Jump to content

zBeeble

Member
  • Posts

    23
  • Joined

  • Last visited

Posts posted by zBeeble

  1. So... I'm working on hand compiling Cura --- I want more recent versions running on FreeBSD, and there seems to be no reason why there shouldn't be.

     

    w.r.t. that, I'm trying to deal with an error (coming out of protobuf) that is failing due to the compiler being c++11 rather than c++14.  I've got it inserting the compiler argument -std=c++14, but I also see it inserting -std=gnu++11

     

    For the time being, I've pulled this out of the FreeBSD ports world and am trying to manually get this going.  Once I understand this, I can more aptly make the port work.

     

    Anyways... that's the long and short of it.  If there is a more apt forum, please direct me.  If not, let's roll up our collective sleeves.

  2. So... Doing even better today.  I'll link in a project file.  So... the Otyugh here is best printed on it's side.  In general, the teeth are not something you want to support (easy to break) ... but printing it in this orientation on the deck --- only a small part is initially on the build plate.  Now, I think there's a problem with using a raft ... so I instead I just decided to have something that is all supported.  It works amazingly well.

     

    Watching it print, as I sometimes do, there's an obvious change from the big round shapes to the many little fingers in close proximity to the print.  This is done well enough that different parts of the print are held accurately enough to join.

    CFFFP_Otyugh Updated.3mf

    • Heart 1
  3. Here is a video.  https://nextcloud.towernet.ca/s/M3SibPJkrz9JMmF

     

    What you're looking at is my laptop.  I don't print every day, so it took me awhile to get back into it and to have the bug manifest.  What you're looking at here is a video taken on my phone of the laptop cura just after I hit CTRL-N to clear the plate.  You'll notice me flicking the zoom in and out --- and the "in" always stops at the same point --- which is uncomfortably close.

     

    Then I exit cura, and restart it.  After restarting it, I load some zombies and try the zoom --- fixed again.  This happens often.  CTRL-N is a common cause, but other cura commands cause it too.

  4. So... I'm having a bit of trouble with a miniature I'm printing for my D&D group.  I decided on this attempt to use a raft because on a previous print, a couple of the supports had failed to stick to the build plate.  Anyways... the print was a complete failure.  When I went back to Cura to make some changes, I think that I see JackelwarePrintProblem.thumb.jpg.84f1b6dfadbd8d0cc715801d94b02b90.jpgwhy... there's a curious and large difference between the outline of the model and it's actual position.  This leads to a large gap between the support and the model.  This gap seems to be something along the lines of the thickness of the raft.

  5. So.  firstly, kudos.  Awesome stuff.  All that.  Little bug.

     

    When I use cura, if I perform certain actions, and not every time, like say "new project" the camera position is reset.  That's fine, but what's not fine is that it will now be limited to how far away I can zoom it.  This would be a noise level issue if it weren't for the move mechanic where the arrowheads that perform the action are long enough that they're often offscreen if I don't tilt the camera to some crazy angle to access them.

     

    Right now, my work around is to restart cura.  I'm running 5.2.1 on windows 10 on two different machines.  One has a GTX 3080, the other has both intel integrated graphics and a quadra 2000 --- and the application is set to prefer the 2000.

  6. Actually, reading this again, I have more to add.

     

    On 8/17/2018 at 4:28 AM, SandervG said:

    But thanks for the feedback @zBeeble . I see where you are coming from. If we would enable users to select more fields, there is a fair chance this would ruin most of the collected data. For instance, 10 people would have downloaded Cura from which there are 8 professionals, 6 home users and 9 educators. That is 23 'persona's' from 10 downloads. So instead we hope you'll pick the answer which defines you most. If you have a 3D printer at work (from company budget), but you also use it to print an occasional desk ornament or a present, does not make you less of a professional user. We all learn things along the way, but that doesn't make us an educator or teacher (educational use). It doesn't have to be that black and white, not 100% is not automatically 0%. Hopefully this helps.

     

    So... if you have 10 downloads, maybe you have 2 professional-home-educators, 3 home-professionals, 2 home-educators, and 3 plain professionals.  You're asking people to self identify in 2018 ... which means, like sex (where there are a _lot_ more than 2 kinds now), people do all kinds of wonderful things with their printers.  Once could say your trinary pigeon holes are antiquated and looking at your question as a set of axis is more appropriate for the times.

     

  7. I've attached my new bath knob.  Permissive commons license and already on thingiverse.  I was watching the print (3D printers can be so mesmerizing on a lazy Saturday)... and I think I detected one place where Cura can do better (although I don't know if I know where to twiddle... is this possible with a plugin?  I'm happy to roll up my sleaves if I'm not wasting my time.

     

    I figure this is pretty common.  The hole narrows in the middle.  It's wider on both sizes and there are good reasons not to fillet (basically... the screw supplied with the original part wouldn't fit properly for one).  Your first reaction is that supports would make it print 100% successfully.  I'd say supports are hard to remove in there... but also, the current behavior is so close ... it just seems cura can do a tad better...

     

    If you print the knob with the outside surface down (seems the sensible way), the hole starts at 12mm and narrows to 5.5mm (I know 'cause I made it).  On the first layer of bridging, it starts fine ... going across from the edge of the circule out until it is printing the part of the smaller circle that isn't a contiguous bridge... which continues until it is again.  This dumps that non-contiguous material down the hole and leaves two side bits almost-ok.  On the next pass, we almost get it right where the 2nd bridge layer is 90 degrees to the first ... and you get the square part of the circle-square being almost right... but still, on this pass, more material down the chute.

     

    If you were amazingly smart about this, I'd imagine the best way would be to print this as a series of tangents to the inside circle that began and ended on the solid outside circle.  If you were less smart about it, printing the two 90 degree edge bits on the first pass, you might not be dumping the circular inner walls down the hole.

     

    Thoughts?

     

    BathKnob v6.stl

  8. 12 hours ago, SandervG said:

    That is also me, but a test account I made a long time ago. Never deleted it. 

    Fun fact, this account is registered on my email address from back when we were small enough we could just use our first names. That account has my 'official' email address. Never wanted to say goodbye to that first email address, it is like an ancient artifact ?

     

    But thanks for the feedback @zBeeble . I see where you are coming from. If we would enable users to select more fields, there is a fair chance this would ruin most of the collected data. For instance, 10 people would have downloaded Cura from which there are 8 professionals, 6 home users and 9 educators. That is 23 'persona's' from 10 downloads. So instead we hope you'll pick the answer which defines you most. If you have a 3D printer at work (from company budget), but you also use it to print an occasional desk ornament or a present, does not make you less of a professional user. We all learn things along the way, but that doesn't make us an educator or teacher (educational use). It doesn't have to be that black and white, not 100% is not automatically 0%. Hopefully this helps. 

     

    Doesn't my education count?  Really.  I learned, for instance, what a PID controller is... and how they're quite common in our lives.  A fascinating diversion is how PID controllers were originally constructed as pressurized air devices and used to hold a heading on ships.  See... learning... thus I was educated (allbeit by myself, the internet and the practical experiment of building something that works).

     

    I don't recall all the items in the list on your survey, but I remember thinking that some would imply the others.  That it wouldn't make sense, say, to say you were prototyping without saying you were designing your own things.  Maybe you need people to "rank" their uses.  I'm actually in favour of you guys getting _more_ information.

     

    12 hours ago, SandervG said:

     

    Finally, if you feel you still can't select any of the options because they would do you unjust or would make you feel uncomfortable, there is a 4th option; 'I don't want to share any information'. 

     

    Yeah... sometimes I do.  I've downloaded it multiple times and I'm not sure how many times you want my information.  Sometimes I even want to download it without the use of a browser.  In general, I like to live by the rule that I only answer surveys when I receive some value in kind.  Good news is that you've already qualified in my mind.  Your application is well-done and useful.... easily worth a few surveys and some information.

     

  9. It's entirely possible that I'm not a normal 3D printer user.  I've been curious about the technology for some years, but I"d been to busy to really invest.  Some of you might balk at that statement, but I stand by it --- my very nature is to treat every opportunity as a learning experience.

     

    When I finally had time, I researched things on the internet, ordered things on eBay and set about building my printer from scratch.  Along the way, I learned about many things.  I'm fairly happy with my first effort.  And by that I'm not talking about the first thing I printed, but my first printer.  I definitely plan to build a more sophisticated printer in the future.

     

    And before I get into the meat of the post, I'd like to say that I fully appreciate Ultimaker/Cura for releasing their slicer to the open source community and this post is not IN ANY WAY trying to pour rain on that effort.

     

    That-all-said, the surveys are getting to me.  The choices are not good.  Education _or_ personal _or_ work.  Fairly, I would say all three.  In my short time as a 3D printer owner, I've made prototypes for work, learned much about how 3D printers work and made some birthday presents.  Similarly, I'm asked to choose among prototyping, production, .... and a bunch of things I forget as for what I'm making.  I want to honestly chose about 80% of the answers, not one.

     

    I don't have any problem at all with your survey-on-download approach, I'm just bummed by your survey.  I'm forced to pick random answers that don't really reflect what I'm doing...

     

  10. OK.  I've beaten down a bunch of errors.  There's two things I can't get past.

     

    If I checkout the git tree, and after making, there's no option to make install or run in place.

     

    I have a different tree checked out that has a cura_app.py to try to run... but there I get:

     

    [2:114:414]dgilbert@canoe:~/Cura-master> LD_LIBRARY_PATH=/usr/local/lib/gcc6:/lib:/usr/local/lib ./cura_app.py
    [MainThread] cura.CrashHandler.show [36]: An uncaught exception has occurred!
    [MainThread] cura.CrashHandler.show [39]: Traceback (most recent call last):
    [MainThread] cura.CrashHandler.show [39]:   File "./cura_app.py", line 69, in <module>
    [MainThread] cura.CrashHandler.show [39]:     app = cura.CuraApplication.CuraApplication.getInstance()
    [MainThread] cura.CrashHandler.show [39]:   File "/usr/local/lib/python3.6/site-packages/UM/Application.py", line 340, in getInstance
    [MainThread] cura.CrashHandler.show [39]:     Application._instance = cls(**kwargs)
    [MainThread] cura.CrashHandler.show [39]:   File "/home/dgilbert/Cura-master/cura/CuraApplication.py", line 154, in __init__
    [MainThread] cura.CrashHandler.show [39]:     ContainerRegistry.getInstance().addResourceType(self.ResourceTypes.QualityInstanceContainer)
    [MainThread] cura.CrashHandler.show [39]: TypeError: addResourceType() missing 1 required positional argument: 'container_type'

     

    ideas?

  11. Well... an appimage might run if I fired off a VM running linux.  The linux emulation tends to work fine for binaries that you can run, but the sheer volume of libraries this requires is unlikely to be less work to run under emulation.

     

    Besides... its written primarily in Python.  The C++ part is already ported (and seems like it was easier to port).  I realize that portable software can be difficult to write... but we're dealing with the low-hanging fruit here --- it's written in python.

     

    argh.  I suppose one big problem is that I'm not familliar with cmake and while cmake exists on FreeBSD ... whatever this cmake does is broken.  More tomorrow, maybe.

  12. So... I'm a fairly competent ports maintainer for FreeBSD.  I see that someone has already gotten the CuraEngine ported.  I thus have been working in the direction of getting the Cura graphical front end going.

     

    So... is anyone else working on this?

     

    It almost seems like the job of porting things written in python is made nearly impossible by their build systems.

     

    I think I have all the requirements built... but I'm still getting rather dumb output from cmake.

     

    Help?

×
×
  • Create New...