I would be curious to see the details of the request and response from
Firebug, Wireshark or similar. I have never received anything other
than an image as a response. Since a browser is making the request for
the photo, it should cache it.
Nice idea, but, for many reasons, Twitter couldn’t “retweet” something
that isn’t a tweet at all.
Jason, I’m glad you mentioned this, because we had neglected to ask
for compressed data in TalkingPuffin.
For others wanting to do this (in Java/Scala):
import java.util.zip.GZIPInputStream
conn.setRequestProperty(Accept-Encoding, gzip)
...
val is = conn.getInputStream
val ce
BY ME: http://twitter.com/statuses/retweeted_by_me.xml?count=200 works
just fine.
TO ME: /statuses/retweeted_to_me.xml?count=10 works fine.
User dcbriccetti
I pointed this out on Oct 24th.
http://groups.google.com/group/twitter-development-talk/browse_frm/thread/de91a4569fb10e5d/f8fabf89447a85c9?lnk=gstq=briccetti#f8fabf89447a85c9
Glad you have spotted the trouble.
I like your scripts. Thanks. I like to see other people’s Twitter
code.
Cursoring is working, but it seems wrong to get just 20 at a time.
GET /abdur/research/members.xml?cursor=-1
GET /abdur/research/members.xml?cursor=1316585587587157646
yields the same as
GET /abdur/research/members.xml?count=200cursor=-1
GET
Indeed.
I know every little bit helps when massively scaling, but, speaking
generally, I dislike scrolling through small chunks. Like in the
Twitter Web site just now, looking at lists, I have to scroll in small
chunks to see who’s in a list. (Other than that Twitter’s Web site
support for lists
I’ve learned that wrapping every Twitter Web service call with retry
logic is essential.
http://github.com/dcbriccetti/talking-puffin/blob/master/twitter-api/src/main/scala/org/talkingpuffin/twitter/Http.scala
I doubt I will follow many (or any) lists. But I will scour them
periodically for individuals to follow.
I also recommend using Wireshark, tcpdump or the like to get an
authoritative picture of what’s happening. And if there is indeed a
bug, the output serves as clear proof.
By the way, I changed this code to do ten or fifteen of these at a
time, using multiple threads, and it’s much, much faster.
How can I retrieve more than 20 at a time?
?cursor=-1count=200 has no effect
Welcome the the Twitter Development Talk
I meant to send that to list owner. Sorry.
Google group admins can actually DELETE spam, too, which would be
nice.
Bulk add would be great. Here’s a log from doing individual adds.
Notice there were many over capacity errors. And that it took about a
minute 42 to add about 60 users. If I could do all in one request,
perhaps that would help reduce the over capacity errors.
I’m enjoying lists. Thanks for the
Multiple identical list create requests result in multiple lists,
differing in slug/url/full name.
Perhaps it should ignore, or return an error?
$ curl -u dcbriccetti -d name=Scala http://twitter.com/dcbriccetti/lists.xml
?xml version=1.0 encoding=UTF-8?
list
id29926/id
nameScala/name
Once Twitter posted details of the API in this public forum, that
opened the door to our discussing lists publicly (if it wasn’t open
already). And I’ve been doing so. I even made a video showing
TalkingPuffin all of Robert Scoble’s lists at once on a big screen:
Sorry, that video is here: http://www.youtube.com/watch?v=NUVDP5sX2rA
Copy/paste error.
Thanks, Marcel. I’ve started playing with this and it is working
nicely so far.
It would be nice to have a list count included in the user object so
one doesn’t have to make a separate call for each user to see if that
user has lists. Or something else that solves the same problem.
If anyone’s interested, here’s the commit for TalkingPuffin for
preliminary support for displaying lists:
http://github.com/dcbriccetti/talking-puffin/commit/ea86bfb523d7aea98df6ebae52893c16f53fa542
A picture of the feature:
http://www.flickr.com/photos/dcbriccetti/4015781495/
DELETE http://twitter.com/friendships/destroy/USER.xml
works, but
DELETE http://api.twitter.com/1/friendships/destroy/USER.xml
fails with
You don't have permission to access /1/friendships/destroy/USER.xmlon
this server.
Alex Payne told me several months ago at a Bay Area Scala Enthusiasts
meeting that parallel operations are fine.
Sigh. I wrote a polling solution to fetch the rate limit info before
discovering that the info is in the headers of every HTTP response.
If you’re interested, here’s a commit to TalkingPuffin where I removed
the polling and used the headers instead:
Nalin, I appreciate the ideas. Remember, you can’t DM somebody who
isn’t following you. I want to maintain a fairly complete collection
of filters for the common “noise tweets,” and use it in TalkingPuffin
under user control, and invite other application developers to use it
as well.
Maybe as
It sounds like you need to get the developer of your Twitter client in
here so we can help him.
I don’t know the background, but I do wish they would split that into,
say, sourceName and sourceUrl.
http://github.com/dcbriccetti/talking-puffin/blob/master/twitter-api/src/main/scala/org/talkingpuffin/twitter/TwitterStatus.scala
A Twitter client can do an HTTP get to here:
http://talkingpuffin.appspot.com/filters/noise
and expect lines of plain text like this:
Just joined a twibe. Visit http\://twibes\.com/.*
just joined a video chat at http\://tinychat\.com.*
Please submit your candidate “noise tweets”:
http://bit.ly/NoiseTweets
I’ll review them and add them to the repository.
These should generally be messages of the type of viral advertising—
some service doing something for a Twitter user and then blabbing
about it through the careless or
I detest tweets like these:
just joined a video chat at http://xxx Make your own video chat at
http://xxx #xxx
just joined a twibe ...
I am thinking of starting a repository of regular expressions matching
noise-tweets like these, that Twitter clients could query via a Web
Service, and the
Participating Twitter clients could have a “Noise” button to submit
the tweet being read as a candidate for inclusion in the filtering.
Do take a look and join us if you’re interested. We’re using the Scala
language.
http://TalkingPuffin.org/
I would like to know the same. I have a recently-renamed desktop
application. Twitter’s OAuth support is in beta, no? Doesn’t mean the
same as Google.
I would like “talkingpuffin” and “simpletwitterclient” to map to
“TalkingPuffin”. Twitter, can you help? Thanks.
Try this:
http://simple-twitter-client.appspot.com/
Perhaps this will save time for some of you.
http://briccetti.blogspot.com/2009/04/my-first-scala-web-app-on-google-app.html
and activate
them, and filter the feed using them. But if that’s not gonna happen
right away, I’d like to see if the development community can do
this.
What do you think of all of this, especially the idea of Twitter
client developers allowing users to set a proxy for twitter.com?
Dave Briccetti
http
Hi. I’ve searched around for 1/2 hour or so, and haven’t found an
authoritative explanation of the sizes of pictures, and how to
retrieve them.
It seems that profile_image_url leads to a tiny picture:
You can now run it without building it yourself, via Java Web Start:
http://davebsoft.com/applications/simple-twitter-client. Feedback
welcome, through Twitter, email or my blog.
Thanks, Alex. I’ve added some simple tagging and filtering features.
Here’s a demo video:
http://briccetti.blogspot.com/2009/02/tagging-and-filtering-features-in-my.html
40 matches
Mail list logo
