Veda2.0 Released!


Suppoer for Version control
#1
Hi

At the Danish Energy Agency we are using GIT-version control extensively in the use of the TIMES model. In the old VEDA version, we had to work the versioning outside the VEDA model folders, since the syncronisation watch in VEDA was too sensitive. (GIT only cares about content and not file dates).

Has the new VEDA built-in version control, or is it version-control aware in any way?  If not, is this perhaps scheduled to be included at some point?

Kind regards
Lars Termansen
Reply
#2
(05-10-2020, 04:39 PM)lbtermansen Wrote: Hi

At the Danish Energy Agency we are using GIT-version control extensively in the use of the TIMES model. In the old VEDA version, we had to work the versioning outside the VEDA model folders, since the syncronisation watch in VEDA was too sensitive. (GIT only cares about content and not file dates).

Has the new VEDA built-in version control, or is it version-control aware in any way?  If not, is this perhaps scheduled to be included at some point?

Kind regards
Lars Termansen


Hi Lars, with the new VEDA2.0 we are directly using the GitHub database for our models. I think is working well but maybe Amit can add more on this point. 
Regards,
Maurizio
Maurizio Gargiulo <br />
Reply
#3
Hi Mauri

So you are using git on the model files in the "veda_models" folder? And you keep all the excel-files under version control?

On a git related issue. Version control allows one to compare changes across commits and branches in repositories, to pinpoint problems or when issues have been introduced in a model. In essence excel files are not well suited for this kind of comparison - which originally was developed for text-based file types.

Would it be possible to allow VEDA2 to recognize flat text file formats (eg. csv-files) as a TIMES model, thus easily utilize the strengths of version control? I realize this would mean only one table per file (and no fancy excel formulas and features).

BTW, we currently try to make all the excel veda-tables automatically through R-scripting, where we do all data mangling and setting up model topography, then exporting to excel for VEDA to read. Thus ensuring that no funny errors appear from using fancy excel formulas or cross-workbook-links mess things up.

kind regards
Lars
Reply
#4
I have been using https://www.araxis.com/merge/index.en, which does a pretty good job of comparing Excel files. This can be installed as a comparison tool in git clients.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)