EVENT NAME Epics Collaboration Meeting 2020 EVENT DATE: WEDNESDAY, 21 OCTOBER 2020 EVENT BY: HEINZ JUNKES ----------------------------------------------- EVENT CHAT [2020-10-20 16:05:20] [Mijail Todorovich] Is anyone here using or thinking about using EPICS in industry? [2020-10-20 16:25:39] [john runchey ] As a 6-month epics user here, that was an excellent intro [2020-10-20 16:29:35] [Henrik Carling] Mijail: Depending on the perspective there are already industrial commercial organisations working with EPICS, wither selling EPICS services or "EPICS ready" products - you will hear some of them today. From another perspective, there are quite few industrial organisations today using EPICS. My take on this is that many industries are conservative and anxious to take the step towards industry 4.0, IoT etc. This is why we want to increase the interaction with industry to show this - already existing, powerful, stable, flexible etc - system that can be used also in industrial contexts. [2020-10-20 16:29:43] [Henrik Carling] John: Thanks :-) [2020-10-20 16:30:25] [Markus Kristensson] Was unfortunately late for your talk, will listen to it later! [2020-10-20 16:31:26] [Mijail Todorovich] Thank you, Henrik [2020-10-20 16:40:16] [Kacper Matuszynski] This info about LA Airport made me thinking, do you know if they use it in space? i.e. ISS [2020-10-20 16:41:40] [Ralph Lange] Aerospace needs a *ton* of expensive certifications. [2020-10-20 16:41:59] [Andrew Johnson] EPICS is not used in space as far as we know, that would require a whole different level of testing. You also probably wouldn't want to use generic software like EPICS in such a specific application. [2020-10-20 16:42:21] [Kacper Matuszynski] thanks Ralph, thanks Andrew [2020-10-20 16:43:46] [Henrik Carling] To add a bit of twist, EPICS is widely used in astronomical observatories - so in space, kind of :-) [2020-10-20 16:44:08] [john runchey ] When you say an IOC can be created anywhere and brought onto the network, how are PV name collisions handled? Need to be co-ordinated up front during development? [2020-10-20 16:51:19] [Andrew Johnson] Within a naming domain (generally an IP subnet unless you have additional configuration) you have to decide how you're going to name things across the whole system. If 2 IOCs have records with the same name, clients will connect to the first IOC that responds. [2020-10-20 16:51:52] [Ernest Williams] Awesome talk !! [2020-10-20 16:51:57] [vishnu] Epics for Realtime application [2020-10-20 16:52:17] [Kacper Matuszynski] Hi Bob, that was really great to hear the talk directly from you. [2020-10-20 16:53:09] [Henrik Carling] John: At ESS we use a software tool that generates names for use depending on the context. Usually, when you do a large scale implementation, you need to create a naming convention/policy before you start implementing the control system. So, yes - up-front coordination is needed. [2020-10-20 16:54:45] [Johan Holman] Will the talks be recorded and available afterwards? [2020-10-20 16:55:27] [Prajjwal] @Johan, I heard yesterday that they would be [2020-10-20 16:57:28] [Heinz Junkes] we make the records available at the indico site [2020-10-20 16:57:36] [Andrew Johnson] The talks are being recorded and published, it's the middle of the night in Asia/Pacific where EPICS is also used (Japan, China, Korea, Australia etc.) and we don't want to exclude those users. [2020-10-20 17:07:22] [norihiko kamikubota (kek)] Hi Andrew, I am still awake in Tsukuba Japan. I will tell our members that talks are recorded and will be accessible. Thanks d(^ ^)/ [2020-10-20 17:12:29] [Andrew Johnson] @Kamikubota san: Heinz and Patrick are also providing Q&A sessions at 5pm your time for any questions, but the speakers won't be present then. [2020-10-20 17:14:16] [Freddie Akeroyd] John: most IOCs are developed with the first part of the PV name containing a macro, which can be easily set at ioc start time for example. So you need to coordinate on overall naming as others have mentioned, but it is usually easy to force PV names to be overall unique however the internal records have been named. [2020-10-20 17:19:05] [Maren Hauschildt-Purves] @Andrew: there also were some contributions from Dirk and Ralph in the Q&A. [2020-10-20 17:25:12] [Hitesh Dhola] Hello Freddie [2020-10-20 17:28:47] [Hitesh Dhola] Good Presentation Rok [2020-10-20 17:36:04] [Freddie Akeroyd] Hello Hitesh :-) Hope all is well with you [2020-10-20 17:43:14] [Kazuro Furukawa] 😋😋😋 [2020-10-20 17:45:34] [Hitesh Dhola] Yes Freddie, All is well. Waiting for your presentation. [2020-10-20 17:47:12] [Hitesh Dhola] Then I can go to sleep. [2020-10-20 17:50:30] [Rok Sabjan] Thanks Hitesh! [2020-10-20 18:33:37] [Hitesh Dhola] Very Informative Presentation Freddie. Clearly Describes EPICS integration [2020-10-20 18:36:40] [Steven Hartman] Thanks, Freddie. Well done. [2020-10-20 18:39:10] [michael davidsaver] Freddie, were you referring to this? https://lewis.readthedocs.io/en/latest/index.html [2020-10-20 19:14:13] [Freddie Akeroyd] yes that is it, and our device emulators are available at https://github.com/ISISComputingGroup/EPICS-DeviceEmulator [2020-10-20 19:20:26] [Freddie Akeroyd] @Hitesh, @Steven - thank you. Not sure what happened with the microphone at the beginning, it was working at the pre-conference test and I was sure I hadn't touched the settings, but it seemed to need me to reselect it... -----------------------------------------------