Content Mgmt Systems?

Pipelines
#1

I’m trying to streamline our asset creation workflow with a database, and am looking towards a Content Management System (CMS) to admin the thing.

What are you guys using?

Things I’m looking for:
[ul]
[li]Database listings for each asset: exporter tags used, shader settings, textures used, data archetypes, etc. We have this data, but it’s not all in one place.
[/li][li]To that end, import and collate all the assets’ XML files into database format. Maybe use XSL or somesuch to convert Programmer tags to Designer-readable English, and back out for publishing.
[/li][li]Synch with Perforce to get latest assets (middle of the night?) then update database. Perhaps update db at export time (if not too slow)?
[/li][li]Create thumbnails of Max files, TGA/PSD/DDS/etc., SWF UI elements, etc.
[/li][li]Search: Game designers can see what assets exist already for re-use, or as starting points for new assets.
[/li][li]An editable Form for each asset: designers can ask for changes, ask for new assets based on existing assets, etc. They can see all the decisions needed for each asset, and make them.
[/li][li]Approval chain when submitting changes to existing assets, or when asking for new assets. Dept heads approve, or reject with some text, automated handoff to next decider, etc.
[/li][/ul]

I know it’s a lot to ask, I’ll need to cobble together a few different tools.

Currently looking at MS InfoPath/SharePoint, but totally willing to check out other systems. Found this interesting site: CMS Critic

Basically at its heart our CMS will be a method to educate Designers (and Artists) about all the decisions that need to be made (and implemented) for each asset, and help avoid wasted work or reworking existing assets. The Wiki alone, although helpful, isn’t cutting it.

Any experience to share about similar systems would be appreciated!

#2

The frontend aspect of this is completely up to you and your needs- really discuss with designers and managers what they will need to see and how. I don’t have any experience with this so I cannot say, to that regard. You may be able to go with a proprietary app, I don’t imagine it’d be too hard to write. But I don’t know the options, it’d be worthwhile to look around.

You should probably collect all data on a p4 checkin. Only issue I can see is if Max is closed, to get some data it would need to open it up- so you may want to write out the needed data during saving or exporting. I’d suggest avoiding a ‘batch’ approach, as trying to track all of a project’s assets via some set of rules is often, unfortunately, impossible- or you will be doing lots of extra work. Better to make it an automatic part of the pipeline. Batch approaches can work very well for some things (such as rendering a library of videos of all animations), but that is a much smaller task on a subset of files- the more you want to track, the less likely this batch approach will work without constant headache and revision.

#3

Pretty much all your specifications are met by a great piece of software named “Tactic”. It was just implemented by Digital Domain in their pipeline and it seems to work great for their needs. It’s completely written in Python and can be customized extremely easy. We are working in porting it to Max now and if you are interested in a demo let me know.

Here is the website for it: http://www.southpawtech.com/

#4

Definitely interested, sounds promising. Just sent in a query via the website contact form.

I’m starting simple, just to get something working, a web-based asset-workorder form with all the designer’s choices. I basically need to simplify the technical gobbledygook so designers can discern what it is they need to make decisions about. Then build from there.

Integration with Max would be helpful, but would also need access outside Max since designers/engineers/etc. don’t all have Max seats.

#5

Hi Eric,

This seems a rather structured approach and one I’m definitely keen to be explored at my company. However, to put a slight downer, designers tend to be less helpful in asset management. Quite often at Blitz, they will request through the bug reporting database to their partnered artist what they require and as a last resort will pouch assets from another level, even though the artistic style is different!

I’m unsure how the process of converting programmer-centirc tags can be of use when a designer potentially has free reign over what the artists or programmer should implement into a level they have sole control of.

Perhaps what you are seeking is more for the Game Director/Producer to start from, rather than the actual prodction team. Apologises if I’m off-topic on this one.

#6

No these are good points, understood and agreed. This was a problem that needed to be solved one way or another, since we’re working on a platform with tight memory constraints.

We hit a wall recently where the designers had less characters than they wanted in a level because the level was inefficient, taking up too much mem. Also we were seeing impossible projections for the time that production would need in order to finish all the content that was asked for.

We realized we all need more information about what the tradeoffs are, what assets we can reuse, and how we can streamline, all while keeping focus ultimately on what’s fun for the player.

This information sits in the heads of just a couple people on the team, myself included, but it needs to be clarified and revealed to the Creative Director, the Lead Designer, and the rest of the design team as well.

#7

I discovered a GoogleGroup for Maya, and there’s a conversation regarding content management here:

http://www.mail-archive.com/[email protected]/msg00375.html

Specifally, SVN for Maya. Just throught I’d throw this into this thread to see if it’ll be of any help. We really need to tempt the contributors from this GoogleGroup onto TAO :wink:

#8

I’ve had most experience with Perforce. It does the job of dealing with the data extremely well, unfortunately it lacks the visual representation of that data, relationships and dependencies between different assets and their relationship to the project itself.

I had a very brief experience with Tactic about a year ago, it showed a lot of promise and the team was very responsive and went to great length to help us integrate it into XSI pipeline. So I would definitely take a look at it as an option.

#9

Well, thought I’d give you an update. We had to put off integrating a new system like TACTIC, simply because it would have required taking coders away from other essential systems, bending it to work with our system. The simpler solution is usually the best way anyhow.

The solution we came up with is for a Tech Artist or Engineer with a deep knowledge of the game’s systems to take a Designer’s level design proposal and check it against existing systems, flagging in red every bit of needed behavior that doesn’t have code support yet, or any needed new content that could possibly re-use existing content (particles, sound, etc.).

We’ve also been taking some time to update our wiki with Design-centric resources that illustrate the existing choices they have for various effects or asset types.

Once the Design team saw the vetting of that first level (with all that red!), and saw all the choices they have, it made it a bit clearer what they need to do. It’s also helping the Creative Director & Design Lead make decisions about which new code or content they want to put the studio’s time into, and what could be dropped.

We’ll probably still go with an asset system, and TACTIC looks great on paper, so we’ll take another look once we’re in pre-pro for the next one.

Thanks again for everyones’ input and ideas.