Getting error on missing vs2010 redis package file(which is installed). Maybe you should link wrapper statically?
Also, just putting that file to the same folder as openscad_cl works fine, but still...

Had to downgrade to 2.64 r51026 (although I only tested 2.66 SVN and 2.66 current official release.)

Running under 2.64 with no problems.

Tracked it down to viz.py
the command bpy.ops.transform.translate(value=(0,0,z_dim/2.0))

Sorry - I am no blender expert.

Regards,
Brendan

/tmp/render.crash.txt

Blender 2.66 (sub 0), Revision: 54697

backtrace

/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xea6a9c]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xea6cda]
/lib/x86_64-linux-gnu/libc.so.6(+0x364c0) [0x7f480976e4c0]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(WM_keymap_list_find+0x12) [0xeb0442]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(WM_keymap_active+0x28) [0xeb1078]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(initTransform+0x12b) [0xf5626b]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xf570cb]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xf57187]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xec042b]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(WM_operator_call_py+0x6e) [0xec0d6e]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0x16db100]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalFrameEx+0x6727) [0x2976bd7]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalCodeEx+0x81e) [0x29703fe]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0x28f04f0]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyObject_Call+0x5a) [0x28ca96a]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0x28e1cad]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyObject_Call+0x5a) [0x28ca96a]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0x2926f31]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyObject_Call+0x5a) [0x28ca96a]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalFrameEx+0x17a0) [0x2971c50]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalFrameEx+0x6694) [0x2976b44]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalCodeEx+0x81e) [0x29703fe]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyEval_EvalCode+0x3b) [0x29704ab]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(PyRun_FileExFlags+0xb9) [0x299aba9]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0x16d09ce]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xea7b60]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(BLI_argsParse+0xc7) [0x1731547]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender(main+0x21a) [0xea8efe]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed) [0x7f480975976d]
/usr/local/bin/blender-2.66-linux-glibc211-x86_64/blender() [0xde51

"Add the directory of the OpenScad executable to your search path." No idea how to do that on my computer so every other step seems to fail.
I can't run make_machine.py
so I can't make machine.scad
so conf/config.scad has no variables
so fixing-block.scad won't compile correctly.

I only need one tiny part, and I know how to edit these files, but I need to generate an entire printer somehow.

Would having an example machine.scad be difficult? I don't know why an openscad project need python at all.

I'm sorry if this is reading as angry, but I really just need one tiny part and I'm very frustrated. Also my girlfriend is yelling at me because I didn't tell her that she needed to save her work, so it's my fault that her last 3 hours are lost.

Здравствуйте, пытался загрузить stl файлы пишет нужен puch доступ.?

Hello,

Not sure how else I can contact you, but do you still offer/sell the kits and if so how much would one cost (UK) or would it be cheaper for me to get the parts myself?

Do you have a template for the frames and everything else? Or a manual I can follow for it as I can't seem to find one. Or a parts list.

Regards
Thomas

It would be really nice to get the translate , rotate and instance data out of the _assembly modules...

With this data the Mendel90 could be rebuilt in blender from this and output STLs.

I am unable to print the X-idler_bracket.stl and x_motor_bracket.stl because they do not slice with a nut trap in the top (using Slic3r or Cura or ReplicatorG). According to blogs this is an OpenScad problem. The collection of source files for Mendel90 does not seem to contain the scad versions. Where can I find the .scad source for these so I can add a nut trap another way?

I see that the E3D branch does not have the GT2 pulley changeset. Looking through the change set it seems like this is just a materials change and does not affect any printed parts (minus the removed T5 printed pulleys). Is that correct?

E.G. Buddaschnozzle.

It would require moving the the third bearing off centre, moving the extruder back a few mm and making the hole bigger.

Hey, I noticed that inkscape was opening the GUI and not doing PDF conversions, so I did some digging. I'm running Inkscape 0.48.3.1 r9886 (Mar 29 2012) and Python 2.7.3.

Apparently adding the -z option tells inkscape to not open a GUI, so I added that. But the big thing I noticed was that Python was ignoring the CLI arguments altogether. I tested this by running the CLI string:

inkscape -f mendel/sheets/frame_base.svg -A mendel/sheets/frame_base.pdf -z

.. manually on the command line. That would work fine, but when Python executed the same thing, the GUI would open and nothing would happen.

This behavior was caused by the shell = True argument to Popen. The documentation states:

The shell argument (which defaults to False) specifies whether to use the shell as the program to execute. If shell is True, it is recommended to pass args as a string rather than as a sequence.

Changing it to shell = False made inkscape work as expected.

Here is my diff:

diff --git a/InkCL.py b/InkCL.py
index 4db94da..a1042dd 100644
--- a/InkCL.py
+++ b/InkCL.py
@@ -8,7 +8,7 @@ def run(*args):
     for arg in args:
         print arg,
     print
-    run = subprocess.Popen(["inkscape"] + list(args), shell = True, stdout = subprocess.PIPE, stderr = subprocess.
+    run = subprocess.Popen(["inkscape"] + list(args) + list(' -z'), shell = False, stdout = subprocess.PIPE, stder
     out,err=[e.splitlines() for e in run.communicate()]
     return run.returncode, out, err
'''

Hopefully that's useful to someone.

Attempting to run bom.py I get an error. Not sure how to proceed.

In case anyone is interested in generating a bom csv, here is a little diff to sort out the columns and such. I just copy and paste this code into bom.py, above the "Printed:" line, then I generated the BOM, then I cut and pasted the csv into another file and imported into a spreadsheet.

diff --git a/bom.py b/bom.py
index 9ca4639..0b96ddb 100755
--- a/bom.py
+++ b/bom.py
@@ -65,6 +65,35 @@ class BOM:
             print >> file, "%3d" % self.vitamins[part], description

         print >> file
+        print >> file, "CSV:"
+
+        if breakdown:
+            longest = 0
+            for ass in self.assemblies:
+                name = ass.replace("_assembly","")
+                longest = max(longest, len(name))
+            for ass in sorted(self.assemblies):
+                name = ass.replace("_assembly","").replace("_"," ").capitalize()
+               print >> file, " %s|" % name,
+            print >> file
+
+        for part in sorted(self.vitamins):
+            if ': ' in part:
+                part_no, description = part.split(': ')
+            else:
+                part_no, description = "", part
+            if breakdown:
+                for ass in sorted(self.assemblies):
+                    bom = self.assemblies[ass]
+                    if part in bom.vitamins:
+                        file.write("%2d|" % bom.vitamins[part])
+                    else:
+                        file.write("  |")
+            print >> file, "%3d|" % self.vitamins[part], description
+
+
+
+        print >> file
         print >> file, "Printed:"
         for part in sorted(self.printed):
             if breakdown:

Quick and dirty, and not a bit elegant, I know. But maybe it saves someone time.

Given we can generate individual not-in-BOM stls :)

Can you put to git sources for multipart panels like :

http://www.thingiverse.com/image:110689

When running under windows, bom generation fails because lines are terminated with CRLF.
When i change bom.py:103 s = line[7:-3](ignore one extra character), everything else seems to work.
Conditionally ignoring CR should fix this issue and allow working under windows

Is it possible to have the mendel sheets re exported as 2005 dxf format as the holes come out as 21 line cuts which drastically pushes up the price if you want to get these lasercut

There is a small Issue when creating your own Mende90 with different frame material. The Postion of a few drillholes is affected by the thickness of the choosen material. So when I create my own Mendel90 with MDF12 plates the drillholes for the fixing blocks moving a little bit up or exactly, away from the edge (around 1mm). This makes it difficult to build the frame.

Hey! :-)
could you maybe add a github repo description (plus tags), and also describe in the README on top what this project is (about)?
would be appreciated!
robin

The stls build fine, but I get this error when the bom gets generated. This is using today's git (3554174) but I've had it before, and have never successfully built a bom. This is an untouched repo, no changes to configuration files.

./make_machine.py mendel
Generating bom ...
Traceback (most recent call last):
File "./make_machine.py", line 17, in
make_machine(sys.argv[1])
File "./make_machine.py", line 10, in make_machine
boms(machine)
File "/home/james/Mendel90/bom.py", line 122, in boms
main.assemblies[stack[-1]].add_part(s)
IndexError: list index out of range

After editing your .bash_profile as I mentioned here: [https://github.com//issues/47], the generated BOM files contain only headers, and no materials listed. Only empty headers, as shown here:

Vitamins:

Printed:

Any interest in a Makefile based build system?

The advantage would be the ability to not recompile/rebuild files that haven't changed. I think right now ./make_machine.py always rebuilds everything, and it takes quite a while to run, even on a very modern/powerful machine.

Just curious.

Copied a conf file, didn't make any changes. Compile fails.
Linux Mint 17
Python 2.7.6
OpenSCAD version 2014.01.29

mysturdy_E3D_config.scad.txt

openscad.log.txt

When running render.py I've got a lot of errors:

seva@sevasoft:~/src/Mendel90$ python ./render.py monsterap
monsterap/render
blender -b  utils/render.blend -P utils/viz.py -- monsterap/stls/x_belt_grip.stl monsterap/render/x_belt_grip.png
read blend: /home/seva/src/Mendel90/utils/render.blend
Error: Node type Material Output doesn't exist anymore, removed
Error: Node type Diffuse BSDF doesn't exist anymore, removed
Error: Node type Material Output doesn't exist anymore, removed
Error: Node type Mix Shader doesn't exist anymore, removed
Error: Node type Diffuse BSDF doesn't exist anymore, removed
Error: Node type Glossy BSDF doesn't exist anymore, removed
Error: Node type Material Output doesn't exist anymore, removed
Error: Node type Emission doesn't exist anymore, removed
monsterap/stls/x_belt_grip.stl
monsterap/render/x_belt_grip.png

7.260000228881836
Traceback (most recent call last):
  File "/home/seva/src/Mendel90/utils/viz.py", line 64, in 
    load_stl(stl)
  File "/home/seva/src/Mendel90/utils/viz.py", line 30, in load_stl
    ob.material_slots.data.active_material = bpy.data.materials[mat]
AttributeError: 'bpy_prop_collection' object has no attribute 'data'

Blender quit

Blender version:

seva@sevasoft:~/src/Mendel90/monsterap$ blender -v
Blender 2.58 (sub 0)
    build date: 2011-08-01
    build time: 03:31:53
    build revision: unknown
    build platform: Linux
    build type: 
    build c flags: -g -O2  -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE64_SOURCE -fopenmp  -msse2  -msse -pipe -fPIC -funsigned-char -fno-strict-aliasing  -Wall -Wcast-align -Werror=declaration-after-statement -Werror=implicit-function-declaration -Werror=return-type -Wstrict-prototypes -Wno-char-subscripts -Wno-unknown-pragmas -Wpointer-arith -Wunused-parameter -Wwrite-strings -Wno-error=unused-but-set-variable
    build c++ flags: -g -O2  -D__STDC_CONSTANT_MACROS -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE64_SOURCE -fopenmp  -msse2  -msse -pipe -fPIC -funsigned-char -fno-strict-aliasing  -Wall -Wno-invalid-offsetof -Wno-sign-compare
    build link flags: -pthread
    build system: CMake
seva@sevasoft:~/src/Mendel90/monsterap$ 

Row 25 of main.scad should read:

use <ATX_PSU_brackets.scad>

instead of the current:

use <ATX_psu_brackets.scad>

The current way probably works ok under Windows due to Windows file names being case-insensitive, but it breaks under Unix-style file systems.

After printing the wades block for the e3d hot end I'm seeing that the hole where the hot end is inserted into the block has a flat spot or is "keyed" yet the hot end itself has no flat spot and is perfectly round.

I feel like I'm missing something but I cannot see how this is supposed to fit together. What am I doing wrong?

Hi guys. When compiling my machine i get that error.
The only files generated are the BOMs and just ..sheets/frame_base_dxf.scad which i cannot render in openscad because gives error.

What can it be?

I'm using Ubuntux64 13.04.

Thanks!

While printing out sheets for a Sturdy90 I realized the drill points seem to be off for the Frame Right sheet. The Sturdy sheet doesn't seem setup for ATX and Melzi.

I scaled the Mendel90 to have a build area of 24x28x22cm (x,y,z) and there were some problems with the Dibond drawings. (1)The vertical edge of the right stay has a bracket attaching to the air. You can see it in the rendered SCAD. (2)The square tube on the right goes exactly over the holes for the wire ties, and over the holes where the wire exits (through the base) to the power supply. (3) The left stay has a square-bottomed slot which besides the fact that it is not millable with a round router bit, appears to have no purpose. (4) The left stay attaches on top of (not between) the brackets for the main gantry piece. It was necessary to trim a triangle out of the corner of the left stay in order to attach it.

I'm going to make a version of the sturdy with 3/4" MDF and instead of using the fixing blocks I want to just put a dado in the base for the gantry and stays. I also want to put dados on either side of the gantry and base for the stays but that dimension seems less critical.

My plan was to simply increase the height of the gantry by the depth of my dado and reference the drill template from the top.

I'm trying to decipher the scad files to see where I can echo/print out the offset of the gantry from the front or rear of the base as well as the left and right stays. Can you provide any guidance to get these dimensions?

I was also planning on just reducing the screw count etc in my head but a way to simply exclude all the fixing blocks would also be helpful.

Hey,

I noticed the dibond directory today after doing a git pull, so I tried to build it.

Unfortunately, this command:

openscad -o dibond/sheets/frame_base.dxf dibond/sheets/frame_base_dxf.scad

yields this error:

Current top level object is not a 2D object.

OpenSCAD version 2012.05.12
on Ubuntu 12.04 LTS

Wish I knew more about openscad so I could fix. I guess openscad thinks it's a 3d model instead of 2d.

Not sure whether it is temporary problem with the pdf manual in here or with certain stl files that cause the problem.

The Dibond version crashes the extruder motor into the into the top z-bar clamp.

This has happened to me a couple of times now. When homing the z-axis while at X-min the extruder motor hits the top z-bar clamp. This prevents the limit-switch from being hit so the motors won't stop running. This causes the z-bar clamp to move out of place and messes up the calibration of the z-length. Steps will also be missed and this causes you to lose the levelling calibration. Also if you're not quick to turn the power of the other z-bar clamp will also be hit. I've replaced them both by now.

Here's a photo of this in practice:

I don't know if this is a known issue all ready, but I've opened up an issue here because today I found out this bug is also in the models:

Set these three lines almost at the top of main.scad:

X = 1 * X_travel / 2; // - limit_switch_offset;
Y = 0 * Y_travel / 2; // - limit_switch_offset;
Z = 1 * Z_travel;