The software engineering lifecycle: How we built the new Dropbox Plus
In this technical retrospective (2020), @mattgerstman shares practical architectural lessons learned from running, scaling, or migrating production software under lean constraints.
Core Principles for Independent Builders:
Defaulting to simpler architectures (such as monolithic frameworks, SQLite, or managed VPS) prevents early engineering bankruptcy.
Evaluating true compute and egress costs reveals that high cloud bills can often be slashed by 80%+ through judicious re-platforming.
Code you write at 2 AM as a solo engineer must be easy to debug six months later without complex microservice dependencies.
Community Debate Themes:
HN engineers debated the tradeoffs between cloud convenience and raw server economics, database indexing strategies, and resisting premature architectural complexity.
CATEGORY Frugal Engineering →
HISTORICAL ERA Pandemic Bubble (2020–2022)
COMMUNITY SCORE ★ 135 points
TOTAL DISCUSSIONS 150 comments
This line from the linked post is a great summary of exactly how I think Dropbox's product direction has taken a turn for the worse.
This is no sleight against the author (who seems to be doing great work; I enjoyed this read!) or the engineering team. Just like, the general direction of the company. This line also stands out:
I can't speak for all paying users, but the value I derive from Dropbox is how it syncs fast and "just works". I do _not_ find new value in popups and notification badges about features that other companies have already perfected (much like I don't need to go anywhere else for file syncing).
I realize I'm not representative of the userbase as a whole, I'm just frustrated.
I love some of the Dropbox features, but the thing that annoys me EVERY time I log into their website is that it wants to try and shove team plans in front of me.
I actually logged in just before posting to check if it was still there and yep. Giant popup in the corner of the window[1].
I have dismissed these types of notices multiple times. What I don't understand is WHY. Surely it can't be a good idea to harass every one of your individual users to become a business user.
[1]: https://www.dropbox.com/s/hibz1zkohjnc2ly/Screen%20Shot%202020-09-17%20at%207.36.50%20am.png?dl=0 (yes it did just sync that screenshot to Dropbox).
I have been a long-time Dropbox+ (and now Dropbox Family) user. I don't mind so much that it is more expensive than the paid personal G/M offerings. It (usually) just works for me, however over the last year annoyances have crept in. Specifically the new File Explorer interface. It is incredibly slow and inefficient and I curse myself every time I open it instead of through the native Windows File Explorer. Hope that gets fixed soon. That annoyance alone is probably not enough to drive me to another service, assuming other features don't get degraded. Paying 200$ for a product that fights you is not something i plan on doing for a long time...
My two cents, thanks for the developers, Dropbox works perfectly and it's hands down the best place to store one's digital life.
I am a paying gsuite user, and nevertheless I pay for Dropbox because Google is not even close to where Dropbox is in this area. How to verify this: try to resync every couple of minutes a file of 500mb or more with only few changes on Drive with the sync client, then do the same with Dropbox. Delta syncing in Dropbox is still unbeaten, even by those that claim they have it (Google doesn't to my best knowledge, I think Microsoft introduced it a while ago, but doesn't work as good, try the same test).
I've been disappointed with CPU/battery usage recently. When adding a lot of photos, I have to plug in my laptop--otherwise the battery dies.
When uploading a 100MB file on a gen8 Thinkpad:
- Dropbox uses 6 minutes of CPU time, basically maxing out a core for the entire upload
- offsite scp uses 100x more efficient than Dropbox
- In wall clock time, Dropbox is 4x slower.
I imagine their sync problem is very very hard, but but pushing bits up for a large file should be much more efficient.