9:08 p.m.
Hi all,
I have created a draft[1] of the features and goals which I'd like EKG to
have. I went through the EKG archives a bit and tried to glean out whatever
has already been requested. I hope I have covered all of them. Many of the
ideas are inter-dependant and some may be outright non-viable however I
thought it better to get everything out in the open.
Also I'd like to ask what is the intended user for this tool. I have assumed
it is for anyone who wants to know about what's going on in a ML in a
nutshell- especially so for a new user to the community who'd like to know
how things work in the ML without wanting to browse through tons of
archives. It will also be beneficial for the "bosses" who want to know where
the ML(and the community) is heading.
Regards
R Dhushyanth
[1] https://fedoraproject.org/wiki/User:Dushyanth/draft-EKG
2:24 a.m.
On Tue, Mar 9, 2010 at 8:38 AM, Dhushyanth R <r.dushyanth(a)gmail.com> wrote:
I have created a draft[1] of the features and goals which I'd
like EKG to
have. I went through the EKG archives a bit and tried to glean out whatever
has already been requested. I hope I have covered all of them. Many of the
ideas are inter-dependant and some may be outright non-viable however I
thought it better to get everything out in the open.
Which are the ideas which you think are non-viable ? I couple of the
ideas on your page seem to have an overlap in terms of themes/aims
with Fedora Insight, but that could be cursory.
What would be interesting to know is what you think is achievable
within the GSoC timelines should this proposal/idea be accepted and,
what are the high priority features you intend to put in your
deliverable (ie. if these are not done, your GSoC application could be
deemed to be incomplete) ?
Does your proposal/idea overlap/depend upon any other existing piece
in Fedora Infrastructure ? If so, a statement about that would be good
to read.
Also I'd like to ask what is the intended user for this tool. I
have assumed
it is for anyone who wants to know about what's going on in a ML in a
nutshell- especially so for a new user to the community who'd like to know
how things work in the ML without wanting to browse through tons of
archives. It will also be beneficial for the "bosses" who want to know where
the ML(and the community) is heading.
I'd say that you have got it somewhat nailed in terms of intended user.
--
sankarshan mukhopadhyay
<http://sankarshan.randomink.org/blog>
11:54 a.m.
Hi,
On Tue, Mar 9, 2010 at 1:54 PM, sankarshan <foss.mailinglists(a)gmail.com> wrote:
On Tue, Mar 9, 2010 at 8:38 AM, Dhushyanth R <r.dushyanth(a)gmail.com> wrote:
> I have created a draft[1] of the features and goals which I'd like EKG to
> have. I went through the EKG archives a bit and tried to glean out whatever
> has already been requested. I hope I have covered all of them. Many of the
> ideas are inter-dependant and some may be outright non-viable however I
> thought it better to get everything out in the open.
Which are the ideas which you think are non-viable ? I couple of the
ideas on your page seem to have an overlap in terms of themes/aims
with Fedora Insight, but that could be cursory.
On the non-viable(more correctly: controversial) front, I was thinking
some of the information which can be gathered with EKG may intrude on
user privacy. eg. Tracking behaviour like active hours, free time, etc
may be a tad too much information. But then again once someone posts
it is "open for everyone". So I was hoping to know where to draw the
line
What would be interesting to know is what you think is achievable
within the GSoC timelines should this proposal/idea be accepted and,
what are the high priority features you intend to put in your
deliverable (ie. if these are not done, your GSoC application could be
deemed to be incomplete) ?
The initial data collecting part will be relatively easy. The tough
part will be testing to make sure that the results are truly
useful(and accurate representations). I was hoping to discuss this
more before jumping the gun. The risk of making it a "google weekend
of code" or a "google year of code" are quite large.
Does your proposal/idea overlap/depend upon any other existing piece
in Fedora Infrastructure ? If so, a statement about that would be good
to read.
Both Fedora Insight(FI) and EKG want to "to keep up with the goings on
of the community without following a dozen different blogs and
_mailing lists_ " but there is a difference: FI seeks to provide a
summary of contents available in different places in one place. EKG is
for plain simple statistics.
I do view EKG as a valuable tool useful for FI content writers. So I
would definitely like to know about more ideas which can help them.
I'll mail the fedora-marketing-list to know how EKG can do more for
them.
--
sankarshan mukhopadhyay
Regards
R Dhushyanth
2:54 p.m.
2010/3/9 Dhushyanth R <r.dushyanth(a)gmail.com>:
Hi,
On Tue, Mar 9, 2010 at 1:54 PM, sankarshan <foss.mailinglists(a)gmail.com> wrote:
>
> On Tue, Mar 9, 2010 at 8:38 AM, Dhushyanth R <r.dushyanth(a)gmail.com> wrote:
>
> > I have created a draft[1] of the features and goals which I'd like EKG to
> > have. I went through the EKG archives a bit and tried to glean out whatever
> > has already been requested. I hope I have covered all of them. Many of the
> > ideas are inter-dependant and some may be outright non-viable however I
> > thought it better to get everything out in the open.
>
> Which are the ideas which you think are non-viable ? I couple of the
> ideas on your page seem to have an overlap in terms of themes/aims
> with Fedora Insight, but that could be cursory.
On the non-viable(more correctly: controversial) front, I was thinking
some of the information which can be gathered with EKG may intrude on
user privacy. eg. Tracking behaviour like active hours, free time, etc
may be a tad too much information. But then again once someone posts
it is "open for everyone". So I was hoping to know where to draw the
line
You just said it. Once you post on the mailing lists that are public,
that info is public. Lack of privacy isn't necessarily a bad thing in
some respects either, provided it's done in a fair basis. It's a good
way of being transparent, so your boss knows what you're up to as much
as you do.
Violations of privacy come into play when you combine this with
something Fedora Infrastructure keeps private, say the IP address you
used to register your FAS account with. This is where you have an
expectation of privacy. It's also information that may be interesting
to some people, but for other reasons that may or may not be dubious.
Hence there's a reasonable expectation that it should remain private
across the board.
It's a tricky debate, but for the purposes of EKG, i wouldn't worry
about it too much. I'm always up for discussion about it, even so.
> What would be interesting to know is what you think is
achievable
> within the GSoC timelines should this proposal/idea be accepted and,
> what are the high priority features you intend to put in your
> deliverable (ie. if these are not done, your GSoC application could be
> deemed to be incomplete) ?
>
The initial data collecting part will be relatively easy. The tough
part will be testing to make sure that the results are truly
useful(and accurate representations). I was hoping to discuss this
more before jumping the gun. The risk of making it a "google weekend
of code" or a "google year of code" are quite large.
When you put together your application, you end up defining the work
you have to do for 12 weeks. Based on my experiences dealing with
coding quirks and so on, you can actually fill 12 weeks time with
something you expect to do in a week. Creating a prototype of EKG is
easy. Creating something that works reliably and handles quirks is
fundamentally harder. It's doable though, don't let me scare you.
-Yaakov
4:41 a.m.
Hey Dhushyanth,
2010/3/9 Dhushyanth R <r.dushyanth(a)gmail.com>:
Hi all,
I have created a draft[1] of the features and goals which I'd like EKG to
have. I went through the EKG archives a bit and tried to glean out whatever
has already been requested. I hope I have covered all of them. Many of the
ideas are inter-dependant and some may be outright non-viable however I
thought it better to get everything out in the open.
Let me add one more point that is very interesting to alot of people.
One more statistic that people would like to know is where are people
from. Namely, location is always important, but also, and you can get
this from the domain from the email address, is this person a
volunteer or is this person from a specific company. Companies make
financial decisions about this, and this could also affect the amount
of money Red Hat gives Fedora for support each year.
Also I'd like to ask what is the intended user for this tool. I
have assumed
it is for anyone who wants to know about what's going on in a ML in a
nutshell- especially so for a new user to the community who'd like to know
how things work in the ML without wanting to browse through tons of
archives. It will also be beneficial for the "bosses" who want to know where
the ML(and the community) is heading.
Yup, pretty much anyone who needs statistics to make decisions. I kind
of looked at it like a statistics engine.
One other thing i wanted to throw at you is a bit of a broader
perspective. Mailing lists only tell a part of the story. We've also
been interested in developer activitiy in source control systems,
forums, wikis, and other places where people communicate. None of
these are hard fast requirements, but the ability to analyse it is
useful.
Finally, since i had to drop development for my new job, i've heard of
a couple of other tools that do similar things, that people have
written for reasons such as a thesis. Since i think totally new code
written for GSoC has the worst chance of being viable, we're going to
have to have a seperate conversation, apart from the 'aims', about
where you want to start with code, which projects you want to utilise
to reach these aims, and so on. Perhaps Max can comment a bit on this
too.
-Yaakov
5198
days inactive
5198
days old
4 comments
4 participants
participants (4)
-
Dhushyanth R -
sankarshan -
Yaakov Nemoy -
Yaakov Nemoy