Menu

Bug or feature? Part adding to the partlist after the step in buffer exchange

Artius
2016-11-22
2016-12-22
  • Artius

    Artius - 2016-11-22

    Hi. I'm trying to use LPub3d instead LPub. But I have a problem with using buffer exchange. Issue is that the Part, which flying to the model in the Step, contained in buffer, then stands on its place in Step after buffer, but its displayed in partlist as in step contained in buffer so in the next step. But I want to display this part in PLI only in step contained in buffer, like it realised in LPub.
    How can i solve this problem? Maybe I should configurate something in the preferences?
    Thanks.

     
  • Trevor Sandy

    Trevor Sandy - 2016-11-24

    Artius, I'm not sure to completely understand the behaviour you expect.

    Can you prepare screenshot(s) indicating what you expect (just use OK or KO to indicate what you do and what you do not expect)

    Send me by email the screenshot(s) and a sample of you model file producing the unexpected behaviour.

    You can see my email address in LPub3D Help=>About menu.

    With this information, I can come back to you with the required configuration to achive the behaviour you expect.

    Cheers,

     
  • Johann

    Johann - 2016-11-26

    Good evening
    Was the problem understood?
    I noticed the same thing.
    Will the problem be fixed?

    Regards

     
  • Trevor Sandy

    Trevor Sandy - 2016-11-26

    Johann,

    Indeed. Artius prepared an excellent bit of informaton to demonstrate the different behaviour.

    2 scenarios along with my findings are prresented below:

    V1 differences in behaviour between LPub and LPub3D (see model file attachment test.mpd)

    step1_LPub.jpg

    step2_LPub.jpg

    step1_LPub3D.jpg

    step2_LPub3D.jpg

    My findings (see model file attachment test valid.mpd)

    The example test.mpd presents the same results in LPub and LPub3D - 3 added parts in Step 2, so 3 parts in the PLI for step 2.

    The correct display for step 2 is as represented in the screenshot step2LPub3D.jpg.

    Explanation: In step 2 there are 3 added parts: 3004.dat (line 69), 3003.dat (line 70) and 3003.dat (line 86). These 3 parts are correctly represented in the step 2 PLI.

    From the description, it looks like part 3004.dat (line 69) is actually the same part as 3004.dat in step 1 (line 66) and; therefore, should not exist in step 2.

    In the model file test.mpd, If you remove line 69 you will get the expected result described.

    However, the sample file test.mpd is not well formatted for managing BUFEXCHG across steps.

    I have attached an updated version of the sample file (test valid.mpd) showing valid use/pattern for BUFEXCH which works in both LPub3D and LPub.

    Note: If you are testing with LPub, do not forget to clear all the generated file caches by deleting the LPub folder and its contents created where your model file is stored.

    V2 differences in behaviour between LPub and LPub3D (see model file attachement test-v2.mpd)

    LPub3d_step1-v2.jpg

    LPub3d_step2-v2.jpg

    My findings:

    The expectation described for step 1 is not possible in LPub3D - it is also not possible in my version of LPub (LPub 4.0.14).

    In both my LPub3D and LPub environments, and according the their source code, the behaviour of sample file test-v2.mpd is the exact same. The PLI is present in step 1 and step 2 because in both steps parts are added (line 48 and 64 in step 1 and line 68 in step 2). Step 2 is correctly showing 1 part in the PLI (line 68) and the second added part in step 1 (line 64) is successfully excluded from the BUFEXCHG. Therefore, the final result is 2 parts in Step 1's PLI and 1 part in step 2 PLI. This is the behaviour I am expecting and experiencing in both LPub3D and LPub.

    While it is ok, and perhaps more compact, to employ the PLI..IGN meta command against the non-flying part as shown in test-v2.mpd, this is not how I normally see the BUFEXCHG/IGN commands used. Usually, the IGN part is the flying part and its associated helpers (arrows etc...). This is the pattern I shared in the sample model file test valid.mpd.

    If using LPub 4.0.0.14 as I am, I believe the results are misleading users to believe the part in line 68 is actually excluded from the PLI. In fact the behaviour shown is more likely because LPub is not well managing its generated cache files. You can validate the definitive behaviour by deleting the LPub folder and its contents generated where you store your model file and re loading the sample model file.

    To further demonstrate the definitive behavior of LPub (and equally LPub3D) I expect and I am experiencing, I have prepared a short video using LPub (and test-v2.mpd) available here.

    Let me know if you have any questions.

    Cheers,

     

    Last edit: Trevor Sandy 2016-11-26
  • Johann

    Johann - 2016-12-22

    Hi
    Sorry for the late reply.
    I use another code for bufferexhg. I use the same code like hollywood.it.

    Here the code:

    1 2 -40 0 120 1 0 0 0 1 0 0 0 1 819.dat
    0 STEP
    0 BUFEXCHG A STORE
    0 GHOST 1 25 0 -24 120 1 0 0 0 1 0 0 0 1 3003.dat
    0 GHOST 1 25 -40 -84 120 1 0 0 0 1 0 0 0 1 3003.dat
    0 MLCAD ARROW 4 -40 -24 120 0 1 0 -1 0 0 0 0 1 13 6 12 40 3 30 1 2 16 16
    0 MLCAD SKIP_BEGIN
    3 4 -40 -12 120 -46.5 -27 120 -40 -21.4286 120
    3 4 -40 -12 120 -33.5 -27 120 -40 -21.4286 120
    3 4 -40 -21.4286 120 -43 -24 120 -37 -24 120
    4 4 -43 -24 120 -43 -64 120 -37 -64 120 -37 -24 120
    0 MLCAD SKIP_END
    0 STEP
    0 BUFEXCHG A RETRIEVE
    1 25 0 -24 120 1 0 0 0 1 0 0 0 1 3003.dat
    1 25 -40 -24 120 1 0 0 0 1 0 0 0 1 3003.dat
    1 25 -80 -24 120 1 0 0 0 1 0 0 0 1 3003.dat
    0 STEP

    This code only works in LPub 4.x,not in LPub3D.

    Regards Johann

     
  • Allard van Efferen

    Working with the latest version 2.0.20 this issue mentioned above is still present.

     
  • Allard van Efferen

    Besides I try to avoid messing in the code and only using the user interface. Through the user interface only I can't get it to work. even after cleaning the cache.

     

Log in to post a comment.