Re: Request for discussion: Is our Sprint more of a Mini-DebConf? What to have next?
Hi Sascha,
On 15/06/16 14:01, Sascha Steinbiss wrote:
> Hi Steffen and all,
>
>> I was peeking into the one or other scientific collaboration of mine to
>> invite them to our next Debian Med sprint. But I could not really tell
>> much about it, yet. Besides deciding where to convene next (which to my
>> recollection is decided by someone saying loudly that he/she wants to
>> host us ... hello?)
> FYI as a side note: I have reached out to informatics people here at
> Sanger during the regular campus-wide Informatics Group Meetings by
> giving a brief talk about Debian Med and the history of the sprints and
> also trying to get an idea of who would support such an event in-house.
> In the discussion afterwards indeed some upstream developers agreed that
> this would be a good idea and would benefit the institute. However,
> there was little initiative from any particular group to approach me
> afterwards and find out how to make it work together; it was merely
> suggested to contact the Genome Campus conference centre (which might be
> on the more expensive side compared to our previous venues).
> I can try to poke some people to bring the topic back to the table.
Fantastic campus, fantastic people, just that carp-loaded bit of water does
not really qualify. Why don't you start a second theme and have it during
the summer, one that could more easily then start migrating abroad. It
could for instance be "attached to upstream" both in the sense of data
providers and possibly a conference for those traveling from far?
Also "automate big clinical chemical data" (ABCCD) comes to mind :o)
>> we should possibly also lean back a bit and decide if we want to
>> change the one or other thing. This could start with the name. In my
>> perception what we are having is not so much a problem-focused
>> Sprint.
> Yes, I agree in principle.
>> So, what do you all think? Would it help to officially come up with
>> some extended Sprint format? What would you change? Nothing? More
>> talks? More time?
> If the number of participants supports this, maybe some pre-allocated
> time for custom break-out sessions, to allow for some in-depth mentoring
> or bug squashing without missing group discussions or talks?
> What do you think?
Most of us seem to be professional bioinformaticians. If we want to do
bug squashing we can basically do that any time via skype and github.
I sense particular value in learning whom to actively contact for help when
something is difficult to track down if this is not upstream. It would be
very nice for all the infrastructure bits, though, especially for the CI
additions that you in particular I have seen to contribute much for.
What if what you referred to as "bug squashing" we apply on data processing,
i.e. workflows for best biological insights? Particularly for a large
data site like
yours it may be very nice to have something intertwining Debian with CWL
for education, documentation and production.
>> Should there possibly a second Debian Med meeting on another
>> continent than Europe?
> Sure, if there is enough interest -- especially for newcomers from
> overseas :) For now there have only been a handful of people traveling
> that far. Not sure if more would come if there was an event closer to them.
> However, I myself would probably only be able to make it if the travel
> costs are moderate as my employer won't cover that.
We could have Sanger this year, and maybe The Cancer Genome Atlas
in the next? Maybe we could find some way to report a bit more on what
we talked about, i.e. the things that are not in papers: problems in not
always homogeneous biological big data and how to spot or circumvent
them. We may find quite an audience also outside of Debian for such
extra level of confidence. Eh, we have Charles at RIKEN, who is badly
overdue to join - here or there.
>> With the same focus? Or more medical?
> Personally I probably wouldn't be too interested in medical or research
> applications as I tend to focus more on the technical/infrastructure
> aspect. But that's just me, YMMV. I can imagine that it might make sense
> to attract the non-core-Debian crowd as well.
>
I fell out of the Debian GSoC after the organisers killed accepted projects
of mine since they felt it was not close enough to Debian. I am still
furious
thinking about it. If we start paying people to travel overseas so that we
can help biomedical research to avoid false positive/negative findings
because of erroneous data handling, then tend to think that this indeed
beyond the scope of core Debian and we should find other ways to
finance ourselves - as much as we today want to avoid any such split
from Debian. We are so few that even charging for the participation to
allow someone to travel would not help so much and break a bit of how
we interact so far.
Best,
Steffen
Reply to: