Hi Thomas,
Thanks for posting. I might need some more help to try and troubleshoot.
I’ve taken a quick look at the gh script, it looks fine.
You’ll note in the header of the generated ifc file (open it in notepad or refer gh text panel), a time stamp is included, /* time_stamp */ ‘2021-12-24T10:41:48’,
Note that if I download a file from your test environment, it’s from December last year. Something is amiss there, and the streamed data isn’t being revised, hence the same file. If you use csdiff (or some other text comparison tool), you can see there is no difference in the files, including the time stamp.
I tested in the latest version of the plugin, I haven’t downgraded to the old version shape diver uses, but if I move a slider in your gh script, I can see the streamed text panel changing with each revision. Do you see this with the older version of the ggIFC plugin?
With respect to increasing file sizes, this suggests data not being expired correctly (if files always increase). Note that I certainly have improved/fixed this over the past year or two, so the version of the plugin on shapediver might be problematic. If you reduce the number of gridlines, I would expect a smaller ifc file. It’s ok to have gaps in the line numbering. Note that creating an IFC file in grasshopper is a little complicated than the typical immutable objects generated in grasshopper. There are multiple relationships an object can have in IFC, not a simple parent-child in a one way hierarchy that GH wires represent. I’m still considering if a top-down or bottom up approach can work (particularly if details of GH2 emerge) but it’s part of the reason that the stream component is in isolation (without an upstream dependency on all components that generate the IFC data set).
I do see a reduction in the file with latest build, and I don’t identify any “obsolete” data that persists.
Sorry this isn’t more helpful, let me know your thoughts and observations.
Cheers,
Jon