[Info-vax] OpenVMS in the future, Open sourced or Closed? Intent is to keep it...
mcleanjoh at gmail.com
mcleanjoh at gmail.com
Tue May 5 19:20:40 EDT 2015
On Tuesday, May 5, 2015 at 2:59:10 AM UTC+10, clairg... at gmail.com wrote:
> Open Sourcing VMS
>
> As Steve pointed out, the technical obstacles to overcome are massive. However, the even bigger obstacle is buy-in from a significant percentage of the customer base otherwise it is a non-starter. If you were to seriously pursue an open source strategy you must start the conversation with customers early on and I can tell you that conversation will go nowhere fast.
>
> The traditional notion of "open sourcing VMS" is extremely unlikely to ever happen. But, I think it might be fruitful to pursue a way to incorporate the ideas and code of those who wish to contribute, via a process controlled by VSI. That would maintain a definable governance by a funded organization dedicated to the future of the product. If that means it is proprietary, so be it. It could be a way to reach a workable compromise between the open and proprietary extremes and gain many of the benefits of both.
>
> I'm open to ideas along this line. A clearly-defined process would be needed which we don't have time for right now but over the next year or so maybe we could experiment a little to see what might be possible.
>
> Years ago it seemed like the only way for VMS to have a future was to make it open source and have a community of developers take it forward. Things have changed. Let's figure out the best way to take advantage of the new opportunity.
Clair, I think there's two types of possible software contributions to VMS.
One way is to contribute complete utilities, such as the kind of software found on the Freeware CD's.
The other way is to contribute routines that developers might find useful. I'm porting something to Linux and finding that a lot of routines are already available (although often the problem is finding them!). Sure VMS has the low level routines, system services, LIB$ routines etc., but I'm thinking that the "contributed" routines will sit between these low level routines and the application.
cheers
John
More information about the Info-vax
mailing list