[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Newsgroup Home]
[news.eclipse.technology.eep] Re: more ideas
Wayne Beaton wrote:
I think that it's probably more of a problem of granularity. Most
projects do provide examples, they're just too big and tend to solve
"real" problems, thereby introducing too much complexity to be really
helpful as a learning tool. On the other side of the coin, some projects
provide snippets that are very often too small, or constrained.

Right. I tend to think now what's better articles or examples. :)
For instance, I don't think EMF has any examples that come without articles. Maybe the role of example
is just to look good? And if someone is to learn from example, then write article for that example.

Agreed. Having a single home for examples should be good for the
community as a whole.

great! let's make it a EEP wiki page. During the weekend I'll go over projects and see if they have any examples.
A wiki "Example" category could span over different project's example pages.

One of the bigger challenges we're going to have is managing it all.
Like every other project, EEP won't be able to accept every line of code
that's contributed. Contributions will have to come with some amount of
development resources and some dedication to maintaining the examples.
The proposal document discusses an archival process for out of date
examples, but we'll have to come up with a strategy for identifying
what's out of date and doing something about it.



hmm but looking e.g. at Articles - do you feel they're up-to-date? Anyone is updating them?
Sometimes I see there scary 3.0 screenshots :) however content is still absolutely fine.
So it'd be wonderful if everyone could maintain their examples, but realistically only some will do.
Archiving out of date examples sounds reasonable. But like articles, classics is sometimes better than fresh stuff :)

But other question, how new projects will get into EEP? During Equinox call, first thing I heard was sigh that all contributions will require IP reviews. And who'd like to spend precious IP time for such a small thing as examples? But maybe EEP could have a "buffer" for contributions, with minimal requirements (code should work, show something, have minimum javadocs, be simple & minimal). Then monitoring examples popularity EEP could decide to polish more popular examples, send them for IP reviews and put in EEP CVS. Sounds a bit like incubator proposal, but I imagine this buffer/incubator/whatever rather as a catalog of available examples. Contributor would fill simple form to submit an example and that's it, he's in.
Like on live.eclipse.org or EPIC.

btw, Equinox will discuss a place to contribute examples and demos on next confcall: http://wiki.eclipse.org/Equinox_Minutes_-_20080226 - might be worth to bring EEP topic too.

Oops. Missed the call. Did it get discussed?

Yes, equinox needs examples, but where should they live and how should they look - it's still open question.
I feel Equinox people are too busy now with other cool things now and it's somewhere at the end of their lists.
There was created bug https://bugs.eclipse.org/bugs/show_bug.cgi?id=220450 to continue the topic.