Date
1 - 3 of 3
Open Source OPC UA
Jan Kiszka
Hi all,
does anyone already have some experience with OSS OPC UA implementations, specifically for resource constrained devices? We stumbled over http://open62541.org which seems like being on the right path. There are more https://github.com/open62541/open62541/wiki/List-of-Open-Source-OPC-UA-Implementations but not all look that attractive for embedded devices, due to language, license or main drivers. I think, eventually, CIP should look into the topic free and standard-conforming OPC UA as well because it is gaining a lot of traction as THE protocol for data exchange in automation industry. Regards, Jan |
|
Wolfgang Mauerer <wm@...>
Am 29/11/2016 um 20:36 schrieb Jan Kiszka:
does anyone already have some experience with OSS OPC UAthere's another tiny implementation from some of the Ostwestfalen-Lippe (OWL) group of people in Germany, but since it targets devices with only 10s of KiB of RAM and ROM, I don't think it's suitable for our purposes. As for language and main drivers: Do we have an agreement or an idea in the group what suitable languages are, and what characteristics main drivers should have? Considering the size of systems we're targeting, I personally would not see C (and possibly C++) as the only admissible language. Best regards, Wolfgang
|
|
Jan Kiszka
On 2016-11-30 11:41, Wolfgang Mauerer wrote:
Am 29/11/2016 um 20:36 schrieb Jan Kiszka:If you have a link (and the project has a proper license), maybe send adoes anyone already have some experience with OSS OPC UAthere's another tiny implementation from some of the Ostwestfalen-Lippe PR to the open62541.org to extend their list. If we only had higher languages, that would be a major limitation. It's not so much about their size, more about their amount of dependencies and their suitability for embedded processors (just having "fun" with Go on Intel Quark...). But also having bindings for Java, C#, python, JavaScript, Go, you-name-them, would be valuable. So I would recommend to start with a common core - thus a native one - and add those bindings to them. Later on, if there is added value, we could also consider alternative core implementations. Regards, Jan
|
|