“Choosing San Francisco in 2020 is like choosing Java in 2010... Proven, but makes it harder to do simple things and doesn’t give you an advantage over incumbents” -Balaji S. Srinivasan
Arguably minimal fuss & stable. Java code written today will still run without change in 5 years. Can't say the same of other languages.
What other language would you recommend for a large eng organization of 1000+ ppl?
For performance-critical sections, you can plug in C/C++, wrapped in Python APIs.
A 500K SLOC project in Python might be unwieldy, though.
Unfortunately there's no great way of managing all of those dependencies for build and deployment.
That’s only half true now that Java actually has a steady release cycle. There were plenty of breaking changes from 8+.
This is in contrast to the legendary backwards compatibility of 1.4 to 1.5 to 6 to 7 to 8.
A5 y.o. usually has a limited command of the natural language yet.
I've always found that Java particularly shines in team settings in a way that a lot of individual productivity languages like python and Ruby fall down atleast.
Obviously there is a broad spectrum of other languages between and beyond them, but which do you prefer?
I like Go and Rust the most these days. Not because they're perfect by any means, but they avoid a lot of traps of the other languages and scale well for the whole lifecycle of actual software development in teams.
Java is bad for a few reasons. The JVM makes everything an object on the heap, which means you're always in a war with your garbage collector. You have to run in bigger machines with plenty of memory headroom. There's a lot of backward compatibility cruft, like the fake generics. The worst part is the people though. The language is too mainstream but not hard enough to weed out the bad coders, and they make a mess out of the code base.
"We’re in an interesting transitional era where San Francisco / Delaware is obviously* wrong but incorporating on-chain and going full remote is not yet obviously right."
(i.e. he's advocating using blockchains to define your legal entity and building a distributed team worldwide for your employees.)
https://twitter.com/balajis/status/1149073637325979648?lang=...
The blockchain dream (which still needs a lot of tooling and infrastructure) is that you fill out a form once, incorporation and annual maintenance is basically free, and most of the mundane legal stuff is enforced by software for very small fees. He also mentions nation-states as a configurable parameter: basically this would mean that moving your company's legal jurisdiction would be as simple as updating a form field, and nations would have to compete to retain businesses.
It's a shell company of "Northwest Registered Agent", a shadowy company that has dozens of shell companies for each state and does SEO to get them to the top when you search how to register a company in these states. They're not reputable and you should absolutely stay away from them in the future. If you see onlineaccount.net in any of the source code for the page, or if it directs you there after login, it's the same shell company.
You also actually do need a registered agent, but probably not that one. I incorporated through Clerky (which is reputable, YC uses them), but it costs much closer to $1000 than $50, and then Clerky has a partnership with CSC (Corporation Services Company) to act as the registered agent. CSC itself will try to upsell you on various services that you don't technically need.
YMMV but you are right that they aren’t a government agency.
They provide a legal agreement template between founders as well as registered agent services.
Legal contracts are not self-executing but carry the weight of the legal system.
Smart contracts are self executing but if a mistake is made or the contract is buggy it is essentially moot legally.
Frankly, this smells a lot like some straight up sovereign-citizen-tier pseudolaw. If you actually try to do business this way, you're going to be in for a rude awakening when the actual court system still ends up exercising jurisdiction over you.
I was in SF for the deliberate rolling blackouts in 2001
EDIT: it’s different in the reason, but it’s equally preposterous to have such badly managed infrastructure
This is quite different, and the rolling blackouts were statewide.
Edit: The rolling blackouts were the result of letting Republicans draft hamfisted deregulation legislation. The resulting greed saw Texas energy companies fuck California about as hard as they could. What's happening now is also related to greed in that it's cheaper for PG&E to shut off power to prevent wildfires than it is for them to maintain their equipment. PG&E burned down huge swaths of California over the past couple years so they're being a bit more cautious now. The big difference is that while most inhabited parts of the state will want reliable power, the risk of fire is not equal. So where previously we all got fucked by PG&E's greed, now it's just those situated where PG&E runs high voltage transmission lines near combustible materials.
As fun as it seems to be for folks to shit all over San Francisco, neither the rolling blackouts of the early 2000s nor the preemptive blackouts this week have anything to do with city or county management.
the result of letting Republicans draft hamfisted deregulation legislation
CA Republicans were powerless to "draft" anything.
You need to go back a bit further than that, back to Republican governor Pete Wilson and Republican author of the bill Jim Brulte. Wilson knowingly set up his successor (Gray Davis) to fail.
Republican governor Pete Wilson and Republican author of the bill Jim Brulte.
Again, Democrats controlled both the Assembly and Senate throughout the process, and ever since. Bill contents are amended at will both in and out of committee, and "gut-and-amend" (removing a Bill's language wholesale and replacing it with that of a completely different bill to avoid rules about committee review) is so common, that term was coined for it.
[0] https://en.wikipedia.org/wiki/California_electricity_crisis#...
The author of the bill was Senator Jim Brulte, a Republican from Rancho Cucamonga. Wilson admitted publicly that defects in the deregulation system would need fixing by "the next governor".
As for the Brulte bill, not one Democrat voted against it in either house (77-0 in the Assembly, 39-0 in the Senate). It was shepherded through by Senator Peace and Assemblymember Martinez. Peace continued to claim parentage until after the resulting disaster.
The biggest bullshit is that now that PGE knows they’re too big to fail, they don’t even have to give a shit when they piss people off. If North Berkeley residents sued in a class action suit the state of California would probably just bail them out.
The real issue here is that PG+E is responsible for wildfires caused by power lines, and they don’t have many ways to deal with it. They aren’t allowed to raise rates to pay for more anti-fire measures. They can spend their fire prevention budget smarter, but at some point the gains there are maxed out. One of the few things they can do is to shut off power during times of high fire risk.
They don't need to. They could cut back or eliminate their dividends, cut back executive compensation, stop buying back stock, etc., etc.
They cut back their line crews ages ago (plenty of cost savings there), and if San Bruno is any indication, PG&E hasn't bothered to keep records (additional cost savings) for quite a while.
1) Those options, similar to spending the fire protection budget more smartly, also cap out at some point.
2) One of the reasons things like executive compensation get out of control is because they are a rounding error vs. the overall corporate spending. Cutting that back to the bone is not likely to be materially helpful.
3) "cut back or eliminate their dividends, ... stop buying back stock"
They may as well shut down the company (which is what they are doing on a temporary basis). There is no reason for them to be shouldering increased risk if the reward is reduced compensation. If that is the way the game gets played only idiots would be willing to own and manage a power company. It is more comforting to think that power companies are being managed by intelligent and rational people.
As I said PG&E makes a really great case for municipal power.
There is no reason for them to be shouldering increased risk if the reward is reduced compensation.
Fire prevention reduces risk. Conversely, this reduced risk scenario you're talking about has resulted in PG&E going bankrupt twice this century.
As for exec compensation and buybacks, PG&E spends a few hundred million dollars annually buying back stock. The fire prevention budget doesn't need to be infinite, but $200-300 million would go a long way. KTVU identified three SVPs who make around $500,000 annually. Yes, that's not a ton of money compared to PG&E's income, but a couple extra million could easily pay for brush/fuel removal.
https://ycharts.com/companies/PCG/stock_buyback
http://www.ktvu.com/news/despite-bankruptcy-pg-e-executive-g...
Edit: Per the KTVU article, that half a mill per SVP is the base salary, so bonuses and non-cash compensation aren't included.
So if a power pole is owned by the government it suddenly becomes non-flammable? That isn't based on the laws of physics.
There is a strategy that makes countries fantastically wealthy: only do things that are profitable. Running the energy grid for no profits is not a path to broad-based success, it only results in taxpayers paying for things that have no marginal value. If it makes sense with electricity then there is no logical reason not to try the same strategy with gasoline, food and shelter. At some point the damage goes critical and Venezuela happens.
Also, buybacks aren't primarily a form of executive compensation; I'm no expert but they look like a work-around for double-tax issues in the American income tax law. A buyback is a method of transferring money -> stock holders and in this context is basically just a dividend.
I've no idea what is going on in California, obviously, but strategies like 'nationalise it!', 'stop the capitalists raising prices!' and 'Stop people profiting from their investments!' are just classic arguments that hurt everyone in the long term. Literally caused societies to collapse in extreme cases.
Those aren't natural monopolies.
I just quoted the base salary (a.k.a. cash) of three SVPs.
The article I linked to stated very early on that it was discussing base salary.
For example, the CEO of PG&E before 2019 was https://en.wikipedia.org/wiki/Geisha_Williams .
Here are the PG&E filings to the SEC: http://investor.pgecorp.com/financials/annual-reports-and-pr...
2017: http://s1.q4cdn.com/880135780/files/doc_financials/2017/annu...
2018: http://s1.q4cdn.com/880135780/files/doc_financials/2017/annu...
2019: http://s1.q4cdn.com/880135780/files/doc_financials/2019/05/2...
Ultimately, if something in a company goes wrong, the buck needs to stop at the CEO and if not CEO, the board of directors. In the end I guess only we shareholders are to blame.
For lower performance between 2016 and 2018, the CEO got a raises each year (snippet from 2019 page 81):
Name and Principal Position: Geisha J. Williams(a) Chief Executive Officer and President, PG&E Corporation
year salary bonus stock award total
2018 1,079,167 0 6,400,078 1,600,003 0 40,341 170,253 9,289,842
2017 991,667 0 6,500,168 0 0 996,810 108,575 8,597,220
2016 695,833 0 2,250,072 0 610,594 519,983 87,748 4,164,230
And what makes you think you can't find a competent manager for $250k a year?
Can you name some much easier jobs that pay that much?
You'll always be outbid for the absolute best, but I see no reason you wouldn't have plenty of viable candidates when you're offering >98th percentile income for a management job.
Also being "extremely type A" doesn't sound like a difficulty, really. And I don't see how it justifies 99th percentile pay either.
There are 24 year olds at LinkedIn and Facebook making $300k
And yeah, you clearly haven’t been keeping up with dev pay scales at big companies if you think it tops out at 150k. I knew people at $300k total comp a few years out of school, and senior engineers can make far more.
> And yeah, you clearly haven’t been keeping up with dev pay scales at big companies if you think it tops out at 150k.
Okay. I don't know any of these devs and just did a search.
SVPs at big companies run entire divisions of thousands of people. They make it to those levels by being workaholics and politically adept.
You might be lucky in the honeymoon period where the company is still getting entrenched, but once that is over and getting them out again is not that easy, you will gravitate to minimal service, especially in areas that require investment, and being bled dry and fleeced for wathever they can.
Large risks that would require preventive maintainable or modernization to mitigate will be ignored as the company knows they will be bailed out should it come to that.
And all this because of the neoliberale dogma that the private sector is somehow always 'better' than the public sector.
In reality large. organization of equal size have the same efficiency problems, regardless of them being public or private sector. And before you point to a leaner small private champion in a competitive field you should systemically add the cost of the dozens or even hundreds of competitors they are 'beating'.
The advantage you have when things are not in the hands of the private sector is that you can prioritise ongoing quality and value of service provisioning over maximisation of short term profit extraction.
Twice, since deregulation. Zero times, in the century before deregulation.
For those following along at home, after blowing up a San Bruno neighborhood PG&E diverted $100 million of money earmarked for safety. Where'd that money go? Executive bonuses and dividends. They have plenty of money for improvements and there is zero need for rate increases.
https://www.sfgate.com/bayarea/article/PG-E-diverted-safety-...
The problem isn't deregulation alone, it's that generation was stripped from distribution, but retail sales wasn't decoupled from distribution.
According to Texas Coalition for Affordable Power (TCAP): "deregulation cost Texans about $22 billion from 2002 to 2012. And residents in the deregulated market pay prices that are considerably higher than those who live in parts of the state that are still regulated. For example, TCAP found that the average consumer living in one of the areas that opted out of deregulation, such as Austin and San Antonio, paid $288 less in 2012 than consumers in the deregulated areas."