So much this!
asyncio was the main selling point for me, but in general, why not follow the language?
I never really understood the "rather stay with py2.7" thing.
I get it with big old monolithic applications. You don't "just" rewrite those, but
_every new python project_ should be done with the latest stable release.
Is anyone starting their PHP projects on PHP4?
Any new node projects in 0.10?
Of course not, that would be moronic.
Paying LastPass user here.
Not sure how this is going to go down.
TBH I'm hoping that nothing will change. Yes, the UX might not be the best in the world, but to me, the important thing is availability and security (probably not in that order).
A browser extension and a decent android app is what I need, and I already have that.
Althought that might be your preference, I'm fairly certain very few people feel the same.
I too want a phone that "works", but I'd never pay $1000+.
I can live with some sacrifices if i can get a Good Enough phone for ~$400.
And imo, OnePlus One is fits the bill quite well.
I was sceptical about no SD card and the non-removable battery, but when the battery lasts for days, and i have 64GB eMMC and online backup, I found I don't need those things anyway.
The opposite. To little supply for the demand.
They used the invites to "throttle" the purchase rate to be able to keep up.
They probably thought an invite system was a better aproach than a weekly free-for-all aka. SuiciDDOS against the webshop.
Hopefully, this time supply will be significantly higher, meaning the invite system will be removed as soon as they know they can keep up with demand.
aka: Please spend an hour in transport so you can spend 10 minutes fixing the issue while we breath down your neck and constantly ask stupid questions that won't help us solve anything.
Just wanted to drop a comment on the awesomeness of your app.
Hacker News 2 is by far the best Hacker News app, not just on Android, but on all mobile platforms i've tried (so, iOS, Android and Windows Phone)
Awesome work you are doing.
Mine expects you to know virtualenv and how it works, and also that you explicitly install it. Same goes for dependencies.
My needs where to avoid sourcing the virtualenv, so i wrote the tool to fix just that.
Also, i built mine to resolve symlinks and be usable as the shebang in your python scripts, which is pretty neat when writing command line tools.
Downside is, ofc, that compared to just running something inside a virtualenv shell, vpython takes forever (up to 100ms-ish on my machine) to find the virtualenv, and set the correct env variables.
Most of my projects are long running though (webservers, number crunchers, etc), so the extra miliseconds it takes to load is really not that important.