Arthur Pemberton wrote:
On Wed, Feb 25, 2009 at 7:37 AM, <makkalot@xxxxxxxxx> wrote:
On Wednesday 25 February 2009 05:43:16 am Arthur Pemberton wrote:
Hi,
And actually implemented it better than I had in mind.
My name is Arthur Pemberton, and I was just browsing through the
Google Summer of Code 2008. And it looks a lot like this idea that I
had in mind, except that i did actually start coding it, except for
starting a file tree [1].
?
So I am curious as to if working on this to kind of merge in my
(incomplete) idea would be a welcomed GSoC project.
So here is the idea I had:
First of all, I named my idea LiCo (short for Linux Commander). The
idea was to make point and click admin types comfortable with managing
groups of machines (so forgive any proceeding Windows analogies). I
was going to create a UI, the only part I actually started, with PyQt
and use python over SSH to command n number of Linux machines to do
simple operations: 'update machine', 'sync time', 'copy remote file'
Func uses XML-RPC over SSL which is kind of better than ssh because it already
does all the serialization things between end points. There is already a UI
(web-based) FuncWeb written in Python and uses Func API directly.
Definitely, doing it Func's way bypasses a lot of problems as far as I can see.
However a
UI written in PyQT maybe interesting also u can help improving the current
one.
Agreed.
Not agreed. Most folks don't want a local client installed to admin
applications.
If you do find this interesting, Python-GTK would probably be a nicer
option as some of us actually use PyGTK but not QT.
Also u can add new modules to Func as u said above 'update
machine', 'sync time', 'copy remote file' (i think we have them) if u want to
make some contribution.
Agreed
We already have modules for interacting with yum and copying files.
Current modules are mostly written for fedora,redhat
machines having different modules for other platforms is also and interesting
idea.
I can try, however, I am a Fedora/RehHat/CentOS user myself.
I can at least look into if its feasible, however the current approach
seems to fulfill all the needs.
Personally I wouldn't look at this unless you had a compelling reason to
do so.
Anyways, if the devs are interested and willing to mentor this as a
GSoC project let me know. If not I'll be looking through your code and
see how things are done as 'LiCo' was to be a long term learning
project for me -- so your code would be very educational.
I'm not a core developer but last year i was in GSOC2008 and spent some great
time with Func and its developers so, it maybe better to contribute to Func
project instead of reinventing the things ...
I like to code, but I like to reuse code to get things done even more,
so I would like to contribute either way.
I will get a formal proposal written for the sake of GSoC. In the even
that is accepted, I'll spend the summer playing with the code.
GSOC proposals are open already?
I don't think we've even put call for ideas up this year.
To be clear, I'm probably /not/ mentoring Func for GSOC this year -- I
will probably have a slot open for Cobbler. Ask alikins.
So we'll have the slot for 1 student at most -- there's also some rumor
that Google, due to 'economic factors' may have less GSOC slots for
certain orgs
this year, which I hope isn't the case. Something to be aware of though.
-
_______________________________________________
Func-list mailing list
Func-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/func-list
