The line actually changes - it was recorded twice, once with each emphasis. Coppola expressed regret about the decision, since he also considered just using one take of the line and allowing the audience to imagine the different inflections.
Maybe modern languages don't have tuples because modern programming language designers don't have any|much|enough experience with tuples? If I wanted to play with tuples, what language(s) would I have to use?
Python, or any of the functional programming languages: OCaml, Haskell, probably Scala. Tuples are really just syntactic shortcut for immutable lists, so Clojure would be appropriate as well, even though it makes no distinction between a tuple and a list (e.g. "rest" parameters to a variadic function are a list).
"Another role Nathoo takes on with founders and startups is an accounting advisor. She’ll ensure that every company incorporates in the state of Delaware, and if they haven’t done this, she’ll help with that process. Y Combinator and most investors will only invest in companies that have been in incorporated in Delaware, and many founders don’t know this. Nathoo says that of the current class of 47 startups at Y Combinator, only one company’s incorporation documents were problem-free when joining the program."
If Kirsty were willing to publicly share any part of her accumulated knowledge in this area, I'm sure I wouldn't be the only one to appreciate it.
When we agree to fund companies, they fall into one of three buckets:
1) not incorporated at all
2) Incorporated outside Delaware and / or as an LLC
3) Incorporated in Delaware already
Those in 1) are by far the easiest to deal with - we have a standard process to get everyone set up so that going forward there won't be any problems.
Those in 2) start to get a bit more complicated and we have to work with the founders to convert to a DE C-Corp. Sometimes that means just starting anew with a new company and sometimes, if there is too much corporate history, converting the companies. This takes up some time and depending on the original state can be costly and time-consuming. But it all works out in the end...
Those in 3) are the ones that are often the hardest! There can be problems around only some of the paperwork being completed or signed, founders don't have vesting on their stock, uneven stock splits between founders (a strong indicator of future founder breakups), needlessly complicated cap tables, not enough stock authorised for us to buy our shares - the list goes on. All this can be fixed too!
The founder that says to me "we're incorporated in Delaware so you can invest in us straightaway" is usually the one that becomes one of the most complicated companies for us to invest in.
My advice would be that if you're applying to YC, then don't incorporate unless there is a specific reason to. It is much easier and cheaper for you in the long run to use our process.
uneven stock splits between founders (a strong indicator of future founder breakups)
I've heard lots of stories about this issue, but I think you're the first person to actually have a statistically significant amount of data to back up what you're saying.
Can you elaborate on this point a bit? In particular, I'm wondering if it's the uneven stock split which is the problem, or if that and founder breakups are merely both symptoms of an underlying issue -- say, different levels of commitment from the founders, or unequal status levels.
Or put another way: If a team applies to YC and says that they plan on a 67/33 equity split, would you convincing them to change to a 50/50 split improve their chances of success, or are they still at a disadvantage compared to teams which originally planned on a 50/50 split?
The reasons I hear often for uneven stock splits are because one founder came up with the idea, or has been working on it for a month longer etc etc. When a company is in its absolute infancy, this seems like a logical conclusion but what about when the company is 5 years old? What we see is that even after only 3 months of YC when the founders are all working as hard as each other under stress and often the idea bears no resemblance to the original idea, that this starts to become a problem. The founder with less stock starts to feel like this is not such a good deal for him / her and it can lead to problems.
There are other reasons for uneven stock and as you mention, different levels of commitment or unequal status levels cause problems too. This is something that we would seek to understand more during the applications process when we see it and to try to make sure the founders have really thought through whether this is what they want. The key to a lot of this is open communication between the founders.
Of course, there are some situations where an uneven split does work. An example would be a founder has a mortgage and a family to support and therefore takes more salary in exchange for less stock.
We do not insist on an even split in any situation but I do always make sure that the founders think through their decision carefully.
There's one other source for this data - Noam Wasserman at HBS did an extensive study of startups. His data (in The Founder's Dilemmas) shows that companies that make a "quick and easy" decision to split 50/50 received significantly lower first round valuations than teams that spent more time debating the split (although it then made no difference if the long-negotiated split was equal or unequal). He also believes anecdotally (but doesn't have data) that the quick-equal teams are less successful long term.
Great book, incidentally - this is from page 163, I highly recommend the whole thing.
I heard once that the circles-on-circles construction of the motion of the bodies of solar system when Earth is taken as its center is merely the Fourier transform of the ellipses that describe the solar system with the Sun as its center.
Unfortunately, I never saw the actual derivation, so that idea is basically hearsay.
I think you misheard. The epicycle system is, in essence, an n-termed complex Fourier series (not transform). The point that the person who told you this was trying to make was that any path can be represented using enough terms in a complex Fourier series, so it's a fairly bad system for deriving physical laws.
Really short version: I worked on correcting the extreme acidity my genetic disorder promotes and I consumed a combo of high brine sea salt, coconut oil, glyconutrients and organic yogurt. I did a lot more than that but that is a good place to start, assuming you are hoping to act on the info. Organic yogurt has more varieties of flora than other yogurts and coconut oil does good things for gut flora.
By "extreme acidity" do you mean in your gastrointestinal system or elsewhere? (I've heard people talk about blood-level acidity before.) Was there a test that let you track/monitor acidity as you made changes?
I have a form of cystic fibrosis. It is well established that it makes all the tissues generally more acid but it is not really treated under conventional medicine.
I know of people who tested urine or saliva to monitor ph levels but I did not. I monitored it symptomatically. I could feel when my gut was more acid and it eventually became clear that lung inflammation was directly correlated.
As for blood ph, the body works really hard to keep that within a narrow range. Otherwise, you die. And pretty quickly. The body will strip your bones of calcium to try to keep blood ph stable. If you have an acidity issue, it is best to assume it is systemic rather than localized.
Squeak Smalltalk bootstrapped itself (back in 1996) by implementing its virtual machine in Slang, a subset of Smalltalk that can essentially be pretty-printed into C.
