chrispardy.dev
https://chrispardy.dev
Dad of 2, lifelong learner, I build stuff, mostly software.
Director of Engineering at Newfire Global Partners
He/Him pronouns.
138 posts
100 followers
193 following
Regular Contributor
Active Commenter
comment in response to
post
That being said I took a game concept that I'd noodled on 10 years ago and completely fleshed it out in about 1 hour. I took an idea I had for a story to tell my kids before bed and have a 16 chapter book outlined in a similar time.
These are still very much my ideas, they've just been refined.
comment in response to
post
I need to consider carefully what I'm saying to make sure it understands since it tends to produce very detailed responses, that's good but it takes away from the natural feeling.
It's also way to nice - it's always saying how good my ideas are when really they are not.
comment in response to
post
If I was working an idea with someone over slack it would be messy. Lots of back and forth lots of discussion in threads. Fundamentally the base unit of the conversation would be very short messages, partially formed ideas that you share to help crystalize.
Instead it feels more like email
comment in response to
post
Serious question, do you need to iron those shirts when you arrive at a hotel or are you a demi god?
comment in response to
post
My understanding was their complaint was: why do they need to show a passport and 2 other docs to get a real ID?
Which is a valid complaint if a real ID was just a proof of citizenship / residence status like a passport is. But it's also a state issued id which includes a residency aspect.
comment in response to
post
I'm 1 to 2 levels connected from some of the most senior people at Draft Kings on LinkedIn, yet frequently got their recruiters reaching out to me (via LinkedIn) asking me if I've heard about Draft Kings?
Like what do you think I talked about with your boss's boss at his kids birthday?
comment in response to
post
As long as they're not getting married under my roof.
comment in response to
post
Question: if I'm an arachnophobe is that ok?
comment in response to
post
Zalenski needs to invite Trump to Kyiv to finish signing the Mineral deal, after all Biden made the trip so why can't Trump?
comment in response to
post
I think I've been to that hotel, I had the sinking suspicion it was run by cannibals and the shower products were seasoning me.
comment in response to
post
Paul Simon? Because if it's Boy in the Bubble it's just dated references that rhyme strung together at this point.
comment in response to
post
If the code is written by AI why is it writing source code and not assembly? Why are we reading it? At which point why do we care about quality? At some point we need to start thinking of AI as a compiler that turns Human language into software. 3/3
comment in response to
post
We need to start thinking of Copilot as the driver in a programming pair. As the navigator you the engineer should be saying "remove duplicate code".
2/3
comment in response to
post
Quebec is basically a cheat code for people in the US and Canada to travel to Europe without crossing an Ocean.
comment in response to
post
Reminder that the only NATO ally to have been attacked and invoked article 5 is the US.
So maybe spending more isn't the answer?
Or maybe we should spend more, but diplomacy and soft power have better outcomes than stockpiling bullets.
comment in response to
post
I think it has to be a consensus.
comment in response to
post
I think this is the fundamental trust that needs to go with the safety net not guardrails approach. The whole pipeline is there in case of a failure not to catch bad actors.
I'm a pretty good dev but there are times when I'm in a rush or just off and that safety net has caught me.
comment in response to
post
The only question is if a court will find the admin in contempt or will they say 🤷 nothing we can do since they won't comply.
comment in response to
post
That being said if your process had changes waiting a day for a code review, and then getting rejected 2+ times you should take a look at that process and address the root cause of the issues. I've always been able to find improvements without needing to say "code review is just too hard".
comment in response to
post
I think there's also immense value in the artifacts that asynchronous communication tools necessarily leave behind, because they give the future team members the ability to rebuild context. So the time spent waiting is paid back in full by simplified discovery.
comment in response to
post
While generally the approach of put everyone in a room either real or virtual does reduce process overhead it has other impacts you need to manage, such as timezones, flexible working accommodations, and adapting to team member communication styles or abilities.
comment in response to
post
The thing is there's no way around the "waiting". If your tests take 5 minutes and developers run them locally or if they take 5 minutes and they run remotely then you still wait 5 minutes.
That leaves you two options - cut steps or reduce process overhead.
comment in response to
post
HaveThePhillyStreetPolesBeenGreased.com
comment in response to
post
New things do it differently and you realize there's a "better" way to do these things.
That said I still really like Redux and Styled Components.
comment in response to
post
Let's see what happens when the 20 something year old "hotshot coders" get into the old Fortran systems that are probably prevalent.
comment in response to
post
Backflow is only a problem if you've changed context. I think in practice what you and I would recommend is similar, I've just seen incredible value in code reviews being done by people outside of the pair/mob. When done right it has a very high ROI.
comment in response to
post
Remember the argument that because all the Grandparents had been willing to sacrifice so much during WWII they would be willing to sacrifice their health so people didn't have to do lockdowns because Freedom!
comment in response to
post
Snow day
comment in response to
post
Thanks!
comment in response to
post
Practically this requires 2 things:
1) Someone without that context - so your pairing partner doesn't count.
2) The people with the full context to still have it in memory - so reviews must be done quickly.
A process that prioritizes code reviews and gates work on having them fulfills this. 4/4
comment in response to
post
By asking someone who doesn't have the context to look at the artifacts, derive a context from them and then understand the problem you're checking if you captured enough context for the problem at hand. 3/4
comment in response to
post
That context is made up of different things you've tried, maybe some research you've done, todays weather, etc.
A fraction of that context is ultimately going to get captured in whatever artifact you produce - in some cases you may capture 90% but in some cases you're going to get 5%. 2/4
comment in response to
post
Also just generally none of these checks should lead you to scrap work and redo it. It's more about refinements. So a rejection represents a fractional work item.