Re: Basic EPICS bootstrap (rant)

[Simon Rose via Tech-talk <tech-talk@aps.anl.gov>]

I think that some of the reason that I suspect most people run into is the customisation of dependencies in EPICS. For example, does asyn depend on calc? sequencer?

 

A lot of this ultimately depends on your specific needs, and there isn’t necessarily a one-size-fits-all dependency graph.

 

On top of that, there is the question: who maintains all of this extra packaging setup? 😃

 

Cheers,

 

Simon

 

From: Tech-talk <tech-talk-bounces@aps.anl.gov> on behalf of Marco Filho via Tech-talk <tech-talk@aps.anl.gov>
Reply to: Marco Filho <marco.filho@ess.eu>
Date: Wednesday, 8 October 2025 at 08:16
To: Graeme Winter <gw363@cornell.edu>, Han Lee <jeonglee@lbl.gov>
Cc: "tech-talk@aps.anl.gov" <tech-talk@aps.anl.gov>
Subject: Re: Basic EPICS bootstrap (rant)

 

Hi!

Might be a silly question, but expanding on this I wonder why is there no .deb package for EPICS? Is there any particular reason or did we just steer as a community towards other packaging methods?

Best regards,

Marco

---

From: Tech-talk <tech-talk-bounces@aps.anl.gov> on behalf of Han Lee via Tech-talk <tech-talk@aps.anl.gov>
Sent: Wednesday, October 8, 2025 07:43
To: Graeme Winter <gw363@cornell.edu>
Cc: tech-talk@aps.anl.gov <tech-talk@aps.anl.gov>
Subject: Re: Basic EPICS bootstrap (rant)

 

Hi Graeme,

 

Sorry for the duplicate email, I realized the links were broken.

I've put together a simple example that might be helpful. Please check the following repository:
https://github.com/jeonghanlee/EPICS-env-distribution

You will still need to install the correct library packages. In this case, you can find them here:
https://github.com/jeonghanlee/pkg_automation

If the pre-built binaries do not support your "arm64" architecture, you can build them yourself using the instructions at:
https://github.com/jeonghanlee/EPICS-env

I hope this helps!

Best,
Han

P.S. You might also find a slightly older training resource on organizing EPICS IOC environments here:
https://jeonghanlee.github.io/epics-trainings/

 

On Tue, Oct 7, 2025 at 4:35 PM Han Lee <jeonglee@lbl.gov> wrote:

Hi Graeme,

 

I've put together a simple example that might be helpful. Please check the following repository:

https://github.com/jeonghanlee/EPICS-env-distribution

 

You will still need to install the correct library packages. In this case, you can find them here:

https://github.com/jeonghanlee/pkg_automation

 

If the pre-built binaries do not support your "arm64" architecture, you can build them yourself using the instructions at:

https://github.com/jeonghanlee/EPICS-env

 

I hope this helps!

 

Best,

Han

 

P.S. You might also find a slightly older training resource on organizing EPICS IOC environments here: https://jeonghanlee.github.io/epics-trainings/

 

On Tue, Oct 7, 2025 at 11:56 AM Graeme Winter via Tech-talk <tech-talk@aps.anl.gov> wrote:

Good afternoon

 

Working through 

 

Getting Started epics-controls.org

 

 

It really feels like this is something which could be made _much_ easier by writing a kind of bootstrap script but also by the configure process being able to pick up stuff from the environment rather than having to hard-code every location multiple times in every configuration file (where the defaults are of course where the author does _their_ epics development) - the need for them to all correctly cross-reference is also an irritation

 

Before I scrape another few mm off my already suffering sanity - 

 

- _has_ anyone got a bootstrap script which will clone / install all the necessary / regular support modules

- am I holding it wrong / reading the wrong getting started guide

 

I note I can conda install epics-base but this _only_ gives epics base

 

That rant to one side, also, does anyone know how to address a missing symbol issue like

 

Undefined symbols for architecture arm64:

  "_pvar_func_vxi11RegisterCommands", referenced from:

      _cryo1_registerRecordDeviceDriver in cryo1_registerRecordDeviceDriver.o

ld: symbol(s) not found for architecture arm64

 

The symbol is indeed missing and would appear to be from asyn/vxi11/drvVxi11.c but does not appear to be built

 

This got me thinking I should re-do the NN steps to get back to here which then got me wondering about process itself…

 

Thanks Graeme

--

Jeong Han Lee, Dr.rer.nat.

Staff Scientist and Engineer

Lawrence Berkeley National Laboratory

https://orcid.org/0000-0002-1699-2660

--

Jeong Han Lee, Dr.rer.nat.

Staff Scientist and Engineer

Lawrence Berkeley National Laboratory

https://orcid.org/0000-0002-1699-2660