“The best edit is often a complete re-write. And maybe it’s time to start writing lots of things from scratch.”
Peter Thiel (actually a notes essay from Peter Thiel’s CS183: Startup – Class 13 lecture).
As the URL states, I’m passionate about OSS. But as passionate as I am for current-day OSS, I’m even more passionate about what they could become. OSS can do some very cool stuff that seriously benefit the organisations that use them, but nobody will claim I’m getting too picky if I say that they still have lots of room for improvement.
So what if we were to take a Ctrl-Alt-Delete mindset to OSS and start re-writing it from scratch without the baggage of years worth of product upgrades? What type of changes would you look to make? Evolutionary or revolutionary?
It seems that the industry as a whole is currently taking the more evolutionary path, but I’m not sure that this is the path to the future of OSS (ie without the current-day, industry-wide flaws).
We are mid-way through a tectonic shift in the world of electronic communications and OSS are a pivotal component of that shift, as described in my Market Research Report, “The Changing Landscape of OSS.” And mid-way through a tectonic shift sounds like the perfect time to propose revolutionary change to OSS too.
So over the next week, we’re going to look at some of the big-ticket complaints about OSS and take a control-alt-delete perspective on them:
- The integration tax is too high
- Our requirements keep changing
- What are the tangible benefits or insights
- We don’t have enough skilled people
- Our data keeps deteriorating
- What language are you using
- [Ed: inspired by Roger’s suggestion] The cost (and complexity) of change is too high
I’d love to hear your big-change thinking on these topics and the other issues that you feel are potentially damaging the OSS brand (ie in general, not particular brands of OSS).
5 Responses
It takes too long to get changes made.
It costs $xxx,xxx for “you guys” to get out of bed to make a change.
Hi Roger
Great additional points!
Keep watching as the series unfolds because your points deeply intertwine with the stories in the upcoming blogs!
Hi Ryan – now that you’ve brought my attention back to it; this short note was meant to be imply you need a point (7) the cost of change is too high – which I suspect might have been the motivation for the Peter Thiel quote.
Too often in implementation, the energy that is left to address non-functional requirements after eliciting functional requirements is spent focused on manifest (performance, reliability, availability) and operational (throughput, manageability, security, serviceability, testability) qualities. Developmental qualities (realizability, planability) and evolutionary qualities (scalability, maintainability, extensibility, flexibility, portability, reuseability) receive little focus, even though they will come to dominate the cost of system ownership – and more importantly for this complaint, the cost of system change.
See http://smartweb.sourceforge.net/qualities.html – although I’m certain there is an IEEE model out there somewhere that this has been drawn from.
Hi Roger,
You make a very persuasive point! Funnily enough, that’s kind of what I was getting at in the more recent post which borrows your original comment. Check it out here https://passionateaboutoss.com/rebuilding-the-oss-boat-during-the-voyage/
Thanks for the sourceforge link. I hadn’t seen this one before, but will definitely build a future blog around it!
PS. I’ve also now made an update to this “complete re-write” post to include your point about the costliness of change