Continuing on from my recent remarks about the deterioration of LinkedIn, and my earlier thoughts on personal CRM as a non-LinkedIn, I’ve requested a download of my LinkedIn data. I wanted to take a look at what is included in it.

As I remembered from an earlier download the provided contact list contains the name, current role and date of connecting, but no links to the corresponding profiles. That renders the list of names more or less useless, if you would actually want to take your data and move on. However, going to the overview of my network on the LinkedIn site I can get my entire network shown in a single list. This overview used to be paginated, but now the network page is an endless scroll. It takes a bit of scrolling to go to the bottom of the list of a few thousand connections but then I had all my connections shown on a single page. Having saved that html file I can now strip out the links to profiles and add them to the list of connections in the data download. How I can make all that downloaded material useful as input for a personal CRM system is still an open but interesting question.

I installed delta.chat on my phone, to play with, nudged by Frank’s posting. It’s a E2E encrypted chat application with a twist: it uses e-mail as infrastructure. You set it up like an e-mail client, giving it access to one of your e-mail accounts. It will then use your e-mail account to send PGP encrypted messages.

So it’s actually a tool that brings you encrypted mail without the usual hassle of PGP set-up. Because it uses mail, you can find your messages in your regular mail archive (but encrypted), and you can contact anyone from the app if you have an e-mail address. The first message you send will be unencrypted (because you nor the app knows if the receiver has delta.chat installed), afterwards it will be encrypted as the app will have exchanged public encryption keys. Using e-mail means it’s robust, it doesn’t suffer from ‘there’s noone on here’ and there’s no silo lock-in. It also doesn’t need your phone number. It does ask for access to your contacts, which I denied as it is not at all a given that people will run delta.chat with the e-mail addresses they normally use.

I’ve tied it to my gmail address for now (ton dot zijlstra at gmail, ping me on delta.chat if you use it), because I wanted to have an easy interface to check what is going on in my inbox, and I have gmail on my phone anyway (even if I don’t use it for anything). I may switch over to a dedicated e-mail address later.

Some screenshots to illustrate:

Screenshot_20210218-090559_Delta Chat
How my initial exchange with Frank looked in Delta.chat


How my message to Frank looked in my mail. As it’s the first message it was unencrypted.


How I received Frank’s reply, which has an encrypted attachment.


The encrypted attachment when opened in a text editor shows it’s PGP.

I haven’t explored whether I can export my keys from Delta.chat. If you can’t, without Delta.chat I have no way of opening them. It’s a local tool only, so I suspect I might be able to get access to the keys outside of the app.

Slideshare is being integrated into Scribd as of tomorrow. To avoid falling under the TOS of Scribd, per their own suggestion you need to delete your account.

I always thought Scribd was Slideshare’s more evil sibling, even if I don’t remember precisely how I arrived at that conclusion. I deleted my own Slideshare account last week. Tonight I deleted my company’s Slideshare account as well.

My company also has a Scribd account, from the time when you couldn’t really upload regular documents to Slideshare yet. It went unused for the past 4 years or so, but we do use some embeds on our site.

I tried to also delete that Scribd account tonight, and immediately ran into the type of dark patterns that justify my existing perception of Scribd.

First Scribd does not allow you to download all of your own content. Read that again. We had 43 documents on Scribd, I could download 20 of them, and then downloads simply stopped working, and a banner appeared suggesting I open a monthly subscription. They had a 30 day free trial, so I went that route, and downloads then resumed. After downloading all our content, I deleted all content from our company’s Scribd account.

Second, I ended the free trial subscription as well, which does the Facebooky thing of having to confirm 3 times or more you really want to cancel (we’re so sorry to see you go, are you sure you don’t want to change your mind, if you click this we’ll pull a very sad face… etc.)

Third, after deleting our content I wanted to delete our account and could not find a deletion button. I had to duckduckgo how to delete a Scribd account, and on their own help page found that I could not delete our account until a subscription has been cancelled (which I did) and it has reached its end date. Until then there’s no deletion button visible. This means I can delete my account only on October 23rd, when the 30 day trial subscription ends that I already cancelled, and which I only entered into because they wouldn’t allow me to access my own content otherwise!

Good riddance, in short. Or rather: good riddance, in 30 days. Added it to my task list so I don’t forget.

This morning I set out to download all my Slideshare content. As Slideshare is becoming part of Scribd this month, I’m shutting my Slideshare account down (and will shut down both the Slideshare and Scribd accounts of my company as well).

Yesterday I downloaded the CSV file you get when you go to Slideshare ‘data export’ feature, which turns out is nothing of the kind. That CSV contains the download links, web urls, titles, dates and statistics of all your presentations. I thought that was useful, as the statistics provided insight in the utility of Slideshare.

I wrote a script that read the CSV file. First to take the Slideshare filename and add its publication year and month in front of it, like YYYY-MM-my-presentation-name. Then to call the listed download URL and save the results to YYYY-MM-my-presentation-name in my Downloads folder. That way I would have the downloads in chronological order, and be able to easily see the differences betwen similarly named presentations (I presented a lot about Open Data over the years!) in my file system. I watched my Downloads folder fill up nicely with the expected downloads, and congratulated myself on my AppleScripting skills…..

Then I noticed the downloaded files were at most a few kilobytes, which wasn’t at all expected as my presentation decks easily are a few dozen MBs. I should have tried this earlier at the start, but opening a Slideshare downloadlink I realised it wasn’t a link to a downloadable file directly but to a web-interface that then started the download in the background after a few seconds, and after prompting you to confirm the download. So I hadn’t downloaded 132 presentations just now, but 132 web pages with a download prompt.

Apparently Slideshare expects you to lift each of those downloadlinks from their CSV file, open it in the browser by hand, and then manually confirm each download. However if you go to your account page ‘My Uploads’ you can in quick succession click the download button for the dozen presentations presented there, and use the pagination buttons to move to the next dozen, and repeat.

Their ‘data export’ in other words is worse than their regular account interface.
The crappiness of this ‘functionality’ definitely is a great cultural fit with their new owner Scribd though.

Having clicked Download 132 times, I then deleted my account.

Next steps are moving the downloaded files to a web accessible folder on one of my hosting packages, and adapt my blogpostings that have a slideshare embed to point to that folder.

As LinkedIn has sold Slideshare to Scribd (Slideshare’s more evil twin), and the practical handover happening on September 24th, I am preparing to close down my Slideshare account. As part of that I’m downloading my material on Slideshare. The first step is getting a CSV file from them that lists all the download URLs for my slides. It also provides some statistics with those download links, so for archiving purposes I’m adding some of those stats here.

My usage of Slideshare was always intended for two things: 1) have a way to embed my presentations in my blog and for others to view them, 2) have a place that can store those files, 3) allows others to download those files. Those last two reasons were way more of an issue to solve when I started using Slideshare in 2006. Hosting packages back then were generally too small to also host presentations, both in terms of bandwidth and storage. The first reason still is an issue: having a decent viewer to show these files on a website.

My first Slideshare was in December 2006, my last November 2019, so thirteen years exactly. I uploaded 132 presentations so about 10 per year on average, but in reality it was much less spread out:

2006 1
2007 6
2008 13
2009 17
2010 32
2011 24
2012 14
2013 10
2014 6
2015 0
2016 1
2017 1
2018 3
2019 4

The peak years were 2008 through 2013, which coincide with becoming self-employed and doing a lot of awareness raising for open data. From 2014 most of my presentations were for my company, and I posted much less under my own account. (I also will need to download the material from my company’s accounts before the 24th as well).

My 2 most downloaded presentations form an interesting combination:

  • My 2008 presentation at Reboot in Copenhagen (332), that I remember very much (and that I recently converted into Notions)
  • A 2010 presentation on FabLabs (259) that I gave to an engineering company (says the description) for an internal workshop, but I have no immediate recollection of doing that. (Checking my 2010 calendar just now I do remember, seeing the client’s name)

The total views for my 132 presentations were 292708 (2217 on average)
The three most viewed presentations were:

  • My 2010 Lift Marseille, France, talk about FabLabs, 11338 views
  • My 2010 brief remarks on private sector open data during Open Data Week in Nantes, France, 8242 views
  • My talk at PolitCamp Graz, Austria in 2008, the event where I got interested in open data, but this one was about social media use w.r.t. political communication, 8009 views

The three presentations that were mostly viewed in embeds were:

  • My 2010 Lift Marseille, France, talk about FabLabs again, 7157 views in embeds, or about half of total views
  • My 2013 opening keynote for a software company’s European customer event, 3285 embed views
  • My 2012 workshop on open data as policy instrument, at the Dutch national open data conference, 3055 embed views

Given that Slideshare for me was about allowing downloads, and providing embeds, let’s look at those totals. Thirteen years with 132 uploaded presentations come out at 2286 downloads and 51633 embedded views. It’s not nothing obviously, but one can wonder if it is something worthwile enough to allow thirteen years of third party tracking.

Nick Punt writes a worthwile post (found via Roland Tanglao) on “De-Escalating Social Media, Designing humility and forgiveness into social media products

He writes

This is why it’s my belief that as designed today, social media is out of balance. It is far easier to escalate than it is to de-escalate, and this is a major problem that companies like Twitter and Facebook need to address.

This got me thinking about what particular use cases need de-escalation, and whether there’s something simple we can do to test the waters and address these types of problems.

And goes on to explore how to create a path for admitting mistakes on Twitter. This currently isn’t encouraged by Twitter’s design. You see no social reinforcement, as no others visibly admit mistakes. You do see many people pilig onto someone for whatever perceived slight, and you do see people’s reflex of digging in when attacked.

Punt suggest three bits of added functionality for Twitter:

  • The ability to add a ‘mea culpa’ to a tweet in the shape of “@ton_zylstra indicated they made a mistake in this tweet”. Doing that immediately stops the amplicifation of those messages. No more replies, likes or retweets without comments. Retweet with comment is still possible to amplify the correction, as opposed to the original message.
  • Surfacing corrections: those that have seen the original tweet in their timelines will also get presented with the correction.
  • Enabling forgiveness: works just like likes, but then to forgive the original poster for the mistake, as a form of positive reinforcement.

I like this line of thinking, although I think it won’t be added to existing silo’d networks. This type of nudging of constructive behaviour as well as adding specific types of friction are however of interest. Maybe it is easier for other platforms and newer players to adopt as a distinguishing feature. E.g. in Mastodon.