31–40 of 45 posts
I think a lot of "Pivots" are still in bereavement over Pivotal and Pivotal Labs being gone.
Earlier quoted context omitted.
Oh so you mean mandatory pairing (which does away with the deep thinking required for some algos) and requiring "clean code" and other Uncle Bob BS doesn't contribute to actually scalable and efficient code? Color me shocked
I used to think the active pairing was just a fad or crazy talk. Then I went and worked at Pivotal and learned it first hand. The way they did it works.
Though for the times I paired, it does help for easy/medium problems, but for me it is absolutely a no-go for things you need to "sit and think about" (check docs, or keep more than two or three things in your head at the same time)
Earlier quoted context omitted.
I used to think the active pairing was just a fad or crazy talk. Then I went and worked at Pivotal and learned it first hand. The way they did it works.
I believe you. Though for the times I paired, it does help for easy/medium problems, but for me it is absolutely a no-go for things you need to "sit and think about" (check docs, or keep more than two or three things in your head at the same time)
This gets combined with how they "fire" people there. If your pair isn't keeping up and contributing enough, then people stop wanting to pair with them and they get rotated out to another team or entirely. That effectively means that the person sitting next to you, can actually help you.
It does get really intense because if you're having a hard day or week say in your personal life, you still have to be on your game or you have to explain your situation as best as you can. But the thing is, everyone who walked into the door of Pivotal, knew the rules ahead of time. You were there, because you wanted to be there, not because it was just another job.
I liken it a lot to joining the military, which is what it felt like. A bit cultish too. All-in-all, I had a great experience there, learned a ton, and it changed my thinking forever.
We haven't been planning to release it commercially, but if you're interested in trying it out, please reach out via singlebrook.com/contact and I'd be happy to set up a test account for you. If you happen to really like it, we can probably figure out an arrangment for you to use it.
Man, PT was so good. Linear is okay, but I find it really slow - more than once I've opened an existing issue and started editing it, only for a few seconds later the actual content to pop in and make a mess.
If your cycle ends with a bunch of work undone, it rolls into the next cycle. If that cycle already a full capacity of work planned, well, now your cycle is at 150% or whatever. Repeat. The solution, presumably, is to stop creating new cycles.
But what PT would do, and seamlessly at that, is rebalance the sprints. Your project manager wants to move a number of stories to the top of the backlog? Great, but now that item projected for three sprints out is now four sprints out. The velocity and the projections were extracted from reality, not mandated from on high.
I asked Linear if they could add this -- a cycle rebalance. I was told, no, this simply isn't something they want to do, but maybe I could do it with zapier. Trying to get them to understand why I wanted this was like trying to explain air to a fish. It simply did not compute.
What is your use case? (how many users, for what type of project?). Asking because tracking building a CRUD app is different than building an OS, a movie, a video game, building, etc..
> People say Linear is fast but it's nothing compared to how well Pivotal worked. My company switched off Pivotal Tracker because it would slow to a crawl and require several seconds (!!) to load the page, with individual actions causing a DOM cascade that frequently hung browsers. Maybe it worked at small scale but it definitely didn’t work at a large scale.
Oh so you mean mandatory pairing (which does away with the deep thinking required for some algos) and requiring "clean code" and other Uncle Bob BS doesn't contribute to actually scalable and efficient code? Color me shocked
That said, pairing has its ups and downs. I think as other folks have said, you miss out on being able to go deep on problems that require ruminating. You also lack overarching reviews that help to keep architecture clean. The code might be clean, but the architecture might get really messy really fast.
https://buttondown.com/storytime-team/archive/
Tracker's basics are pretty simple but getting the details right takes some finesse. Its multi-user update model is deceptively sophisticated -- being very responsive without ever losing data is a bit tricky, and you're probably making some tradeoffs you'd rather not make. The drag-and-drop behavior also took us a while to get exactly how we wanted it.
Pivotal was far from a perfect company (if there can even be such a thing to begin with), but sadly, a lot of good things were lost in its latter days. This was one of them. Are there particular attributes, behaviors, or properties you feel are important that you can't find elsewhere? I see you mentioned latency, for example — what else is key to you?
We have a page dedicated to what was special. You can read the descriptions in the hero section here: bye-tracker.net