On 17/07/16 03:40, oldk1331 wrote:
I would have to look at github bugtracker to form an opinion.
I particular what is better in github compared to sourceforge?
Hi Waldek, I can see that you are a hardcore oldschool guy :) GitHub is the toy
for young people nowadays. (I can't believe you have 0 experience with github?)
GitHub's advantages:
The biggest advantege of github is its popularity. 100 times popular
than sourceforge
I think. SymPy, Julia, Rust use it, handle 10000 bugs without
problem. Even LLVM
is discussing about github recently.
Github bug tracker also has email support. I'm not sure if it can be
integrated with
mailling list.
I guess we are only talking about the bug tracker here but I suspect
this is all tied in to the wider git philosophy? (code stored in
differential form???).
For me the important question is how do you want to work and then what
tools support that is a secondary question.
For instance, I use git/github but I don't really use it in the way that
it is intended. So when I use it in this way there is little difference
between using git and sf.
I may be working on a new capability for my code but while I am doing
that I may find a bug fix or correct a comment or tidy up some code.
Often any changes will be a mixture of these things and would be very
hard to categorise. So I just make changes as I go and upload to github
after every session.
From what oldk and Ralf say I gather that I should be switching
environments (rebase) every time I find a new issue. I can see the
advantages of this and it would (sort of) track the dependencies.
However there seem to be a lot of overheads associated with this way of
working, I'm sure I will keep forgetting to change environments and I
suspect I would have to spend a lot of time resolving conflicts
internally as well as with wider FriCAS code.
For oldk and Ralf this seems to have become the natural way of working
and I'm sure at that stage the overheads start to diminish and the
advantages become apparent, you seem to have found a way of working and
a set of tools that work for you.
So my questions are:
* Are you really only talking about the bug tracker? Would the real
advantages come from changing the wider management of bugs and features?
* Is this the direction that all potential contributors would have to move?
* Would there be any guidance about which things would be split out into
feature sets (degree of granularity and so on)?
I must admit, given that I have a finite set of brain cells, I would
rather use them to try to understand cohomology but if this is the
direction that the FriCAS project wanted to move I would go along with it.
Martin B
--
You received this message because you are subscribed to the Google Groups "FriCAS -
computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/fricas-devel.
For more options, visit https://groups.google.com/d/optout.