Re: [Texmacs-dev] git & texmacs

[Massimiliano Gubinelli]

> On 19. May 2021, at 23:00, Giovanni Piredda <pireddag@posteo.de> wrote:
> 
> 
> Am 19.05.2021 um 10:58 schrieb Massimiliano Gubinelli:
>> it adds some background to our discussion of moving some of the development 
>> of TeXmacs to a git-based workflow, possibly on github.
>> 
>> Max
> 
> The development of free software on github leaves me uneasy.
> 
> I see the draw that the platform has on many people and the consequent 
> benefit for each project (general feeling in addition to the graph in the 
> first post of the linked ocaml discussion thread); on the other hand it is 
> delegating part of the organization of the project to a commercial enterprise.
> 
> Mixing of things done "for free" and commercial work can be seen as healthy 
> (e.g. GNU license)---I by and large agree with that; I do not like 
> integration and delegation. The very great success of delegation-integration 
> (see e.g. Q&A sites and social networks) brings consequent benefits for each 
> user (organized access to the knowledge and the energy of others), but I am 
> not yet convinced that one should follow up.
> 

you are right and this considerations has indeed to be done. I do not have 
strong opinions about it but moving all the development to github it is maybe 
something we do not want to do.

We could keep the project on savannah (which can use either git or svn) and 
continue to use the savannah's bug tracker. I'm not sure savannah accepts pull 
requests from outsiders, we should check. If PR on savannah are not possible 
then we could still have a presence on github to collect PR and propagate them 
to savannah. In this way we remain essentially independent from github.

It should also be mentioned that with a distributed system like git we have 
less risk of commit to a remote server, indeed on git the "blessed" copy of the 
software is the one on Joris' computer :) 

> On the technical benefit of using git instead of the current svn I do not 
> know anything and I cannot comment.
> 

One main difference is that svn is centralized and git is decentralized: 
everybody possess a full copy of the repository, including all the history.

> I agree with Max that having contributions marked with the name of their 
> authors helps develop the community. Publishing each contribution with 
> authors' names is also standard in science; in the case of free software, 
> people would be able to list their contributions in their CVs.
> 

This aspect is what is more important to me right now. Otherwise the current 
state of affairs is as good as any other alternatives. The problem of svn is 
that we loose track of the contributions from people which are not yet in the 
organisation. 

I do not have hard data to support my impressions, so I would like to hear also 
from others what is their opinion and the advantages/risk of moving to a 
different version control system.

Max


> Giovanni
> 
> 
> _______________________________________________
> Texmacs-dev mailing list
> Texmacs-dev@gnu.org
> https://lists.gnu.org/mailman/listinfo/texmacs-dev