I have been meaning to learn Go at some point.
2 Likes
Have paying customers, or get grants from agencies. With more expensive work, they become very similar, but at lower levels you will likely need to focus on working with individuals and businesses, or getting a few large grants for work.
Keep in mind that by the time youâre hiring more people you just hire in the skill sets you need to expand so you donât need to get great expertise in those areas, just functional enough to hire people to do them at some point.
In the end I want to know at least a little about a lot of things but I think I really want to dive deep with concurrent programming and parallel algorithms.
That subject area is too big to âdive deepâ into. The first person to get any of it right was Dijkstra, so it isnât reasonable to to all of it. And here, too, there is a big difference to the math theory, the computer science theory and experiments, and the applied. This is an area that I find once you get an intuition for it, that should be enough to suit your needs. But getting adequate intuition takes a few years.
Honestly its really hard to say things about the future. I am still a student in university. I just know that my last semester or two I finish my BS by taking two graduate level courses. One goes over multi threading and the other parallel algorithms.
It may not sound like multi threading but I have worked for the guy that teaches it.
1 Like
So many dam options and I want to know as much as possible
1 Like
If possible, Iâd take those at the same time if I were you. Theyâll cover a decent breadth and introduction to the subject. But get through those before you get your mind too set on the subject.
1 Like
So I have been studying a range of classic and modern encryption standards and had the thought, Why not play a little game on the forum. An encrypted treasure trail of sorts.
I would upload a file like an image, the file may have an encryption, it may have a key hidden in the file. Who knows. But people would have to evaluate what they think needs to be done to find the next key or break an encryption. Obviously wouldnât use encryptions that would take longer than say a few days to brute force or a few hour for other methods of cracking unless there is a key somewhere else. None the less, Anyone interested in the idea?
That could be a lot of fun and what better way to learn about encryption and some problems/downfalls of it than breaking them xP
peeks in this thread
oh hi
I didnât see you all there behind this huge paradigm shift happening in C++
2 Likes
It can work if you can prevent people from perverting it
2 Likes
Yeah that could be fun, but maybe release a few as this many geeks in one place it should fall pretty quick.
1 Like
so you want to run an L1T ARG?
iâve actually been planning something similar for a time.
PM me if you want to collab on one.
1 Like
This guy should be in Java land, not C++. Iâm not taking a 8X performance hit on my applications for the sake of his design philosophy.
There are situations where these data structures will give you better performance. Did you watch the video?
Yes. But the ones he used in the general case are slower. Between the increased coding complexity, yet greater reliance of compiler optimizations required, and that the instructions deviate so far from machine instructions it is better to just understand what you are designing for. I donât see a strong enough case for these being added to the STL to justify themselves.
I do research programming. I fight each cycle, memory copy, and cache miss to have something compute in a reasonable time. I plan out my libraries carefully, and check them afterwards. This gets too far away from âyou donât pay for what you donât getâ. This looks more like syntactic salt for structuring programs.
Right now, what the STL needs is a graph, and to branch into dynamically weighted versions of its structures, Not immutables.
I know they are useful for some things. They hold some of the greatest promise for implicit multi-threading compiler support. But as this is presented, there is so much extra required just to make up for understanding and carrying out what is needed for your programs that it isnât justifiable to the extend of STL inclusion.
For instance, there has been kicking around a proposal for an alternative to inheriting common ancestors and instead having a number of supported characteristics. Like supporting incrementing, equality, hashing, and so forth. This is a more lightweight, dynamic, and representative model for generalized interface support. Needs polishing, but is promising. It is also simpler, easier, and general. This is something that should make it into the 2020 standard. It has a stronger case for consideration than splitting the STL for immutables support.
Agree. While in theory it should work, most people who partake in a work environment that implements agile arenât working in an environment that implements agile. Especially when the development team arenât meeting expectations that are beyond reasonable and someone has to direct members
1 Like
Depends on how difficult I make it not to mention I doubt more than a few people here would have any clue where to start
I am a intern for a small consulting company called Source Allies for this summer and I have to say that I see the positives of it but there seems to be the negative where people have different opinions of how much abstraction to use and that causes problems when not everybody communicates.