fewer aircraft [SOLVED]

[karla]

Could we discuss and agree a reasonable number of aircraft that ought to be made available In the next formal release of FGFS? Aircraft take up almost half of the download and it makes sense to concentrate on a smaller number of good quality aircraft that are part of the main package with others available in a database if desired. Many new incomplete aircraft are rolling off the development lines even now when there are dozens of old ones needing fixing, upgrading or improving.

FGFS with, say, 30 really good aircraft would be much more attractive to new users than the current situation of over 300 aircraft that appear to be a feast at first sight but often disappoint when you realize that many are old and undeveloped with incomplete cockpits, basic undercarriage, gaudy paintwork, poor aerodynamic features and so on. I appreciate that many of the aircraft were created years ago and suited old PCs but I suggest we should move on and transfer many FG aircraft to an archive where they may still be easily accessed, played with or developed as people wish.

Obviously we will never all agree on a list of aircraft because we all have different tastes - so someone should take the lead and select a few good examples of major classes of aircraft to include in the next release. Also, it should be made easy for users to locate and download all the other aircraft from a suitable database. One could argue that we already do this with scenery where we default to one particular region of the world but can easily download and fly in other regions.

I recommend this as the easiest way of improving the quality of our simulator package.

[Hooray]

I am not sure if directly saying which aircraft should be included and which ones shouldn't, is such a good idea. If you check out the mailing lists, you'll see that there is usually a poll where people are voting for certain aircraft that are to be included in the next release.

Aircraft change over time. So, maybe it would be more "future proof" to simply come up with a list of criteria for aircraft that seem suitable for inclusion?
We could then maintain such a list of different criteria using the wiki. And whenever a new release is being planned, people can refer to the wiki to come up with a list of suitable aircraft.


What do you think?

[HHS]

Hmmm....

It was discussed very often.

The principale behind this project is to be completly OpenSource. That means even unfinished aircrafts have to distributed.
We already vote everytime about 12-20 aircrafts which are worth for the base package including all different types like helis, jets, props, GA, military....

I would suggest to split up the downloads for a release between alpha, in developement and beta(finished or usuable). So people can decide which they want to try, and take a look and see much better how far developement is.

But I'm against a totally exclude of unfinished aircrafts!

[SkyWlf77]

Not that anyone is going to pay attention to this posting (as it has been ignored before), BUT:

Before we can really consider doing something like this, one thing has to change. There has to be standardization of certain subjects. What this means is that a defined set of rules needs to be established to determine when an aircraft meets "Development", "Alpha", "Beta" and "Released" stages. Right now, "Alpha", "Beta" and "Released" have basically no meaning whatsoever as it's just a title that each aircraft author gives to their aircraft and it could mean literally anything at all. Without some defined set of guidelines as to what stage an aircraft is in, confusion reigns and disappointment is inevitable. I have downloaded several aircraft in "Beta" and "Released" stages and have been sorely disappointed as they really don't belong above "Alpha" stage at best and possibly even still in "Development" stage.

There are other areas that need standardization implemented, but I'll refrain from those at this time as this topic is about aircraft inclusion.

-Jason

[Hooray]

Before we can really consider doing something like this, one thing has to change. There has to be standardization of certain subjects. What this means is that a defined set of rules needs to be established to determine when an aircraft meets "Development", "Alpha", "Beta" and "Released" stages. Right now, "Alpha", "Beta" and "Released" have basically no meaning whatsoever as it's just a title that each aircraft author gives to their aircraft and it could mean literally anything at all.

That's actually a pretty good point, this has been discussed a number of times recently (for example, see Aircrafts frustration or Aircraft Metadata)- and even stuart mentioned that we need to come up with a list of guidelines for determining an aircraft's status tag:

Subject: Feature Poll
One way you could easily contribute would be to submit patches to CVS setting the "status" flag on each aircraft accurately. Yes, it'll require learning a bit about CVS, and XML, but that would be a fine contribution. You'd have to decide on appropriate criteria for the status flags, but that isn't impossible.

I added those suggestions to the wiki volunteer page recently: http://wiki.flightgear.org/index.php/Vo ... tus_fields

Without some defined set of guidelines as to what stage an aircraft is in, confusion reigns and disappointment is inevitable. I have downloaded several aircraft in "Beta" and "Released" stages and have been sorely disappointed as they really don't belong above "Alpha" stage at best and possibly even still in "Development" stage.

Frankly, I am not sure if it is even being realistic trying to accurately describe any aircraft just by using one single "status tag" for the whole aircraft, how would you want to account for different states of completion (fdm, 3D model, cockpit, texturing, effects, checklists, tutorials etc)??

So I agree, the next logical step would be to determine a set of rules/guidelines for declaring an aircraft's status.

[zakalawe]

I have a medium-term plan to do the technical parts of this in the next six months - essentially adding information to the -set.xml files to better describe aircraft. We already have some data field, and adding more is easy, and keeps the data 'close' to the aircraft. Agreeing the keys and values will be a challenge, but anything would be better than the current situation. I had planned to enforce 'license' and 'development status' fields, but reading this page has made me realise there's other useful options, such as 'is-military', 'has a cockpit', 'supports IFR' and so on. Some of these keys can still be optional , of course.

Once the data exists, some nice people (hopefully not me!) need to write tools to drive the GUI in fgrun, and on the flightgear.org aircrafts page, automatically - which is basically an XML processing script. Then FGRun can download aircraft on demand, and the world will be a much happier place.

By the way, karla/Don, the ultimate plan is (probably) to ship exactly *one* aircraft with the download - the C172P - and have every other aircraft be downloaded on demand, or in a 'pack' (where a pack might be 'gliders', 'ww2 fighters', 'production quality aircraft', or any grouping anyone cares to create). This will make the Linux distributions much happier to include FlightGear

[simbabeat]

Why not have one or two planes in the base package and make it very clear that more can be downloaded from the .org. I think the main problem karla is bringing up here you guys is the download size, not quality. Everybody is going to go download a bunch of planes anyway.

Cheers

[zakalawe]

Why not have one or two planes in the base package and make it very clear that more can be downloaded from the .org.

Well the idea of extracting the aircraft 'catalog' automatically, and making it available to FGrun, is that you won't have to download aircraft manually - FGRun will show all the planes in the catalog, and if you select one that isn't installed, FGRun can download it (which will be fast, since individual planes are mostly small, with a few exceptions) before it launches FG. This is also how updates will work, at which point we can stop shoving all the aircraft into one giant unwieldy Gitorious repo.

The above scenario doesn't work for some setups / situations of course - if we were building a DVD image then absolutely you'd include more aircraft, and similarly if you want a package for people with reduced or restricted internet connectivity, you need to do something different - but it's also pretty easy to implement - there's no tricky coding in parsing some XML and HTTP-GETing files.

[karla]

Hey zak' just loading one aircraft with FGFS is probably the best solution; it may also encourage some of us to help develop the chosen flagship aircraft even further because we all want FG to be more widely recognized and "as real as you want it". I like the idea of FGRun (or FGo!) downloading any or all aircraft of whatever status on demand but a simple database could also do the job.

I agree with Hooray that we could add more detail to the http://wiki.flightgear.org/index.php/Table_of_models such as ratings for: liveries, cockpit, documentation, etc etc that would indicate the quality of the aircraft. Not all will agree with all the categories, nor criteria, nor ratings but a simple 0 to 5 scale on, say, pilot documentation could show whether the documents: exist at all, whether there's just a check list or whether there's a 150 page manual. The status - and quality - of all aircraft could then be easily seen.

This topic may may have been discussed before - but we can do something simple and effective right now that will tempt new users to our game.

[Thorsten]

It would be at least useful to tag the aircraft which plainly do not work with the current binary and remove them to a separate database.

[VicMar]

Hi Guys,
I can't help but put a spanner in the works on this topic.

As the C172p is the default aircraft, it is a shame that the brilliant quality of the model seems to have not been tested properly. I say that because using the C172p in FG version 2.0.0 on a Mac results in very low fps in 'Chase' and 'Helicopter' views.

I would suggest that any aircraft rating also be related to testing across ALL operating systems. Please remember that not everyone has a G5 with a gazillion mb of RAM.

Cheers,
Vic

Discussion moved to this thread

[Gijs]

Karla, I don't really get your initial problem. Past FlightGear distributions (atleat the Windows ones) only included about 10 aircraft. The rest could be downloaded from the website. Your "FGFS with, say, 30 really good aircraft would be much more attractive to new users than the current situation of over 300 aircraft" sentence makes me wonder if you aren't mixing up the websites download page, Git and the base packages...

Having said that, I agree that:
- individual aircraft should provide better status/completeness reports (and thus the webite's downloadpage).
- the downloadpage should have a filter so users can choose what status/levels to display.

But as always, this takes time and effort, not many are willing/able to contribute...

Cheers,
Gijs

[karla]

Apologies; I didn't realize that only ten aircraft were included in the Windows (and I guess Linux and Mac) distribution when I downloaded the CVS/Git with the 300+ FG aircraft in all states of completeness. So my concern that new users would be put off by too many undeveloped aircraft was mistaken. Good.

I am thinking about getting involved in the aircraft Wiki database sometime to help people more easily assess the qualities of aircraft on offer - but not just yet because: I want to help a little more with Tu-154B, possibly improve the Cub, fixing EGKK ground network, adding lighting to AI aircraft, aim to also add features to the A320 and also help a little in the KSFO updating exercise.

Therefore I'll head the thread as solved.