How should we review & select?


#1

Bonjour,

There is just one submission right now but … it’s been up for a few hours only :wink: Do people have suggestions on how we should organize to review & select the proposals?

I tend to think we should wait November 26th, 2018 and maybe discuss that in person after browsing the submissions. They are the material we will have to work with, we have no choice in the matter. Depending on what we have, we will be able to make a sensible selection.

I don’t feel strongly about that though, if someone has prior experience or ideas, please speak up!

Cheers


About the Decentralized Internet & Privacy category
#2

It has been discussed live via mumble (or so) last year.

Sure we should prepare this meeting by reviewing the submission ourselves prior to it.

If there is too much submissions, we could divide the preliminary reviews.

In my case, I was on the submission side, and my submission was refused because the project I wanted to speak about had no website, nor available code online yet at the moment of the review (sad). I think I wont submit the request this year despite the fact that Meta-Press.es do meet the last-year requirements via https://www.meta-press.es. I had no funds for this project yet, so did not advance enough.


#3

There’s ICT24 cascading funding for search & discovery. I think you should submit a proposal. Prospect is ~30K EURO for 6 months.

Regarding proposal review, I wish we used this forum, with a SC-private category where we can discuss things among us without fear of breaking egos. For each proposal we can create a topic by pasting the link into the Topic title line so it’s linked already in the topic list. Then we can vote on topics and comment to discuss.


Pentabarf registration & permissions
#4

The alternative is to use the forum as proposed by @how or use the pentabarf rating tab. You can practice with the fake submission I created, as shown below:

Once we’ve decided on the criterion for selection, each of us could assign a rating to each talk, and remarks to explain, if relevant. And the final selection can be based on that.

I’m torn because the pentabarf way does not allow for a discussion but the forum would involve a lot of copy/paste. That’s not the only tradeoff but you hopefully get why I’m still unsure :wink:

What do you think?


#5

In support of @how’s proposal I created the private review category which should be visible only to the FOSDEM2019 group in which I added @how, @Siltaar, @Natouille, @pierre, @emgrasmeder and myself. The template for each topics in this category contains a poll for us to rate the proposals and I created one example based on my fake submission as follows:

  • create new topic
  • paste the URL of the proposal in the title
  • create topic
  • edit topic to change the title to show the proposal title instead of the URL

With that we can decide which is best (pentabarf or forum)


#6

Return of my years inside Sudweb conference experience: to vote for submissions, we had a +2 / -2 system

  • +2 : I want to see that conference
  • +1 it seems good
  • 0 no opinion
  • -1 doesn’t seem interesting
  • -2 no way for that subject

As we had only 8 slots during the day, we were allowed only 8 “+2”.

Do we have some datas about last year ? Like the number of subjects submited ?
And how many slots do we have ?


#7

I implemented that in the proposed category (and edited the topics created by @how accordingly). It looks good. What do people think?


#8

IIRC we had about 40-50 submissions (privacy & decentralized altogether). Sine we’ve kept the same structure as last year (20min presentation) we have a total of 16 slots.


#9

@how I’ve allocated time to carefully review / sort / discuss next week, starting 27, i.e. right after the deadline. I really like what you’ve done with referencing the proposals in a dedicated category, it helps :slight_smile:

My (informal) goal on the 27 is to spend a few hours reading through all submissions carefully to get a sense of what we have to work with, globally. From this impression I’ll be able to advocate for criterion to sort the best proposals. Not sure I’ll succeed but that’s the intention anyway.

I’m not a big fan for delaying the deadline but it would make sense to selectively accept late proposals (it is technically possible to submit after the deadline) if we all agree it’s worth the disturbance.

@Natouille @Siltaar @emgrasmeder @pierre what do you think?


#10

Looks like we have 41 proposals. Very much looking forward to reading them this evening! @how will you collect all of them and create matching posts in the forum? If you’re short on time, I can do the rest: just let me know where you stopped :slight_smile:


#11

I added two more yesterday.


#12

I finished reading & understanding all proposals. I’m a little bit dizzy but here are a few thoughts:

  • we have a majority of good proposals
  • there are some big names which is good but not to the point where it would flood the schedule
  • a large majority of talks are about technology (software, protocols) addressing decentralization and/or privacy challenges and I discovered a few I did not know existed
  • there are a few non technical proposals but less than I anticipated

I’ll sleep on this to digest and think of selection criterion. Thanks @how for the amazing work organizing the forum to help with the selection!

P.S. I added a few URL links to pentabarf when they were missing.


#13

Thanks! I followed a few links to get an idea of what was there.

Also I started tagging stuff. When there’s no question about the decision I add the #accepted tag so we can track our candidates.

I’m pretty sure we can cluster something around #gdpr.

I didn’t see any ActivityPub round table proposal, which is a shame. I really think we should have it.

I’m also wary about proposals asking for more than 30 minutes.


#14

For the record, the private conference call to decide on selection criterion is being scheduled.


#15

@Natouille @Siltaar @emgrasmeder @pierre you have been granted elevated permissions on the forum so that you’re not blocked in any way when editing massively, including URLs multiple times and other anti-spam blockers associated with new members. Sorry if that u already: I should have thought about that in advance.


#16

Reviewed some already.
Thanks @how for the amazing work.

When I rate, I try not to read other votes before.

I’m also biased toward librehosters :slight_smile:
I’ll finish voting tonight in the train .

Have a nice day!


#17

Ok, I had fun voting with you tonight, but well, there are a lot of candidates.
I think some are obvious to discard, and some are obvious to keep.
I’m afraid I gave too much +2.
And I’m totally biased to what I want to see. No so called “objectivity” here :slight_smile:
I didn’t look at the person name usually so I don’t know if I accepted more woman or man.(But I guess more men applied, so :confused: )
(And I didn’t do my homework in promoting the call to diverse people, I was too busy lately, sorry :confused:)


#18

@loic, I’m looking into querying the database directly to extract polling results directly so we can visualize the first 20 slots and see if we want to merge anything from the top 16 to make more space for other stuff we’d like to see.

Could you add the data-explorer plugin? You would do something like this over ssh:

Add to /var/discourse/containers/app.yml in the plugins section:

- git clone https://github.com/discourse/discourse-data-explorer.git

Then rebuild the container:

cd /var/discourse
./launcher rebuild app

#19

We’re using OVH OpenStack cluster which is experiencing troubles. I’ll try to do that tomorrow. I checked all votes and we need to wait until @Natouille @Siltaar and @emgrasmeder get a chance to cast their vote.