[darcs-users] RFC: GSoC proposal: Distributed Issue Tracking in Darcsden

[Kẏra]

Max Battcher <admin <at> worldmaker.net> writes:

> Certainly from my perspective it makes sense for darcs to tackle some of
the more academic projects and crazy feature ideas solely because it is just
about the only source control system to be in
>  a good position to do so. 
> As to this particular proposal, I would be interested to see what becomes
of it.
> If I may offer suggestions, I think this would be a great opportunity to
also bring in work from elsewhere in the theoretical space (such as some of
the hubbub from the Operational Transforms world).
> In fact, I would suggest that maybe a better prototype than just using
normal file merging in darcs would be to build a real-time example using
JSON objects and built on something like share.js (http://sharejs.org/):
then extrapolate down into how darcs would
>  handle that flow. Maybe there might be some interesting lessons from
libraries like share.js that could apply to darcs. Maybe the JSON-based
approach of web OT libraries might provide insight into the possibilities of
more generic document stores for darcs...
> Sent from my wireless telegraph. Full stop.
> 
> 
> From:
> Eric KowSent:
> ‎3/‎21/‎2014 11:38To:
> VikramanCc:
> Darcs mailing listSubject:
> Re: [darcs-users] RFC: GSoC proposal: Distributed Issue Tracking in Darcsden
> 
> 
> Hi Vikraman,
> (I'm afraid I don't have feedback for you specifically, but just
> general musing on GSoC)
> Exploratory proposals like this are cool in that they explicitly put
> Darcs in the interesting-research realm of version control (aside from
> being a usable system with a friendly UI). We have some experience
> doing exploratory projects (Petr's primitive patches 3 work)
> From a practical standpoint, I'd be a bit nervous about whether or not
> we could deliver an actually usable issue tracker, and whether we
> would want to maintain something like a darcs issues in the codebase;
> on the other hand, part of me thinks that it'd be good for Darcs to
> wholeheartedly embrace its Out-Thereness and dive into projects like
> this.  For example maybe the work on trying to fit an alternative set
> of patch commutation rules would force us to discover lots more
> interesting holes in patch theory (or at least force us to reorganise,
> refactor, clean up the Darcs library even more).
> So what do people think about exploratory GSoC projects?
> All other things being equal, how do we want to allocate things?
> Should we focus on short-term achievable bits of progress (to darcs
> and the ecosystem in general).
> Or should we develop a general project framework for letting our freak
flag fly?
> On 21 March 2014 14:44, Vikraman <vikraman.choudhury <at> gmail.com> wrote:
> > Hello everyone,
> >
> > I am proposing a GSoC project with darcs. The idea is to improve the
> > issue tracker in darcsden, by implementing distributed issue tracking
> > using patch theory. Any feedback is welcome!
> >
> > You can read my proposal in markdown here[0], and a pdf version here[1].
> >
> > [0]http://hub.darcs.net/vikraman/darcs-gsoc-proposal/browse/proposal.md
> > [1]http://dev.gentoo.org/~vikraman/misc/darcs-gsoc-proposal.pdf
> >
> > --
> > Vikraman

I personally would love to see this built into darcs:
http://bugs.darcs.net/issue2363

Github is popular due to its social features and unilaterally providing
everything a free software project would need (wiki, issue tracker,
etc). The problem with github is that it is a centralized service as a
software substitute (SaaSS) and largely proprietary. 

There are distributed issue trackers like Simple Defects:
http://syncwith.us/sd/

There are also DVCS' with built in distributed issue-tracking, wiki, and
web-interface such as fossil: https://www.fossil-scm.org/

If darcs could work on implementing this, it would be poised to become a
free software and distributed alternative to github. The wiki itself
could easily be powered by darcs, and there should be a way for the
issue tracker to be as well.