What's the best method to make a collaborative development in WinCC OA?
We use client-server project with manually loading panels and scripts from client to server project.
Collaborative development in WinCC OA
- fmulder
- Posts:330
- Joined: Wed Feb 03, 2010 9:46 am
Re: Collaborative development in WinCC OA
Not a simple question, and not a simple answer.
In general, we use an SVN repository. One guy changes a panel, and all other guys do an SVN update. We also determine who is working on certain panels or scripts so that we do not collide and overwrite each others work. This works quite well for scripts and libraries.
We also add SVN version numbers to scripts and libraries and display them in LogViewer and an about box.
We use TortoiseSVn and do all checksins/updates from Windows. ( I do not like the SVN integraton in OA)
The database is something different. Basically each of us works on his own area and his own datapoints. He can then make an ascii import file that he commits. Others wil get the newer file when they update.
I'm working on a script to automatically import those DPL files because people will tend to forget them. I plan to compare the file time against the time at which the file was last imported.
Share the fun
Frenk Mulder
In general, we use an SVN repository. One guy changes a panel, and all other guys do an SVN update. We also determine who is working on certain panels or scripts so that we do not collide and overwrite each others work. This works quite well for scripts and libraries.
We also add SVN version numbers to scripts and libraries and display them in LogViewer and an about box.
We use TortoiseSVn and do all checksins/updates from Windows. ( I do not like the SVN integraton in OA)
The database is something different. Basically each of us works on his own area and his own datapoints. He can then make an ascii import file that he commits. Others wil get the newer file when they update.
I'm working on a script to automatically import those DPL files because people will tend to forget them. I plan to compare the file time against the time at which the file was last imported.
Share the fun
Frenk Mulder
- Piterskiy
- Posts:46
- Joined: Wed Jul 09, 2014 12:44 pm
Re: Collaborative development in WinCC OA
Thanks for your useful answer!
- xavi.aran
- Posts:63
- Joined: Thu Apr 21, 2011 1:21 pm
Re: Collaborative development in WinCC OA
We use the same method. We use SVN tortoise for panels, scripts, etc... for Database only one persons has the responsability of update and create datapoints or datatypes.
We test in local and later test un server with a Client.
Best Regards,
We test in local and later test un server with a Client.
Best Regards,
- fmulder
- Posts:330
- Joined: Wed Feb 03, 2010 9:46 am
Re: Collaborative development in WinCC OA
Let me add another very usefull tip.
We have several datapoint types that we use for our scripting. These are all datapoints that have no configs. For example:
- A datapoint per user interface that we use for navigation
- a datapoint with notes
- a datapoint with tasks
we use a scrpit function to create or alter these datapoint types. This means:
- developer A implements the script library for the 'Notes' functionality
- This script wil automatically create or alter the datapoint type that it needs
- developer A does a 'commit'
- Developer B does an update and gets the new library
- His datapoint type gets adjusted automatically when he starts his user interface
- This means that we do not need the Ascii manager and that your datapoint is always up-to-date
This method works very well for us
We have several datapoint types that we use for our scripting. These are all datapoints that have no configs. For example:
- A datapoint per user interface that we use for navigation
- a datapoint with notes
- a datapoint with tasks
we use a scrpit function to create or alter these datapoint types. This means:
- developer A implements the script library for the 'Notes' functionality
- This script wil automatically create or alter the datapoint type that it needs
- developer A does a 'commit'
- Developer B does an update and gets the new library
- His datapoint type gets adjusted automatically when he starts his user interface
- This means that we do not need the Ascii manager and that your datapoint is always up-to-date
This method works very well for us