Is this only true for US-based companies? Where I'm from, I rarely see use of subscription services (especially ones that incur per-seat costs). Developers here tend to earn roughly the same number, but in a different currency, which makes all software and services several times more expensive. This results in preference of OSS alternatives and free tiers whenever possible. We use Discord for communication, Gitlab for source control, vscode for writing code, and... honestly, what else do you even need to run a software company anyway?
At my last startup, the total monthly subscription cost was the $1 Bitwarden tier (granted, we were tiny). And yes, a Ryzen system bolted to a $10 lackrack was entirely sufficient for our hosting needs.
In US, I’m seeing the culture of hacking your own hardware dying. Everyone wants to use stuff that takes 5 mins setup (long term rent costs don’t matter), doesn’t need any wires or screws, or maintenance - only what is in my wallet. That applies to most startups, and now started also propagating up to larger companies (although those have more stringent controls).
In terms of developing software - it’s the same there as hardware - noone wants to deal with their own databases or servers anymore, so it’s all cloud/aws, and now they don’t want to deal with OS anymore, so it’s all dynamo/cloud functions/lambdas etc. Try developing software for all this offline, without being wired into a giant cloud company infra.
it's as if everyone is working towards a quick exit rather than long term sustainability. Few startups seem to care about profitability and bootatrapping, it's all about hypergrowth, mostly using investor money.
Gitlab is a good example, we had to leave the free tier in order to get features that Gitlab specifically chose to be in the paid tier because their absence is a big blocker when you grow... So we pay $100/mo/seat on Gitlab.
We also use at least 10 paid general-purpose tools that each fill a real need (we don't keep tools we don't use):
- Notion for knowledge management
- Airtable to manage operations data (not customer data)
- Hubspot for Sales and Marketing
- Lattice for HR
- Spendesk for expenses management
- Quickbooks for accounting
- Typeform for various customer feedback
- Zoom for remote work and link between our offices
- GSuite for everything in it
- Figma for UI design
- Recruitee for recruitment management
And finally, add also at least 10 paid Engineering-specific tools:
- Gitlab of course
- AWS for hosting of course
- Cloudflare for DNS stuff
- Sentry for error reporting
- Crowdin for internationalization
- FullStory for UX studies
- Apollo engine to get decent analytics about our backend
- Webflow for our Website (Not for the webapp of course)
- Pagerduty for operations
- We used Cloudcraft at some point to have an idea how our AWS stack look like
I could go on... But in the end, you end up with ~10 Saas tools peer seat, with costs from $2 to $100 per tool. And it is valuable, we pay them because they are worth it. But yes this is an interesting trend...
What do you use Gitlab for at the Gold tier that you don't get at Silver? The price jump for us was just too much but we did spring for Silver after dropping self-hosted Jenkins, self-hosted npm, Jira, and Github.
Look at the history of google. Enclosures made of LEGO, and just laying motherboards on sheets of cardboard in shelves rather than buying “server pc”s. This kind of attitude plus the quality of their search, was what got them enthusiastically funded before they had any idea of revenue.
(They took so long to figure out the revenue model that they almost had the money taken back by the VCs, and had a CEO thrust upon them)
Lackracks have been around since the early 2000s at least, probably earlier than that, even. There's a whole range of Ikea hacks that exist (and dedicated websites for them). Highly recommend perusing the wide range of options if you like the Lackrack!
Is that a country with good internet connections for consumers and small companies? Here in Germany I wouldn't dare to run a lackrack server in the office, because the internet connection is crap almost everywhere.
Because they didn’t have a functional telecom monopoly at the end of the Cold War, so built all new infrastructure, not just wiring plant but organizationally as well. While the old western countries had strong monopolies or cartels with regulatory capture. Of course not every one of the eastern states got this, and some “entrenched” systems worked out fine too (e.g. South Korea and Singapore).
But in the case of Eastern Europe, where neoliberal ideology gripped most of the newly independent states, this is one of the few examples, but quite a good one, of where that ideology did in fact work out exactly as intended, for good.
I think it wa a disaster for Russia, didn't really happen in Ukraine, Belarus, Moldova; meh for Romainia, Bulgaria, pretty good for some of the Warsaw countries west of those (DDR a special case). Anything can be exploited (communism or complete liberalism/libertarianism and everything between) but the people who did the best* seemed to have taken less an ideological view and more of a pick and choose with a slight lean against the old system. Places like Czech for example.
I'm talking about the 1990s and early 00s.
* I wound't agree with them all, even the successful ones...but as none of them are my country it's not really any of my business except to learn via observation.
Quite the opposite - the GDP rose dramatically, the standard of living too.
> I'm talking about the 1990s and early 00s.
That's the problem: these policies have a delayed effect. They were inacted in early 90s, and the country reaped all their benefits in mid 00s, while they were already scaled back and revamped. And for the last 5-10 years of stagnation have been the direct result of the latter.
At my last startup, the total monthly subscription cost was the $1 Bitwarden tier (granted, we were tiny). And yes, a Ryzen system bolted to a $10 lackrack was entirely sufficient for our hosting needs.