Author: iandick

Weeknotes 2023.7: Sprung

Milder weather, almost springlike, and the crocuses are out around Glasgow. I don’t mind the cold but the slightly lighter mornings and nights are a big thing for me. Arriving home from work not in the dark makes such a difference. Also helps with the walks.

For a few years now I’ve got into the habit of doing a daily walk irrespective of weather. If I didn’t I’d be camped in front of some form of screen for the day. In fact there were days of just a couple of hundred steps which is so poor. So in December and January the walks before and after work are in the dark – its fine really but that little bit of light makes all the difference. The walks are also important for me, not just from a physical health perspective but also mental. A good way to prep or unpack the day.

Org Design

Spent some of this week thinking of my current and wider teams organisation design. Conway’s Law is a well known organisational concept in software design but feel its often overlooked. I stumbled on Chris Fleming’s post this week which references Conways Law but expands on it – some points to consider as Fleming concludes, sometimes there’s no easy answer.

Also had a team member resign this week. Lots to unpack with that one. Most memorable for the wrong reasons was a discussion on health and safety at work. Sobering.

Ticked Off

Mastodon is still working well for me – less politics and news, lots of tech, media and geek. Feels like early twitter. Musk’s paid blue tick nonsense was bad enough, but the removal of 2FA via SMS without much guidance on using alternatives is jaw dropping. Bigger surprise was news today that Meta is bringing in paid verification. For $12 a month (or $15 on iOS) you too can get a blue badge on Instagram or Facebook along with increased visibility and priority support. Your username will need to match an official government supplied ID document and your profile picture must include your face to be verified. Am I missing something or is this just a clone of Musks moves at Twitter? I’ve stopped using Twitter and Facebook and Instagram is heading the same way so this will have little impact on me directly but services now charging for what was free, like SMS 2FA, does not feel right, especially for improved security and protection.

Consuming

  • Started watching Shrinking on Apple TV+ and I’m not convinced. Same team as Ted Lasso and can see similar dynamics in a different environment, but the character interaction is contrived and confusing.
  • Still enjoying the Steam Deck. A real mix of new, smaller indie games and emulation. Emu Deck is highly recommended – unlocks so many great games.
  • Watched Empire of Light and Devotion this week. Both fine films and had some parallels around historical race issues…but they also were strangely flat in pace. Neither seemed to change gear at all.
  • Final shoutout to Poker Face – really enjoying the weekly “how to catch them”. Natasha Lyonne is fantastic.

Weeknotes 2023.6: In the Red

As I put a bit more focus into the blog I’m bringing back a weekly status type post. For a few years I did a bit of link sharing at the end of the week along with what I’d been up to and a little bit on tv and movies. That fell away but I did focus on private week notes. This was in a Day One work journal and I found it a great way to clear my head at the end of the week – what happened, what worked, what do I need to do into next week and that would flow into Todoist. The prompts for each Day One entry were:

  • What happened this week?
  • What was hard?
  • What leadership did I see?
  • What did I experiment with?
  • What inspired me?
  • What do I need to take care of?
  • What am I looking forward to next week?

The last question may feel contrived but it was to get my headspace into a more positive place. All too easy to focus on negatives. I’ve got 2-3 years of week notes in that format…but why private? Partly because of who I work for and the type of work we do, but the real reason is I’m an introvert and many things I want to pour into a journal type post, especially around work, needs to remain private or I’d never write it down, get it out the system, refocus for the following week. 

However I miss the weekly blog post so it’s back, now titled weeknotes with a mixture of personal musings and updates from the last week.

Bunny

A wee bunny in the Glasgow Botanics  

Spent some time with other head of functions at the Radission Red in Glasgow. An away day that we really should do more of as it’s a chance to collaborate away from daily work pressures and focus on each of our functional needs, look at strategy, how we can support each other and some other bigger picture needs. I always come away pretty energised from these sessions even though I do have to step out of my comfort zone to fully participate but this time was different. Felt a fraud in the room, quite a bit of negativity too – maybe a sign of how things are working our career wise? Shared that with a couple of trusted voices and imposter syndrome was the feedback…but maybe coupled with being an incomplete leader. There’s a handful of team issues that need attention and I must carve out time in the coming week to get the issues either resolved or at least moving in a direction even if it’s not positively.

Something I have focussed on is keeping weekends work free. I say it every year and by mid Feb I’ve usually broken out the laptop and done a few hours over the weekend. So far so good for 2023, and I’ve been using the time to catch up on so many films that I’ve missed through the last few years. I’ve watched more through Jan than the first six months of 2022. I’ve watched almost all the Bafta and Oscar buzz worthy films and favourites are Aftersun and TÁR. 

Happy

Poster just of Byres Road

Things I found interesting this week

  • ChatGPT Is a Blurry JPEG of the Web – OpenAI’s ChatGPT has got lots of headlines, more so this week as Microsoft makes bets that ChatGPT will allow Bing to better compete with Google, and this post resonates in describing the power and also the challenge of relying on ChatGPT. Depending on how you interact with the model, answers will come back as authoritative and in many cases will be 80-90% accurate at best. In amongst the response can lurk inaccuracies – especially for more typical search engine type questions – thats my experience anyway. The coding examples tend to be either right or wide of the mark. As an aside, one of the team at work asked a few weeks ago about using ChatGPT – sure – as long as you can understand, document and support the code. I hadn’t thought to try basic arithmetic in ChatGPT but as the article explains, and you can try for yourself, the results get bad quickly. There’s no doubt that ChatGPT is a powerful tool but it’s not AI…yet. This is a nice companion piece – AI is Not the Problem.
  • GPT in 60 Lines of NumPy – long and technical post on how to setup GPT using a v2 model. Hats off to OpenAI for open sourcing so much of their material already.
  • Finally caught up on Black Bird on Apple TV+. Six parts, pretty grim subject matter but the cast and script was top notch. Sepideh Moafi stole every scene she was in.
  • Black Panther: Wakanda Forever – Black Panther is one of my favourite Marvel films but like many of the phase 4 titles this was…OK. Too much CGI, characters I didn’t care for but they did do well respecting the loss of Chadwick Boseman. Probably not helped by having one of the major plot lines spoiled before watching.
  • Scottish politics has been entertaining for years. The news that MP Stewart McDonalds emails had been hacked by a Russian group was interesting but not sure there was much proof of who had done it. Roll forward a few days and Craig Murray claims to have McDonalds emails and will publishing material that is in the public interest. Mmm. How did he get the emails? If there was anything interesting surely as a journalist he’d publish the story first? We’ll see what if anything he publishes in the coming weeks.
  • Shift Happens: A book about keyboards looks to be a definitive book on keyboards and one I couldn’t resist. It’s expensive but the books look so good with over 1300 photos. There’s more detail on the  books contents on its official site. 
  • The Rest is Politics has been a favourite pod since the middle of last year. Alastair Campbell and Rory Stewart’s unlikely bromance is a fair piece on what makes the podcast a must listen.

(F)acts, (A)ssumptions, (B)eliefs

Here’s an activity ANY team can do to stay more aligned.

It is fabulous. It is simple. But consider how often you and your team are misaligned on the foundational things underpinning your work.

FAB – Facts, Assumptions, Beliefs

1. Start a document

2. List incontrovertible FACTS relevant to your work.

For example:

  • Net Revenue Retention for [Profile] is currently ###%
  • [Competitor] recently launched [Product]
  • There’s a relationship between the housing market and interest rates

Provide links to the data/insights that support these facts.

3. List ASSUMPTIONS relevant to your work, along with something to indicate the current level of certainty. Indicate the relationship between the assumption and your current work.

For example:

We are operating on the assumption that [Profile] will not adopt the new product extensions in 2023 but will gradually adopt them in 2024.

  • Implications: decreased investment in 2023 with a ramp into 2024.
  • Confidence: medium-high (though falling)
  • Type: Operating (revisit on #/##/####)

We assume that a lack of expertise in [Segment] will hamper the shift to [New Technology]. 

  • Implications: New product initiatives centered around the expertise gap in [Segment]
  • Confidence: low (but increasing, with active research)
  • Type: Testing (report research on #/##/####)

Provide links to the data/insights that support these assumptions.

Note the difference between operating assumptions—things we are currently operating based on and will revisit—and assumptions we are testing.

4. List your BELIEFS. 

“Wait, aren’t those assumptions?” Yes, and no. Beliefs are foundational assumptions. They may span years or decades. By calling them beliefs, we are also acknowledging contrarian and untestable assumptions.

e.g.

  • We believe in a sea-change shift to [some practice]
  • We believe [technology] will eventually become a commodity

5. After establishing this document, establish a ritual of regularly revisiting FAB. 

  1. Keep a history of the doc
  2. Review and update regularly
  3. Encourage your team to make notes in the document and batch up the feedback for the next meeting.

Bonus 1: FAB is fractal. There are company-wide FABs and team-level FABs. In an ideal world, you make all of these public and accessible.

Bonus 2: Note how some things remain stable (hopefully), while other things change all the time (not necessarily a bad thing). If EVERYTHING changes all the time, you should explore that. Why?

From https://cutlefish.substack.com/p/tbm-4652-facts-assumptions-beliefs

Linear Roadmaps are Misleading

Linear roadmaps are misleading without a crystal ball for seeing the future. A roadmap that recognizes the existence of risk as time goes on is more honest. But an effective PM needs to anticipate possible branches, too – and create clear criteria for following each path.

The second roadmap is a very good start, because it recognizes the existence of uncertainty. But that’s just the first step – the team should be actively working to reduce uncertainty. Make a plan: this is what we need to learn by this milestone to decide what to do next.

By being explicit about our plans for the future, we are not increasing risk or reducing agility. We open our ideas to critique and experimentation before they are set in stone, so that they can evolve with our understanding of opportunities.

It’s crucial to set success criteria at the beginning of the project: we know that we were successful if this metric is increased by that amount. But you should also have a plan for when it hasn’t, and find out before the last minute whether or not it’s working.

The temporal component in diagram 3 is virtual. The steps are not timeboxed: we know that hypothesis A will be followed by different experiments depending on whether it was confirmed, or we know we may achieve our OKR and will need to do something after, but we don’t know when.

The straight lines might be misleading; the dots are not meant to represent every milestone, merely those decision points that are known to be likely ahead of time. You may wander for some time before you reach them!

The one reason you may want to put features in the branches is if you’ve been pressured by senor management to acknowledge their pet idea. Clearly show that while it’s within the possibility space, there are conditions for reaching it. If they’re not met, it won’t happen.

The great thing about strategic roadmaps is that working backwards from “what do we need to learn” helps easily tie your work to tangible goals. The definition of “good” becomes “does is help us answer the question?” rather than “did management like it?”

The roadmap tweet lit up on LinkedIn again so it’s time to publish a small clarifying update.

I’m seeing readers focus on the former aspect of the branching paths: “how do we make a decision?” But IMO the latter aspect is far more important: “how do we get where we want to go?”

From https://twitter.com/pavelasamsonov/status/1296818042928861184?s=61&t=rHrwF0m-o8nX1nk6oIOp6A

Bookmark Overload

I’ve reached bookmark bankruptcy. Too many apps, too many bookmarks, I can’t find articles I’ve saved and what’s most galling is it’s all my fault. So how did it get this bad when there are so many excellent services out there for bookmarking?

Bookmarks

Mostly it’s my constant searching for something better mixed in with some FOMO. There’s always a better service, a new way to automate from iOS or Mac, a brand new read-it-later app that get people super excited…and so I need to give it a go. Enough!

A bit of backstory. For years I was a happy Instapaper customer but Pocket came along and for me become my goto read-it-later service. At the same time I used del.icio.us for bookmarks as I bounced between browsers and operating systems. It made far more sense to save to a web service than tie back to an app locked on a platform. And then I fell for Pinboard ran by Maciej Cegłowski. It was paid for, private, stable and I loved the popular bookmarks. So my read later process was basically internet/rss/twitter -> pocket as a temp/read later holding area -> pinboard for a permanent bookmark that is tagged and searchable. Perfect…so where did I go wrong?

Goodlinks came out in 2020 and looked to be an iOS and Mac alternative to Pinboard. Pinboard was also seeing some reliability issues and didn’t have a great iOS client. So I imported all the Pinboard bookmarks into Goodlinks and started using it….but not exclusively…and there was the mistake. I no longer had one source for all my bookmarks. I also stopped doing the catching up after throwing content into Pocket. Not so much read-it-later but dump-it-and-forget. I also was still putting links directly into Pinboard that I wanted to keep and I fell into a really bad habit of open tabs in all my browsers. So Safari would have around 100 open tabs but that was on the Mac. On the iPad it would range 40-60 and iPhone around the same…and they were all for different sites. Madness!

Roll forward to an extended break through the end of 2022 and I played with the new hot read-it-later services, Matter and Readwise. Both are excellent tools and offer many advantages over current tools I use but these are paid for services at around $8 a month so I need to make sure I’m getting value out of them…and meanwhile I’ve still got my broken process in play that is costing me time and effort. I’ve even started dumping links into Notes – more madness.

Fresh Start

The plan is to clean up and remove all the bookmarking app’s I have installed or access via the web and get everything into Pinboard. I’ve already cleared out around 100 links from Goodlinks and then wiped all it’s bookmarks. Next I’m going through Pocket – this is taking quite a while as there are hundreds of links stuck in that tool. Once I’ve got Pocket cleaned I’ll be removing the app and service from Mac and iOS and deleting my account. Next will be cleaning up anything in Matter and Readwise and again shutting down my accounts. Finally – clean the browsers of the tat that’s build up there. That will give me a clean and up to date set of bookmarks in one source – Pinboard.

At that point it’s decision time. I’m tempted to switch to using Goodlinks only but Pinboard has a lot of integrations via Ifttt and also shortcuts thanks to the Pins app. Much will depend on Pinboards stability which seems to have improved again and what I want to automate out of my reading and bookmarking. But thats for another day – at least I’ll have one true source for all my bookmarks that is accessible and portable…until there’s another new service launched.

Farewell Twitter

Third party Twitter clients are no more. Over the last 10 days the most high profile Twitter clients were blocked from accessing Twitter. No notice, no communication, no updates to customers, just a broken experience for thousands of users. A few days after the block Twitter confirmed the apps were blocked as “Twitter is enforcing its long-standing API rules. That may result in some apps not working.” No mention of what API rules were being broken and still nothing directly to any of the application developers.

Andy Bain spotted that the Developer Agreement was changed on Jan 19th, 6 days after the block was enabled. Twitter has added you will not “use or access the Licensed Materials to create or attempt to create a substitute or similar service or product to the Twitter Applications;”. So you can no longer legally make a third party Twitter app.

Twitters altered developer agreement

What a kick in the teeth to classic apps like Tweetbot and Twitterrific. It’s no surprise that API access has been blocked. Twitter are in real trouble, more so since Musk took over. Users using third party apps don’t see adverts and promoted tweets, which looks to be the only way Twitter will make money and survive and even that looks unlikely. Craig Hockenberry, one of the developers behind Twitterrific, wrote an amazing post that sums up his and many others feelings right now. Twitter had already burned through a lot of developer goodwill through the last decade but the shitty way Musk killed these app’s without any acknowledgement or notice is the final straw for many. That Twitterrific spawned the word tweet, was the first to use a bird icon and drove much of what Twitter grew to be makes the Developer Agreement changes all the more sickening.

Tweetbot Farewell

Both Iconfactory and Tapbots have removed their app’s from stores and posted farewells. These developers are in a horrible position with many customers now asking for a refund, something which could seriously hurt these small firms. If you do have an existing subscription then don’t be tight and ask for a few £££’s to be returned. Swallow it and turn your frustration towards Musk and Twitter. Ask yourself if this is a platform worth investing time and effort on. For many years my experience of Twitter was Tweetbot. I’d look in on the official app and it was junk. It made the service second rate for me. No third party app? No Twitter. For me it’s that simple.

I’m really enjoying Mastodon and finding it a much richer and more pleasant experience than Twitter. Some of that is down to people I follow, some of it down to there being less official accounts for news, politics etc. It’s also been fantastic to see the Mastodon opportunity being grasped by so many third party developers. They don’t have to fear a corporation restricting access at a whim and there’s been an amazing growth in the maturity of the apps in the last two months. Reminds me of the early iOS days where a new app could come out of nowhere and quickly usurp the current best in class.

Third party Mastodon clients

My Mastodon app of choice is Ivory from Tapbots and there’s only two things I’m really waiting for in that client. A Mac version with timeline syncing and LET ME PAY YOU SOME MONEY TAPBOTS. It’s time to take it out of beta. Hope to see more of you on Mastodon…and I’ll leave you with a picture of Ollie which sadly is now a historical figure on my shelf.

Ollie