[twitter-dev] Sign in with Twitter - Using iPhone
Is there any way to use "Sign in with Twitter" from withing iPhone? I want the user of my application to be logged in using Twitter username/password. I am looking for similar functionality as "Facebook Connect for iPhone" provides.
[twitter-dev] Re: Help needed, to search photos in twitter
On Jun 4, 11:23 pm, Chad Etzel wrote: > Hi Chethan, > > I wrote the TwitPicGrid page you cite. I just do a simple Search API > search and append " twitpic" to the query. Since all twitpic posts > have "http://twitpic.com/x"; in the tweet, this search will get the > links in the result set. Then, since most people describe the content > of the picture in their tweet, there is a relatively high probability > that the keyword in the search will be featured in the picture > somehow. > > So basically, if you want twitpics of "lunch", just search twitter for > "lunch twitpic" and parse out the twitpic links in the results. > > -Chad > > On Thu, Jun 4, 2009 at 9:05 AM, Chethan wrote: > > > Hello, can anyone please tell me how to search for photos in twitter/ > > twitpic, (ex:http://www.twitpicsearch.com,http://tweetgrid.com/twitpicgrid), > > it would be really helpful if you could provide me with some sample > > code, I am using PHP and jQuery > > > Thankyou Thank you Chad, TwitPicGrid is really cool. Thanks for helping.
[twitter-dev] Re: Help needed, to search photos in twitter
On Jun 5, 7:09 am, Jonathan wrote: > Hello - > > I developed the application Twitcaps (http://twitcaps.com) and I can > tell you that what I am using in my search API calls is "twitpic.com", > "yfrog.com" or "twitgoo.com" (or any other arbitrary image provider, > img.ly, etc). I found that by including the ".com" at the end, I wound > up with more real image URLs and less mere conversational mentions of > "twitpic" to parse through. > > Unfortunately, I don't have any good sample code for you. I'm running > on Grails/Groovy. > > -jonathan > > On Jun 4, 7:05 am, Chethan wrote: > > > Hello, can anyone please tell me how to search for photos in twitter/ > > twitpic, (ex:http://www.twitpicsearch.com,http://tweetgrid.com/twitpicgrid), > > it would be really helpful if you could provide me with some sample > > code, I am using PHP and jQuery > > > Thankyou Thank you jonathan. It was really helpful
[twitter-dev] Re: Has anyone created a ping bot account?
A ping bot account that sends current time every secends.
#!/bin/sh
MSG=`date`
# Interval in seconds
INTERVAL=1
while [ 1 -lt 3 ]
do
curl -u TWITTERUSER:TWITTERPASS -d "status=${MSG}"
sleep ${INTERVAL}
done
Hope you get the idea.
--
A K M Mokaddim
http://talk.cmyweb.net
http://twitter.com/shiplu
Stop Top Posting !!
বাংলিশ লেখার চাইতে বাংলা লেখা অনেক ভাল
Sent from Dhaka, Bangladesh
[twitter-dev] Re: Larger Users Not Returning Follower Data
Jesse, If the implementation is to make that a preference which is turned off by default (no DM by non followers) that users can toggle, then i am totally for it. As you point out its then the users responsibility to clean their inboxes if they get hit by spam after turning the feature on. So for what it counts, I'm all in favor allowing DMs from non followers if its a preference users can control. Sean On Thu, Jun 4, 2009 at 10:54 PM, Jesse Stay wrote: > Sean, why not let the users decide that though? If I enable the option for > my account it's my responsibility to weed out the spam. If I don't want the > spam then I won't enable it on my account. Giving users multiple options is > a good thing. > > Jesse > > > On Thu, Jun 4, 2009 at 2:31 PM, Sean Scott wrote: > >> Just speaking from a user perspective, I'd love to see that debate about >> opening DM to senders who you are not following to the community as a whole >> or a representative subset of them. By opening DMs to non-followed >> twitters, it would be way to easy for spammers to start spamming via DMs. >> From a user perspective i don't see a compelling argument for opening DMS >> to folks i do not follow. >> Off my soap box >> >> >> On Thu, Jun 4, 2009 at 3:23 PM, Doug Williams wrote: >> >>> We would like users to be judicious with their following habits and only >>> follow users who contribute value to their timeline. This justifies the >>> following limits we impose. >>> We are aware that many users would like to accept all incoming directs. >>> This, along with the quid pro quo following to build community, capture the >>> majority of the use-cases for auto-following. We are discussing internally >>> how to best approach these two uses within the bounds of the product we are >>> trying to build. At this time we have nothing to report but know we are >>> actively thinking about these ideas. >>> >>> Thanks, >>> Doug >>> >>> >>> >>> >>> >>> On Thu, Jun 4, 2009 at 2:12 AM, Jesse Stay wrote: >>> Also, how do you recommend we deal with the larger users that would like to follow back their followers? With the hard limit of 1,000 follows per day, there is no way they'll ever catch up, as some of them have more than 1k new followers per day as is. If this limit were more dynamic based on the size of the user that would be nice. Capabilities to follow people in bulk may also help. Of course, I think many of these would no longer need to follow back if they could just have the option to enable anyone to DM them if they choose. I think that's the underlying cause to want to auto-follow for most people. The only other cause is for an additional token/feeling of community, although I think many would be willing to forgo that if they had the ability to just allow everyone to DM them - it feels good to have someone you admire follow you back, even if it's not 100% sincere. Jesse On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: > Yes, that's what appears to be happening. My experience starts at > around 500K+. I'm okay with waiting with my script if you guys need to > take > longer to retrieve the info. Or if you'd prefer we paginate I'll start > doing that as well. Maybe a hard limit of 200K and you have to Page to > get > above that? > Jesse > > > On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: > >> I've heard that list sizes greater than 150K-200K start to return >> timeouts at higher rates. Although I'd enjoy hearing first-hand >> experiences >> and recommendations. >> Thanks, >> Doug >> >> >> >> >> On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: >> >>> In my case specifically it's the Social Graph methods. I didn't >>> realize you had paging available now. Is there some logic as to when I >>> should expect to page and when I can just rely on the full result? >>> Jesse >>> >>> >>> On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: >>> What methods in particular are you referring to? The social graph methods now support paging so retrieving all of that data is now possible, where it used to throw 502s. It does however require a bit of application logic to assume when paging is necessary (e.g. large follower counts). Additionally, we are making changes to the databases which cause latency that result in periodic 502s. We are not able to give definitive ETAs on these fixes due to priorities that change as unforeseeable critical needs arise. More specificity would be beneficial. Do you have a replaceable bug, problem, or suggestion that you would like to discuss? Thanks, Doug On Wed, Jun 3, 2009
[twitter-dev] Re: Larger Users Not Returning Follower Data
Sean, why not let the users decide that though? If I enable the option for my account it's my responsibility to weed out the spam. If I don't want the spam then I won't enable it on my account. Giving users multiple options is a good thing. Jesse On Thu, Jun 4, 2009 at 2:31 PM, Sean Scott wrote: > Just speaking from a user perspective, I'd love to see that debate about > opening DM to senders who you are not following to the community as a whole > or a representative subset of them. By opening DMs to non-followed > twitters, it would be way to easy for spammers to start spamming via DMs. > From a user perspective i don't see a compelling argument for opening DMS > to folks i do not follow. > Off my soap box > > > On Thu, Jun 4, 2009 at 3:23 PM, Doug Williams wrote: > >> We would like users to be judicious with their following habits and only >> follow users who contribute value to their timeline. This justifies the >> following limits we impose. >> We are aware that many users would like to accept all incoming directs. >> This, along with the quid pro quo following to build community, capture the >> majority of the use-cases for auto-following. We are discussing internally >> how to best approach these two uses within the bounds of the product we are >> trying to build. At this time we have nothing to report but know we are >> actively thinking about these ideas. >> >> Thanks, >> Doug >> >> >> >> >> >> On Thu, Jun 4, 2009 at 2:12 AM, Jesse Stay wrote: >> >>> Also, how do you recommend we deal with the larger users that would like >>> to follow back their followers? With the hard limit of 1,000 follows per >>> day, there is no way they'll ever catch up, as some of them have more than >>> 1k new followers per day as is. If this limit were more dynamic based on >>> the size of the user that would be nice. Capabilities to follow people in >>> bulk may also help. >>> Of course, I think many of these would no longer need to follow back if >>> they could just have the option to enable anyone to DM them if they choose. >>> I think that's the underlying cause to want to auto-follow for most people. >>> The only other cause is for an additional token/feeling of community, >>> although I think many would be willing to forgo that if they had the ability >>> to just allow everyone to DM them - it feels good to have someone you admire >>> follow you back, even if it's not 100% sincere. >>> >>> Jesse >>> >>> >>> On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: >>> Yes, that's what appears to be happening. My experience starts at around 500K+. I'm okay with waiting with my script if you guys need to take longer to retrieve the info. Or if you'd prefer we paginate I'll start doing that as well. Maybe a hard limit of 200K and you have to Page to get above that? Jesse On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: > I've heard that list sizes greater than 150K-200K start to return > timeouts at higher rates. Although I'd enjoy hearing first-hand > experiences > and recommendations. > Thanks, > Doug > > > > > On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: > >> In my case specifically it's the Social Graph methods. I didn't >> realize you had paging available now. Is there some logic as to when I >> should expect to page and when I can just rely on the full result? >> Jesse >> >> >> On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: >> >>> What methods in particular are you referring to? The social graph >>> methods now support paging so retrieving all of that data is now >>> possible, >>> where it used to throw 502s. It does however require a bit of >>> application >>> logic to assume when paging is necessary (e.g. large follower counts). >>> Additionally, we are making changes to the databases which cause latency >>> that result in periodic 502s. We are not able to give definitive ETAs on >>> these fixes due to priorities that change as unforeseeable critical >>> needs >>> arise. >>> More specificity would be beneficial. Do you have a replaceable bug, >>> problem, or suggestion that you would like to discuss? >>> Thanks, >>> Doug >>> >>> >>> >>> On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: >>> I was discussing this with Iain, and have also talked about it with Damon, so I know I'm not alone in this. I am having huge issues retrieving follower and friend data for the larger users (1 million+ followers), most of the time returning 502 Bad Gateway errors. I know there are a few of these users getting really frustrated about our apps not being able to retrieve data for them. Is there a plan to fix this? Is the API team aware of this? Any ETA by chance? Thank
[twitter-dev] Has anyone created a ping bot account?
Has anyone created a ping bot account that sends tweets every minute or so? I could create my own... but something about a wheel, and reinventing... -Chad
[twitter-dev] Re: random sampling of users....do we know anything about user id range?
I'm hoping that Twitter counts "users" when reporting their numbers, and not "accounts". The reason being that I've signed up probably... 5 accounts myself (main, API testing, business, etc, etc). I'm not sure how many the average user signs up, but it's definitely on average more than 1. I'm wondering if IDs are sequential for users afterall. Not using system generated primary keys if I remember right puts more strain on the system as it has to check uniqueness and generate a number (not hard, but still), and most of Twitter is all about scaling and speed as I see it. Otherwise it seems on creating a new user, they are taking the last ID, and adding an artibrary number to it (1d20?) for the next user ID. 45M seems like a lot of users, but I could see there being that many "accounts". perhaps On Jun 4, 11:44 am, Nick Arnett wrote: > On Wed, Jun 3, 2009 at 7:13 PM, TechRavingMad wrote: > > > > > There are a little over 44.5 million twitter IDs as of right now > > (10:10pm cst 6/3/9) with what seems to be about 10 being added every > > second. > > However, Twitter has been quite clear about not saying if status IDs > correspond to the actual number of statuses, so I'd guess that they're > equally circumspect about whether or not the number of user IDs corresponds > to the number of users. In other words, we can be sure there are not more > than 44.5 million users, but we don't know how much lower the actual number > is. We don't know if all IDs have been used... and even Twitter doesn't > know how many of those IDs belong to the same users. > > I would think that if one wants a random sample of users, one would have to > propose a selection method and ask Twitter if there's any reason that it > would introduce a selection bias... and hope that they are willing to reply. > > Seems to me that the biggest problem would be to include "quiet" users, > since only those who post in public become visible. > > NIck
[twitter-dev] Re: Help needed, to search photos in twitter
Hello - I developed the application Twitcaps (http://twitcaps.com) and I can tell you that what I am using in my search API calls is "twitpic.com", "yfrog.com" or "twitgoo.com" (or any other arbitrary image provider, img.ly, etc). I found that by including the ".com" at the end, I wound up with more real image URLs and less mere conversational mentions of "twitpic" to parse through. Unfortunately, I don't have any good sample code for you. I'm running on Grails/Groovy. -jonathan On Jun 4, 7:05 am, Chethan wrote: > Hello, can anyone please tell me how to search for photos in twitter/ > twitpic, (ex:http://www.twitpicsearch.com,http://tweetgrid.com/twitpicgrid), > it would be really helpful if you could provide me with some sample > code, I am using PHP and jQuery > > Thankyou
[twitter-dev] Re: Forgive Forgive Forgive - XML 2 CSV XSL
http://snippets.dzone.com/posts/show/3701 Gets it done in Ruby. I haven't tested it, but it looks right and simple On Jun 4, 4:04 pm, Mark wrote: > I would like to take the follower .xml output and convert to csv. > > From googling I believe I need to create an XSL or XSD and then use > MSXML > > example: > > MSXSL C:\twitterout\adventuregirl69.txt C:\Users\mark\Documents > \twitteratom2txt.xsl –o out.xml > > I don't know xsl trying to learn from examples, does anyone have > any .XSLs they created for Twitter's API responses? > > Would like to create flat file output from xml response. > > Goal is to tie a zip code to a twitter profile like @geofollow
[twitter-dev] Re: how to deal with true ??
I don't the Twitter web allows for more then 140 characters to be posted anymore. On Sat, May 30, 2009 at 04:27, Jochen Kaechelin wrote: > > Let's say I have a update with more then 140 characters. > Twitters webfrontend will cut the status and add "..." which > points to > >http://twitter.com/userename/status/xxx > > where I can see the complete update. > > Twittelator, ..., ... and my own client only show the three dots > without > and additional information. > > Is there a way to get the remaining characters or must I show my users > the > link to http://twitter.com/userename/status/xxx ? > > Thanx > > > -- Abraham Williams | http://the.hackerconundrum.com Hacker | http://abrah.am | http://twitter.com/abraham Project | http://fireeagle.labs.poseurtech.com This email is: [ ] blogable [x] ask first [ ] private. Sent from Madison, Wisconsin, United States
[twitter-dev] problems with "trend current" api method
I have a problems with the some API methods: specifically, "trend current" , "trend daily, and "trend weekly" methods.For starters, my code will retrieve the JSON string containing the "trend current" json string. Using Curl, the Json string response is captured and decode into a php array using the json_decode() function in php. My problem seems to come from some of the array keys that were generated; specifically, the timestamps array keys , for example, ['2009-5-14] or ['2009-5-14 4:49:15]( these are just examples). I know they are generated from 'Unix' timestamps, and I can generate them independently. however, my question is: How can I duplicate and access these array keys? Any help would be very much appreciated
[twitter-dev] Re: Enable ability to block apps via Twitter or the API
Sorry, but I still cant agree on why asking for a API key on the normal API
cannot solve this. A whole application can be banned/throttled/controlled
using the API key if needed this way. At present applications register and
gets API keys anyway, so all this will do is add an extra layer of
authentication on API calls.
I think this is more an obsession with OAuth. ;-)
Nayeem
2009/6/2 Doug Williams
> Chad is correct. Until we have everyone pushed through a funnel where API
> keys are required or applications can be deduced (as with OAuth) we have no
> way of knowing which application actually sent an update or DM in some
> cases. Furthermore, we don't have the notion of tweet level spam reporting.
> Currently users are only able to flag accounts a spam through "@spam
> @username" or "d spam @username" updates.
> So, until we develop tools to deal with spam on a per tweet-basis and have
> every application going through a pipe that we can control, application
> blocking is not a valuable use of our resources.
> Thanks,
> Doug
> --
>
> Doug Williams
> Twitter Platform Support
> http://twitter.com/dougw
>
>
>
> On Tue, Jun 2, 2009 at 12:45 PM, Chad Etzel wrote:
>
>>
>> No, it can't be required. Worse yet, it can be spoofed w/ basic auth,
>> so a "blocked" app could just change it's source parameter and appear
>> as something like TweetDeck.
>>
>> -Chad
>>
>> On Tue, Jun 2, 2009 at 3:36 PM, Developer In London
>> wrote:
>> > Couldnt the app-id be made a required parameter for the API calls? That
>> way
>> > it can still work with basic auth.
>> >
>> > 2009/6/2 Doug Williams
>> >>
>> >> Floated the idea. Until we funnel everyone through OAuth (that means no
>> >> Basic Auth) this really isn't possible. It's something we'll keep in
>> our
>> >> back pockets for the long-term.
>> >> Great suggestion though, Jesse.
>> >> Cheers,
>> >> Doug
>> >> --
>> >>
>> >> Doug Williams
>> >> Twitter Platform Support
>> >> http://twitter.com/dougw
>> >>
>> >>
>> >>
>> >> On Tue, Jun 2, 2009 at 8:01 AM, Carlos wrote:
>> >>>
>> >>> agreed, I'd like this as well.
>> >>>
>> >>> On May 31, 6:52 pm, Jesse Stay wrote:
>> >>> > Not going to name names, but there are a few really noisy apps out
>> >>> > there
>> >>> > right now. It would be really nice if, via either the API (my
>> >>> > preference as
>> >>> > it would be less work on your part and fits well with my app), or
>> the
>> >>> > UI,
>> >>> > you enabled users to block receiving Tweets generated from specific
>> >>> > apps.
>> >>> > This would then punish the app developers for creating spammy apps
>> and
>> >>> > not
>> >>> > the users themselves for just using what was put out there, making
>> it
>> >>> > much
>> >>> > less of a mess to control. Facebook does this, as does FriendFeed.
>> >>> > Any
>> >>> > chance you could enable this (please???) for Twitter?
>> >>> >
>> >>> > Thanks,
>> >>> >
>> >>> > @Jesse
>> >>
>> >
>> >
>> >
>> > --
>> > cashflowclublondon.co.uk
>> >
>> > ("`-''-/").___..--''"`-._
>> >`6_ 6 ) `-. ( ).`-.__.`)
>> >(_Y_.)' ._ ) `._ `. ``-..-'
>> > _..`--'_..-_/ /--'_.' ,'
>> > (il),-'' (li),' ((!.-'
>> > .
>> >
>>
>
>
--
cashflowclublondon.co.uk
("`-''-/").___..--''"`-._
`6_ 6 ) `-. ( ).`-.__.`)
(_Y_.)' ._ ) `._ `. ``-..-'
_..`--'_..-_/ /--'_.' ,'
(il),-'' (li),' ((!.-'
.
[twitter-dev] Re: Larger Users Not Returning Follower Data
To provide a slightly different user perspective, I think it would be a nice user-level setting to be able to "accept DMs from users I do not follow" That said, I also understand the potential complexity and performance issues such a setting could present, so I'm not expecting it any time soon, but I'm just injecting my opinion into the conversation. On Thu, Jun 4, 2009 at 14:31, Sean Scott wrote: > Just speaking from a user perspective, I'd love to see that debate about > opening DM to senders who you are not following to the community as a whole > or a representative subset of them. By opening DMs to non-followed > twitters, it would be way to easy for spammers to start spamming via DMs. > From a user perspective i don't see a compelling argument for opening DMS > to folks i do not follow. > Off my soap box > > > On Thu, Jun 4, 2009 at 3:23 PM, Doug Williams wrote: > >> We would like users to be judicious with their following habits and only >> follow users who contribute value to their timeline. This justifies the >> following limits we impose. >> We are aware that many users would like to accept all incoming directs. >> This, along with the quid pro quo following to build community, capture the >> majority of the use-cases for auto-following. We are discussing internally >> how to best approach these two uses within the bounds of the product we are >> trying to build. At this time we have nothing to report but know we are >> actively thinking about these ideas. >> >> Thanks, >> Doug >> >> >> >> >> >> On Thu, Jun 4, 2009 at 2:12 AM, Jesse Stay wrote: >> >>> Also, how do you recommend we deal with the larger users that would like >>> to follow back their followers? With the hard limit of 1,000 follows per >>> day, there is no way they'll ever catch up, as some of them have more than >>> 1k new followers per day as is. If this limit were more dynamic based on >>> the size of the user that would be nice. Capabilities to follow people in >>> bulk may also help. >>> Of course, I think many of these would no longer need to follow back if >>> they could just have the option to enable anyone to DM them if they choose. >>> I think that's the underlying cause to want to auto-follow for most people. >>> The only other cause is for an additional token/feeling of community, >>> although I think many would be willing to forgo that if they had the ability >>> to just allow everyone to DM them - it feels good to have someone you admire >>> follow you back, even if it's not 100% sincere. >>> >>> Jesse >>> >>> >>> On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: >>> Yes, that's what appears to be happening. My experience starts at around 500K+. I'm okay with waiting with my script if you guys need to take longer to retrieve the info. Or if you'd prefer we paginate I'll start doing that as well. Maybe a hard limit of 200K and you have to Page to get above that? Jesse On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: > I've heard that list sizes greater than 150K-200K start to return > timeouts at higher rates. Although I'd enjoy hearing first-hand > experiences > and recommendations. > Thanks, > Doug > > > > > On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: > >> In my case specifically it's the Social Graph methods. I didn't >> realize you had paging available now. Is there some logic as to when I >> should expect to page and when I can just rely on the full result? >> Jesse >> >> >> On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: >> >>> What methods in particular are you referring to? The social graph >>> methods now support paging so retrieving all of that data is now >>> possible, >>> where it used to throw 502s. It does however require a bit of >>> application >>> logic to assume when paging is necessary (e.g. large follower counts). >>> Additionally, we are making changes to the databases which cause latency >>> that result in periodic 502s. We are not able to give definitive ETAs on >>> these fixes due to priorities that change as unforeseeable critical >>> needs >>> arise. >>> More specificity would be beneficial. Do you have a replaceable bug, >>> problem, or suggestion that you would like to discuss? >>> Thanks, >>> Doug >>> >>> >>> >>> On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: >>> I was discussing this with Iain, and have also talked about it with Damon, so I know I'm not alone in this. I am having huge issues retrieving follower and friend data for the larger users (1 million+ followers), most of the time returning 502 Bad Gateway errors. I know there are a few of these users getting really frustrated about our apps not being able to retrieve data for them. Is there a pl
[twitter-dev] Re: Larger Users Not Returning Follower Data
On Thu, Jun 4, 2009 at 1:23 PM, Doug Williams wrote: > > We are aware that many users would like to accept all incoming directs. Sounds like a checkbox in your profile to me.
[twitter-dev] API Changes for June 4, 2009
Hi All, You can now take saved searches anywhere you go thanks to Matt's efforts: - Feature (REST): Added access to saved search data: saved_searches, saved_searches/show, saved_searches/create, and saved_searches/destroy. See also: Google Code Issue 605: http://code.google.com/p/twitter-api/issues/detail?id=605 See also: API Documentation: http://apiwiki.twitter.com/Twitter-API-Documentation Thanks, Doug -- Doug Williams Twitter Platform Support http://twitter.com/dougw
[twitter-dev] Re: Search API results return more info
This would probably heavier for Twitter to add since the to_user info is already included in the status info but the profile_image_url would have to be looked up to be added. It should be easy once the APIs merge. You should open an issue in the Google Code project so Twitter can keep track of the request. On Mon, Jun 1, 2009 at 12:27, Coderanger wrote: > > The search results already have "profile_image_url" which is the > "from_user" profile image, but could the "to_user" have its profile > image url included as well. This avoids me making multiple calls and > hitting your server unnnecessarily. > > It would clearly only be included/filled when there is a "to_user". I > think the bandwidth overhead would be minimal, if not it could be > added thru an extra parameter "&extended_user_data=1" or > "&include_to_user_profile" so it could be selectively used if > required. > > This addition enables me (http://coderanger.com/twitcher) or other > apps to properly attribute the search results like a normal user > status. > -- Abraham Williams | http://the.hackerconundrum.com Hacker | http://abrah.am | http://twitter.com/abraham Project | http://fireeagle.labs.poseurtech.com This email is: [ ] blogable [x] ask first [ ] private. Sent from Madison, Wisconsin, United States
[twitter-dev] Re: Larger Users Not Returning Follower Data
Just speaking from a user perspective, I'd love to see that debate about opening DM to senders who you are not following to the community as a whole or a representative subset of them. By opening DMs to non-followed twitters, it would be way to easy for spammers to start spamming via DMs. From a user perspective i don't see a compelling argument for opening DMS to folks i do not follow. Off my soap box On Thu, Jun 4, 2009 at 3:23 PM, Doug Williams wrote: > We would like users to be judicious with their following habits and only > follow users who contribute value to their timeline. This justifies the > following limits we impose. > We are aware that many users would like to accept all incoming directs. > This, along with the quid pro quo following to build community, capture the > majority of the use-cases for auto-following. We are discussing internally > how to best approach these two uses within the bounds of the product we are > trying to build. At this time we have nothing to report but know we are > actively thinking about these ideas. > > Thanks, > Doug > > > > > > On Thu, Jun 4, 2009 at 2:12 AM, Jesse Stay wrote: > >> Also, how do you recommend we deal with the larger users that would like >> to follow back their followers? With the hard limit of 1,000 follows per >> day, there is no way they'll ever catch up, as some of them have more than >> 1k new followers per day as is. If this limit were more dynamic based on >> the size of the user that would be nice. Capabilities to follow people in >> bulk may also help. >> Of course, I think many of these would no longer need to follow back if >> they could just have the option to enable anyone to DM them if they choose. >> I think that's the underlying cause to want to auto-follow for most people. >> The only other cause is for an additional token/feeling of community, >> although I think many would be willing to forgo that if they had the ability >> to just allow everyone to DM them - it feels good to have someone you admire >> follow you back, even if it's not 100% sincere. >> >> Jesse >> >> >> On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: >> >>> Yes, that's what appears to be happening. My experience starts at around >>> 500K+. I'm okay with waiting with my script if you guys need to take longer >>> to retrieve the info. Or if you'd prefer we paginate I'll start doing that >>> as well. Maybe a hard limit of 200K and you have to Page to get above that? >>> Jesse >>> >>> >>> On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: >>> I've heard that list sizes greater than 150K-200K start to return timeouts at higher rates. Although I'd enjoy hearing first-hand experiences and recommendations. Thanks, Doug On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: > In my case specifically it's the Social Graph methods. I didn't > realize you had paging available now. Is there some logic as to when I > should expect to page and when I can just rely on the full result? > Jesse > > > On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: > >> What methods in particular are you referring to? The social graph >> methods now support paging so retrieving all of that data is now >> possible, >> where it used to throw 502s. It does however require a bit of application >> logic to assume when paging is necessary (e.g. large follower counts). >> Additionally, we are making changes to the databases which cause latency >> that result in periodic 502s. We are not able to give definitive ETAs on >> these fixes due to priorities that change as unforeseeable critical needs >> arise. >> More specificity would be beneficial. Do you have a replaceable bug, >> problem, or suggestion that you would like to discuss? >> Thanks, >> Doug >> >> >> >> On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: >> >>> I was discussing this with Iain, and have also talked about it with >>> Damon, so I know I'm not alone in this. I am having huge issues >>> retrieving >>> follower and friend data for the larger users (1 million+ followers), >>> most >>> of the time returning 502 Bad Gateway errors. I know there are a few of >>> these users getting really frustrated about our apps not being able to >>> retrieve data for them. Is there a plan to fix this? Is the API team >>> aware >>> of this? Any ETA by chance? >>> Thanks, >>> >>> @Jesse >>> >> >> > >>> >> > -- Sean Scott cell: 612.867.8133 portfolio: http://www.flickr.com/photos/92876...@n00/sets/72157613990263453/ profile: http://www.linkedin.com/profile?viewProfile=&key=2242610 blog: http://www.twofortyeight.com/ other: http://twitter.com/kalisurfer
[twitter-dev] Forgive Forgive Forgive - XML 2 CSV XSL
I would like to take the follower .xml output and convert to csv. >From googling I believe I need to create an XSL or XSD and then use MSXML example: MSXSL C:\twitterout\adventuregirl69.txt C:\Users\mark\Documents \twitteratom2txt.xsl –o out.xml I don't know xsl trying to learn from examples, does anyone have any .XSLs they created for Twitter's API responses? Would like to create flat file output from xml response. Goal is to tie a zip code to a twitter profile like @geofollow
[twitter-dev] Re: Debug mode?
That's the whole point, though. I'm still at the point where I'm trying to *get* an oauth token. On Thu, Jun 4, 2009 at 13:51, Abraham Williams <4bra...@gmail.com> wrote: > > > On Tue, Jun 2, 2009 at 22:53, @jigglyonee wrote: > >> >> I'm passing, hopefully, the correct parameters. A typical query string >> looks like this: >> >> >> http://twitter.com/oauth/request_token?oauth_consumer_key=9CCTnLpstYI8RIxGE7yhQ&oauth_token=&oauth_signature_method=HMAC-SHA1&oauth_timestamp=1244001043&oauth_nonce=401a5dc6-8750-4e60-890e-7ec30cdcf6a3&oauth_version=1.0&oauth_signature=badstdQUfnlCweKJMoIohTnfKxw >> > > The oauth_token has no value. > > > -- > Abraham Williams | http://the.hackerconundrum.com > Hacker | http://abrah.am | http://twitter.com/abraham > Project | http://fireeagle.labs.poseurtech.com > This email is: [ ] blogable [x] ask first [ ] private. > Sent from Madison, Wisconsin, United States -- Internets. Serious business.
[twitter-dev] Re: Sreaming API: /follow closing immediately
On Thu, Jun 4, 2009 at 5:51 PM, John Kalucki wrote: > > Note that Hosebird limits logins by user, not by resource. If you are > consuming /gardenhose and /follow from the same account, the first > connection may be thrown off by the second connection. Ah, well that would certainly explain another problem I was seeing. Thanks for the clarification :) -Chad
[twitter-dev] Re: making API call to update status
If an application is created using read-only an account grants access and then the application changes to read-write the user needs to grant access again in order for the application to post updates. On Thu, Jun 4, 2009 at 00:13, avinash srivastava wrote: > looks like while setting up your application you have asked for read > permission only. Check for access type in your application settings on > twitter and make it read and write. > > > Avinash > > > > On Thu, Jun 4, 2009 at 10:20 AM, Nasir wrote: > >> >> Hi, >> >> Can any one help me to make api call after authentication using OAuth- >> php. Its gives an error (Read-only application cannot POST) >> >> Thanks >> > > -- Abraham Williams | http://the.hackerconundrum.com Hacker | http://abrah.am | http://twitter.com/abraham Project | http://fireeagle.labs.poseurtech.com This email is: [ ] blogable [x] ask first [ ] private. Sent from Madison, Wisconsin, United States
[twitter-dev] Re: Regarding Twitter/OAuth PHP libraries constructors
It makes sense to me. Part of why my library is like this is because the
Fire Eagle class and the OAuth code is like this. One issue with build
isAuthenticated into library is they then have to be aware of however you
are storing the access tokens/etc. I would probably be easier for you to
extend the class and add your own isAuthenticated function.
Abraham
On Wed, Jun 3, 2009 at 16:57, Tim wrote:
>
> Hello,
>
> (note: this question might actually be more a general programming
> question rather than specific to Twitter)
> I have been looking at a few PHP libraries to use OAuth with Twitter
> (thanks to Abraham and Jason amongst others) and noticed that most (if
> not all) of them have a constructor more or less like that:
> function __construct($consumer_key, $consumer_secret, $oauth_token =
> NULL, $oauth_token_secret = NULL)
> called with 2 or 4 arguments depending on where in the OAuth process
> we are. (access token already known or not)
>
> That usually means that the status of the OAuth process is to be
> managed in the controllers when creating the object.
> If you take the example from Abraham's code:
>
> http://github.com/abraham/twitteroauth/blob/6b28e9e689d44a569576f726777e9986fa67729c/example/index.php
> there's a lot of session variables management (granted, this code is
> provided as an example and advertised as "not for production"...)
>
> Basically, I feel that I'd rather do something like:
>
> $twitter = new TwitterOAuth();
> $twitter->getPublicTimeline();
> if ($twitter->isAuthenticated()) { $twitter->getFriendsTimeline();}
> etc.
>
> with the logic of if the access token is known being handled by the
> class itself.
>
> Since all the libraries I have seen didn't go this way (and were
> written by people who most probably know what they're doing a lot more
> than I do), I am wondering if I make any sense.
>
> Thanks for your help,
>
> Tim
>
--
Abraham Williams | http://the.hackerconundrum.com
Hacker | http://abrah.am | http://twitter.com/abraham
Project | http://fireeagle.labs.poseurtech.com
This email is: [ ] blogable [x] ask first [ ] private.
Sent from Madison, Wisconsin, United States
[twitter-dev] Re: Sreaming API: /follow closing immediately
There was a problem today with the authentication database used by one Hosebird servers. This caused an unlucky proportion of connections that require an elevated Role to be rejected. This has been fixed. Note that Hosebird limits logins by user, not by resource. If you are consuming /gardenhose and /follow from the same account, the first connection may be thrown off by the second connection. -John Kalucki Services, Twitter, Inc. On Jun 4, 2:26 pm, Chad Etzel wrote: > I am playing with the follow methods, and they have been acting funny > all day. Lately it just closes immediately. Right now I am just > testing with curl: > > $ curl -v -d @follow.dathttp://stream.twitter.com/follow.json-uuser:pass > > * About to connect() to stream.twitter.com port 80 > * Trying 128.121.146.231... connected > * Connected to stream.twitter.com (128.121.146.231) port 80 > * Server auth using Basic with user 'jcnetstreamr1'> POST /follow.json > HTTP/1.1 > > Authorization: Basic amNuZXRzdHJlYW1yMTpyY2s0OG1udA== > > User-Agent: curl/7.15.5 (x86_64-redhat-linux-gnu) libcurl/7.15.5 > > OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5 > > Host: stream.twitter.com > > Accept: */* > > Content-Length: 49 > > Content-Type: application/x-www-form-urlencoded > > > follow=6633812 17281594 19667049 3841961 17243913HTTP/1.1 200 OK > > < Content-Type: application/json > < Content-Length: 0 > < Server: Jetty(6.1.14) > * Connection #0 to host stream.twitter.com left intact > * Closing connection #0 > $ > > I've tried this out on 2 servers, one on the east coast, and one on > the west coast. > > Content-Length is 0, so I guess curl just closes immediately. > > Other times it takes about 30 seconds to get the Response headers back: > > HTTP/1.1 200 OK > < Content-Type: application/json > < Transfer-Encoding: chunked > < Server: Jetty(6.1.14) > > The PHP script I've written to connect to the follow methods is acting > even weirder. I'm using the same basic code as I have to use the > /spritzer stream, which seems to work ok. > > Anyone else with similar behavior on the /follow streams? > > -Chad
[twitter-dev] Sreaming API: /follow closing immediately
I am playing with the follow methods, and they have been acting funny all day. Lately it just closes immediately. Right now I am just testing with curl: $ curl -v -d @follow.dat http://stream.twitter.com/follow.json -uuser:pass * About to connect() to stream.twitter.com port 80 * Trying 128.121.146.231... connected * Connected to stream.twitter.com (128.121.146.231) port 80 * Server auth using Basic with user 'jcnetstreamr1' > POST /follow.json HTTP/1.1 > Authorization: Basic amNuZXRzdHJlYW1yMTpyY2s0OG1udA== > User-Agent: curl/7.15.5 (x86_64-redhat-linux-gnu) libcurl/7.15.5 > OpenSSL/0.9.8b zlib/1.2.3 libidn/0.6.5 > Host: stream.twitter.com > Accept: */* > Content-Length: 49 > Content-Type: application/x-www-form-urlencoded > > follow=6633812 17281594 19667049 3841961 17243913HTTP/1.1 200 OK < Content-Type: application/json < Content-Length: 0 < Server: Jetty(6.1.14) * Connection #0 to host stream.twitter.com left intact * Closing connection #0 $ I've tried this out on 2 servers, one on the east coast, and one on the west coast. Content-Length is 0, so I guess curl just closes immediately. Other times it takes about 30 seconds to get the Response headers back: HTTP/1.1 200 OK < Content-Type: application/json < Transfer-Encoding: chunked < Server: Jetty(6.1.14) The PHP script I've written to connect to the follow methods is acting even weirder. I'm using the same basic code as I have to use the /spritzer stream, which seems to work ok. Anyone else with similar behavior on the /follow streams? -Chad
[twitter-dev] Re: Larger Users Not Returning Follower Data
We would like users to be judicious with their following habits and only follow users who contribute value to their timeline. This justifies the following limits we impose. We are aware that many users would like to accept all incoming directs. This, along with the quid pro quo following to build community, capture the majority of the use-cases for auto-following. We are discussing internally how to best approach these two uses within the bounds of the product we are trying to build. At this time we have nothing to report but know we are actively thinking about these ideas. Thanks, Doug On Thu, Jun 4, 2009 at 2:12 AM, Jesse Stay wrote: > Also, how do you recommend we deal with the larger users that would like to > follow back their followers? With the hard limit of 1,000 follows per day, > there is no way they'll ever catch up, as some of them have more than 1k new > followers per day as is. If this limit were more dynamic based on the size > of the user that would be nice. Capabilities to follow people in bulk may > also help. > Of course, I think many of these would no longer need to follow back if > they could just have the option to enable anyone to DM them if they choose. > I think that's the underlying cause to want to auto-follow for most people. > The only other cause is for an additional token/feeling of community, > although I think many would be willing to forgo that if they had the ability > to just allow everyone to DM them - it feels good to have someone you admire > follow you back, even if it's not 100% sincere. > > Jesse > > > On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: > >> Yes, that's what appears to be happening. My experience starts at around >> 500K+. I'm okay with waiting with my script if you guys need to take longer >> to retrieve the info. Or if you'd prefer we paginate I'll start doing that >> as well. Maybe a hard limit of 200K and you have to Page to get above that? >> Jesse >> >> >> On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: >> >>> I've heard that list sizes greater than 150K-200K start to return >>> timeouts at higher rates. Although I'd enjoy hearing first-hand experiences >>> and recommendations. >>> Thanks, >>> Doug >>> >>> >>> >>> >>> On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: >>> In my case specifically it's the Social Graph methods. I didn't realize you had paging available now. Is there some logic as to when I should expect to page and when I can just rely on the full result? Jesse On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: > What methods in particular are you referring to? The social graph > methods now support paging so retrieving all of that data is now possible, > where it used to throw 502s. It does however require a bit of application > logic to assume when paging is necessary (e.g. large follower counts). > Additionally, we are making changes to the databases which cause latency > that result in periodic 502s. We are not able to give definitive ETAs on > these fixes due to priorities that change as unforeseeable critical needs > arise. > More specificity would be beneficial. Do you have a replaceable bug, > problem, or suggestion that you would like to discuss? > Thanks, > Doug > > > > On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: > >> I was discussing this with Iain, and have also talked about it with >> Damon, so I know I'm not alone in this. I am having huge issues >> retrieving >> follower and friend data for the larger users (1 million+ followers), >> most >> of the time returning 502 Bad Gateway errors. I know there are a few of >> these users getting really frustrated about our apps not being able to >> retrieve data for them. Is there a plan to fix this? Is the API team >> aware >> of this? Any ETA by chance? >> Thanks, >> >> @Jesse >> > > >>> >> >
[twitter-dev] Re: Debug mode?
On Tue, Jun 2, 2009 at 22:53, @jigglyonee wrote: > > I'm passing, hopefully, the correct parameters. A typical query string > looks like this: > > > http://twitter.com/oauth/request_token?oauth_consumer_key=9CCTnLpstYI8RIxGE7yhQ&oauth_token=&oauth_signature_method=HMAC-SHA1&oauth_timestamp=1244001043&oauth_nonce=401a5dc6-8750-4e60-890e-7ec30cdcf6a3&oauth_version=1.0&oauth_signature=badstdQUfnlCweKJMoIohTnfKxw > The oauth_token has no value. -- Abraham Williams | http://the.hackerconundrum.com Hacker | http://abrah.am | http://twitter.com/abraham Project | http://fireeagle.labs.poseurtech.com This email is: [ ] blogable [x] ask first [ ] private. Sent from Madison, Wisconsin, United States
[twitter-dev] pics, videos, files, etc
Hello all, This is more of a twitter api question, once removed, but wanted to get everyone's experiences with attaching media like twitpic, yfrog, etc via their API. I have used twitPic, but how is yFrog? Does anyone offer a whole package, files, vids, pics in one api? Thanks Peter
[twitter-dev] Re: Help needed, to search photos in twitter
Hi Chethan, I wrote the TwitPicGrid page you cite. I just do a simple Search API search and append " twitpic" to the query. Since all twitpic posts have "http://twitpic.com/x"; in the tweet, this search will get the links in the result set. Then, since most people describe the content of the picture in their tweet, there is a relatively high probability that the keyword in the search will be featured in the picture somehow. So basically, if you want twitpics of "lunch", just search twitter for "lunch twitpic" and parse out the twitpic links in the results. -Chad On Thu, Jun 4, 2009 at 9:05 AM, Chethan wrote: > > Hello, can anyone please tell me how to search for photos in twitter/ > twitpic, (ex: http://www.twitpicsearch.com, http://tweetgrid.com/twitpicgrid), > it would be really helpful if you could provide me with some sample > code, I am using PHP and jQuery > > Thankyou >
[twitter-dev] Max_id bug or just misunderstood semantics?
A message from my logging system: request was http://search.twitter.com/search.atom ?max_id=2012539218&q= %22susan+boyle%22&rpp=100 response was Range -"susan boyle" - from 2,012,539,246 to 2,013,509,906 The exact request sent out was: http://search.twitter.com/search.atom?max_id=2012539218&q=%22susan+boyle%22&rpp=100 but all the statuses returned were higher than 'max_id' This has been happening since early wednesday (or perhaps late tuesday, June 2) Is this a change (intended or otherwise) in the semantics of max_id on the search API?
[twitter-dev] Language code in search query parameter.
Hello, Is it possible to use the language parameter as part of the "q" parameter in a search? Im thinging something along: "/search? q=ombudsman%20lang:se". This would let users save searches in their various clients, without the clients having support for the specific langage parameter. Thanks, Richard
[twitter-dev] Re: Debug mode?
Thanks everyone! I will try all these suggestions
1. The hueniverse guide is where I got started. It's actually incredibly
useful.
2. I will definitely try the explorer.
3. Thanks, Dossy! I've been staring at those for so long that I didn't see
the multiple GETs (I'm almost positive that I didn't copy and paste that
wrong, so I'll check my code). Also, I'll check the &'s.
On Thu, Jun 4, 2009 at 05:12, Dossy Shiobara wrote:
>
> On 6/4/09 12:07 AM, JDG wrote:
>
>>My signature base string for that looked like:
>>
>>GET%26http%3A%2F%2Ftwitter.com%2Foauth%2Frequest_token%26GET%26http%3A
>>%2F%2Ftwitter.com%2Foauth%2Frequest_token%26oauth_consumer_key
>>%3D9CCTnLpstYI8RIxGE7yhQ%26oauth_nonce
>>%3Db52968e8-7145-4c91-8066-563d37a1107f%26oauth_signature_method
>>%3DHMAC-SHA1%26oauth_timestamp%3D1244001043%26oauth_token%3D
>>%26oauth_version%3D1.0
>>
>
> If that is truly what you're signing, there are two problems I see:
>
> 1) You are URI-encoding the ampersand separator ("&" -> "%26"). Don't.
>
> 2) You have the method and URL in the signature string twice. Don't.
>
> Those are the immediate problems I see, anyhow.
>
> --
> Dossy Shiobara | do...@panoptic.com | http://dossy.org/
> Panoptic Computer Network | http://panoptic.com/
> "He realized the fastest way to change is to laugh at your own
>folly -- then you can let go and quickly move on." (p. 70)
>
--
Internets. Serious business.
[twitter-dev] Help needed, to search photos in twitter
Hello, can anyone please tell me how to search for photos in twitter/ twitpic, (ex: http://www.twitpicsearch.com, http://tweetgrid.com/twitpicgrid), it would be really helpful if you could provide me with some sample code, I am using PHP and jQuery Thankyou
[twitter-dev] Re: Typo and Problem with Twitter's Redirect
Hi there, I just finished a patch that will delay showing the link for 5 seconds. If your site responds and the user is connected in less than that they won't have a chance to click the link. If the page is still displayed for some reason after 5 seconds the link will appear. There is still a possibility of double-submit but this should lessen it. Does anyone see a problem with this approach? Thanks; – Matt Sanford / @mzsanford Twitter Dev On Jun 2, 2009, at 2:24 PM, Abraham Williams wrote: That is an issue when users are on slow internet connections or the target site is really slow. Not sure that I would call it a bug... but it does need a solution. The http request from the auto redirect hits the target sites callback but before html is downloaded and the browser redraws replacing the link text from twitter the link is clicked on again causing a second http request to hit the targets sites callback. A fix could be using the same style of javascript on the link as is used on the authorize button. On Tue, Jun 2, 2009 at 15:25, Francis Shanahan > wrote: Lately I've noticed once the user "Grant"s on the Twitter oAuth page, the "Redirect" page from Twitter is showing up a little longer. There's a typo "dosen't" on this page by the way. The bigger problem though is that a redirect is taking place but the browser hasn't reacted and the user has a chance to click the "click here if you're browser doesn't redirect" link. I'm not sure, but it seems like the token has already gone out to the target site. If they clicks that link it introduces a bug since the target site gets the oAuth token twice and tries to use it in exchange for an access token. You can test it at http://tweetarun.com Anyone else noticed this? Is this a bug? or is something else going on? -fs -- Abraham Williams | http://the.hackerconundrum.com Hacker | http://abrah.am | http://twitter.com/abraham Project | http://fireeagle.labs.poseurtech.com This email is: [ ] blogable [x] ask first [ ] private. Sent from Madison, Wisconsin, United States
[twitter-dev] Re: random sampling of users....do we know anything about user id range?
On Wed, Jun 3, 2009 at 7:13 PM, TechRavingMad wrote: > > There are a little over 44.5 million twitter IDs as of right now > (10:10pm cst 6/3/9) with what seems to be about 10 being added every > second. > However, Twitter has been quite clear about not saying if status IDs correspond to the actual number of statuses, so I'd guess that they're equally circumspect about whether or not the number of user IDs corresponds to the number of users. In other words, we can be sure there are not more than 44.5 million users, but we don't know how much lower the actual number is. We don't know if all IDs have been used... and even Twitter doesn't know how many of those IDs belong to the same users. I would think that if one wants a random sample of users, one would have to propose a selection method and ask Twitter if there's any reason that it would introduce a selection bias... and hope that they are willing to reply. Seems to me that the biggest problem would be to include "quiet" users, since only those who post in public become visible. NIck
[twitter-dev] Twitter4J 2.0.7 released
Hi all, Twitter4J 2.0.7 is available for download. http://yusuke.homeip.net/twitter4j/en/index.html#download It is(or will be) available at the Maven central repository. http://repo1.maven.org/maven2/net/homeip/yusuke/twitter4j/ Snapshot builds can be found at: http://yusuke.homeip.net/maven2/net/homeip/yusuke/twitter4j/ This is a maintenance release with no new features. Release Notes - Twitter4J - Version 2.0.7 - HTML format Bug [TFJ-161] - OAuth AccessToken missing screen name Task [TFJ-159] - getUserTimeline(String id, int count, long sinceId) should be deprecated [TFJ-160] - fix build.xml to include twitter4j.properties Cheers, -- Yusuke Yamamoto yus...@mac.com this email is: [x] bloggable/twittable [ ] ask first [ ] private follow me on : http://twitter.com/yusukeyamamoto subscribe me : http://yusuke.homeip.net/blog/
[twitter-dev] Re: Debug mode?
On 6/4/09 12:07 AM, JDG wrote:
My signature base string for that looked like:
GET%26http%3A%2F%2Ftwitter.com%2Foauth%2Frequest_token%26GET%26http%3A
%2F%2Ftwitter.com%2Foauth%2Frequest_token%26oauth_consumer_key
%3D9CCTnLpstYI8RIxGE7yhQ%26oauth_nonce
%3Db52968e8-7145-4c91-8066-563d37a1107f%26oauth_signature_method
%3DHMAC-SHA1%26oauth_timestamp%3D1244001043%26oauth_token%3D
%26oauth_version%3D1.0
If that is truly what you're signing, there are two problems I see:
1) You are URI-encoding the ampersand separator ("&" -> "%26"). Don't.
2) You have the method and URL in the signature string twice. Don't.
Those are the immediate problems I see, anyhow.
--
Dossy Shiobara | do...@panoptic.com | http://dossy.org/
Panoptic Computer Network | http://panoptic.com/
"He realized the fastest way to change is to laugh at your own
folly -- then you can let go and quickly move on." (p. 70)
[twitter-dev] Re: Debug mode?
Similar to Tim's suggestion, I'd also check out http://www.hueniverse.com/hueniverse/2008/10/beginners-gui-1.html ... shows step by step conversion values which you can use to pinpoint where your library might be having problems. On Jun 3, 10:27 pm, Tim wrote: > Maybe this OAuth Explorer would help you checking the different step > and the parameters of your > query:http://sevengoslings.net/~fangel/oauth-explorer/ > > Also, you might want to try an existing OAuth library in parallel with > yours to see that it works the same. (even though you apparently tried > this already) > > Hope this helps, > > Tim > > On Jun 2, 8:53 pm, "@jigglyonee" wrote: > > > Hi all, > > > I've been trying to write a JavaScript OAuth module in order to write > > my own twitter client (I know there are libraries out there, but it's > > for my education as well as fun). > > > I'm stumped, unfortunately, right at the beginning -- trying to get a > > request token. I've tried GETting and POSTing > > tohttp://twitter.com/oauth/request_token > > to no avail -- it always returns "Failed to validate oauth signature > > and token". > > > I'm passing, hopefully, the correct parameters. A typical query string > > looks like this: > > >http://twitter.com/oauth/request_token?oauth_consumer_key=9CCTnLpstYI... > > > My signature base string for that looked like: > > > GET%26http%3A%2F%2Ftwitter.com%2Foauth%2Frequest_token%26GET%26http%3A > > %2F%2Ftwitter.com%2Foauth%2Frequest_token%26oauth_consumer_key > > %3D9CCTnLpstYI8RIxGE7yhQ%26oauth_nonce > > %3Db52968e8-7145-4c91-8066-563d37a1107f%26oauth_signature_method > > %3DHMAC-SHA1%26oauth_timestamp%3D1244001043%26oauth_token%3D > > %26oauth_version%3D1.0 > > > I'm fairly sure that the signature is being encoded correctly, as the > > HMAC-SHA1 algorithm returns the same in my library and in the > > javax.crypto libraries. I'm encoding using the key > > "&" since I don't have a token secret yet. > > > My question is twofold. 1) does something look amiss with this? 2) if > > nothing's obvious, is there any sort of debug mode for the APIs where > > we can get some more detailed error messages? > > > Thanks! > > > Josh
[twitter-dev] Re: Larger Users Not Returning Follower Data
Also, how do you recommend we deal with the larger users that would like to follow back their followers? With the hard limit of 1,000 follows per day, there is no way they'll ever catch up, as some of them have more than 1k new followers per day as is. If this limit were more dynamic based on the size of the user that would be nice. Capabilities to follow people in bulk may also help. Of course, I think many of these would no longer need to follow back if they could just have the option to enable anyone to DM them if they choose. I think that's the underlying cause to want to auto-follow for most people. The only other cause is for an additional token/feeling of community, although I think many would be willing to forgo that if they had the ability to just allow everyone to DM them - it feels good to have someone you admire follow you back, even if it's not 100% sincere. Jesse On Thu, Jun 4, 2009 at 3:07 AM, Jesse Stay wrote: > Yes, that's what appears to be happening. My experience starts at around > 500K+. I'm okay with waiting with my script if you guys need to take longer > to retrieve the info. Or if you'd prefer we paginate I'll start doing that > as well. Maybe a hard limit of 200K and you have to Page to get above that? > Jesse > > > On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: > >> I've heard that list sizes greater than 150K-200K start to return timeouts >> at higher rates. Although I'd enjoy hearing first-hand experiences and >> recommendations. >> Thanks, >> Doug >> >> >> >> >> On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: >> >>> In my case specifically it's the Social Graph methods. I didn't realize >>> you had paging available now. Is there some logic as to when I should >>> expect to page and when I can just rely on the full result? >>> Jesse >>> >>> >>> On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: >>> What methods in particular are you referring to? The social graph methods now support paging so retrieving all of that data is now possible, where it used to throw 502s. It does however require a bit of application logic to assume when paging is necessary (e.g. large follower counts). Additionally, we are making changes to the databases which cause latency that result in periodic 502s. We are not able to give definitive ETAs on these fixes due to priorities that change as unforeseeable critical needs arise. More specificity would be beneficial. Do you have a replaceable bug, problem, or suggestion that you would like to discuss? Thanks, Doug On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: > I was discussing this with Iain, and have also talked about it with > Damon, so I know I'm not alone in this. I am having huge issues > retrieving > follower and friend data for the larger users (1 million+ followers), most > of the time returning 502 Bad Gateway errors. I know there are a few of > these users getting really frustrated about our apps not being able to > retrieve data for them. Is there a plan to fix this? Is the API team > aware > of this? Any ETA by chance? > Thanks, > > @Jesse > >>> >> >
[twitter-dev] Re: Larger Users Not Returning Follower Data
Yes, that's what appears to be happening. My experience starts at around 500K+. I'm okay with waiting with my script if you guys need to take longer to retrieve the info. Or if you'd prefer we paginate I'll start doing that as well. Maybe a hard limit of 200K and you have to Page to get above that? Jesse On Thu, Jun 4, 2009 at 12:26 AM, Doug Williams wrote: > I've heard that list sizes greater than 150K-200K start to return timeouts > at higher rates. Although I'd enjoy hearing first-hand experiences and > recommendations. > Thanks, > Doug > > > > > On Wed, Jun 3, 2009 at 9:19 PM, Jesse Stay wrote: > >> In my case specifically it's the Social Graph methods. I didn't realize >> you had paging available now. Is there some logic as to when I should >> expect to page and when I can just rely on the full result? >> Jesse >> >> >> On Wed, Jun 3, 2009 at 9:56 PM, Doug Williams wrote: >> >>> What methods in particular are you referring to? The social graph methods >>> now support paging so retrieving all of that data is now possible, where it >>> used to throw 502s. It does however require a bit of application logic to >>> assume when paging is necessary (e.g. large follower counts). Additionally, >>> we are making changes to the databases which cause latency that result in >>> periodic 502s. We are not able to give definitive ETAs on these fixes due to >>> priorities that change as unforeseeable critical needs arise. >>> More specificity would be beneficial. Do you have a replaceable bug, >>> problem, or suggestion that you would like to discuss? >>> Thanks, >>> Doug >>> >>> >>> >>> On Wed, Jun 3, 2009 at 7:22 PM, Jesse Stay wrote: >>> I was discussing this with Iain, and have also talked about it with Damon, so I know I'm not alone in this. I am having huge issues retrieving follower and friend data for the larger users (1 million+ followers), most of the time returning 502 Bad Gateway errors. I know there are a few of these users getting really frustrated about our apps not being able to retrieve data for them. Is there a plan to fix this? Is the API team aware of this? Any ETA by chance? Thanks, @Jesse >>> >>> >> >
