HN Explorer

39 comments

[ 3.0 ms ] story [ 94.0 ms ] thread
michaelcampbell
The trailing close paren is becoming part of the bit.ly links. (At least on IE and Firefox, for me.)
mrduncan
Twitter has always included trailing parens in links (even though, it seems more often than not they're not supposed to be). On the bright side, bit.ly ignores them so you still get the originally intended link.

I'm pretty sure trailing square brackets are ignored - anyone know for sure?

jmillerinc
Thanks for catching that. I've changed the code to use square brackets instead of close parens.
michaelcampbell
It's doing it on the square brackets too. =) It's not a huge deal; as another person replied to me, bitly seems to ignore them and get you where you're going.
oscardelben
This is actually useful. Care to share how you implemented it?
jmillerinc
It's a Python script I run every 5 minutes with cron. It uses BeautifulSoup to scrape the newsyc front page, Tweepy to send out out tweets, and sqlite3 to keep track of which stories have already been tweeted.
mike-cardwell
Why use an sqlite DB when you can use the twitter feed its self to check whether or not a story has been posted?
jmillerinc
Because twitter is unstable, and the api is rate-limited.
virrgil
Could you share the source please?
protomyth
I like how you included both links. The problem I had with the other twitter feed is that it linked only to the original story and not to Hacker News.
jmillerinc
Thx. The tweets will also contain an attribution if the submitter lists a twitter handle in their HN profile.
ck2
Clever but how about a real RSS feed of those so I don't have to go to an RSS feed, then twitter, then bitly, then the original link?

Especially since with the shortened URL I don't really know where I am going.

w1ntermute
Clever but how about a real RSS feed of those so I don't have to go to an RSS feed, then twitter, then bitly, then the original link?

Yes, use Readability to parse the pages and show the full text of the article directly in the RSS feed.

jmillerinc
The Readability parse is a good idea, but implementing that would need more time than I can spare right now. If anyone wants to do the legwork on that, please contact me and I'll be happy to put it in the feed.
dschobel
Well done!

Any idea how difficult it would be to support an arbitrary user customizable point threshold?

jmillerinc
I'd have to create multiple twitter feeds for different point thresholds. If there's a particular one you have in mind, let me know.
Hopka
Something like 100 - so it's only 2 or 3 stories per day would be nice.
joe_the_user
Here's a challenge.

Create an app that lets you request an average of n stories per day and translates that request into a threshold.

dschobel
Yep, that or create a simple web-app where users can log in (using OAuth preferably) and configure the number of stories they see a day or the karma threshold and then spin up a custom RSS feed for them.

Might be a fun weekend project...

jmillerinc
Done. @newsyc100 (also have @newsyc50)
royrod
Nicely done, Jeff, useful!
conesus
How about if the original link is small enough to fit in without truncation (bit.ly), then show the original link as well as a bit.ly link to the comments.

I hate losing the domain information in bit.ly links, and the @hackernewsbot (http://twitter.com/hackernewsbot) does a nice job of showing the original URL when possible.

petercooper
http://twitter.com/newsycombinator is also good if you want anything that hits the front page (surprisingly it has almost 14,000 followers.. which seems sky high compared to the number of HN users to me.. :-))
avar
If it isn't too much trouble a version of this for 40 points (and maybe 60) would be great. I find 20 to be way too much traffic in my Twitter feed, so I unsubscribed from it.
credo
I just a newsyc20 tweet that said (via my twitter-id)

Figuring out the HN poster's twitter id and then including that in the tweet makes newsyc20 even more interesting