Remote workers are wasting their time proving they’re working
vox.com
vox.com
If they can't be confident employees are productive because they can't see them working, how on earth did they measure productivity before? Weren't they merely measuring how many people they could see working and not whether they actually produced anything?
Maybe we should schedule a weekly meeting to discuss this problem further.
Fully async is possible but it's really hard. You get much slower feedback and much longer iteration times. Instead of ad hoc collaboration and course correction, you're dealing with playing tag and over the fence throws.
This is about as good as deciding George Costanza must be working hard because he's always angrily furrowing his brow.
If someone looks like they are slacking (not busy( but I can’t tell if they are delivering enough, then the problem isn’t that they aren’t busy.
If there are teams that need fat trimmed it’s HR, PM, and management. THE tool necessary to check if software eng are at work exists; source control. Rather than having managers do data entry in Jira, read your source control logs into unified place for them; Google sheets, or some dashboard. Their data entry jobs are not producing business functionality. IT is not the cost center anymore but the backbone of business.
I can manage all my HR needs from a slick website now. I have not connected with an HR person since I was hired into my current gig 3 years ago.
Given how simple it is to build complex software stacks now, the endless hiring of middle management has just increased the gamification of office work. They have little productive work to do and engage, unintentionally maybe, in distracting efforts to appear on top of things.
Frankly; who cares if past orgs used to in-office life can’t survive this. That’s a free market. Managers keep chanting “disrupt!” and similar shallow sound bites. Keep following orders; disrupt away the unproductive sycophants who went into tech management the last 10-15 because of the boom itself, with zero ability to produce.
And us boots-on-the-ground devs tend to largely ignore what management plans, as we actually know what is and isn't possible with the libraries we're allowed to use and the time allotted.
The project I'm working on has been "behind schedule" for months, is super time sensitive, and I've just been ripped away from it to do a completely different, important, time sensitive project. My manager STILL complained when I didn't show up to TWO DAYS of "Agile training" that was just watching useless powerpoint presentations about how "the SCRUM MASTER" serves everyone and is a magical being that "provides value" and other pointless Rally As Agile cargo cult.
Heaven's sake, the agile manifesto is like 100 words!
Careful. I don't commit every experiment I run, and some features or bugs require quite a bit of experimentation. I once had a client who refused to pay for a day on which I did a lot of work, but didn't commit it to source control.
This can lead to a lot of little scripts like `list_thingies.sh` or `get_thingy_by_id.sh`, so that if something goes weird I can easily provide a small repro-script to another person with some confidence it'll be helpful even if their company has a very different tech-stack.
However all those unpredictably one-off experiments and tools for inspecting remote data shouldn't be part of whatever I check in for our production code, not least because they often have some credentials embedded in them. (Especially when the thing being tested is whether we're authorizing our client correctly.)
That said, we've been discussing how we might want to keep them in a separate "random exploratory shit just in case it's relevant again someday" repo.
You definitely should. If they're useful, not just to you but also to others, then you want them backed up and accessible. Just remove those credentials.
Also it’s illegal to not pay at least minimum wage for “piecemeal work”.
I refuse to take hourly contract work. Half now, half upon delivery. Customers need to put some real skin in the game.
Real talk; open source ML is on the way that will end digital content creation jobs. Intelligent deployment of infra is here and will get better; bye devops/secops.
If you’re suckling at the teet of the past you’re missing out on a lot of cool projects. Stable diffusion is just the start of smart open source software. I’d reconsider your path altogether if you’re envisioning a long career in software and avoiding cross training.
They've done a good job of setting up eternal rent-seeking establishments; gonna be impossible to get rid of them now.