TS2 SDK - A thought for ACES

[75027]

Dear ACES developers,
Since TS2 is using the same architecture as FSX, I'm assuming that it will have a similar SDK that will be provided to users in the product at RTM. As far as I know, the beta testing of an SDK has always occured either at the same time or later than the beta testing of the "main" part of the product.

I'd like the ACES team to consider providing the TS2 SDK for testing (to whatever user-community SDK beta groups they care to use) at a much earlier date than when they plan to start beta testing the main product within the user community.

Here's why: As I mentioned in another thread, I believe the quality of the creative tools delivered in the product (scenery tools, mission tools, etc) will be instrumental in the success or failure of the product.
If I'm correct in thinking that the majority of the scenery, AI, and mission code that will be in TS2 is already in FSX, then a good portion, if not a majority, of the TS2 SDK already exists in the FSX SDK - and so is not beta at this point at all. Only the TS-specific functionality would be new.

Giving the TS user community, though the user-community SDK beta test team, (which would be a subset of the larger TS community beta team) an opportunity to provide critical feedback at a very early stage - and earlier than MS usually does - will ensure that bugs have a greater chance of being found and rectified by RTM. This is a win-win situation for MS & the TS community.

Since we are dealing only with the SDK here, MS would not have to provide a complete version of the product to these testers. The goal is to find flaws in the SDK tools & provide feedback on those tools.

For example, to reduce the support burden to MS during this time when the developers are very busy coding, there would be no need for the SDK to be provided as an msi file. (a simple zip file would suffice). There would be no need to provide locomotives, rolling stock, or scenery items, etc., as the purpose of the testing is for the community to test the tools to see how easy it is to create these objects. The only requirement would be to include a very small portion of the world: Perhaps an island with a flat section & a mountain, as the FSX demo did, or one of the global routes in Australia or New Zealand onto which we can test the track-laying/tunneling ability of the tools, the ability to test our rolling stock that we create on a short track section, and to test our ability to create scenery, AI, & missions.
At the risk of repeating myself, the goal of this earlier-than usual testing is not for the community to build fully-functioning routes, but to ensure that such routes can be easily built by the tools provided at RTM.

It would also give ACES the opportunity of uncovering bugs related to specific hardware configurations at a much earlier stage.

Bottom line: earlier involvment of the user community in testing the SDK will result in a more robust toolset at RTM that could significantly increase the adoption of TS2 by the TS community immediately after RTM. Plus, more add-on content will be created more quickly post RTM as there will be a cadre of developers who already have many months (maybe over a years) experience with them - which also will add to the appeal of the product to the public in the months after RTM.

That's my 60-second elevator speech. Thanks for listening & I look forward to your reply.

- 75027

[MaxFreak]

Nice idea

And has already been done by Crytek .. Sandbox2 and a copy of the Alpha SDK where released to selected teams ( along with dinner at Crytek HQ ) to weed out any bugs and lend support to mod teams . Also disclosed in a Q & A with Cevat Yerli , Cry-Alex and Ben from Crytek where most of the poly and texture budgets ... a real help for those with an interest in modding for Crysis .

Would be great if some details such as this were available for TS2 some time next year , including the 3D package/s and v#'s that will be supported .

Cheers

~A~

[batroost]

Good idea - I'd volunteer to have a go at modelling a Narrow Gauge route - perhaps a bit of the Ffestiniog - asuming it is NG-capable.

Batroost

[AndiS]

Early community involvement to find out whether all the required functionality is included is always a clever idea. Of course, it is more costly than relying on your in-house experts, but most expert know but one field (i.e., one combination of era and location).

Now that MS have officially announce "we model the world", they have two options: 1) someone at MS learns about all the railways of the world; 2) they ask someone from everywhere in the world whether their system will suit his demand. Clearly, the second approach is a bit more realistic.

Since all the basic capabilities of the sim will show in the SDK, it is a good idea to discuss that piece of knowledge.

If I were MS, I would demand that every tester does a set of certain tasks with the kit and writes an application note about it. Then they get a rich collection (advanced) documentation in the process as a bonus. Sample tasks would be: Install a demonstration setup of all the signaling systems you are aware of on the existing track. Import an engine of each traction type plus a carriage and a freight wagon into the sim, create the physics definition for it; create demonstrator activities for it showing working practice in your domain (era, region, organisation). Of course, a single person cannot do all, either teams cooperate or the tasks are split. In the end, the QA department at Aces should have a proof of concept for each aspect of railway operation for each part in their world-spanning track network.

[75027]

Early community involvement to find out whether all the required functionality is included is always a clever idea. Of course, it is more costly than relying on your in-house experts, but most expert know but one field (i.e., one combination of era and location).

Now that MS have officially announce "we model the world", they have two options: 1) someone at MS learns about all the railways of the world; 2) they ask someone from everywhere in the world whether their system will suit his demand. Clearly, the second approach is a bit more realistic.

Since all the basic capabilities of the sim will show in the SDK, it is a good idea to discuss that piece of knowledge.

If I were MS, I would demand that every tester does a set of certain tasks with the kit and writes an application note about it. Then they get a rich collection (advanced) documentation in the process as a bonus. Sample tasks would be: Install a demonstration setup of all the signaling systems you are aware of on the existing track. Import an engine of each traction type plus a carriage and a freight wagon into the sim, create the physics definition for it; create demonstrator activities for it showing working practice in your domain (era, region, organisation). Of course, a single person cannot do all, either teams cooperate or the tasks are split. In the end, the QA department at Aces should have a proof of concept for each aspect of railway operation for each part in their world-spanning track network.

Thanks for the feedback, batroost, Maxfreak & AndiS. AndiS, you bring up some excellent points. In the interest of keeping my post short (well, short-ish), I skipped over a lot of the rational, cost-benefit analysis, and methodology of the testing that the user community could do. There are a variety of approaches that MS could take & I welcome further comments & ideas from us users as well as, of course, the ACES contingent monitoring these threads!

- 75027