Skip to main content

Notice: this Wiki will be going read only early in 2024 and edits will no longer be possible. Please see: https://gitlab.eclipse.org/eclipsefdn/helpdesk/-/wikis/Wiki-shutdown-plan for the plan.

Jump to: navigation, search

Talk:Orbit/Promotion, Release, and Retention Policies

< Talk:Orbit
Revision as of 06:27, 2 February 2007 by Mober.at+eclipse.gmail.com (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Comments from Martin Oberhuber:

Do we really need multiple different qualified versions in the bundles area? By definition, these have the same content in terms of class files but slightly different meta-info, "markup" or licensing words. I would expect that such "markup" can only get better over time and can't see any reason why anybody would want a "previous-qualified" version. Especially, I cannot see how clients would ever reliably select a particular one from multiple different qualified versions (except referencing exactly one but that's also unreliable since it may be gone at some time).

The only compelling reason for keeping old "qualified" versions of bundles would be to be able and reproduce old builds exactly as they were. But for this, just keeping 3 qualified versions is not sufficient -- rather than that, clients would need to advertise what they put into an official release. It might be better to plan that clients who need to reproduce an old build should do that by running the Orbit build scripts out of CVS themselves rather than downloading some pre-built bundle. It might be a good idea anyways to make the Orbit build scripts such that they can run on any client's computer.

Also, if I'm not mistaken one of the prime reasons for Orbit was to ensure that all clients of a given library version use exactly the same binary. This means that at release time there can be only one qualified version - "the" version for everyone. No point in allowing anyone use any other qualified version than what's the official Orbit version.

My EUR 0.02

Back to the top