I'm in the same boat, but I use miniconda and install the individual packages I want as I need them, which helps minimize the size. For doing scientific analysis, it is really nice to have (basic) environment which has the basic required packages (matplotlib, numpy, scipy, ipython, etc.) from which I can duplicate and build off of. It works well with VSCode, and unlike other comments, I don't really see any negatives to the CLI. For someone who started not even programming a lot, to doing this in a scientific context (so potentially less rigorous with regards to how things should be perfectly set up, but where it is necessary that they are functional), this was by far the easiest and least complex way to set up all the stuff you need to get started.
I think your point was very well made, but to satisfy the people who comment on the energies required for phase change etc., you could just change it to very small temperature delta/difference, and be perfectly fine.
It is dangerous to speak to others allusions, but I believe both that HN moderators/YC had nothing to do with the strength of the thread, and also that YC or YC affiliated companies occupy more of the HN consciousness than other companies of the same size, space, etc. An analysis would actually be really interesting, but I would bet, and agree with OP, that YC companies are naturally more occurring on the top/active HN lists, and (less analyzable) not always based on the strength of their articles, blogs, actions, but rather their affiliations.
I do scientific programming in both Python and Matlab. The two things that to me are major benefits of Matlab are the ease of setting up your installation, and the documentation. The Matlab documentation is amazing when compared to the numpy/scipy documentation, and is almost reason alone for a beginner to use Matlab. FWIW, the Mathematica documentation is also fantastic.
Some interesting quotes (can be found in other accounts as well):
For Mr. Cook, the focus on Asia “came down to two things,” said one former high-ranking Apple executive. Factories in Asia “can scale up and down faster” and “Asian supply chains have surpassed what’s in the U.S.” The result is that “we can’t compete at this point,” the executive said.
“The entire supply chain is in China now,” said another former high-ranking Apple executive. “You need a thousand rubber gaskets? That’s the factory next door. You need a million screws? That factory is a block away. You need that screw made a little bit different? It will take three hours.”
So, from what I've read, there is also the huge factor of being able to easily source parts, much easier than in the US, and also much easier to change a factory.
Isn't that the idea behind (academic) conferences? Bringing together everyone in a field to meet new people/network, and also learn from talks and poster sessions is the motivation. There is something I've found fundamentally different in being able to spend time with professors or graduate students "trapped" at a conference poster session than reaching out across time zones via an email->potential phone call. The ability to see tons of different research, ask stupid questions, and bounce ideas off of people in real time without any other distractions means that, to me, they usually are much more productive than I expected them to be.
Technical PhDs are funded positions via grants, you most likely cannot just search for available scholarships. Very few technical PhDs are on scholarships, as they're usually more complex than just being funded by your advisor. The typical large research school stipend is 28-35k a year post tax paid as salary (ie tuition is already covered, etc.).
I did Denver -> SF this past summer and it was absolutely unreal. I was extremely happy I did it, the sights were gorgeous, and it was an awesome time to just sit and think while some of the West's most gorgeous scenery passes by.
That's a pretty good question, so I looked things up a little. Casting, especially on pieces this big, is difficult. It has problems with chemical segregation due to temperature gradients in the material, as well as porosity from the differential shrinkage due to the same temperature gradients [1,2]. From [2] "Consolidation of porosity formed in the ingot during solidification and homogenization is the major aim of the initial stage of forging. After that, the material is forged to form the shape of the desired products." The porosity and chemical segregation if I remember correctly, is very bad in nuclear applications, where bubbling and embrittlement would only be accelerated in non-homogenous materials. They are usually 1 piece to minimize welds, and this is just a feeling, but I think replacing the pressure vessel is pretty much currently the same as buying a new reactor [4]. These are very radioactive 1000 ton pieces of solid metal that you need to move and dispose of. Interesting question though, and I'd want to look into it more later.
I'm a fairly recent PhD, but on the whole I had a very positive time. It took some time to figure out what I wanted my research to be, and that was the most stressful part (first two years/in conjunction with a Masters). After that, it became pretty fun, with a very flexible schedule, very talented, driven and supportive people around me, and the ability to work on what I want with a lot of academic freedom.
One thing I emphasize to all the new graduate students/people thinking of grad school is that, of course, pursue research that is interesting, but also to make sure to pick a good adviser, one who's PhD students can vouch for in private. That was key for me and many of the people I know to having a good versus bad PhD.
Yeah, that doesn't help the huge collection of papers I already have. And I often find that the web extensions previously broke, haven't tried it recently. Just being able to download the pdf and have it auto watch a folder/import it to me is an almost completely necessary feature.
This thread has convinced me to try to switch, but one thing that I love in Mendeley is the ability to watch a folder, and have any pdf in that folder automatically imported into my account, which then gets synced across devices. That, and the document detail lookup based on a manually input doi, which fixes author names etc. are the two sticking points. When I add a doi to a pdf in Zotero, there doesn't seem to be a way to look it up and import the relevant information to the file information in Zotero. Does anyone have any way around them?
What form of entertainment or sport can you not reduce like this? You could say something in the same vein about jogging, watching TV, cycling, organized sports, etc.
I'm American and going swimming basically means to go nowhere for as long as the exercise lasts. The entire concept of it is laughably unproductive, resulting in at most sunburns, sports injuries, and further pollution from chemicals used to clean the pools. It's preposterous hedonism.
Very cool tools/perspective, but I think we have pretty different workflows/setups. I love not struggling to install TeX or its packages, and for me Overleaf has taken care of so many things that to reproduce it on my own seems much more effort than I'm willing to do.
I used Overleaf for my PhD thesis without any issues, including a lot of figures, etc. In fact, I almost only use Overleaf when it comes to LaTeX - it takes care of most of the pain points and is a great service that I pay for.
Yeah, not the OP, but if I had sensitive information I don't know what I would do. I know this sounds shill-like, but I absolutely love Overleaf for Latex. It has every package you need, decent templates that you can then customize, and the Dropbox and GitHub integration are great - uploading an image is as simple as saving it in the right Dropbox folder - very useful for when trying to upload a bunch of images or update a figure in Illustrator. I end up using MikTex like 2-3 times a year when I'm on a plane, and then everything syncs once I'm back online. Being able to see who changed what is also great, and much less overhead than trying to convince non-programmers to use git or something like that for the Latex files.
I have had the exact same experience as the OP and if you can convince several PI's with better things to do than reinstall a new version of Word, people at other institutions without MS Office licenses using Libre Office and others who don't use either and convert it back and forth to a Google Doc (thereby breaking all the hyperlinking), I would honestly be baffled.
Now, in collaborative work I use Overleaf so there is no configuration problems with installing LaTeX (Google Docs for LaTeX), and give everyone write access if they want it, I sync it with my Dropbox and have a working LaTeX distribution if I want to edit without access to the network. Those who don't know LaTeX (which is a small number, especially if they don't want to learn the basics to edit the .tex file in the appropriate place) I can convert a version to Word for their edits, which I then incorporate, or they can just markup a PDF. The Overleaf document then has a complete history of edits, Overleaf supports Word-style comments and track changes, and I can be happy in that the other authors aren't going to break something in the Word "backend".
The definition that I have always stuck with is that a sandwich has leavened bread on two sides of whatever is in the sandwich. Therefore, hot dog is a sandwich, a taco is not, etc.
Having spent a decent amount of time learning to be a programmer while doing scientific image analysis in Matlab (shudders from the real programmers), and with a decent amount of time spent in Mathematica as well, I just can't seem to buy into the Jupyter/notebook based programming enthusiasm. The talk linked in the article explains it better than I ever could, but for me, when I am leaving data in memory, it is much more convenient for me to have a completely linear history, ordered by command execution time. In python I have found the best way to do this is writing standard python functions and scripts, and running them in an IPython environment with the %run magic. You have the linear history, git works well on standard .py files, and you can interactively work with the data in the IPython prompt without worrying that something is proceeding nonlinearly. What I find works best is to explore with the data in the live prompt, which gives you interactivity, and then build up slowly a master collection of functions and commands that when run, with a single command, can reproduce the results you got while exploring. Then to come back to the data at a later point in time, you have to run one script file on the raw data. Of course, this is kind of the point of the Jupyter notebook, but I find that often when I want to change parameters, the ability to jump around and redefine things means I do. By moving from IPython to a script/functions I run, I ensure that everything progresses linearly. Idk, just my two cents.
Fun idea - very back of the envelope math is if you could convert energy into strictly breaking CO2 bonds, the amount of energy it takes is 805 kJ/mol (from Google). WolframAlpha says this is roughly the caloric content of a biscuit, or 30 min of sports [1]. This would then separate 22.4 L or one mole of pure CO2.
There are approximately 1E20 moles of CO2 in the atmosphere (very rough math)[2,3]. This could take a while, especially considering efficiency would probably be 10% of this idealized system. However, this filtering idea isn't bad, I don't know of any filter like this for splitting CO2, but something is almost certainly being worked on for just general sequestration of CO2 from other atmospheric gasses.
Anchoring is usually much cheaper/free since you don't get any facilities, whereas staying at the marina costs $$ per night, due to electricity, WiFi, water, restrooms, showers, etc.
A richer hard boiled egg with some texture. I liked it a lot. Ate it with a spoon and some sort of sauce. It was definitely different. Would probably do again.
These are all non-profits that work closely with a university and have basic research as a premise. They are both privately and publicly funded. I think that is different than a private corporation. Having worked in R&D at GM, and currently applying to jobs as I exit graduate school, the trend that I have seen personally as well as talking to many research scientists at private corporations is that R&D is an expense and must be focused on a business case. In discussion with some long term employees at some of the most research-motivated companies, this is in contrast to what private research was 20 years ago. Number of people, funding, and capabilities have been in decline. General Motors used to have an applied math research group and 10x more people in R&D! This led to things like the catalytic converter, the artificial heart valve and more. From what I have seen, nowadays private research will quickly test out a problem over the course of 6 months to maximum of a year, but will not do a 5-10 year project like academia has the freedom to do. These projects also like @gp stated, must have an application and not just be intellectually stimulating. Take it for what you will, this has just been my experience, but there are massively different purposes for current day private vs public research, which didn't use to be the case.