How do you plan your projects?

[SeanG]

This is a question, or discussion starter really, for both the established and beginning developers:

I am most of the way through a modelling project, and every time I think I am done I find something else to add. So my questions are;


- How do you go about planning a project. And do you define your end-result, or set a 'target' for how far you are going to go with a particular model/project before you start?

I am looking for some advice to improve my workflow

Sean

[Firebird]

I think that in my experience in working with people like JY the first thing to do is see how big a scope your project could possibly be and either limit it or break it down into smaller projects.
What I mean here is how many versions of a model are there? Then if you multiply that by the number of weapon/sensor fits that you want to do for each version you come up with how many models you are looking at.

Sometimes it can be quite surprising. A good example of this was JYs Tornado package. It seemed simple enough. The ADV and the IDS models. Then when he investigated more this broke down further. The ADV had two versions, the F2 and the F3. The IDS broke down to 4 versions, the standard IDS, The GR1 for the UK, The ECR for Germany and Italy and finally the GR4 version for the RAF.
As you can see the 2 versions turned into 6.

Then came the weapon fits. This is where mil aircraft really cause modellers problems. It is bad enough that a version can have multiple weapon fits but then you get into different nations having different weapons.
John was still getting requests for weapon fits after he finalized the project. He cut down the project by eliminating very specialised weapon fits and ignoring the relatively short lived F2 and he still ended up with 52 models.

Two versions to 52 models is some massive growth.in scope.

Even his E-2/C-2 project turned into 14 models and the bonus here was that there are no weapon fits needed.

So my advice would be investigate the absolute growth potential of a project and see how many models could be involved in the worse case scenario. Work out whether you can handle that or whether you can eliminate some models or finally whether you can break a project into multiple parts.

If you were to go back and look now at the complete scope of your Wasp project and incorporate the Scout as well how many different models would you possibly end up with? Now work out what you would do if you were starting that from scratch. It would give a guide as to how to scope a future project.

[John Young]

That’s a really good question Sean. I’m not sure if this is what you are looking for, it's quite broad, but it might be helpful. I drafted this off-line before I saw Steve's post, but what he says is concrete advice and amplifies what I've said about model variations.

I’ve never abandoned a project because it has become so complex or because I have lost interest, but I know that happens to developers time and time again, particularly with custom scenery. This is the preliminary process I go through to help me scope a project:

- Pick a potential subject that will interest me.

- Pick a subject that will be worthwhile doing in terms of likely usage. What are people requesting, do designers of scenery here need a particular AI aircraft, what might interest our growing number of painters? I rarely fly in the sims myself so don’t really make use of the modelling for myself.

- Decide which sim version to design for. Downloads are dropping much lower compared to a few years ago, especially for FS9. My Hawkeye/Greyhound packages have produced 685 FSX/P3Dv4 downloads, but only 278 for FS9. In reality the first figure will likely grow faster than the FS9 one as users wait for the package to be incorporated into Matrix. That’s ignoring the future effect of FS2020 of course. For the moment I’m continuing with the 3 sim versions, but FS9 is not really viable any more.

- Research the subject. Wikki is usually my first call, Scramble on the Net is the next to get operators and numbers, including serials if the project is an AI aircraft.

- Scope the number of models that are likely to be needed. Knowledge of load-outs is important in order to plan the texture sheet.

- Check whether scenery could usefully be provided if it’s an AI aircraft project and assess what is available. Decide whether the project will be the aircraft only or a complete AI package with aircraft, scenery and flight plans. I’m tending to move away from the big packages because it requires so much stamina and “factory work” to get there for the 3 sim versions. That’s not to say that others could be invited to join in and contribute.

- Draw the boundaries around the project and only deviate from them for value added reasons that do not increase the effort significantly. For me getting bogged down for months on end can be a drag. I need the variety and change in subject to keep me going. 2 months and 7-10,000 polygons is a good budget for me.

- Having picked a subject, search for the best possible 3 view plans. Look at several offerings and compare the similarities and differences. The Blueprints web site is a very good start.

- Check to ensure that there are sufficient photographs to refer to and collect them as you go. It’s irritating when you come to a part to make and remember a photo but can’t now find it gain. Aircraft “walkaround” web sites are really useful.

- Get stuck in and get through the pain barrier of the first 3 or 4 hours of modelling to see if you can master the basic shapes. If you have got that far, having done all the preliminary work, then continue to the end.

If you want a further breakdown of how I plan within the modelling process, I can go into that.

John

[SeanG]

Thanks guys I left my question quite broad, as I wasn't really sure what I was asking!
I guess I am trying to figure out a workflow from idea through to completion, which is what I do at work for my design/modelling. The big 'take-away' (I sooo hate corporate speak) is coming up with a scope, then trying to stick to it, which I guess has been my downfall to date.

In the past I have tended to work on more than one thing at a time, and consequently ended up with a bunch of nearly-done things which were never finished. I have managed to change that, hence the Wasp went from idea to completion without any distractions!

Thanks again Chaps!

Sean

[John Young]

I never, ever, work on more than one thing at a time Sean, apart from small jobs that people ask me for from time to time. Best to stay focused on the total process of a single project. The best bit is always the modelling at the beginning, then the mapping and painting. The bit that needs the stamina is what follows - making LODs and the "factory mode" of duplicating up, flight planning and testing scenery if provided. That's compounded if you are designing for more than one sim version of course. I like documenting at the end, because it closes the project.

I switched from making custom scenery a few years back because it could often take 1-2 years to complete a project. Very few people can do that without losing interest and giving up. I realised I was in danger of doing that and switched to the smaller time scales of AI projects to give me a new lease of design life.

John

[andypoppens]

That’s a really good question Sean. I’m not sure if this is what you are looking for, it's quite broad, but it might be helpful. I drafted this off-line before I saw Steve's post, but what he says is concrete advice and amplifies what I've said about model variations.

I’ve never abandoned a project because it has become so complex or because I have lost interest, but I know that happens to developers time and time again, particularly with custom scenery. This is the preliminary process I go through to help me scope a project:

- Pick a potential subject that will interest me.

- Pick a subject that will be worthwhile doing in terms of likely usage. What are people requesting, do designers of scenery here need a particular AI aircraft, what might interest our growing number of painters? I rarely fly in the sims myself so don’t really make use of the modelling for myself.

- Decide which sim version to design for. Downloads are dropping much lower compared to a few years ago, especially for FS9. My Hawkeye/Greyhound packages have produced 685 FSX/P3Dv4 downloads, but only 278 for FS9. In reality the first figure will likely grow faster than the FS9 one as users wait for the package to be incorporated into Matrix. That’s ignoring the future effect of FS2020 of course. For the moment I’m continuing with the 3 sim versions, but FS9 is not really viable any more.

- Research the subject. Wikki is usually my first call, Scramble on the Net is the next to get operators and numbers, including serials if the project is an AI aircraft.

- Scope the number of models that are likely to be needed. Knowledge of load-outs is important in order to plan the texture sheet.

- Check whether scenery could usefully be provided if it’s an AI aircraft project and assess what is available. Decide whether the project will be the aircraft only or a complete AI package with aircraft, scenery and flight plans. I’m tending to move away from the big packages because it requires so much stamina and “factory work” to get there for the 3 sim versions. That’s not to say that others could be invited to join in and contribute.

- Draw the boundaries around the project and only deviate from them for value added reasons that do not increase the effort significantly. For me getting bogged down for months on end can be a drag. I need the variety and change in subject to keep me going. 2 months and 7-10,000 polygons is a good budget for me.

- Having picked a subject, search for the best possible 3 view plans. Look at several offerings and compare the similarities and differences. The Blueprints web site is a very good start.

- Check to ensure that there are sufficient photographs to refer to and collect them as you go. It’s irritating when you come to a part to make and remember a photo but can’t now find it gain. Aircraft “walkaround” web sites are really useful.

- Get stuck in and get through the pain barrier of the first 3 or 4 hours of modelling to see if you can master the basic shapes. If you have got that far, having done all the preliminary work, then continue to the end.

If you want a further breakdown of how I plan within the modelling process, I can go into that.

John

Does this you mean you won’t develop for FS9 anymore cause that’s what I use

Let me share with you why I still use FS9 it was the first one I ever purchased every time I have sat in the cockpit of Alcock and Brown's Vimy the Spirit of St Louis Amelia Earhart's Vega and the Winnie Mae I have wondered how trapped in this cramped cockpit did they survive

[John Young]

I'm afraid Alcock and Brown is not a substantive reason for me continuing with FS9 for much longer Andy. I'll wait until FS2020 is released though before making that decision. You never know, maybe FS2020 will become the dominant sim and I can reduce 3 designs each time down to 1, assuming we can provide value-added design for it at all of course.

John

[TimC340]

You can use the FS9 CoF aeroplanes in FSX and P3D....

[andypoppens]

I'm afraid Alcock and Brown is not a substantive reason for me continuing with FS9 for much longer Andy. I'll wait until FS2020 is released though before making that decision. You never know, maybe FS2020 will become the dominant sim and I can reduce 3 designs each time down to 1, assuming we can provide value-added design for it at all of course.

John

Here’s the thing John there’s a lot of ai models that haven’t been made for FS9 yet

[John Young]

No that's not the "thing" Andy. It's no good carrying on making models for a sim that's more and more less used, when the time could be better invested in sims that are far more popular. 288 downloads from around the world, over 4 weeks, of a comprehensive, gap-filling FS9 package like the Hawkeyes and Greyhounds just doesn't justify the huge amount of time spent creating it. 280, or so, of those users will never acknowledge any effort spent on it at all in any case.

I'll continue to provide for FS9 for a little longer, but it's very unlikely I will be designing for FS9 once FS2020 drives the numbers down even further. We have to move on at some point from a sim that's 17 years old.

John

[Firebird]

Andy,
Just to keep you in the loop here. There has been various discussions over the years about what sims should be supported, what sims should be dropped - and if so when.

Once the somewhat massive project by a few guys here to convert everything to P3D here was completed there would in all likelihood have been a change. After all when people like me that didn't want to change as they didn't want to lose AI and/or scenery were presented with a complete solution that could be installed quickly and efficiently there was likely to have been a tipping point on what the base sim and what sims would be supported.

What has thrown that up into the air a little bit is MS2020. At the moment people are waiting to see what extra it gives us and what backwards capability it has. I would imagine that there aren't many people that aren't waiting for its release and reports on it's capabilities before taking a decision that will mean a big change for quite a few members here.

This is what I am waiting for as a user and it is what John is waiting for as a producer. You must make the choice that suits you. What I suggest is that you do the same thing that everybody else is and wait until you have all the information to make that decision.

Anyway this discussion is really for another thread and not the one needed by Sean.

[SeanG]

Right, so back to the original topic

I am still dabbling in both scenery and ai aircraft, and can certainly see the timeframes involved in a comprehensive scenery project.
My focus, at least for now, is on projects for my own interest, and working in FS9 initially (for that is all I know...) but I will be exploring FSX/P3D in the future as the potential audience for purely-FS9 projects decreases.
I will need to learn how to move to the newer GMax game packs, then I may start producing multi-sim versions of my random projects

Thanks once again for the advice, and insights into how you plan. It has helped me focus a little better and with any luck I’ll be able to apply this to get more projects across the line to ‘release’.

Cheers
Sean

[TimC340]

Forgive me Sean - though for the record, the answer to your question is don't get distracted or sidelined!

So, here we go with some industrial-level sidelining.

FS9 is dead, and it's a pain in the arse to develop for. I came in to the MSFS system sometime when it was still an MS-DOS based program. At the time, I was convinced that the Atari ST-FM (I couldn't afford a Mac!) was the way ahead, and Elite was the be-all and end-all of games on a computer. And I was flying aeroplanes for a job.

MSFS was relatively odd in that it was open-source, and so people could add their own stuff to it, and soon enough the addons became way better than what Microsoft provided. But, nevertheless, we always looked forward to the next iteration, and we moved with it. Sometime around FS98, people started to get attached to, and stuck with, a generation of FS that limited them, and their potential customers, to an outdated product. That became exaggerated with FS9, and the move to FSX became impossible for many enthusiasts.

When Lockheed took on the franchise, the limited changes in the early editions of P3D did little to convince FS9 addicts that there was any reason to move on, but P3Dv4 shifted the game somewhat and MAIW (as ACG and UKMil and others had coalesced into) realised that really they had to embrace the changes. Some developers had begun to show that v4 really had opened up new levels of potential accuracy, and the change to 64-bit programming meant that OOM failures, which dogged FSX, were no longer an issue. MAIW took the decision to make all their legacy products FSX/P3D compatible, and all of a sudden we started moving on.

P3Dv4.4 came along, and dynamic lighting and a few other improvements started to make P3D really worth upgrading to. We still had a hard core of FS9 peeps who didn't want to move on, but that was still fairly easy to accommodate. However, now we have P3Dv5 and we're looking forward to MSFS2020. P3Dv5 is already revealing that we have to change a huge amount if we want to take advantage of the new technology. It's unlikely that MSFS20 will have a lot in common with P3D going forward, but there does seem to be some commitment to respecting legacy formats that may mean our older products might work in that sim. How long do we continue to indulge FS9 operators? When do we call a stop to that sim? I know, looking at Avsim, that there's still lots out there for FS9 - and lots that needs to be updated for P3D while that's still a possibility. But we need to look forward, not backwards. FS9, good as it was, is history.

[Firebird]

Tim, Nothing you said there gives me cause to disagree. I think that a lot of what you say is in tune with what I said I replied to Andy. I am not somebody who is going to say FS9 must be supported for all time.

I think here though, we have to stay focused on what Sean asked. At this moment in time Sean has FS9 and FS9 only and wants to learn the best way to plan a project rather than make a statement about sticking with FS9.

Whilst we are in this holding pattern awaiting FS2020 there is nothing wrong with him doing a project now. The worst case scenario is that he completes it and our converters can bring it to P3D for him. The important thing is that he gets the advice and encouragement to proceed down the path. I am not really sure that whatever sim he decides to use in 6 or 12 months is relevant. For me the key is that he learns and then continues developing.

[asdro]

I want to leave also a comment about FS9 discussion.

I am also a 100% FS9 user and i dont think that this thread should be used for FS9 discussions.

Ofcourse i also follow the FS2020 topics and i saw a few trailers, but i am not hyped.

Many of those who are hyped are newbies and those who just care about eye candies.
A coworker of mine who isn t much into aviation and flightsimulation also was fascinated about the trailers and he is one of those who doesnt care about system depth as i do.

I am into flightsimulation since 2002 and for me it is important to see if Microsoft is developing a real game changer in flightsimulation.

But for now many questions are still not answered so that i am afraid that some old mistakes will be repeated by the Sim developers.
Another very important issue is the strategy of Microsoft.
Are they going to develop a Flightsim every two years like they did with FS2000, 02, 04, FSX ect or will they keep this new sim for a longer time?

SCS Software from Praha in Czech have developed in 2012 their Eurotrucksim 2.
Today the company is still developing for ETS2 and many dlcs are still coming.
What they do is constantly changing and improving their systems without developing a new sim every time.
Today ETS2 and the american pendant ATS are one of the most successful simulation softwares on the market.

I am since 2015 a user of MAIW and i love the AI world. Infact i spent more time for improving AI than flying.
I hope that in the future the FS9 still will be supported by the talented model developers and painters. That is what i can say and which is a reason why i still cone to MAIW.

[Firebird]

Everybody has a right to their opinion about FS9.

However, people seem to be losing track about this thread and what Sean needs.
If people want to start a what if/future thread in a general forum that is perfectly fine, but please allow this thread for it's original purpose.