Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mswen
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
37 ms
·
301.
▲
by
mswen
13y ago
Fun read, but there might be a bit of mis-attribution going on. You could recreate the 4-day work week, the benefits and the communication methods and management structure, or lack thereof and that will likely do wonders for employee retent
302.
▲
by
mswen
13y ago
Thanks for the encouragement. Yes, I read once that Ross Perot went through about 70 rejections when first starting EDS. But prior to that he had been one of the top salespeople at IBM. What if he had given up after 69 rejections and gone
303.
▲
by
mswen
13y ago
I can relate. I also live a dual work existence. Contracted/consulting work pays the monthly bills. I am fortunate to currently have one main client that pays well and in a very timely/reliable manner. If careful on the spending s
304.
▲
by
mswen
13y ago
One of the goals of a CS degree might be to secure a well-paying position doing exactly what you are currently doing in as your full-time job. What are your particular goals in completing a CS degree? Also is your exhaustion coming from the
305.
▲
by
mswen
13y ago
The key question in my mind would be how predictable is the flow of WordPress projects? Is the flow steady? Also how much sales effort does it take to land the projects? If the flow is reasonably steady and you know where and how you would
306.
▲
by
mswen
13y ago
I would try the following. Envision success at both places, what would that look like 3 years from now? What would the job duties include? What is the likelihood of promotion? What current skills would have grown significantly? What skills
307.
▲
by
mswen
13y ago
> should we go to their competitor? That this option is on the table at all suggests that what you are developing is a feature that is lacking not only in X's product but in other products that compete with X. So one alternative is
308.
▲
by
mswen
13y ago
Great article outlining some of the issues that make really sound data analysis hard work even though tool sets have progressed substantially. One thing I do is to remind myself to walk through the basics of a new set of data before I start
309.
▲
by
mswen
13y ago
namecheap for both registration and hosting. They have been easy to work with for hosting as someone going through the journey of learning web development on their own. I know I asked 2 or 3 'stupid' questions early on about dep
310.
▲
by
mswen
13y ago
Struck me as an interesting idea. How would you deal with relationships that go bad? Someone you once trusted with editorial rights to your wiki now thinks you are the devil incarnate?
311.
▲
by
mswen
13y ago
The general advice you will get around here is that ideas are cheap, execution is hard. Many who started out believing they needed to protect their ideas end up in the camp of openly sharing them. My understanding is that most angels and V
312.
▲
by
mswen
13y ago
Have you built a website or mobile app yet? Do that first, prove to yourself that you have the skills to build a prototype. This phase requires almost no money, so don't worry about funding yet. Just start building things. What is c
313.
▲
by
mswen
13y ago
Agreed it is just my opinion. And rarely, if ever, will you find all of these skills in the same person. More often it will be a small team of people each with 1 or 2 specialties plus some other areas they are reasonably competent at. Mos
314.
▲
by
mswen
13y ago
Becoming a data scientist isn't a matter of reading journals and blogs. You can get a sense of the field and what is required by reading those sites but becoming a data scientist is years of hard work. You need to develop serious skil
315.
▲
by
mswen
13y ago
Maybe you can find a second job? Freelancing without a specific set of skills is a really tough start. Generally when people hire a freelancer they are looking for a skill set that allows them pass off a complex project/task to an ou
316.
▲
by
mswen
13y ago
Local job search on indeed.com Java Developer - 218 Ruby Developer - 28 Python Developer - 24 If your local job environment is analogous to mine, I would suggest that you pursue a dual track. Route 1: Keep developing your Java skills - even
317.
▲
by
mswen
13y ago
How does this compare with Mozenda?
318.
▲
by
mswen
13y ago
First: if you have been programming since 12 you are probably above average in intelligence. If you are in the USA you should be able to go take a GED exam so that you have the equivalent of your high school degree. You seem well aware that
319.
▲
by
mswen
13y ago
I started a Lean Gains style fasting approach in Oct 2012. My version has me stop eating by 8PM and not eat again for about 15 to 18 hours. I work out right before the end of my fast. I lift weights 3 days a week and run 2 days a week. Th
320.
▲
by
mswen
13y ago
What is the future potential of the business? Do you personally have what it takes to lead it to that potential? Are you a starter, who gets bored after the technically hard parts are solved? Would the proceeds, if invested in a combination
321.
▲
by
mswen
13y ago
That is what I am doing currently to support my start-up. One steady client that pays well and is willing to work with a somewhat flexible commitment from me at about 10 to 20 hour per week. And with careful spending that will pay bills f
322.
▲
by
mswen
13y ago
I am more like an H with significant skills and experience in multiple areas. I kid around that I have about a 4 year attention span for learning and getting pretty good in a new domain. Finding the bridges between domains generally happens
323.
▲
by
mswen
13y ago
Many people talk about the "T" employee - the person who acquires deep knowledge in one area but then acquires skills and experience very broadly but at a much more shallow level. And if it is a serif font - you have a couple oth
324.
▲
by
mswen
13y ago
That is a pretty ordinary structure in corporate life, just worded kind of oddly. Generally you would refer to the 70% as the base and the 30% as the bonus. Bonuses often have 2 or 3 parts. Part 1 - has the company met its quarterly finan
325.
▲
by
mswen
13y ago
I would design a project with the following components. 1. A web crawler/spider that visits several sites every day. While there the crawler should captures data in a systematic way through parsing the DOM, using regular expressions
326.
▲
by
mswen
13y ago
In many ways you are in an enviable position. You can take pride in the fact that you worked hard (at something you apparently didn't like)but you worked hard enough and competently enough that you got paid really well. Second you ca
327.
▲
by
mswen
13y ago
Seems like a logical approach to me. Also once you get the data set built you could algorithmically generate new recipes. If one type of machine learning algorithm generates more pleasing new cocktails you have a winner. Just don't d
328.
▲
by
mswen
13y ago
What I would do is build a crowd-sourced data set of what ingredients go well together and which ingredients people don't like together (rate it from 1 to 10, where 1 means these ingredients should never be used together to 10 means th
329.
▲
by
mswen
13y ago
Maybe I don't understand your situation but unless you have variation you don't stand a chance of developing a predictive model. In the dependent variable you need at least 0, 1. And in independent variables you need variation.
330.
▲
by
mswen
13y ago
An article like this reminds me of some of my own roots in consumer behavior research and classic market research. Formal models of finding feature-to-market fit are not particularly new. One that comes to mind is the Kano Model. I tracke
More ›