I learned a lot of respect for GUI people after taking a class in it this summer. Even as something of a tech savvy person, the amount of stuff in GUIs that I unknowingly expect was much higher than I would have guessed. It was only an undergrad course, but I'm guessing there is/has been quite a bit of serious academic study into them and the psychology of them. For example, the Save button in Word still being a floppy disk and why it's unlikely to change was kind of fascinating. Apparently even if you design the best save button art and put it in the exact same place, people will still be confused about how to save, even though most users either don't know or don't remember what a 1.44MB floppy disk looked like. They just know it's the Save button. Also, building a simple/shitty Spyhunter-esque game was much more complex than I would have imagined... and we used Windows Forms, which is awesome.
I also had to do some work in Scheme and Prolog. I will probably not acknowledge in the future that I have worked with them. And thank you Walt Disney. I was trying to write my own code on my simple webscraper to parse through a specific page's relevant data before i saw xpath and looked up that library. I would not recommend doing it the way that I attempted to do it.
I've done some HMI work and yes there's a lot of work out into really simple stuff. It's super annoying.
The dude who I like to watch on the computerphile youtube channel has his own channel now and he's fun to watch
re: the equifax hack -- can any of you guys explain this a bit more for a layman? particularly the part about panels? http://spuz.me/blog/zine/3Qu1F4x.html it's not really relevant to what i'm studying but i'm fascinated by the apparent phenomenon of big, important companies/institutions not patching known vulnerabilities
Panels are some sort of admin dashboard that are some sort of off-the-shelf solution that they didn't bother changing default logins on. Those REST Endpoints were also not protected at all as you see whoever was getting 200 responses.
One of my classes is Software Design Patterns and it's being taught in C++. I thought that I was at least competent after taking data structures and doing some OO stuff in C++ but the language remains... harsh and complex. The language allows you to build abstract classes with "virtual" functions, except they don't have to be virtual at all because of the magic of templates and pointers. Spoiler it's pretty cool, though
So I've started the first of the two yearlong projects that I have to do in my program. Myself and three other dudes are developing an Android app/game that overlays a grid of square cells onto Google Maps and users can "claim" cells to build territory based on their GPS location. Nothing super cool but it's more about the process than anything else. There are strict requirements about following Agile development with the Scrum board and all that. None of us have any experience with building stuff for mobile (specifically Android), so if anyone has any obvious do's or don't's, I would love to hear them. We decided to use Xamarin because the school teaches courses primarily in C++ and then C#, so we all have at least a few classes' worth of experience in C#. At least that's the plan, because other than the C# stuff, none of us know anything about what we are getting ourselves into. The professor/overseer explicitly advised against making a game because they suffer from "wouldn't it be cool if..." syndrome, so we'll see how this shit goes.
I did some tutorials with Xamarin and worked with people in my previous job who use it. It seemed like no one really liked it as it doesn't necessarily save you that much time unless you use Xamarin.Forms (iirc) which is cross platform but isn't as flexible as using some native OS stuff. It's possibly gotten better since it was bought by Microsoft over a year ago, there's some good videos on Pluralsight if you have a subscription.
Xamarin is a good choice if you are set on using C#. For JavaScript, I would go the NativeScript/Angular 2 route.
Latest round of SBIR topics from DOE had some interesting CS related topics. https://science.energy.gov/~/media/sbir/pdf/TechnicalTopics/FY2018_Phase_1_Release_1_Topics.pdf
agile process is so annoying Retro, Planning, Grooming, talks abbout process retros makes me want to shoot myself
If anyone is interested I'm in the middle of developing a project based intro to CLJS. The goal is to take JS devs and expose them to ClojureScript via porting simple JS projects. I'll share the repo once everything is completed
Besides the "coolness" factor of using lots of functional programming, is there any demonstrable benefit to using ClojureScript? Is it just simply that you can write Clojure and transpile it to JavaScript? I really wanted to learn Clojure or Scala at one point, but I never knew if there was any performance benefit (whether compiling/running faster or "easier" to code).
its faster and state management is easier with the use of atoms. really its a way to introduce JS devs into clojure.
taking database in SQL. one of our assignments is to refactor confusing (but still working) queries into something that doesn't take 30 minutes to decipher. i thought naming conventions were important in c++ and whatnot but my god, poorly formed database queries i guess the flip side is if i ever do that kind of work, i could theoretically protect my job pretty well by writing shit that no one else understands
This is why there exists a need for strictly db people in large companies. I will say as someone who has to deal with "big data" (like hundreds of millions of rows of data being collected each day) there is not a better feeling when you figure out how to re-work a query or redesign a table + indices so that things run 10x faster.
He may want a more technical look into BC. http://shop.oreilly.com/product/0636920049524.do http://shop.oreilly.com/product/0636920056072.do
These are the courses I used, extremely informative and challenging. https://www.coursera.org/learn/algorithms-part1 https://www.coursera.org/learn/algorithms-part2
My team for that little android app I'm doing for school did our first real retro last night. At this stage I actually like the process stuff, because it's helpful for a bunch of noobs who 1) have never worked on a project for more than ~3 weeks, and 2) have more or less only worked alone. But I can see how they could end up being like pulling teeth when you have a group of people who know something about software development. We don't do code review or anything like that (at this point, at least) but things got weird when we started talking about what could be done better. I don't think Grooming is part of our process, so I'm not sure what that is
Also we're using the free option on Azure for our database hosting, but it required credit card info to set up the account. Any tips on not fucking that up or just general Azure stuff would be appreciated.
Not sure for Azure, but for the free-tier in AWS you would set up the size and instance type backing your DB and you wouldn't ever be charged for that unless you adjusted those settings. I'm betting it's the same in Azure. For data coming out of Azure into some other part of the web (like your phone for instance) they may charge you a rate based on that but it will likely be extremely small if not completely free.
It is the fucking worst. I used to bash script to get around before I found python and sed can go fuck itself.
By 'do it for me' do you mean show live matching like regex101, or something that actually writes the sequence if I highlight what I'm trying to match?
Even better if people can use the tables as created without having to do crazy joins because one fucking field isn’t captured in a table. Sometimes I really hate the way my company’s edw is setup. Ok and my manager for finding new ways to make things more complex for no reason.
At the same time, there are far more advantages to making sure data isn't unnecessarily replicated within a database. If you update a field, you want to update it once. Not 20 times because you included it in a ton of tables for simpler queries. That's the fastest way to make a complete mess of your data, and is kind of the entire purpose of using a relational database to begin with. Of course there are many non-relational schemas out there, so it's really about picking something that's appropriate for your needs.
Right...and yet while they’ve done that, they do go through and clean it up on a somewhat regular basis, getting rid of unused custom tables and the like. It then the fun stuff like date format not always matching from one table to the next is annoying.
Yeah that sounds like a nightmare. I'm a data scientist, so my end goal is almost always just to get the data into R or Python or SAS as quickly as possible and go from there. My team has basically resorted to using staging tables, where we just put the data in raw, without normalizing it at all, and then doing all of the heavy lifting with our code. In no small part because almost all of the modeling we do is time series, and most time series problems have already been solved in Python (or Pandas anyway), SAS, and R.
I'm not sure which language you're looking for specifically, but I would just Google what you're looking for. Here is a quick GUI online for building out regex expressions. I've never used it though so I can't attest to its validity. http://buildregex.com/
This is 99% regular expressions. Unless you can quickly google a pattern, it's a process of trial and error.