There isn’t much happening on this project. But, there is lots of talk. Oz Linden is talking with management about assigning staff to the project. That discussion should be happening this Friday. Next week we will learn whether he got the help he wants.
Jeans without Alpha Layer
In the mean time new people have been posting in the JIRA item STORM-1716 – Mesh Deformer for tailoring mesh clothing. This has become a rather large JIRA entry. So, users are getting hit with the problem the Lindens had, the thread is just too long to read. The result is important information is over looked by the people just joining the discussion. They are clueless and that annoys those that have followed the project for the last year. That results in lots of bickering and short tempers.
The thread is a good example of why the Lindens made the JIRA Change and closed down the BUG Reports and JIRA in general. Just try to get the latest information out of the 1716 thread. It is a pain and it doesn’t work. I read the thing because I want to distill the information out, know what is happening, and hopefully keep my readers informed.
Qarl Fizz’s Sarcasm
There are times when sarcasm is useful and appropriate. Karl Stiefvater is a programmer and I am afraid his sarcasm may have been lost on the less geeky, more vociferous, and ignorant people in the thread.
Qarl Fizz added a comment – 07/Jan/13 12:33 PM
Guys –I just had an idea.
If, in addition to all the deformer parameters we currently have, we add a string value called “Description”. This parameter will describe how you wish for the mesh to behave with other meshes.
For example, you could say “this item is pants, I wish for it to be worn on the legs over underwear but under a jacket which may be worn on the top.”
The deformer will parse this string and follow the constraints it imposes.
This system is far superior to any system currently proposed. Furthermore, it allows for a completely new creation mechanism – you need only specify how you wish your mesh to appear “a white dress with a lacy top” and the deformer will produce exactly what you want.
I think Karl was getting annoyed by the thread. I am sure some people will miss the note of sarcasm and think that providing a text description of what their mesh is supposed to look like will allow the Deformer to create it. Ignorance is infinite. Unfortunately, the world doesn’t and especially computers don’t work that way.
What brought this on? New posters to the thread wanted to know why the Deformer doesn’t do various things they want and why it doesn’t over come their bad design choices, of which, as best I can tell, they are unaware.
We have already been through discussions on why it does what it does. We have already tried various solutions. Some failed and were removed… failed may be too strong. The code worked, but while some problems were solved others were created. We didn’t get the result we wanted.
Deformer Complcations
Now we have what many of us consider the best version of the Deformer. It is in a Project Viewer and can be used to test the process. In my testing I am pretty happy with it. I’ll admit most of my testing is with skirts. Tops are going to be a pain to learn and do well. I’m not there yet.
You can find the latest version of the Deformer Project Viewer is 3.4.1 (267522). I have a link in the left column for the download. The link leads to a Nov 11th build as I write this, but it will automatically update to the latest version.
The current main viewer is 3.4.3 (268262). Lots of things have been added to the viewer since November. The Deformer has not changed. But, we do need to see how the code works with the latest updates to the viewer. I expect to see a new version next week (3) or the week after (4).
Testing
For making and testing deforming rigged mesh clothes one must use the latest version of the Deformer for the upload, I wrote Deformer not viewer. But, we don’t have version numbers for the Deformer. So, we are stuck with viewer versions. So, effectively we must either know the Deformer change dates or consider only deformable mesh uploaded with the latest viewer, regardless of whether there were changes to the Deformer’s specific code, useable for testing.
Anything uploaded using a viewer prior to the November build is probably unusable for testing. If you are designer, your stuff does NOT have to be remodeled for use with the newer viewer. It just needs to be uploaded once more with the latest project viewer.
BUT…
Models made with much older versions of the ‘Deformer’ may need a special base model added. So, let me explain that.
Special Update
Anything made and uploaded in September 2012 and before must be uploaded with a special process. Or the model has to be adjusted in Blender/3DS to fit the new process that uses variable base shapes. There is an easy way to do this.
Sometime after September the ability to associate a custom base shape with your rigged mesh upload was added. Anything uploaded without that option will NOT deform. The Deformer just won’t see it as deformable.
If you used Avastar, you will probably have some problems for models made before December 2012. They are not going to fit as well as you hoped because of problems with Avastar’s program. You may need to follow these steps too. That problem was fixed in about mid December. Just update your Avastar.
Since September one can now use an XML files exported from the viewer that contains the desired base shape. We simply wear the base shape in SL that we want to design for and export that shape.
To do that look in the menu: Advanced->Export appearance as XML. The resulting file can be placed in the folder with the DAE Collada file you export from Blender or 3DS for easy upload with your rigged mesh. I have other articles on how to get the shape into Blender. That is not important for this process.
Yes, there are permissions issues so it has to be a shape you have rights to. If you have been using the Default Shape for your design, you will be able to skip the next steps and just upload a new version using the option for default female or male. Then your model should work with the Deformer and fit all sizes.
So, the problem as I understand it is too much general community participation in STORM-1716 and not enough involvement by the mesh clothing creators.
As someone who is very interested in seeing mesh clothing that really works I follow STORM-1716 almost daily, it is not hard to follow. I am not a mesh content creator however, so following is really my only way to show my ongoing concern. If there was not a continuous stream of input at the jira LL would be justified in saying no one cares.
It is a catch-22, show your interest and you are cluttering up the conversation. Say nothing and run the risk of LL dropping the project for lack of interest.
I suppose that is one way to look at it.
My point is the number of uninformed people that want to have a say without wanting to educate their self on the issues or the processes used. At this point people jumping in to voice nothing more than uninformed opinions is not going to help. I think this is the problem that frustrated the Lindens to the point they closed the JIRA.
Following a thread daily makes it easy to keep. You are informed. You only have to take in small amounts of new information. But, come into the JIRA cold and try to sort out what is going on. It doesn’t work well and TL:DR sets in. Then those people are misinformed. It is bad enough dealing with the ignorant and having to educate them. Its even worse trying to undo misconceptions and clear up confusion. If someone is not interested enough to read the whole thread and know whats going so they can make intelligent comments, why would I think they are worth the effort and have the interest to justify my time to bring them up to date or fight them to correct their misconceptions? There are more fun and productive things to do.
Input into the JIRA is not likely how the Lindens gauge interest in the Deformer. In 414 days only 115 people have commented, making a total of 539 comments. There are 397 Votes and 261 Watches for a total of 658. For a bit over 2 years people have been discussing how VOTES don’t count in the JIRA and encouraging them to click WATCH. But more people click VOTE than Watch. Our community is NOT well informed.
Oz asked for what the Lab needs to proceed, test clothes. Giving him JIRA entries is not going to help. Lots of people can SAY they are interested and make comments about what they think the Deformer should do. But, I don’t see that as real interest. I suspect the Lindens don’t either. The measure Oz seems to be looking at is the submission of test clothes. We had some. Those same ones just need to be re-uploaded with a new viewer. But, last time I asked Oz I think he had received 1 or 2 garments. I can’t see that overwhelming him as fervent support for the project.
We can say what we think is a show of support. But, if isn’t something the Lindens think is a show of support, it isn’t going to influence them. We can’t change them. But we can change what we do.
Why does something like the material system get featured very prominently on the official blog, but not the deformer project? Aren’t both community initiated?
It’s odd to keep mentioning interest, or lackthereof, and the need for proper testing when most other new features that need testing via project viewers receive blog posts.
I was happy to read from you the other day that Oz was going to see about being able to staff the deformer project somehow, I hope if that happens, that can also include a blog post detailing the deformer project and what kind of test items are needed and how to submit them. Not everyone that’s heard about the deformer and are interested in it are going to check this particular blog; the only place I know of with easy to find instructions on exactly how to help.
“Since September one can now use an XML files exported from the viewer that contains the desired base shape. We simply wear the base shape in SL that we want to design for and export that shape.
To do that look in the menu: Advanced->Export appearance as XML.”
Nalates, these instruction are for the current SL viewer?
If so don’t see ‘Export appearance to XML’ there.
only thing close I see is in the ‘Develop->Avatar->Character Tests->Appearance to XML in the SL viewer. Is that the one?
Ignore my last post, only from the Project viewer do we see this function. Found it.
Appearance to XML is the one. I think I was using the project viewer. Sorry. I should not have assumed they would be the same.
Pingback: Transformational Change | MetaReality Podcast
Pingback: I Am Not Standard Size Part 2 « Colleen Criss's Weblog