Experimental Physics and
Industrial Control System

Mark Rivers via Tech-talk <[email protected]> · Fri, 10 Jan 2020 00:21:21 +0000

> There had not been a new release of synApps for a long time.
> I think synApps was stuck on 5.8 for over a year.  I felt like I had to upgrade half of the modules in synApps 5.8.  
> But again, because I didn't document exactly what I changed, I can't give you a very good answer.

I don't find that using synApps releases is very useful.  As you say the latest synApps release is typically lacking some upgrade or bug fix you want, whether that is in asyn, motor, etc.

I find that using the latest tagged version of every synApps module at any point in time works well, with minimal problems with backwards compatibility.  Yes, there are rare times where your startup scripts might need to change but in general I think the EPICS developers do a very good job of maintaining backwards compatibility whenever possible.

Mark

-----Original Message-----
From: J. Lewis Muir <[email protected]> 
Sent: Thursday, January 9, 2020 6:04 PM
To: Mark Rivers <[email protected]>
Cc: Michael Davidsaver <[email protected]>; [email protected]
Subject: Re: EPICS release series after 7.0: 7.1 or 8.0? (ABI)

On 01/09, Mark Rivers wrote:
> > If EPICS and all of the EPICS modules cared about API and ABI 
> > compatibility, I suspect they would not break it as often, and I 
> > think this would result in an easier time upgrading EPICS and EPICS 
> > modules in that I wouldn't have to change as much source code for 
> > API changes,
> 
> Can you tell me the last time you had to change your source code for an API change?  

It was last April when I had to change my source code that worked with EPICS 3.14 to work with EPICS 3.15 because EPICS 3.15 renamed the errlogRemoveListener function to errlogRemoveListeners and changed its signature.

This isn't a particularly useful data point, though, because it was for a user.  I've remained on the same snapshot of EPICS and EPICS modules for quite some time, so I haven't attempted an upgrade for a while and hence have not been in a position where I might discover incompatibilities.

> Was that a change you think could be been avoided, and how?

Probably, by providing a backward compatible function, but I'd have to look at the change to be sure.

> > and I would have a better chance of avoiding so-called "dependency hell."
> 
> Please describe the last time you had a "dependency hell" with EPICS.

It was probably in 2016 when I was upgrading modules in synApps to make things work with the new quadEM driver for the CAENels TetrAMM, but I'm not sure.  There had not been a new release of synApps for a long time.
I think synApps was stuck on 5.8 for over a year.  I felt like I had to upgrade half of the modules in synApps 5.8.  But again, because I didn't document exactly what I changed, I can't give you a very good answer.

Lewis

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System