reliability of everything PC related switched from not-acceptable to not-all-that bad.[2]
    Open source everywhere
    There are still hold outs who use desktop Linux. I thought I might see some at Open Source Open Society 2015. In the event I think I briefly glimpsed something that resemlbed a desktop version of Linux, but in truth the conference was dominated by devices sporting the Apple logo.
    That last sentence would have enraged the open source community in 2000. Today there’s little of that kind of sensitivity.
    And anyway, Apple’s OS X is, like Linux, a Unix operating system. Large parts of OS X are open source.
    Android is the other operating system widely seen at the OSOS2015 event. While there’s some debate about the relationship between Google, Android and open source, at core the phone operating system is fully open. You could grab the source code and create your own Android fork.
    Even Microsoft and Windows have embraced open ideas in many ways since 2000.
    Sure, it’s not perfect. There are question marks, but while open source never won a pitched battle against proprietary software to conquer the PC desktop, open source DNA is everywhere you look.
    Open source economics
    Perhaps the most grown-up thing about today’s open source is that businesses have learnt how to use it to make money. This was mentioned briefly by Catalyst IT founder Don Christie in the session on when to choose open or closed source software.
    Open source companies make money from wrapping services around free software. It’s not as much money as selling popular proprietary software. Services don’t scale as quickly as a runaway hit app.
    Selling services doesn’t subsidise as many overblown long lunches. Nor does it require as many well-paid salespeople. On the other hand, because many services are charged by the hour, it does mean plenty of good jobs for people with the right skills.

    This turned out to be the case, the economic changes of moving from proprietary to open were devastating for most companies. Many didn’t make it. On the other hand, the possible death of moving to open source trumped the certain death of not moving.
    In hindsight, it’s easy to forget just how flaky most software was at the time. It was also before auto-saving was perfected so crashes that lost a whole day’s work were not unknown.

    Like this:Like Loading…



  • Bill Bennett

Leave a Reply

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

%d bloggers like this: