551 karma · joined November 10, 2012
Anyhow, I'm surprised. I'm a Hardware Engineer myself, so how could I miss that!
A sample entry could be:
company: 'Some Company',
jobs: [
{
dev_type: 'Web/Mobile/Data',
dev_sub_type: 'Frontend/Backend/DevOps/Android/iOS',
visa: 'Required/Not required/Transfer only/Sponsored',
remote: 'Yes/No/Maybe'
locations: [
'SF',
'London',
...
]
},
...
]
All posts could have a METADATA: compressed_json entry that can be processed by the site and displayed/filtered accordingly. Perhaps it could be built manually at the beginning until it catches up.Sure, but the migration was running for 15 minutes after we decided to reboot it. It was most probably trapped in a deadlock. Keep in mind it was just a column rename, no indexes involved. Should be a simple operation:
> Alterations that modify only table metadata and not table data are immediate because the server only needs to alter the table .frm file, not touch table contents. The following changes are fast alterations that can be made this way:
> - Renaming a column, except for the InnoDB storage engine before MySQL 5.6.6.
A while back we also migrated a column, the encoding was set to ISO-8859-10 (nordic countries? seriously?), even though our encoding is UTF-8 everywhere! Had to take a DB snapshot, launch a new instance and re-run the migration (there were performance issues with the instance as well, that's why we re-created it).
These are just a few of the stories I've experienced. I didn't contact support after seeing that similar issues were completely ignored in the AWS forums.
Full setup:
- 2x m3.large app servers
- 1x m3.large RDS instance running MySQL 5.6.22, InnoDB, no read replica, no Multi AZ. Other issues were seen in an m3.medium with Multi AZ. We preferred a non-crashing m3.large Single AZ than a hanging m3.medium with Multi AZ. Seems that there are problems regardless of the instance's size.
- Rails 4.0.11 backend, ActiveRecord migration
- The table had about 1000 entries, but heavily used (users table). To give you some context ours is a B2B service, most of the users (90%) use our site from 9 to 5, never leave the page. So accessing the users database is fairly common.
Our experience with RDS hasn't been that smooth: lots of sudden performance issues, table locks and whatnot. DB reboot or even full restore is a once a month thing.
https://jlhonora.github.io/stackexchange-predictions/
I didn't know this data existed, and was a bit turned down when I saw the data was in JSON format. It looks much better with a graph! Check it out and tell me what you think. Any suggestions?
You're not the first person to tell me that. They always tell me they're kidding, but I guess there's some truth to it :) . Please, could you elaborate on your thoughts?
When Antu came home we had an open space for him, where he could run somehow freely. Then winter came and we had to bring him inside our apartment. After a while, I built a wheel for him (not the same one as he has now) and he ran like crazy. Now when we put him in the open space he prefers to run on the wheel. I always wonder if he really enjoys it or we just messed him up. It seems that it is the former:
http://animals.io9.com/wild-animals-sneak-onto-laboratory-ru... http://www.theguardian.com/science/grrlscientist/2014/may/21...
I wanted two things from this project: get it up and running as quickly as possible and learn Go. That's why I programmed an MVP in Python while I finished the backend in Go. I figured that if I waited for the Go code to be finished I may have never shipped it.
I'm still going to use Python for the raspberry Pi, since I don't know any serial port libraries for Golang (well, a quick google search shows me an option: https://code.google.com/p/goserial/) and anyways that part is already working fine.
But if I find the time then yeah, I'd totally rewrite it entirely in Go. But there's many things I'd do before that:
- Live streaming with a Pi NoIR camera (WebRTC + a paid CDN I guess).
- Visualization and data analysis. I'm thinking github-style punchcard for the running activity, and when you click a day you can analyse the minute-by-minute laps.
- Correlation with other sensors. I have temperature, humidity, and ambient light sensors, which could be use to correlate the activity or even predict it.
- Public API.
Check him out at https://twitter.com/runhedgie
This project is a combination of hardware and software. I'm using a Raspberry Pi, a custom-built wireless node based on Arduino, Python, Redis, and a Go API for data analysis. I wanted to build almost everything from scratch, to really see what's happening in every part of the system, so I've done from PCB design and soldering to struggling with Go HTTP routers (I gave up and used Gorilla).
But it's been fun watching the hedgehog interact with the real world.
https://github.com/jlhonora/iot https://github.com/jlhonora/iot-go
I'm already a startup founder at http://satelinx.com and http://beetrack.in , apart from teaching Computer Architecture. I do have experience making and selling hardware, what I don't have is free time :)
I'm preparing a more technical description about the system (I designed and built the hardware myself), I should post it on the feed over the weekend.
As for the dynamo, definitely! But I'd rather let the hedgehog run as smoothly as possible.
And I'm trying to keep up with my jogging as well :)
The activity graph (shown on the twitter feed) is for that same night, which seems to be quite consistent over time. There are no peaks either, so I think it should be accurate. After all, he does run all night.
The phrases are random, check the code for them here: https://github.com/jlhonora/iot/blob/master/rpi/twitter_mana... . The "mood" for the phrases depends on the ran distance.
The graph I posted was generated with matplotlip and I tweeted it manually. When I finish struggling with D3 I'll put an on-demand graph with the data.
If you liked this article perhaps you'll be interested in a project I've been working on. My wife has a Hedgehog (yeah, like Sonic) and he's pretty shy. He also has a wheel and runs ALL night.
To feel that he interacts a little more I connected a wireless node with a magnetic switch so I could count how many laps he runs every night, and therefore the equivalent distance. Every morning he tweets how much he ran, just like RunKeeper :)
Check out his twitter account here: https://twitter.com/runhedgie
And, of course, the code is hosted on github: https://github.com/jlhonora/iot
I just finished this project yesterday, but I have many more ideas to add. I'd like to correlate running activity with temperature (hedgies are pretty sensitive to cold weather) and ambient light. I'd also like to make a public API so that anyone can analyze his activity patterns.
So, stay tuned and follow!
Fortunately, they haven't made any comments on the mobile app connection/data storage yet :) . That's a bigger problem I guess, since we need the phone's internal DB to store some data (it needs to work offline). I mean, they are secure enough, but given our client's previous records I don't think they'll share our opinion.
Really overpricing it is something I didn't think of. We were just thinking of charging the man-hours and a little extra.
> 1. Why do you not want to do it?
Costs of maintenance, portability issues, etc.
> 2. Why do they want to do it that way?
Privacy concerns.
Oh, and, thanks!