Shapeways site update of August 5th 2010

Discussion in 'Official Announcements' started by arno, Aug 5, 2010.

  1. arno
    arno New Member
    Hi All,

    Today we have had another release.

    New volume calculation

    This release has as biggest feature the enabling of the Mesh Medic unify functionality. This implies that the volume calculation is updated, as announced in this blogpost. Sorry it took a bit longer as expected, but we wanted to test this feature properly before enabling it.

    MeshMedic will now try to merge multiple shells that overlap into one resulting that there won't be a problem anymore with overlapping volumes.

    In rare cases, MeshMedic will not be able to do this properly. Eg. when testing this feature we found one file that MeshMedic couldn't unify. This was because it had over 10.000 separate shells. When you try to upload such a model, you can now get an error message, I hope you can understand that.


    ShareThis automatically attached on blogposts

    A sharethis plugin for the blog is created so you can also share blogposts you think are interesting.


    "Add all materials" action to material restrictions function

    An 'add all material' link for the restriction list on the product details page is added. You can now put all materials on the restriction list with the new link and click on the trashcan to add a material for his product.
    We hope this new link will make it easier to handle our growing offer of materials.


    Besides these features, we also did some bugfixes and tweaking:

    Description box not available for co-creator models

    The description box was not available for co-creator models. This has now been fixed.


    Change product detail page title

    The title of the product page is changed to: <product name> by <username> on Shapeways


    No overlay icon shown for T&C approval for API uploaded file

    No icon was shown for T&C approval when a model was uploaded via the upload api. This has been fixed.


    Product search: next page makes sort go to relevancy

    When you clicked on the next page of a sort, the sort shifts back to (it's default) relevancy, regardless of the chosen sort. This has now been fixed.


    Kind regards, Arno van Rossum
     
  2. stannum
    stannum Well-Known Member
    Are the changes done for the new unification process, used to calculate the volume now, related in any way to color model errors? Because a model that in the past was printed 3 times (different images maps, same mesh) now is unprintable.

    The error mail says:
    "Thin wall <3 mm, Part too big, Non cleanable, Bad/Open Edges, Inverted normals,
    Holes, Shells,"

    The image attached to the email shows an error that looks like faces are now split in a per image basis, which was no problem before (multiple images, non overlapping UV). No idea why the other errors were reported at all, the item dimensions are the same, shape is rather simple, etc.

    And while we are with this topic, are there any warranty that future prints will be rejected when previous passed? Just to avoid investing any more time in a rather simple model that is only giving headaches, and if sold, probably cause unhappy customers.

    Thanks in advance.
     
  3. arno
    arno New Member
    Hi,

    Firstly I would like to mention that we do not run meshmedic for color files yet. We are still testing that to iron out the last wrinkles. So meshmedic has not been of any influence to your model.


    Secondly, I assume you are referring to your model 'color quick reference', which was ordered this month? That was a new upload, not another print of an existing model.

    All models get manually checked before their first print (new upload means first print). Recently we indeed had to become more strict in models needing to follow the design rules as there were to many printer failures.
    (see eg the explanations in this thread where someone was really upset when their newly uploaded model was rejected for printing.
    - note: please check the explanations, not the rant :confused ... seems the model was scaled down before uploading so the rejection was justified -

    Your model in question (cqr) has walls of 1mm thick which does not follow the design standards for full color sandstone. I would say 'kudos for the printer guys' that they were able to print such a model before without braking it, or the print run crashing :rolleyes:

    I hope this explains what is happening?

    Kind regards, Arno
     
  4. stannum
    stannum Well-Known Member
    The issue was already solved by email. The initial reported error shown in the image provided was separate faces, not thickness. The email text listed all kind of unrelated errors, making it all more confusing.

    It was a "reprint" in the sense that all previous uploads share the same mesh, and just change the images or how the images are applied to the mesh.

    About the 1mm thing... it's well protected by a 3mm skeleton. And probably that was the reason it printed without crashes before, the thin zones are not vital to be handled and are pretty simple on themselves.

    If that is not allowed any more, then please advertise sandstone is considered "all wall". Tutorials mention 0.1mm detail and "avoid spikes", but if those are too soft limits then you better update the docs so people design chunkier things from the start.
     
  5. dizingof
    dizingof New Member
    Arno,

    When i turn a model to "show only" it still remains on sale... :rolleyes:


     
  6. arno
    arno New Member
    Hi,

    Can you let me know what model please? So I can check if there is a problem or not.

    Please do know that the server is currently busy generating a price-material cache (related to the release this morning with the new material selection). As the server is very busy, other processes we handle via the jobqueue (like processing model uploads, and for sale notifications) are trickling through very slowly today.
    I suspect this is the cause ...

    Kind regards, Arno
     
  7. lsuess
    lsuess New Member
    Does that mean when the multiple-shell-merging will be applied to *.wrl & *.x3d files too then files with more than 10000 atoms modeled as unmerged spheres
    will become impossible? :(
    (recently uploaded one with ~1400 but that might grow)