My work flow has always to push directly.At one time that was the standard way 
to do it.
I have been in this project for 20 odd years.

I have done a few pr for ease of people to look at or test with. I usually push 
these direct after but not always.

There has been no formal change to prs only. From memory it was discussed but 
never agreed on.

In this case push direct or pr would not of changed what happened. I would have 
rebased and force pushed to pr and then merged.

The real problem seems to be it was a surprise that I was going to push my own 
case. Which as i state is beem standard forever and noone has complained.

In my mind why wouldn't  I ? I'd  fixed all issues with the only person who 
wanted changes. I said I was read to put this code in. There were mo conflicts. 
I still don't know exactly what the problem with the branch is.

Now I would be more direct about my intension to push next time for sure, so 
I'll own that.

The reason I did a pr is so other could look at it. Hans and I went through 
quite a few changes and compromise. I thought the process worked well.

The issue seems to be conflicts in code in gmoccapy. There are going to be 
conflicts if we are working on the same code. I have fixed conflicts many times 
in this process. This is the process in a healthy project.

Then you guys 'beat' me up on github, pulled my code and locked everyone out of 
pushing with no formal discussion.

I take it from that, that if my code cause your code conflicts that is not 
allowed but if your code causes conflicts in mine that's fine.

In 23 years I think I've made 1 big mistake. It happens from time to time other 
have too.

Most of my work is in guis, particularly qtdragon .I dont want to have to find  
a buddy to push my work every time.
Sent from my Galaxy
Finally please dont think that since most of you do something, that that is now 
policy.
If you want to change policy let's discuss and try to agree and certainly no 
beat someone up for doing something that you could see  was normal for them.

Thanks Chris


-------- Original message --------
From: andy pugh <[email protected]>
Date: 2026-09-17 6:17 a.m. (GMT-08:00)
To: EMC developers <[email protected]>
Subject: Re: [Emc-developers] branch protection turned on

On Thu, 17 Sept 2026 at 13:44, Luca Toniolo <[email protected]> wrote:

> I ask because most of my direct pushes have been minor doc fixes and such, 
> where a PR is mostly clicking through ceremony.

Do you push directly to master? Or via a PR? It's perfectly possible
to create a PR and then merge it yourself for trivial /
uncontroversial changes.

Historically we had our own git server, and those devs that had access
pushed their changes directly. Those that did not have access would
have to email a patch to a core dev to be reviewed and (maybe) merged,
We then moved to GitHub and those same developers retained the ability
to push directly. This made things simpler and more transparent for
contributors without push access.

We haven't previously insisted on PRs, and I quite often push to wlo
without a PR.

I think it's probably better to go via a PR, mainly for consistency,
even if it is self-merged by the developer.

--
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1912


_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to