adament

Does this remain true if the waste producers were forced to factor in the “true” storage cost of the waste?

Could you not enforce that functions tagged can only call other functions with the tag and must not call panic?

This would not protect against stack overflow or similar environmental panics.

OM System (the camera division of Olympus sold off as a separate company) recently announced their OM-1 camera which together with a some of their expensive lenses got an IP53 rating. Which is still a less than many smartphones.

JAX recompiles functions every time you call them with an array of a new shape, i.e. if called with an array of shape (7,10) and then one with shape (17, 12) the function is compiled twice. This is generally fine for deep learning and some other numerical applications where you do the same computations again and again over arrays with the same shape. But for data exploration like Pandas, in my experience, your data shapes are different with each call, so the repeated recompilations make it unattractive.

In Numba it only recompiles for dimension changes i.e. if shape changes from (7, 10) to (2, 10, 12). However numba does not integrate AD so if you need that, JAX is probably your best bet unless Enzyme matures and you want to look into integratig it with numba.

Yes but some people ask questions rather than try to look for answers in the documentation, since it is perceived lower effort than looking in the documentation themselves. It is quicker to ask a question than look through the documentation, especially if you are not sure about the right keyword for your query. Plus if you ask a question you get to define the criteria for a valid answer to include an example that almost exactly matches your situation.

Edit: I feel like I should point out that I agree that RTFM is an incredibly offensive term, but some form of answer like: “it is covered in the documentation” actually conveys one important bit of information: The answer can actually be found in the documentation.

I understood the parent post to suggest dual licensing where the commercial license also gives you paid support. I.e. if you expect support pay up by buying a commercial license. Though I see no reason to conflate the two things. I would suggest in this case not even having a public issue tracker, and make your open source offering a take it or leave it proposition.

I think MegaHAL is (mostly?) just a markov model, and I think that was not exactly new when I looked at it around 2002. As I recall it was easier to distinguish from a human than Eliza, since it had a greater probability of spouting nonsense, but it was still amusing and fun to play with.

Personal anecdote: I read one of Ray Kurzweils books in school back then, completely misunderstood how neural networks worked and ended up making a markov model chatbot with single word states.

This has also been my experience, but to me the detriment of verboseness is not only a matter of writing code, it also means there is a lot of noise when I try to read the code. I guess it might just be a personal preference whether we prefer explicit and typically more verbose or terseness at the cost of more implicit behaviour / magic? At least I am unfamiliar with a language which is both explicit and has no boilerplate.

Thanks, I am sorry for taking your time.

No it was a comment trying to indicate that I found your question odd, and ask why you think your question is useful? Do you believe there is a single notion of brand new that can be applied across all categories? Is the age for brand new milk the same as for software or for scientific results or items of clothing? Or do you believe that for the categories of software and cryptographic theory the notion of brand new is equivalent?

Frankly in my reading of your question you come across as very arrogant, where you use the guise of a “serious question” to show off your knowledge cryptography.

It seems weird to me to gauge someone’s understanding of “brand new” for cryptography software by measuring against primitives and constructions. To me at least, those are not the same thing. Even if a piece of software contains cryptography I will still also evaluate its age as a piece of software simply as a proxy for maturity and stability of the feature set.

Thank you! I was unfamiliar with both age and Cwtch. From what I can tell, Cwtch is also a linear messaging system. Are you aware of any software offering secure non-linear (hopefully threaded) messaging, i.e. a secure e-mail replacement? It does not have to be MIME, SMTP, IMAP based like PGP, but preferably support for similar branching conversations and archiving and hopefully with support for multiple users. I love Signal but I find that finding old messages, or groups with more than a few people and branching conversations is a lot less pleasant than e-mail. And thus Signal is not currently a replacement for e-mail for me but a great addition.

Note that both the FSF, OSI and DoD (in your own link) agree that in practice the Open Source Definition and Free Software Definition are equivalent. So if you believe your license is open source but not free software, then you are very probably wrong.

Let us first agree on what we are arguing. You proposed that:

> For most software projects, there should be a way to buy a commercial support license.

> For many, such a license should be required for use in any org which is over X employees.

I interpret this to mean that: You impose in the license of the software that if you are an organization over X employees, you are not entitled to use and distribute the software under that license instead you must hold the commercial license or you must in addition to the normal license hold the commercial support license.

In the first case(where the normal license grant does not apply to sufficiently sized orgs) I believe you impose a restriction of use on the software in a specific field of endeavor in contradiction with section 6. in the OSD.

In my view it is clear that if you require an additional commercial support license for some users but not everybody you violate section 7. as the rights to the program is restricted for some parties (orgs with over X employees only has those rights if they also acquire a commercial support license):

“7. Distribution of License

The rights attached to the program must apply to all to whom the program is redistributed without the need for execution of an additional license by those parties.“

Now it might very well be that you had another mechanism in mind for imposing your requirement of “Orgs over X employees must hold a commercial support license”. In that case it might very well be a different mechanism in the definition that conflicts with your mechanism, but I cannot see how you can achieve your aim of imposing a requirement on usage of the software for some users without contradicting the spirit of the open source definition which says that you must grant the same rights to all users of your software.

I cannot find in your link [2] what you mean by the phrase: “In practice, there are many "Open Source" licenses that allow paid licensing [2]”. Usually when people speak of paid support licensing for open source (or free software) that is an OPTIONAL support license or a separate commercial license(see the QT license options) that you acquire the software under, and it does not conflict with either free software or open source software. My opposition is only your stipulation that the commercial license be required for some users of the software, and it is this requirement part that I claim conflicts with the Open Source Definition. Note that it is perfectly valid to charge people to obtain your open source software from you, but you are not allowed to impose restrictions on their ability to redistribute that software (including source) to others for free, hence this is probably not the solution you are looking for?

But the GP wrote: “For many, such a license should be required for use in any org which is over X employees” emphasis mine. If this requirement comes from a clause in the license then it violates the open source definition https://opensource.org/osd because it imposes a restriction on the use of the software.

I do not know Java very well, but from the article I got the impression that in Java you could use List<?> as a reference in non-generic code to a List of arbitrary type. I.e. you could assign: List<?> variable = new ArrayList<Integer>(); That is not possible with std::vector<std::any>> in C++ without converting and allocating a new vector. I am not familiar enough with the C++ standard to say it definitively but similar rules apply for vectors of pointers see https://godbolt.org/z/no3r67P5z

If you have a need for lightweight cryptography today, why wait for the NIST competition? Ascon is already the primary recommendation in the lightweight category from the CAESAR competition, and now also a finalist in the NIST competition.

Being a pedestrian I get the raging tic for loud vehicles easily without any metal cage assistance. And sure I try to not get annoyed by their insane noise pollution and disruption of pleasant conversations or enjoyment of sound in a large radius. But fundamentally I believe that as the driver of an unnecessarily loud vehicle you* are an arsehole.

*you as in the driver, I am aware that the parent poster is not necessarily driving a loud vehicle.

But it should be easy based on this example to include correct HTML tags in the script which the regular expression will emit. Or if you want to recognise HTML tags in the script, you can easily obfuscate construction of in the script using string concatenation.

Does the proposed regular expression really handle embedded script content correctly? From my limited understanding of HTML, pretty much only </script> counts as closing the script contents and everything else is treated as part of the script.

They are explaining to you that in the post you replied to, the part prefixed “>” was a quote from the article and your original post reads as if you thought that part was an expression of the comment authors opinion. The only part of the post you replied to that was not from the article was:

> As far as I can tell, the L in "FLOSS" is typically "libre," not "Liberal." I feel like the author is vastly and misleadingly inflating his relevance in this post.

Yes but it is used as a function for transforming vanilla option prices into a more natural unit: their implied volatility rather than their dollar price. Option traders think of option prices in terms of the implied volatility not the dollar price of an option. So in that sense Black-Scholes is used in the pricing of almost all vanilla options on stock-like things. Even though the price from a market maker usually comes out of a more complex model that (attempts to) ensures the quoted prices are free of arbitrage in the cross-section of strikes and across maturities. Usually some form of either stochastic volatility, local volity or LSV(local stochastic volatility) model.

If you are pricing and risk-managing complex derivatives books then Pandas is not the right tool for the job. Pandas does not include derivatives pricing models, daycount conventions, schedule generation and algorithmic differentiation which are all basic ingredients in a modern derivatives pricing framework. In general what you are concerned with here is computing the fair value of each instrument usually as an expectation in a simulation or as a solution to a partial differential equation, and the first derivatives and sometimes second-order derivatives of the fair value with respect to the model parameters. Pandas makes a lot of work in empirical finance nicer, data cleaning, time series work, etc. But when pricing and risk-managing a derivatives book as a market maker you are less concerned with inferring the future from the past, and more with interpolating current data in such a way that you do not open yourself to arbitrage.

I worked as a model quant for a year a couple of years ago in a bank in the team responsible for developing and implementing the derivatives pricing models. Everything was written in (a lot of) C++, but I suspected and still suspect that you could get the same performance by leveraging one of the off-the-shelf deep learning libraries with a lot less code. Maybe you would have to write a few central components in C++ and assembly to get the exact same performance. I am glad to see someone actually following through on that idea.

I am not sure if it counts as providing a benefit to society but I think the Dell private equity process(taken private in 2013 and reintroduced to the public market in 2018) is generally considered a success from a business standpoint. As mentioned in sibling comments, a lot of private equity investments are an attempt to do something similar.

Yes but exactly because the decentralized currency is much more difficult to control and explicitly circumvents parts of current banking regulations, I have a hard time fathoming that policy makers and central banks would advocate decentralized currencies. The post that I was replying to seemed to suggest that CBDCs would be decentralized or in some way inherently different from the centralized digital settlement of transfers that exists in many countries today, and I want to understand why that should be the case? Why would CBDCs work more like decentralized cryptographic currencies than the current financial system?

I am not passing a judgment on whether decentralized currencies are good, but I do not understand why central banks would move in that direction.

I agree. In which case I think the question as posed in a sibling thread is central: How do CBDCs differ from the current digital settlement system already available at several central banks? The sibling thread mentions the Bank of England and the Danish central bank has a similar digital settlement system. Is it “just” a fancy term for a technical upgrade or is there a substantive difference in function?

While I am not familiar with the story you are referring to, it does not sound weird to me. Given that QE works by the central bank buying risky assets (initially government bonds), presumably the banks received the QE money by selling higher yielding(given that even during QE the rate curve was mostly upward sloping) government bonds to the federal reserve than the interest rate they earned by depositing said money at the federal reserve. This would seem an irrational move from the banks unless they had some additional use for the cash compared to the government bonds?

I think if we want to argue that central banks will soon be pushing for CBDCs to let people have accounts directly with the central bank we need to argue why they at the same time are opposing* narrow banks, which to me seems to be a different technical tool achieving an equivalent goal?

* At least the Fed seems to be: https://www.bloomberg.com/opinion/articles/2019-03-08/the-fe...

How? Maybe it is because I live in Denmark where I would claim we are close to a de facto digital currency: a large group(most people under 40) use credit cards or a phone based payment solution for all purchases, use a phone based solution for settling smaller (<3,000$) amounts between persons, and use online banking platforms for transferring larger amounts. We are charged negative interest rates on deposits if they total more than 50,000$. Sure underlying this we have a physical currency, but for many it is more a historical curiosity than a practical consideration in day to day life.

While I understand the technical differences, I cannot see why from a policy perspective the government or the central bank would be interested in a cryptocurrency like solution? Rather with the current system and Know Your Customer regulation for the banking system, they have tighter control with money transfers since they are all ultimately between bank accounts tied to named entities. I think it is more likely that we abolish physical cash than migrate to a more decentralised currency.

It also depends on what topology you put on your space, while the derivative operator is defined on all of C^k, that does not make it continuous. In fact as the GPs example shows, the topology you put on C^k and C^{k-1} must be so that uniform convergence does not imply convergence in C^k (which differs from many peoples intuitive notion of convergence of functions) or cos(x/k) converges to 0 in C^{k-1} (which is just plain weird).