A fun read! My brother actually lives on Lord Howe Island along with his wife and 2 daughters. I’m visiting them in a few weeks too actually. I will share this tidbit with them haha.
About 2 years ago I was intrigued by distributed systems, particularly distributed consensus algorithms. Going down the rabbit hole of replicated logs, I decided to build my own Kafka event streaming platform. Since then, Kerala has been my hobby side project.
There’s still so much left to do:
- topic partitioning
- file-based storage engine (currently in-memory)
- finish client streaming apis
- much more, checkout the roadmap on the Github.
Considering we’re talking about distributed state-machines, I am curious why you would choose Paxos, a more complex algorithm over Raft. Raft, to my knowledge provides the same guarantees as Paxos, and is simpler to understand and implement.
I am no expert on Paxos - just hoping for an explanation.
Paxos itself and not something built on top of Paxos like multi Paxos is in fact super simple compared to raft.
Where it get complicated is trying to build a practical replicated log system using Paxos. Raft just happen to clearly and completely define this use case.
What SDPaxos or EPaxos try to achieve is good performance over WAN.
Something that Raft and any Paxos variant that rely on a stable leader are very bad at.
This can’t be easily added to raft because the main reason the raft algorithm is simpler is because it assume a stable leader in every operation except leader election.
Absolutely aware of dilution and all - it's more a matter of principle. I've worked my ass off and surely should have some entitlement to the company of helped build.
I was originally offered stock options in a staff pool - but I'm not happy with that. That will dilute so much and be worth nothing eventually. They're aware of this which is why we're discussing equity.
I do have some upper-hand in that they simply cannot afford to replace me - they'd need to employ 2-3 new people plus all overheads of getting them up-to-speed with our process, adjusting to team dynamics etc..
How many people have you worked with who have required 2+ people to replace them when they left? Personally, in about 10 years "professional" experience, I've worked with 0.
I'm willing to accept that perhaps I've worked with non-exceptional people (I am one myself). But really consider whether your employer would actually hire 2+ people to replace you. Your domain knowledge is worth something and the cost of switching is high, but if your employer is paying below market, would they really pay below market for 2+ new people?
What I'm trying to get at is the fact that your perception of your value might be significantly different from your employer's perception. And if that's the case, you really don't have much leverage.
This doesn't seem to jibe with reality^. Most of the literature [0] [1] [2] suggests that the first 10 employees split 10% of the equity (with founders receiving 50% and the rest left over for VC and later employee pools). Y Combinator for instance only takes 6%. Sequoia is said to take 30%. Suggesting that an employee (any employee) is worth the same as YC to the markets seems like a hard sell.
Saying there should be a standard number for the first employee is like saying there should be a standard valuation for every startup. Sure, YC throws out a fairly generous standard valuation, but they are choosing top notch startups and trying to cultivate an entrepreneur first mindset.
To roll into any random startup and say, “I guess 1 percent seems fair” is silly. 1-3 percent is fairly standard for the first employee of a well-backed (YC / VC / Quality Angel) startup, where professionals are taking a calculated risk because the payout could be huge. They are really overvaluing every startup, but know that 1 out of 10 will pay off enough to make it up.
One percent probably isn’t fair for someone taking a reduced salary at a “startup” financed by someone’s uncle targeting a niche market. The limited upside will never pay enough to justify the money lost. If the startup isn’t financed by professionals, I don’t see 1-3 percent and reduced market as reasonable. Even then, I think it’s questionable and really depends on the product, market, founders, how far along things are, and what you bring to the table.
Certainly the value you provide and what you can negotiate are the critical parts, but I feel that a lot of startup pay is exploitative, taking advantage of younger workers who don’t understand what is going on, that the real payday in a startup is the equity and they probably aren’t getting a large enough percent to justify the risk. Why do startups payout so handsomely for founders and investors, but employees are supposed to be learning and doing it for the love?
All those investors are getting preferred stock, liquidation preferences, etc. Not all equity is treated the same and common stock is generally last in line.
For a 6 person startup where someone has made significant contributions, will continue to make them, and has taken a pay cut (in an economy with full employment), 5% does not seem unreasonable to me. If he's taken a pay cut, they probably don't even have funding yet, which means he should ask for more since it's going to be diluted massively later anyway.
Obviously if you don't ask for it, you won't get it. If you don't think you're worth 5%, why would they?
