Initial List of Nodes for Further Discussion

Feel extremely excited about progressing with the project.Thank you to you and our Team.

1 Like

Mine is listed under Helsinki (as that’s where I would run the Betanode) so there’s a small amount of coverage on my part.

Pleased to be on the list and can’t wait to get stuck into seeing how Elixxir is going to shake things up!

3 Likes

I think that, although one would of course aspire to have a uniform distribution of good quality nodes across the globe, it is actually necessary to factor in the unequal number of applications from each region and their variable quality. For instance, there were about 240 applications submitted from USA+Canada versus about 140 from Western Europe + Africa. So it seems reasonable to split American applications into two groups. It is true that the proportion of shortlisted applications from USA+Canada was about 16%, close to the average across the globe, whereas it was about 8% in Western Europe+Africa, but this could have to do with the relative viability of the applications from each region.

There is also the question of geographical variation in service demand: you want to minimize network latency across the globe, for which you need to spread the nodes, but you also need to consider that users could be clustered in particular regions (such as North America).

2 Likes

Croatia here, ready to experiment. I look forward to trying to run this stuff on my BlockAndMortar-io mini-PCs among other hardware.

1 Like

Thank you for considering my node in this list. It’s really an honor. I’m very excited to hopefully make the final cut and begin participating with everyone else in the community in growing the next big thing in crypto. And thank you to everyone who voiced their support for me. I won’t let you down!

1 Like

Well spotted. I see your point. In addition, some of those countries are supposed to be more protective of individual privacy rights (as are also Switzerland and Iceland). But the fact is that there were only 4 Swedish, 3 Danish, 3 Finnish and 1 Norwegian applications. Only one from Finland was shortlisted, meaning that 9% of Scandinavian applications were shortlisted. This is below the 15-16% average, but I guess the completeness, forum engagement and specification requirements played a role in the relatively low acceptance rate. But it’s up to us the community to check apparently underrepresented regions and argue in favor of promoting unselected good-quality applications from them (will do that myself).

2 Likes

Here is a list with the number of submitted and shortlisted applications per country, ordered by the number of submitted applications. I extracted it from the published applications in the forum with unambiguous geographic data; be aware that there might be errors.

A priori, I interpret cross-country differences in the percentage of shortlisted applications as the result of quality differences and of the goal of spreading nodes, but this may serve us to identify under-represented countries where we could try to “rescue” overlooked good quality applications.

Anyway, I think the Elixxir’s team is doing a terrific work, this is a difficult multifactorial optimization problem!

Country Submitted Shortlisted % Shortlisted
United States 210 35 16.67%
Russia 85 4 4.71%
United Kingdom 42 2 4.76%
Canada 31 5 16.13%
Australia 25 8 32.00%
Germany 25 2 8.00%
Turkey 25 1 4.00%
Brazil 21 2 9.52%
China 21 4 19.05%
France 17 5 29.41%
Ukraine 15 2 13.33%
India 14 3 21.43%
Spain 14 3 21.43%
Netherlands 13 1 7.69%
Singapore 12 4 33.33%
South Korea 11 4 36.36%
Nigeria 9 1 11.11%
Switzerland 9 1 11.11%
Belgium 8 1 12.50%
Romania 8 1 12.50%
Indonesia 7 1 14.29%
Japan 7 2 28.57%
Portugal 7 1 14.29%
Austria 6 0 0.00%
Czech Republic 6 0 0.00%
Ireland 6 0 0.00%
Greece 5 2 40.00%
Italy 5 1 20.00%
Mexico 5 0 0.00%
Slovakia 5 3 60.00%
Bulgaria 4 1 25.00%
Croatia 4 1 25.00%
Cyprus 4 0 0.00%
Luxembourg 4 0 0.00%
Moldova 4 1 25.00%
Slovenia 4 0 0.00%
Sweden 4 0 0.00%
Vietnam 4 0 0.00%
Belarus 4 0 0.00%
Colombia 3 0 0.00%
Denmark 3 0 0.00%
Estonia 3 1 33.33%
Finland 3 1 33.33%
Israel 3 1 33.33%
Lithuania 3 1 33.33%
Philippines 3 0 0.00%
Andorra 2 0 0.00%
Costa Rica 2 2 100.00%
Hong Kong 2 1 50.00%
Kazakhstan 2 0 0.00%
Malta 2 0 0.00%
Morocco 2 0 0.00%
Pakistan 2 0 0.00%
Poland 2 0 0.00%
South Africa 2 0 0.00%
Taiwan 2 1 50.00%
Thailand 2 0 0.00%
United Arab Emirates 2 1 50.00%
Venezuela 2 0 0.00%
Argentina 1 0 0.00%
Cayman Islands 1 0 0.00%
Ecuador 1 0 0.00%
Georgia 1 0 0.00%
Ghana 1 0 0.00%
Jamaica 1 0 0.00%
Lebanon 1 0 0.00%
Malaysia 1 1 100.00%
Mongolia 1 0 0.00%
New Zealand 1 0 0.00%
Norway 1 0 0.00%
Panama 1 1 100.00%
Somalia 1 0 0.00%
Swaziland 1 0 0.00%
Uzbekistan 1 0 0.00%
Zimbawe 1 0 0.00%
7 Likes

From this analysis it is clear that the African continent is largely excluded from the initial shortlist.
Although there are 15 applications from Africa (only 1 chosen), some of these applications come with reputable history of running nodes on other high profile upcoming ledger projects. Geographic diversity provides resilience to the network and Africa should get more attention in this regard with nodes physically located within the respective geographic region.

Elixxir should relook at Africa if they are serious about decentralisation.

3 Likes

we will vote together so that one of them can execute a node :guiño:

1 Like

Should change Geographic Bin 3: Western Europe, Africa to Geographic Bin 3: Western Europe, North Africa

1 Like

Those who proposed the application are also in their duty to share more data and propose ideas, if they are not informed and do not worry now that it is the important step, perhaps at the time of executing they will not either…

4 Likes

It is true that there seems to be only 1 preselected application from Africa out of 17. That makes about 6%, which is below average but not so different from the preselection rate in other countries and regions, for instance the 5% in Russia or the 5% in the UK. I’m sure the Elixxir’s team have tried not to under-represent the African continent, we’d need to look carefully at the relative viability of individual applications before reaching conclusions. Anyway this is what we are supposed to do now: if you find African applications that you think should be promoted flag them to the team for consideration. The list is not final.

4 Likes

Thanks for running the numbers, and thanks to everyone participating in this conversation for raising these issues. As you guys have been doing, please continue to flag node applications that you believe are worthy of inclusion in the BetaNet.

The boundaries of the “geographic bins” are partly organized around the number of applications received from each region, but mostly to minimize network latency. If you look at maps of international fiber optic cables (especially those scaled to bandwidth), most of South America is connected most directly to the Eastern US, and most of Africa is connected most directly to Western Europe.

Overall the Elixxir team has been delighted with the number, diversity, and quality of node applications. Many members of the community deserve a lot of the credit for your help in spreading the word about what we are building. But there are clearly gaps that we will work to fill as Elixxir grows and matures – Africa and Scandinavia are good examples of this.

8 Likes

Thank you @lonewolf for the analysis and Thank you @PeterSomerville for clarifying !

The more we scrutinize the better the end result.

3 Likes

South Africa is well positioned and will be able to accommodate the latency requirements to run an Elixxir node. And will represent the Southern African part of the continent.

I tally 773 Beta Apps from your list, very cool. Thank you for laying it out. I had hoped to make the time to make a map and color grade each App with a circle around the Apps chosen for Beta. To take a real good look at where each is located. And then have gotten side tracked constantly this week. I also wanted Elixxir to green light every App. Why not is the real question, it’s our time and $? I had purposed this on Discord and PeterSomerville responded that it may be discussed, but haven’t heard back about that, yet.

3 Likes

Good idea @KILLth, I also started preparing a map of submitted versus shortlisted applications, superimposed on the global optical fiber cable network, but haven’t managed to finish it either. I also noticed that we don’t have privileges to post images here, even if the editor allows uploading them. If I manage to complete the graph I may share it with you on discord for you to check, we may try to post it there (and/or here through the admins).

As for the green light for all applications, the team mentioned in a post that they have limited resources for technical support. Also I guess that at least early betanet tests may be more efficient if run on a relatively small network, under more controlled conditions for debugging and optimization. Maybe they will scale up progressively during betanet.

4 Likes

As I couldn’t do it here, I’ve finally posted in Elixxir Discord’s server (nodes channel) the map of submitted versus shortlisted applications against the global ocean optical fiber network.

5 Likes

5 Likes

Here you go – I pulled your map from the Discord channel (which I would encourage everyone on here to join).

I appreciate the work you guys did in building this map. One important point to add is that some of those undersea cables are much “fatter” than others, capable of handling much greater bandwidth.

3 Likes