Why R should never move to git

classic Classic list List threaded Threaded
34 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Duncan Murdoch-2
On 25/01/2018 7:44 AM, Joris Meys wrote:
> Hi Duncan,
>
> I can see that branch on your github. Remember that you have to reload
> the github page to see the latest additions to your repo. It doesn't do
> that automatically.

Thanks, that was the issue.

Duncan Murdoch

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Duncan Murdoch-2
In reply to this post by Gábor Csárdi
On 25/01/2018 7:44 AM, Gábor Csárdi wrote:

> On Thu, Jan 25, 2018 at 12:34 PM, Duncan Murdoch
> <[hidden email]> wrote:
> [...]
>> but that branch doesn't show up in the Github web site.
>
> It is right there:
> https://github.com/dmurdoch/manipulateWidget/branches
>
>> Any suggestions?
>
> Personally I would suggest to call it master, because it is just
> easier. Your master should correspond to the upstream master, and you
> can do your own stuff in other branches.

That makes sense, but I don't see a way to rename branches on Github.  I
did see a way to make it my default branch, but there's a scary warning:

"Are you sure?

Changing your default branch can have unintended consequences that can
affect new pull requests and clones."

To answer yes to this I would have to say "I understand, update the
default branch", and since I don't understand the consequences, I didn't
click it.

Duncan Murdoch

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Iñaki Úcar
2018-01-25 16:07 GMT+01:00 Duncan Murdoch <[hidden email]>:

> On 25/01/2018 7:44 AM, Gábor Csárdi wrote:
>>
>> On Thu, Jan 25, 2018 at 12:34 PM, Duncan Murdoch
>> <[hidden email]> wrote:
>> [...]
>>>
>>> but that branch doesn't show up in the Github web site.
>>
>>
>> It is right there:
>> https://github.com/dmurdoch/manipulateWidget/branches
>>
>>> Any suggestions?
>>
>>
>> Personally I would suggest to call it master, because it is just
>> easier. Your master should correspond to the upstream master, and you
>> can do your own stuff in other branches.
>
>
> That makes sense, but I don't see a way to rename branches on Github.  I did
> see a way to make it my default branch, but there's a scary warning:
>
> "Are you sure?
>
> Changing your default branch can have unintended consequences that can
> affect new pull requests and clones."
>
> To answer yes to this I would have to say "I understand, update the default
> branch", and since I don't understand the consequences, I didn't click it.

Changing the default branch just moves the HEAD pointer (on GitHub) to
another branch. It would be equivalent to ssh into GitHub's server, cd
into the repo, and enter "git checkout anotherbranch". So not a big
deal in principle, but anyway this doesn't make any difference in this
case.

Also note that the name is not important: "master" has no special
meaning as "trunk" does in svn. It's just a convention, no more.

The executive summary of this thread would be:

1. Keep the default branch synced with upstream (this is usually
"master", but it could be another), and use other branches to fix
things and submit PRs.
2. If you mistakenly committed and pushed something to it, no problem:
just fetch a fresh copy of that branch from upstream and call it by a
different name, and we're back in 1.
3. GitHub can't be blamed about how git works. ;-)

Iñaki

>
> Duncan Murdoch

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
If you ever need to document issues / coding recipes related to GIT / SVN:

* I could pick the commands from e-mails.
* Any documentation you send me.
* Or books (http://shop.oreilly.com/product/0636920022862.do), web pages,
etc..

And create a wiki / documentation page in any platform, in order to help.

Best,
Juan

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
I do not mind investing as much time as necessary :-)

> If you ever need to document issues / coding recipes related to GIT / SVN:
>
> * I could pick the commands from e-mails.
> * Any documentation you send me.
> * Or books (http://shop.oreilly.com/product/0636920022862.do), web pages,
etc..
>
> And create a wiki / documentation page in any platform, in order to help.
>
> Best,
> Juan

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
In reply to this post by Iñaki Úcar
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Suzen, Mehmet-3
In reply to this post by Duncan Murdoch-2
This might be off topic, but if R-core development ever moves to git,
I think it would make sense to have its own git service hosted by a
university, rather than using
github or gitlab. It is possible via https://gogs.io/ project.

Just for the record.

Best,
-m

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Gábor Csárdi
While this is a very hypothetical argument, you could at least explain
_why_ you would think so.

If you were thinking about the unlikely event of GitHub / GitLab
closing business, that is _not_ such a big to any active project that
is hosted there.

Gabor

On Tue, Jan 30, 2018 at 11:07 PM, Suzen, Mehmet <[hidden email]> wrote:

> This might be off topic, but if R-core development ever moves to git,
> I think it would make sense to have its own git service hosted by a
> university, rather than using
> github or gitlab. It is possible via https://gogs.io/ project.
>
> Just for the record.
>
> Best,
> -m
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Suzen, Mehmet-3
Gabor, I was just pointing out options. I think it is more of a policy
decision than a technical one. For example, the very mailing list we
are using is run by ETH Zurich with Martin Maechler. But it can well
be run on google groups. Maybe this list should also move to google
groups, it is unlikely that Google would shut down google groups soon.

Best,
-m

On 31 January 2018 at 00:26, Gábor Csárdi <[hidden email]> wrote:

> While this is a very hypothetical argument, you could at least explain
> _why_ you would think so.
>
> If you were thinking about the unlikely event of GitHub / GitLab
> closing business, that is _not_ such a big to any active project that
> is hosted there.
>
> Gabor
>
> On Tue, Jan 30, 2018 at 11:07 PM, Suzen, Mehmet <[hidden email]> wrote:
>> This might be off topic, but if R-core development ever moves to git,
>> I think it would make sense to have its own git service hosted by a
>> university, rather than using
>> github or gitlab. It is possible via https://gogs.io/ project.
>>
>> Just for the record.
>>
>> Best,
>> -m
>>
>> ______________________________________________
>> [hidden email] mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

barry rowlingson
In reply to this post by Duncan Murdoch-2
On Tue, Jan 30, 2018 at 11:07 PM, Suzen, Mehmet <[hidden email]>
wrote:

> This might be off topic, but if R-core development ever moves to git,
> I think it would make sense to have its own git service hosted by a
> university, rather than using
> github or gitlab. It is possible via https://gogs.io/ project.
>
> Just for the record.
>
>
Let the record also state that *gitlab* is an open source project and can
be downloaded and self-hosted, like gogs, but unlike github.

Barry

PS I've been running a gitlab instance for my group for a couple of years
on a private server.

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Suzen, Mehmet-3
On 31 January 2018 at 16:18, Barry Rowlingson
<[hidden email]> wrote:
>>
>
> Let the record also state that *gitlab* is an open source project and can be
> downloaded and self-hosted, like gogs, but unlike github.


Good to know. Nice one: https://github.com/gitlabhq/gitlabhq

Best,
-m


> PS I've been running a gitlab instance for my group for a couple of years on
> a private server.

Is it a smooth ride so far?

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
Yes, indeed Gitlab GUI Core Code is Open Source (Libre / Community
Edition): https://gitlab.com/gitlab-org/gitlab-ce

> But his instructions required command-line git, and my main claim is that Github is not good enough to do the kinds of things I want to do and R Core would need to do.
>
> My other claim is that git is too hard to use.

I'm sure that Git Command Line Recipe Documentation can solve this
issue, Gitlab, in particular, has a wiki in which this kind of issues
could be documented. Also Git cheat-sheets might prove useful. In
addition, any feature request could be done in Gitlab Issue Section
(See above), or if that does not still does not convince, other
options could be considered, such as Bitbucket
(https://bitbucket.org/), etc.

In addition, the Git Repository:
* Could be self-hosted in the University Servers (Just as SVN actually
is nowadays).
* Be accessed either by the Command Line or the Graphical User
Interface (As users prefer).

The main reason motivating the move to the GIT Repository, as said
before, is that it would to allow individual users or companies from
the R Consortium to do pull requests based on issues for improving
base R code.

Indeed, in some years from now I would like to help to improve base R
myself, maybe re-writing some parts of the code in C++, fixing bugs,
or who knows :)

Kind regards,
Juan Telleria

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
> So I created a branch within my fork, and committed the change there. But Github provides no way to create a pull request that only includes the new stuff!
> Every attempt I made would have included everything from both bug fixes.

I have been doing some tests, and I think that this issue can be
easily addressed with Bitbucket GUI (https://bitbucket.org/product),
which is free for Open Source Projects
(https://www.atlassian.com/software/views/open-source-license-request)
and seems easy to use (So it follows the K.I.S.S. principle).

Just another alternative instead of Gitlab for self-hosting... no worries.

Best,
Juan

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Reply | Threaded
Open this post in threaded view
|

Re: Why R should never move to git

Juan Telleria Ruiz de Aguirre
In reply to this post by Juan Telleria Ruiz de Aguirre
I attach the Github Flow for teams and projects with regular deployments:

https://guides.github.com/pdfs/githubflow-online.pdf

https://guides.github.com/introduction/flow/

Tips:
* Always!!!! Do pull requests based on branches (never on the master).
* Keep your Fork Synchronized with the Upstream Repository.

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
12