On Thu, Jul 5, 2012 at 6:22 PM, Cunningham, Robert < [email protected]> wrote:
> I had started thinking along similar lines, and I just realized that git > seems to be becoming a vital utility for Fossil. The export/import cycle > appears to be useful in multiple scenarios encountered by Fossil users. > Yes, but to qualify that a bit: this applies to a vocal minority of users (the not-fully-satisfied users are almost always the loudest). When i worked in retail (many years ago) the statistic which was drilled into us time and time again was: on average, each happy customer tells 3 friends, whereas each unhappy one tells 7. i.e. for each git user complaining about imperfections in the import/export process, there are 2-3 users who are presumably either "satisfied enough" with it or not using it (so they're satisfied with it by default ;). In 4.5 years of use, i've never once needed any sort of import/export features for fossil. They are intended as a migration path either in or out of fossil, and are certainly not intended to be used as a pipeline back and forth for a single repository as part of a normal workflow. That's downright masochistic, IMO, a trait the git users might have inadvertently picked up from using git too long. > There has also been much discussion about adding git-like features to > Fossil. > Lately it's mostly been eye-candy stuff like the timeline in CLI mode. Stash was, AFAIK, "copied" from git, bit that's just a case of good old fashioned "feature migration" (it is a nice feature, regardless of where it came from), just like WIMP[2]-style GUIs we are all using right now (on all sorts of OSes). Regardless of where the idea comes from, if it fits nicely and adds value then it's potentially a candidate for "migration" to other environments. **** > > Would it be useful to think of Fossil becoming a git fork, where the core > git functionality (at the capability level) is sandwiched between a > low-level database and the high-level Fossil UI? > > ** ** > > If Fossil and git get any closer, they’ll be family. > Yes, they are conceptually family, but family members fight amongst themselves more often than with anyone else ;). - Immutable vs. mutable history are fundamentally opposed world views which cannot exist in the same space[1]. Wherever one of those policies is in place, the other cannot possibly be. - Fossil is designed for small, tightly-controlled teams and does not support the "waterfall" or "multi-level octopus" modes of git. e.g. fossil does not support having a chain of repositories leading multiple steps away from the original, forming a chain of "lieutenants." NOT that git's model is bad, it's just different and implies different dynamics. - Fossil is _easy_ to use. i have never once managed to hose a fossil repo (except that one time a stray global text replace corrupted my _FOSSIL_ file, but that was my own fault), but at least half the time i get write access to a git repo, i somehow end up getting it into an unusable state. A true story: 19 months ago i had one of the libgit2 maintainers at my desk for over an hour trying to figure out what the heck i had done to his repository (a cross-company project we worked on together) and how to straighten it out. **** > > Thoughts? > Let the flame wars begin ;). Just keep in mind that we ARE on the fossil list ;). [1] = http://www.w3.org/DesignIssues/Axioms.html "The assertion that the space of URIs is a universal space sometimes encounters opposition from those who feel there should not be one universal space. These people need not oppose the concept because it is not of a single universal space: Indeed, the fact that URIs form universal space does not prevent anyone else from forming their own universal space, which of course by definition would be able to envelop within it as a subset the universal URI space." (i really want that on a T-shirt.) [2] = WIMP = Windows, Icons, Menus, and Pointers -- ----- stephan beal http://wanderinghorse.net/home/stephan/ http://gplus.to/sgbeal
_______________________________________________ fossil-users mailing list [email protected] http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
