The Global ARCHICAD Community

Stay informed. Get help. Share your knowledge.

Discussions about using GRAPHISOFT's tools (API DevKit) for independent software developers

Moderators: ejrolon, Barry Kelly, Karl Ottenstein, LaszloNagy, gkmethy, MOREH Tamas, Akos Somorjai, Ed Brown, Mihály Palenik, Tibor Lorántfy, rmasaki

#308622
The short answer is no.

The Python experiment only returns the main structure of a limited number of elements such as walls, morphs, arcs, text, and zones and it does not return the needed auxiliary structures associated with them. So somewhat worthless at present. I hope to see a great improvement in version 24 but we'll see?

Keep in mind that the original Python Experiment was only issued to judge user acceptance and not as a practical application. It remains to be seen if Graphisoft will allocate the resources necessary to fully implement the API in Python. At present, it appears NOT? That's why it is important to convey your wishes directly to your local representative. Otherwise it may be 10 years before Python in Archicad is practical.
#308677
poco2013 wrote:
Mon Mar 23, 2020 12:40 pm
The short answer is no.

The Python experiment only returns the main structure of a limited number of elements such as walls, morphs, arcs, text, and zones and it does not return the needed auxiliary structures associated with them. So somewhat worthless at present. I hope to see a great improvement in version 24 but we'll see?

Keep in mind that the original Python Experiment was only issued to judge user acceptance and not as a practical application. It remains to be seen if Graphisoft will allocate the resources necessary to fully implement the API in Python. At present, it appears NOT? That's why it is important to convey your wishes directly to your local representative. Otherwise it may be 10 years before Python in Archicad is practical.
Thank you, Poco2013. I hope that we will have these functions in the next version.
#308698
Ralph Wessel wrote:
Tue Mar 24, 2020 10:03 am
Alternatively, you can get all this info through the ARCHICAD API.
Absolutely. And that may still be a good idea to explore since even with the new Python version, a thorough understanding of the Archicad data structure(s) will invariably be necessary.

However, I think the intent of the OP was to question if the Python experiment will ever have any practical value as is yet to be implemented and if it is worth bothering about? Python’s obvious advantage is its ease of use and speed in programing. Particularity in comparison to Graphisofts’ C++ API which is overly cumbersome and complicated. The downside is, of course, again speed. A ‘long used’ saying is that interpreted Python is 5-10 times slower than C++. So, I don’t think CadImage has much to worry about. But, for short programs, a delay of a few seconds is tolerable
.

IMHO, the Python experiment is promising but not acceptable, at present, since it was demonstrated in the first Beta to have significant limitations and several “killer” bugs which has yet to be acknowledged by Graphisoft or even to accept input on.

Hopefully Graphisoft is working on making a viable product feature and not just some marketing “candy”.