Oolite Bulletins

Thargoid wrote:

But I also have to wonder as we're yet again having quite a fundamental shift in coding practice for OXPs to follow

I don’t see how “all the old stuff will continue to work” constitutes a fundamental shift.

Development is winding down. There are less new features and major changes, and more stuff being postponed (like new planets). But releasing the first “stable” JS interface with known problems is a bad idea, and the various different ways of dealing with equipment constituted a serious problem.

Ahruman wrote:

Thargoid wrote:

But I also have to wonder as we're yet again having quite a fundamental shift in coding practice for OXPs to follow

I don’t see how “all the old stuff will continue to work” constitutes a fundamental shift.

Development is winding down. There are less new features and major changes, and more stuff being postponed (like new planets). But releasing the first “stable” JS interface with known problems is a bad idea, and the various different ways of dealing with equipment constituted a serious problem.

I was wondering, at what point will you consider the work on the trunk to be "done" so that 1.74 can be released, is there a certain goal/landmark you're waiting for? Just curious.

An obvious landmark would be that the known bugs are fixed—at least those that have been introduced after 1.73.4 was released.

Ahruman wrote:

Thargoid wrote:

But I also have to wonder as we're yet again having quite a fundamental shift in coding practice for OXPs to follow

I don’t see how “all the old stuff will continue to work” constitutes a fundamental shift.

Development is winding down. There are less new features and major changes, and more stuff being postponed (like new planets). But releasing the first “stable” JS interface with known problems is a bad idea, and the various different ways of dealing with equipment constituted a serious problem.

Depends if we're going to be playing the usual depreciation follow-the-leader at some point, either now or in the future.

Along similar lines, is anything planned to capitalise on the previous one (player --> player.ship)? I recall a lot of talk and discussion about that separation allowing distinction between the player and their ship (to allow multi-ship operation or similar) but as of now it's just been a reclassification of a few JS functions.

Fundamentally I guess it all boils down to wondering if there is any roadmap for planned changes in the upcoming few versions, and where they are heading towards the next stable release. If we had more visibility of that then things would be less like trying to hit a moving target with the OXP coding and maintenance.

Thargoid wrote:

Along similar lines, is anything planned to capitalise on the previous one (player --> player.ship)? I recall a lot of talk and discussion about that separation allowing distinction between the player and their ship (to allow multi-ship operation or similar) but as of now it's just been a reclassification of a few JS functions.

This might really surprise you, but the equipment & missiles changes in trunk are exactly that: stuff that was only possible to do with player & that was moved to player.ship is now possible to do with any ship.

As far as roadmaps go, I can give you the general outline, at least from my perspective: we're trying to remove as many bugs as reasonably possible, improve the interface a bit and - perhaps the single most important thing as far as the next stable release is concerned - improve the scripting side of things.

The last bit covers a multitude of sins - I don't really have a complete list of things being changed for 1.74 (we've had more than 500 commits since 1.73.4) but here are some highlights of what's happening: we're completing the events available for scripts, we're finishing off the legacy ->js migration (e.g. at last we have a system.addShip() function to run in parallel with the _legacy ones - it's more flexible, and easier to use, and should be bugfixed any moment now), introducing a vastly improved mission screen handling, plugging scripting omissions - more things are scriptable now, and more info is available to scripts (eg some of the new ship & player.ship properties: missiles, equipment, manifest, passengers, contracts). As Ahruman said, new developments (changes to stuff) as such are winding down. Pylon handling via script is still not available, and hopefully we'll manage to get that one in, but at the moment the priority is solving the recent trunk bugs.

We might also be updating the js engine to the latest stable mozilla one, but apart from that we shouldn't have anything major happening between now and 1.74.

Hope this answers some of the questions.

Frame wrote:

But I did note that missiles are now an array(which is cool for adding/removing external things on a model)

Yes, you can create special missile-equipment, usable for you ship only. Than create subentities on your ship in the shape of a missile.

Than add a script to the missiles that they remove the mothers subentity on launch and take the position of that subentity itself. (Some scripting like the hammerhead release).

(I just doubt if this is Oolite-like, it's more a fighter game with missiles under its wings.)

And when frangible, the subentity needs a script that if shot off, it removes the corresponding missile from its mothers missile array

The current game plan is this:

• The goal for 1.74 is a stable-as-in-not-crashy release with no major features beyond what we have. There are some nasty bugs to fix, and I want to bring Spidermonkey up to date because using old versions of scripting engines is a Bad Idea with a capital Bad.
• 1.74.x will be a series of beta and release candidate builds.
• 1.75 will be the non-mythical next stable release.

The plan is not set in stone. Given my past history with time estimates, I’m not making any, even for myself.

Cool with me.

Thargoid wrote:

When I said roadmap I was thinking somewhat more generally in terms of trends where things are going rather than specific features, but the above information gives at least a flavour of that. And 1.75 sounds nicely non-mythical to me too.

I don’t think there are any planned features that aren’t already in, although not all work properly yet. Some features happen more or less spontaneously, and sometimes it’s hard to differentiate between features and bug fixes (“random” shader uniforms being tied to the entity personality and a JavaScript error reporting improvement being extended to parse errors, for instance).