> I wasn't so much worried about "breaking" anything as I was about
> giving people the feeling of "bait and switch." Every time we decide
> to include an open source feature in Solaris, but we trim down the
> deliverables based on our own problems, we end up with (at least) bad
> press.
<sigh> Well, I don't think it's all that much more work, if any, so if
Brian's willing, we might as well.
If someone is willing to come on board to fully test, and sign off on the
GUI's functionality, and we can remove the external dependancies for the
commandline tools, I am willing to include this in this project.
Nevertheless, it probably makes sense to just throw in the GUI. The only
downside I can think of is that if you've chosen to minimize your system
by
removing X, or GNOME, then vim won't work anymore. I think that's a
shame,
but since I believe that including the GUI is inevitable, we'll just have
to deal with that. And perhaps work on making the GUI dynamic in the same
way that the language bindings ought to be.
Is there anyway to compile the apps so that vim is not dependant on the
graphical libraries, but gvim is? (Short of having two seperate build
trees) I ask this because we are talking about replacing vi with vim. My
feeling is that vi is the one editor that needs to work on every system
configuration, as it has been the standard editor for Solaris since the
SunOS days.
If nothing else, it answers the inevitable first question: "so, why
> aren't you just replacing 'vi' right now?"
Good point.
I feel that the amount of deliberation, testing and certification required
to replace vi is sufficiant enough to be a seperate project. There is also a
significant possiblity that code changes will be required, to add support
for Sun specific changes to vi. As I stated above, I feel that we need to
resolve the graphical library dependancy issues before making it the
standard. (vi, and view have no such dependancies. Is it ok to add new
depencies?) . (Personally I think we are underestimating the amount of work
it will take to sign of on replacing vi with vim). (Please let me know if I
am being to finicky. I just assumed changes to existing functionality
weren't made lightly. (vs adding new functionality)
As I have said earlier in the thread, if I am going to be the project team,
I need to break this integration into manageable pieces/projects, with
intermediate deliverables. (Or else we could be waiting a very long time for
any vim deliverables.)
IE:
1) Bring in CLI tools (As listed in the proposal draft)
2) Enable Graphical tools while still keeping the base binaries independent
from graphics frameworks.
3) Enable dynamic language bindings
4) Evaluate replacing vi with vim
5) Change Sun Studio to use system vim
_______________________________________________
tools-discuss mailing list
tools-discuss@opensolaris.org
