OpenSees is way too big and huge for simple applications. I guess a lot of times 90% of the features are actually not used by a certain user.
At this time, it is only possible to get the whole package if you want any piece of it. But wouldn't it be nice if we can have a way to customize the software to our needs and adding the new packages when we need it?
I've seen this in MikTex. It will download and install the missing package automatically.
Of course customizing the source code is way too difficult (or maybe not?) But I just wish we can have that nice feature now.
How about a customizable software architecture?
Moderators: silvia, selimgunay, Moderators
YES!
Our initial contributions to OpenSees were done as separate libraries (and they still are, check the UCD Makefile.def...).
I am actually doing just that in last few weeks, got sidetracked by some other business but will be back on it next week.
I also envisioned a simple GUI tools to make you choose what elements, material models, solvers... one wants inside...
Let me know if you want to collaborate on this, I've been talking to Frank and Greg about this for quite some time, need to do something about it very soon for a number of reasons (smaller code is much faster as per som erecent papers, some of the code in OpenSees is (probably) buggy, see reports at
http://geowulf.engr.ucdavis.edu/~jeremi ... n/?C=M&O=D
. . .
There are some computer tools, as it turns out, one probably actually uses less than 2-5% of all code for simple runs...
Best regards, Boris
I am actually doing just that in last few weeks, got sidetracked by some other business but will be back on it next week.
I also envisioned a simple GUI tools to make you choose what elements, material models, solvers... one wants inside...
Let me know if you want to collaborate on this, I've been talking to Frank and Greg about this for quite some time, need to do something about it very soon for a number of reasons (smaller code is much faster as per som erecent papers, some of the code in OpenSees is (probably) buggy, see reports at
http://geowulf.engr.ucdavis.edu/~jeremi ... n/?C=M&O=D
. . .
There are some computer tools, as it turns out, one probably actually uses less than 2-5% of all code for simple runs...
Best regards, Boris