How Russia erased 2000 award records, and how to prevent this in the future
The largest hackathon in Russia is “Digital Breakthrough”. It is a governmental project that has been running since 2019, and over 50 thousand people participate each year. In summer 2023, they deleted their website, which was the source of truth to confirm the awards for at least 2000 podium place holders of various hacks. The reason was reusing the domain name after merging with another hackathon, and they did not care to preserve the data. My team was affected by this as well.
It may seem like a local issue, but it teaches us the proper management and handling of information, so I wrote a checklist to prevent this in the future. It has suggestions for an organizer, admins, journalists, and of course team members.
1. What an organizer should do
Do you manage a competition like this? You would never delete a website with public records, so you think the data is safe under your governance. However, your successor in this position may have a different view of the project, so everything you secure may be demolished.
This is why you should
Spread the valuable data beyond your reach.
1.1. Give paper certificates
In this hackathon, all certificates were digital images and PDFs without digital signatures or any other way to independently verify them:
Obviously, the name on this image can be replaced in minutes.
If paper certificates were issued, they could be better proofs because more legal bodies trust papers.
It’s tempting to go paperless if you lead a large project. However, the proper way is to make only input requirements paperless. In a bureaucratic system, inputs should be simplified, while outputs should be made bulletproof to be used for any purpose imaginable. This is because you never know the input requirements in all the systems where your outputs will be used.
1.2. Push all names in press releases
When you publish press releases, the easiest way is to just put the team names there. That’s brief and focused. However, this leaves the team members unmentioned, and it’s the members who will need the mention in the future, not the teams.
News reports by large media are good proof of awards. Use them to store the team member names for the history. Many of them will cut the names out if the list goes in tens or hundreds, but it will be their failure, not yours. And some will keep the names.
The recognition of people also shows your heart. The competition was established to help people find their way in life and profession and to improve the industry as a result. All competition mechanics and workflows are just the means for that. When dealing with a large number of people, it’s easy to lose yourself in the workflows and to view adherence to them as your purpose. It is not. A true leader always appreciates people. People are attracted to those who see them over systems and goals.
Other managers don’t have a habit of helping people. They optimize for a bigger PR picture, and people are obstacles to them. Such managers have a hard time finding a job in the private sector when their government fires them.
1.3. Publish the names in the channels that are hard to purge
Hardly anyone has a pure intention to destroy data. The competition website was deleted only because it required active maintenance and the domain name was a singular valuable resource. It likely was a mistake in a trade-off, not a plan.
Channels that don’t require active maintenance normally persist even after such disasters. Those are YouTube, Twitter, Telegram, Facebook, and others. In the case of this hackathon, after the renovation, the new content started using the new branding, but all the old posts were preserved.
It’s important that you publish the names of the team members in those channels. This hackathon used Telegram as the primary social media, but they only published the awarded team names:
There may be reasons to not push hundreds of names in press releases, but those platforms allow you to attach a file with names so it does not take visual space in the feed but still is a good proof.
1.4. Make sure the important pages are archived
When a website is shut down, archive.org often keeps a cached copy. In the case of this competition, their robot did not reach the pages that show the winners.
When you publish important data, after some time check if the archive.org robot crawled to those pages. You can manually ask archive.org to save specific pages, but this is not a strategic solution. Instead, inspect the navigation routes on your website and streamline them to simplify crawling. If you rely on manual backups, at some point you will forget to make one.
1.5. Make a simple news feed on the website
The website for the competition had an intricate tree structure: years, then dates (multiple each year), then cases. It’s convenient to navigate but hard to migrate if your successor wants to upgrade the website.
In parallel to the information tree, have a simple news feed with important data. It could be the same texts that you send out as press releases or more granular posts. Then, if your successor wants a new website and it’s too hard to migrate the tree, they will be able to migrate the news entities as a linear collection. It’s better than nothing.
1.6. Give a heads-up before shutting down a website
If you are shutting down a website, give a proper heads-up and reminders often for at least half a year. People may want to archive the pages they need using various tools (see below).
1.7. Migrate data
If you are changing the website engine, write a simple script and migrate all pieces of news from the old one. If the important data are not a linear collection of documents, write a script that converts the old data structure to a linear collection of news articles.
1.8. Give any letters
If someone was affected by a migration, they should be given support letters on demand in any format they need with any details they request (assessment criteria, judge names, etc.).
In the case of this competition, the support under the new management was only able to send me a digital certificate I have shown above, no digital signature, and no paper options.
2. What a system administrator should do
2.1. Resist the decision to shut down the website
If your management wants to destroy the source of truth for valuable data, don’t rush with it. Double-check if this is their intention. Show them this article if they are in doubt.
2.2. Make sure the important pages are archived
See the duty of the organizer above.
2.3. Take a backup copy home
When the new management realizes the important data is no longer available online, they may ask you to recover it. It may take months or even years. The standard backup copy in your organization may not be available by then. If you have a home copy and recover it for them, you will be a hero, and more so if you will not even be working there by then.
I would probably not ask my manager a permission to take a copy home, but it depends on their character. I would have done so even if directly forbidden, just deleted the sensitive data from my copy first.
The purpose of a hackathon is to motivate and help the development of people in the industry. Sometimes, the actions of the management harm that purpose, demotivate the young people who don’t have much to rely on, and lead to bad publicity. In the moments of such decisions, it’s up to you to remember why you are in this industry and why you have chosen this specific project to work on.
3. What journalists should do
3.1. Get down to the team members’ names
When you receive a press release, it’s tempting to only leave the team names and cities. If not diluted by member names, the winner announcement looks more epic.
However, you should remember all the reasons mentioned above for the organizer to show names in press releases. The names are there to back them up for the history. Don’t fail those contestants who worked hard and barely had any sleep. Find a way to both keep the text epic and to keep the names.
3.2. Ask for the names if not given
Many press releases for team competitions will not contain the member names. You should push their press department to give you the names. They may just send you the link to the detailed results page and say you can find the names there. In this case, do so, but also send them this text that explains why the names should be in a press release.
4. What the winners should do
It’s your job to prove your award if you ever need it. No one owes you anything. Don’t trust the organizer and media to do all of the above.
4.1. Get paper confirmations
Even if they only gave you a digital certificate, ask for an additional one on paper. In my case, it turned out they could have potentially issued it, but I asked for it too late when the management has changed. Don’t wait until you need it.
4.2. Archive all pages
Soon after you were awarded, check if archive.org knows that. If it does not, ask it to archive the page with the results.
4.3. Don’t trust archive.org
While archive.org is popular, it often fails for websites with data loaded dynamically with JavaScript. A page may look as if it was archived alright, but stop working soon after. This happened to most of the pages in my competition:
Interesting here is that the snapshot of 2022 loads something from a snapshot of 2023 when the site was already shot down. This is seen in the Network tab. The request did not return an expected JSON but was redirected to an HTML page of the new website, so JavaScript failed.
This is clearly a bug on archive.org, but it affects a lot of other websites as well. For instance, this is a snapshot of download statistics of a PHP package in the official repository, as of 2020:
Note that the numbers at the top are correct for 2020, but the chart data goes all the way to November 2023. In the Network tab, you can see that the page requested the correct snapshot of all.json from 2020, but archive.org answered with a 302 redirect to a snapshot of 2023.
I don’t know the reason. Archive.org did not have a clear way to report a bug. But the serendipity here is the realization of our ill dependence on archive.org, and it’s just as strong as the dependence on the original website that this service was meant to mitigate. It’s good that archive.org failed us at this relatively unimportant issue (it still did not cache the winners page), and not on something leading to a money loss.
The solution is to use multiple archive services that are trusted by the legal bodies around the world, to which your award needs to be proven.
Try the following alternatives:
Unlike archive.org, these services don’t save all websites proactively. Instead, they let you archive the pages you need with them. Someone was smart enough to archive the page with their results in the same hackathon where I failed to do so:
4.5. Ask your local media to publish an article
Your local media are hungry for news. If your team had a national award, they may be happy to write a short note about it, but they don’t necessarily follow niche competitions. Send them the page with the results and hint them to run a story.
4.6. Seek higher visibility during a competition
During and after a competition, journalists interview some competitors and especially winners. Ask them to make sure they put your name and team in the titles. This way, you can prove your team membership if the winners are not announced individually. One of my friends gave a talk for the organizers, and he now has a mention. Seek such options.
4.7. Be a mentor or an ambassador during the next event
In many hackathons, past winners are invited to be ambassadors of future competitions. You promote the competition by showing up for interviews in the media or elsewhere, and the organization promotes you.
5. What good citizens should do
The deletion of 2000 awards from public records is a huge failure. It exposes the lack of checks, balances, and diverse opinions in the organization.
I saw incompetence in all stages under the new management.
A support agent could not understand why an unsigned PDF certificate was insufficient for me. She also said, “I guess the website was shut down because everyone who needed their certificate has downloaded it already”.
Another agent just closed my ticket after a few days without doing anything.
I found a contact of a middle manager, but she could not understand why I wanted the records recovered and reiterated that the website would not be revived, without saying a reason for it.
This incompetence is only possible under a corrupt government, as this hackathon is a governmental project. Good citizens should vote for leaders who would appoint competent managers and would not start a war, which would scare so many worthy leaders and professionals out of the country.
If we fail to do that, the governmental IT projects will turn into something like this:
Never miss a story, follow my Telegram channel: ainkin_com
