Difference between revisions of "Talk:Modeling Certification"

From Second Life Wiki
Jump to navigation Jump to search
(nods on flexi/light, more megaprim disc)
 
(8 intermediate revisions by 4 users not shown)
Line 6: Line 6:
I'd be in favor of certification by submitting a portfolio, but not by submitting to an examination.
I'd be in favor of certification by submitting a portfolio, but not by submitting to an examination.
--[[User:Osprey Therian|Osprey Therian]] 22:35, 29 April 2007 (PDT)
--[[User:Osprey Therian|Osprey Therian]] 22:35, 29 April 2007 (PDT)
:I would like to see some combination of written and hands-on testing.  Although I like the idea of portfolios, I think they speak directly for themselves; plus, how would one confirm who actually built the stuff?
:--[[User:Kim Anubis|Kim Anubis]] 10:53, 4 May 2007 (PDT)
::Could that be done by the Creator and Owner fields in the objects? A combination of portfolio and test might be a good approach. A timed build with the set of skills being employed or as set of smaller tasks each targeting skills may be of interest.
:: --[[User:Donnagh McDonnagh|Donnagh McDonnagh]] 03:00, 17 May 2007 (PDT)
:::Creator and Owner fields won't tell us who really built an object -- we don't know who was driving the avatar when it did the build.
:::--[[User:Kim Anubis|Kim Anubis]] 09:30, 17 May 2007 (PDT)


'''Topics:'''
'''Topics:'''
Line 42: Line 52:
:::--[[User:Kim Anubis|Kim Anubis]] 11:16, 2 May 2007 (PDT)
:::--[[User:Kim Anubis|Kim Anubis]] 11:16, 2 May 2007 (PDT)
::::The current LL statement is that if they cause problems they will be removed without warning, and they're currently tolerated and in use on the mainland and on islands. Until/unless they officially ban them, they're part of the landscape, so I'd rather builders know about them than let someone sell them on the idea without enough information to make a solid judgement.  --[[User:Storm Thunders|Storm Thunders]] 07:32, 4 May 2007 (PDT)
::::The current LL statement is that if they cause problems they will be removed without warning, and they're currently tolerated and in use on the mainland and on islands. Until/unless they officially ban them, they're part of the landscape, so I'd rather builders know about them than let someone sell them on the idea without enough information to make a solid judgement.  --[[User:Storm Thunders|Storm Thunders]] 07:32, 4 May 2007 (PDT)
:::::I contacted the Linden who fields technical questions as a part of LL's Developers Program and asked about this.  He said that until LL officially announces megaprim capability or something similar he thinks megaprims should not be incuded in certification standards, and said that he would not use them in a build for a client.  I'm convinced that megaprims need to come out of the article.  How do you feel about it? --[[User:Kim Anubis|Kim Anubis]] 09:29, 8 May 2007 (PDT)
::::::I think they require some sort of mention - they're a part of the landscape, and without information they look like an easy way around prim limits.  We could simply put in the official Linden line.  "The ability to create new ones has been deliberately removed.  If they are believed to cause problems they will be deleted from a build without warning.  They're currently tolerated on the mainland and on islands."  At least anyone certified will be aware of the Lab's current statement on them. --[[User:Storm Thunders|Storm Thunders]] 08:19, 14 May 2007 (PDT)
:::::::It would be useful for builders to be aware of the giant prims, and how they work. While perhaps not encouraging new builders to use them, their existence and how they work, and their particular advantages would be something I hope builders would know if they are certified. I see the point that LL does not want to encourage (and may want to discourage) their use, they still exist and are (I want to suggest) an important (if historical) aspect of builds.
:::::::--[[User:Donnagh McDonnagh|Donnagh McDonnagh]] 03:08, 17 May 2007 (PDT)


*Prim torture shapes.
*Prim torture shapes.
Line 85: Line 99:


What's the added value to creating multiple certs?  I see the disadvantages - increased financial and maintenance costs of cert program, market confusion among clients/employers - but not the advantages.  A single cert, or a basic and an advanced should give people the information they want when backed up with a portfolio and references. --[[User:Storm Thunders|Storm Thunders]] 07:00, 3 May 2007 (PDT)
What's the added value to creating multiple certs?  I see the disadvantages - increased financial and maintenance costs of cert program, market confusion among clients/employers - but not the advantages.  A single cert, or a basic and an advanced should give people the information they want when backed up with a portfolio and references. --[[User:Storm Thunders|Storm Thunders]] 07:00, 3 May 2007 (PDT)
: The certification is supposed to help an employer distinguish whether a modeler is going to be competent to do their project/job.  Some jobs are far more difficult than others, and some will require specialized advanced skills.  While basic knowledge of working with tiny prims, large-scale builds, attachments, and organic shapes is something every competent modeler should have, far fewer know all the ins and outs of all of these things.  It would be useful to make a distinction.  If we don't break this up somehow, it's either going to be too easy to pass (and therefore not too meaningful) or too difficult (it'd really stink if good builders were filtered out because they weren't generalists in advanced areas).
:I don't think it would have to cost much more or be much more hassle to have basic and advanced levels, or even to include specialties.  There doesn't need to be a separate test.  It could be one test with different sections. There might be (this is wildly speculative) 10 questions or tasks related to tiny prims, for example.  The first three might count toward core certification, the next three would have to be correct or passed in order to make advanced certification, and the rest toward a specialization.
:--[[User:Kim Anubis|Kim Anubis]] 10:51, 4 May 2007 (PDT)
: At the minimum we need a "core modeling competency" certification.  Having an "advanced" certification then follows.  When we've fully defined the competency list it may be clear that there are additional specializations.
:--[[User:Glenn_Linden | Glenn Linden]] 9:44, 9 May 2007 (PDT)

Latest revision as of 09:30, 17 May 2007

Areas like vehicles, buildings, and avatar attachments should remain a marketing and portfolio issue. There's huge stylistic issues with judging those - camera control, proportions, historical accuracy and realism, etc. --Storm Thunders 08:29, 26 April 2007 (PDT)

I agree that it would be better to focus on testing certain abilities, rather than types of objects.
--Kim Anubis 09:37, 28 April 2007 (PDT)

I'd be in favor of certification by submitting a portfolio, but not by submitting to an examination. --Osprey Therian 22:35, 29 April 2007 (PDT)

I would like to see some combination of written and hands-on testing. Although I like the idea of portfolios, I think they speak directly for themselves; plus, how would one confirm who actually built the stuff?
--Kim Anubis 10:53, 4 May 2007 (PDT)
Could that be done by the Creator and Owner fields in the objects? A combination of portfolio and test might be a good approach. A timed build with the set of skills being employed or as set of smaller tasks each targeting skills may be of interest.
--Donnagh McDonnagh 03:00, 17 May 2007 (PDT)
Creator and Owner fields won't tell us who really built an object -- we don't know who was driving the avatar when it did the build.
--Kim Anubis 09:30, 17 May 2007 (PDT)


Topics:

  • Proof that a person can use the edit controls in basic ways. Create, select, move, rotate, copy, delete. Hollow, cut, resize, dimple. Convert one prim type into another. Resize using the edit window, resize by grabbing the points. Choosing colors and textures, aligning textures, transparent/opaque, fullbright. Understanding shiny, bumpmapped, and how alpha affects them. Grid modes.
  • Linking and unlinking. Manipulating linked groups of prims (individually and as a group) - moving, texturing, coloring, resizing. Deleting a prim in a linkset. Knows that a stable linkset can't be a mix of phantom and real, etc. Understanding root vs child prims, and how that affects position, rotation, etc. Number and distance issues with linking. The existance of script tools that let you work around those limitations. The vehicle prim limit.
I'd suggest changing that last one to "prim limit for physics-enabled objects."
--Kim Anubis 09:37, 28 April 2007 (PDT)
  • Knowledge of flexi limitations (only some shapes can be flexi, always phantom)
I'd like to include something about actually knowing how to use this feature, beyond its limitations.
--Kim Anubis 11:16, 2 May 2007 (PDT)
Good point on both - I'll edit the article if it hasn't already been done. --Storm Thunders 07:32, 4 May 2007 (PDT)
  • Knowledge of lighting limitations( 6 sources, no shadows ).
I'd like to include something about knowledge required to use this feature, rather than just its limitations.
--Kim Anubis 11:16, 2 May 2007 (PDT)
  • Alpha flicker and techniques to prevent it. Positioning, "blacksiding", etc.
  • How to make a prim into an attachment. Setting it to an attachment point. Using the edit window's grid mode for attachments. Making a basic one-prim HUD.
Gee, if we have to use the grid, I'm gonna flunk. ;) There's more than one way to do some of these things.
--Kim Anubis 09:37, 28 April 2007 (PDT)
Knowing that the options exist and how to use them is what I'm thinking of. Most of the list I was along the lines of "does X know enough about the topic that they can find/ask/experiment to get whatever they need?" Whether or not one chooses to use them is a personal choice. :) --Storm Thunders 06:50, 30 April 2007 (PDT)
  • Making "mini" prims with different techniques - cutting, dimpling, hollowing with a transparent texture on the outside.
If we do decide to split this certification up into specialties or levels (similar to what's up with the scripting certification), this might be a good section to break out.
--Kim Anubis 11:16, 2 May 2007 (PDT)
  • Oversized prims. How to do the ones that are technique (like flattening a 10M box to create a 15M plane) and best practices with the huge or mega prims that were introduced thru the Havoc hack (phantom is best, don't make them physical, bounding box and hollowing issues, beware of parcel boundaries).
I never bothered with megaprims because early on Linden Lab said they wouldn't be supported. Just went digging through the SL Wiki and the Knowledge Base, and did not find anything saying they're being supported now, but maybe you can turn up something?
--Kim Anubis 09:37, 28 April 2007 (PDT)
There's a fair bit about both the Havok hack prims and methods to make a standard prim appear larger than 10M in SL's forums (specifically Content Creation | Building). --Storm Thunders 06:50, 30 April 2007 (PDT)
It isn't that there's a lack of information on the use of megaprims. It's that Linden Lab announced they would not support the use of these hacked prims (and even broke the ability to create more of them). This is why I don't use them for builds for my clients. I don't think something Linden Lab does not want or support should be included in certification that bears the Linden stamp of approval.
--Kim Anubis 11:16, 2 May 2007 (PDT)
The current LL statement is that if they cause problems they will be removed without warning, and they're currently tolerated and in use on the mainland and on islands. Until/unless they officially ban them, they're part of the landscape, so I'd rather builders know about them than let someone sell them on the idea without enough information to make a solid judgement. --Storm Thunders 07:32, 4 May 2007 (PDT)
I contacted the Linden who fields technical questions as a part of LL's Developers Program and asked about this. He said that until LL officially announces megaprim capability or something similar he thinks megaprims should not be incuded in certification standards, and said that he would not use them in a build for a client. I'm convinced that megaprims need to come out of the article. How do you feel about it? --Kim Anubis 09:29, 8 May 2007 (PDT)
I think they require some sort of mention - they're a part of the landscape, and without information they look like an easy way around prim limits. We could simply put in the official Linden line. "The ability to create new ones has been deliberately removed. If they are believed to cause problems they will be deleted from a build without warning. They're currently tolerated on the mainland and on islands." At least anyone certified will be aware of the Lab's current statement on them. --Storm Thunders 08:19, 14 May 2007 (PDT)
It would be useful for builders to be aware of the giant prims, and how they work. While perhaps not encouraging new builders to use them, their existence and how they work, and their particular advantages would be something I hope builders would know if they are certified. I see the point that LL does not want to encourage (and may want to discourage) their use, they still exist and are (I want to suggest) an important (if historical) aspect of builds.
--Donnagh McDonnagh 03:08, 17 May 2007 (PDT)
  • Prim torture shapes.
This might be in an "advanced" section of the test?
--Kim Anubis 09:37, 28 April 2007 (PDT)
More along the lines of knowing that some prim properties are hidden and exposed depending on the current shape-type of the prim, and that you can manipulate those properties in one shape, then transform the prim to another. --Storm Thunders 06:50, 30 April 2007 (PDT)
The knowledge that there are such abilities is one thing, and knowing how to do them is something else. Perhaps the former should be part of the basic certification, and then the latter should be part of an advanced or specialized certification?
--Kim Anubis 11:16, 2 May 2007 (PDT)
  • Techniques to minimize your prim count.
  • Texturing techniques that influence building. Shadowing, lighting, details, alpha games... texture maps. Textures and lag. Size, tiling, alpha channel issues, formats for upload. Common uses of alpha textures.
I often hire different people to do models and textures -- texturing is really a different specialized area and I'd like to see a separate certification for it. If I need someone to create a texture that will tile, I don't ask a modeler.
--Kim Anubis 09:37, 28 April 2007 (PDT)
There's combo techniques that can make for a better builder. As an example, flexi and alpha textures - flexi prims "flex" differently depending on their z size, but with a couple alpha textures you can create complex objects like those popular fox tails. --Storm Thunders 06:50, 30 April 2007 (PDT)
That doesn't sound like a basic certification to me, though. "Better builder" and use of a bunch of combined techniques sounds like a higher certification level.
--Kim Anubis 11:16, 2 May 2007 (PDT)
  • Balancing and minimizing lag. Textures, shapes, levels of detail, etc...
  • Working with curves and organic shapes.
This area is about to change drastically, with the advent of the new Sculpted Prim. Eventually I suppose the certification will probably cover importing functional Sculpted Prim textures and applying them.
--Kim Anubis 09:37, 28 April 2007 (PDT)
  • Basic script usage, more of a "drop it in and let it run" sort of thing. Animating textures, basic sound use, setting sit targets and assembling pose balls, removing hoverscript and particles, basic doors...
I don't know how much of this should really be included. Although some of them can do it, I don't require all of the modelers working for me to know how to do these things. I'm more likely to either hand them an already scripted door blank, or have the modeler pass the chair model to a programmer who'll add the sit target and sound player (whoopee cushion!).
--Kim Anubis 09:37, 28 April 2007 (PDT)
  • Permissions. Implications of different combinations. Behavior of mixed permission objects & contents. Groups.
--Anna Gulaev 20:37, 28 April 2007 (EDT)

Subdividing

This certification is looking rather complex (and near impossible to attain). How does everyone feel about subdividing it into a core certification and secondary certifications? (Like what we are doing with the Scripting Certification) -- Strife Onizuka 09:39, 30 April 2007 (PDT)

I'd rather focus on content first. I've never seen a complete SL modelling topics list and I think developing one would benefit us. The certs may end up with time limitations or other issues forcing divisions, or we may decide it makes sense to have levels or specialities of certification. --Storm Thunders 06:12, 1 May 2007 (PDT)
Unfortunately, I don't see a horde of builders rushing in to add a lot more to the content list, and I don't see anyone but us chickens deciding how the certification process is going to work, either. If no one else is going to step up, we might as well keep going. I like the idea of levels or specialties. Great builders of large-scale architecture and those who make incredibly fine and detailed jewelry from tiny prims aren't interchangeable, for example, and it would be useful to give recognition to those with advanced skills in their specialties.
--Kim Anubis 11:16, 2 May 2007 (PDT)
If we're going to subdivide into topics like jewelry and architecture, we'll need subject experts. --Storm Thunders 07:00, 3 May 2007 (PDT)
In a project like this, things only really get done if someone picks up the flag and runs with it. You can get alot done if you just decide to do it and then ask questions afterwards. Course if you do it this way expect people to disagree and be sure you address peoples suggestions. The alternative is to wait indefinitely for some amount of agreement in the community to say a design is good. Best to use the spaghetti school of development, if it sticks to the wall you keep it, if it doesn't stick you go back to the stove. -- Strife Onizuka 13:50, 2 May 2007 (PDT)

What's the added value to creating multiple certs? I see the disadvantages - increased financial and maintenance costs of cert program, market confusion among clients/employers - but not the advantages. A single cert, or a basic and an advanced should give people the information they want when backed up with a portfolio and references. --Storm Thunders 07:00, 3 May 2007 (PDT)

The certification is supposed to help an employer distinguish whether a modeler is going to be competent to do their project/job. Some jobs are far more difficult than others, and some will require specialized advanced skills. While basic knowledge of working with tiny prims, large-scale builds, attachments, and organic shapes is something every competent modeler should have, far fewer know all the ins and outs of all of these things. It would be useful to make a distinction. If we don't break this up somehow, it's either going to be too easy to pass (and therefore not too meaningful) or too difficult (it'd really stink if good builders were filtered out because they weren't generalists in advanced areas).
I don't think it would have to cost much more or be much more hassle to have basic and advanced levels, or even to include specialties. There doesn't need to be a separate test. It could be one test with different sections. There might be (this is wildly speculative) 10 questions or tasks related to tiny prims, for example. The first three might count toward core certification, the next three would have to be correct or passed in order to make advanced certification, and the rest toward a specialization.
--Kim Anubis 10:51, 4 May 2007 (PDT)
At the minimum we need a "core modeling competency" certification. Having an "advanced" certification then follows. When we've fully defined the competency list it may be clear that there are additional specializations.
-- Glenn Linden 9:44, 9 May 2007 (PDT)