Re: [twitter-dev] 4294967295

[Dan Checkoway]

Cool, I appreciate the response.  I forgot to mention, although you guys
probably know this by now...originally it was just the firehose on which we
saw those funky values, but lately we've been seeing them in the wild as
well.

Thanks again, Taylor.

Dan

On Wed, Dec 22, 2010 at 10:56 PM, Taylor Singletary <
taylorsinglet...@twitter.com> wrote:

> It's going to be a little bit of time before we can totally prevent these
> values from occurring.
>
> Right now, you should probably just consider this value as "unknown" rather
> than necessarily null, 0, or otherwise. The team responsible for the low
> level component causing the bug has a fix planned, but it can't be applied
> until a few more dependencies are resolved.
>
> Thanks,
> Taylor
>
> On Wed, Dec 22, 2010 at 2:59 PM, Dan Checkoway wrote:
>
>> I just wanted to follow up on this, because the issue continues to happen,
>> and it gets more and more interesting.
>>
>> We've now been seeing user.listed_count coming back as "4294967293" on
>> occasion.  So just to recap, we have now seen these values in the
>> user.listed_count field:
>>
>> 4294967295 (a.k.a. unsigned -1)
>> 4294967294 (a.k.a. unsigned -2)
>> 4294967293 (a.k.a. unsigned -3)
>>
>> twitter4j has worked around this issue no problem, but I'm more than just
>> a bit curious what these values represent.  Should -1 and -2 and -3 be
>> treated to mean anything other than "we don't know what the listed count
>> is"?  What happens if/when -4 starts popping out?
>>
>> I realize this is pretty low priority, but it's still a bug...
>>
>> Thanks,
>> Dan
>>
>>
>> On Fri, Dec 17, 2010 at 7:17 AM, Dan Checkoway wrote:
>>
>>> Check this out...today sometime between 4:01:43 AM PST and 4:01:53 AM PST
>>> (sorry for the ambiguity, those are our "every 10 sec" logging timestamps
>>> for other stuff), we saw the unsigned equivalent of -2 (4294967294) being
>>> sent by twitter in user.listed_count...
>>>
>>>
>>> Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
>>> java.lang
>>> .NumberFormatException: For input string: "4294967294"
>>>
>>> at
>>> java.lang.NumberFormatException.forInputString(NumberFormatException.
>>> java:48)
>>> at java.lang.Integer.parseInt(Integer.java:459)
>>> at java.lang.Integer.valueOf(Integer.java:553)
>>> at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>>> at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>>>  at twitter4j.UserJSONImpl.(UserJSONImpl.java:86)
>>> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:101)
>>> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
>>> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:118)
>>> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
>>> at
>>> twitter4j.StatusStreamImpl.handleNextElement(StatusStreamImpl.java:116)
>>> at twitter4j.StatusStreamImpl.next(StatusStreamImpl.java:89)
>>> at
>>> twitter4j.TwitterStream$StreamHandlingThread.run(TwitterStream.java:529)
>>>
>>> Any idea what's going on and/or when it might be fixed?
>>>
>>> Thanks,
>>> Dan
>>>
>>>
>>> On Tue, Dec 14, 2010 at 8:10 PM, Taylor Singletary <
>>> taylorsinglet...@twitter.com> wrote:
>>>
 Thanks! This is being looked into. I'll update when I have news.

 Taylor

 On Tuesday, December 14, 2010, Dan Checkoway 
 wrote:
 > Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens,
 but here are a handful of timestamps for unrelated stuff that got logged no
 more than 10 seconds *prior* to the 4294967295 error popping out...so
 they're fairly close:
 >
 > Dec 14, 2010 12:34:11 PM PST
 > Dec 14, 2010 1:13:07 PM PST
 > Dec 14, 2010 1:22:48 PM PST
 > Dec 14, 2010 1:27:22 PM PST
 > Dec 14, 2010 1:29:48 PM PST
 > Dec 14, 2010 1:33:36 PM PST
 >
 > Based on the twitter4j stack trace, I can tell you that it was
 *always* user.listed_count that had the funky value:
 >
 > Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
 java.lang
 > .NumberFormatException: For input string: "4294967295"
 > at
 java.lang.NumberFormatException.forInputString(NumberFormatException.
 > java:48)
 > at java.lang.Integer.parseInt(Integer.java:459)
 > at java.lang.Integer.valueOf(Integer.java:553)
 > at
 twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
 > at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
 >
 > Thanks,
 > Dan
 >
 > On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary <
 taylorsinglet...@twitter.com> wrote:
 > Understandable, Dan.
 >
 > Can you tell me the last time an event like this happened?
 >
 > Taylor
 >
 > On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway 
 wrote:
 >> I know this is the weenie answer, but I haven't been able to track a
 >> specific offending JSON

Re: [twitter-dev] 4294967295

[Taylor Singletary]

It's going to be a little bit of time before we can totally prevent these
values from occurring.

Right now, you should probably just consider this value as "unknown" rather
than necessarily null, 0, or otherwise. The team responsible for the low
level component causing the bug has a fix planned, but it can't be applied
until a few more dependencies are resolved.

Thanks,
Taylor

On Wed, Dec 22, 2010 at 2:59 PM, Dan Checkoway  wrote:

> I just wanted to follow up on this, because the issue continues to happen,
> and it gets more and more interesting.
>
> We've now been seeing user.listed_count coming back as "4294967293" on
> occasion.  So just to recap, we have now seen these values in the
> user.listed_count field:
>
> 4294967295 (a.k.a. unsigned -1)
> 4294967294 (a.k.a. unsigned -2)
> 4294967293 (a.k.a. unsigned -3)
>
> twitter4j has worked around this issue no problem, but I'm more than just a
> bit curious what these values represent.  Should -1 and -2 and -3 be treated
> to mean anything other than "we don't know what the listed count is"?  What
> happens if/when -4 starts popping out?
>
> I realize this is pretty low priority, but it's still a bug...
>
> Thanks,
> Dan
>
>
> On Fri, Dec 17, 2010 at 7:17 AM, Dan Checkoway wrote:
>
>> Check this out...today sometime between 4:01:43 AM PST and 4:01:53 AM PST
>> (sorry for the ambiguity, those are our "every 10 sec" logging timestamps
>> for other stuff), we saw the unsigned equivalent of -2 (4294967294) being
>> sent by twitter in user.listed_count...
>>
>>
>> Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
>> java.lang
>> .NumberFormatException: For input string: "4294967294"
>>
>> at
>> java.lang.NumberFormatException.forInputString(NumberFormatException.
>> java:48)
>> at java.lang.Integer.parseInt(Integer.java:459)
>> at java.lang.Integer.valueOf(Integer.java:553)
>> at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>> at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>>  at twitter4j.UserJSONImpl.(UserJSONImpl.java:86)
>> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:101)
>> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
>> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:118)
>> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
>> at
>> twitter4j.StatusStreamImpl.handleNextElement(StatusStreamImpl.java:116)
>> at twitter4j.StatusStreamImpl.next(StatusStreamImpl.java:89)
>> at
>> twitter4j.TwitterStream$StreamHandlingThread.run(TwitterStream.java:529)
>>
>> Any idea what's going on and/or when it might be fixed?
>>
>> Thanks,
>> Dan
>>
>>
>> On Tue, Dec 14, 2010 at 8:10 PM, Taylor Singletary <
>> taylorsinglet...@twitter.com> wrote:
>>
>>> Thanks! This is being looked into. I'll update when I have news.
>>>
>>> Taylor
>>>
>>> On Tuesday, December 14, 2010, Dan Checkoway 
>>> wrote:
>>> > Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens,
>>> but here are a handful of timestamps for unrelated stuff that got logged no
>>> more than 10 seconds *prior* to the 4294967295 error popping out...so
>>> they're fairly close:
>>> >
>>> > Dec 14, 2010 12:34:11 PM PST
>>> > Dec 14, 2010 1:13:07 PM PST
>>> > Dec 14, 2010 1:22:48 PM PST
>>> > Dec 14, 2010 1:27:22 PM PST
>>> > Dec 14, 2010 1:29:48 PM PST
>>> > Dec 14, 2010 1:33:36 PM PST
>>> >
>>> > Based on the twitter4j stack trace, I can tell you that it was *always*
>>> user.listed_count that had the funky value:
>>> >
>>> > Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
>>> java.lang
>>> > .NumberFormatException: For input string: "4294967295"
>>> > at
>>> java.lang.NumberFormatException.forInputString(NumberFormatException.
>>> > java:48)
>>> > at java.lang.Integer.parseInt(Integer.java:459)
>>> > at java.lang.Integer.valueOf(Integer.java:553)
>>> > at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>>> > at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>>> >
>>> > Thanks,
>>> > Dan
>>> >
>>> > On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary <
>>> taylorsinglet...@twitter.com> wrote:
>>> > Understandable, Dan.
>>> >
>>> > Can you tell me the last time an event like this happened?
>>> >
>>> > Taylor
>>> >
>>> > On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway 
>>> wrote:
>>> >> I know this is the weenie answer, but I haven't been able to track a
>>> >> specific offending JSON object down yet, since it only seems to happen
>>> on
>>> >> the firehose, and we're using twitter4j to process that.
>>> >>
>>> >> If we were able to connect to the firehose more than once at a time, I
>>> could
>>> >> easily write a tool to detect and highlight the issue.  Short of that,
>>> I'll
>>> >> try watching the sample stream for a while to see if the same issue
>>> pops up
>>> >> there.  Will report any findings...
>>> >>
>>> >> Thanks,
>>> >> Dan
>>> >>
>>> >> On Tu

Re: [twitter-dev] 4294967295

[Dan Checkoway]

I just wanted to follow up on this, because the issue continues to happen,
and it gets more and more interesting.

We've now been seeing user.listed_count coming back as "4294967293" on
occasion.  So just to recap, we have now seen these values in the
user.listed_count field:

4294967295 (a.k.a. unsigned -1)
4294967294 (a.k.a. unsigned -2)
4294967293 (a.k.a. unsigned -3)

twitter4j has worked around this issue no problem, but I'm more than just a
bit curious what these values represent.  Should -1 and -2 and -3 be treated
to mean anything other than "we don't know what the listed count is"?  What
happens if/when -4 starts popping out?

I realize this is pretty low priority, but it's still a bug...

Thanks,
Dan

On Fri, Dec 17, 2010 at 7:17 AM, Dan Checkoway  wrote:

> Check this out...today sometime between 4:01:43 AM PST and 4:01:53 AM PST
> (sorry for the ambiguity, those are our "every 10 sec" logging timestamps
> for other stuff), we saw the unsigned equivalent of -2 (4294967294) being
> sent by twitter in user.listed_count...
>
>
> Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
> java.lang
> .NumberFormatException: For input string: "4294967294"
>
> at
> java.lang.NumberFormatException.forInputString(NumberFormatException.
> java:48)
> at java.lang.Integer.parseInt(Integer.java:459)
> at java.lang.Integer.valueOf(Integer.java:553)
> at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
> at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
> at twitter4j.UserJSONImpl.(UserJSONImpl.java:86)
> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:101)
> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
> at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:118)
> at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
> at
> twitter4j.StatusStreamImpl.handleNextElement(StatusStreamImpl.java:116)
> at twitter4j.StatusStreamImpl.next(StatusStreamImpl.java:89)
> at
> twitter4j.TwitterStream$StreamHandlingThread.run(TwitterStream.java:529)
>
> Any idea what's going on and/or when it might be fixed?
>
> Thanks,
> Dan
>
>
> On Tue, Dec 14, 2010 at 8:10 PM, Taylor Singletary <
> taylorsinglet...@twitter.com> wrote:
>
>> Thanks! This is being looked into. I'll update when I have news.
>>
>> Taylor
>>
>> On Tuesday, December 14, 2010, Dan Checkoway 
>> wrote:
>> > Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens, but
>> here are a handful of timestamps for unrelated stuff that got logged no more
>> than 10 seconds *prior* to the 4294967295 error popping out...so they're
>> fairly close:
>> >
>> > Dec 14, 2010 12:34:11 PM PST
>> > Dec 14, 2010 1:13:07 PM PST
>> > Dec 14, 2010 1:22:48 PM PST
>> > Dec 14, 2010 1:27:22 PM PST
>> > Dec 14, 2010 1:29:48 PM PST
>> > Dec 14, 2010 1:33:36 PM PST
>> >
>> > Based on the twitter4j stack trace, I can tell you that it was *always*
>> user.listed_count that had the funky value:
>> >
>> > Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
>> java.lang
>> > .NumberFormatException: For input string: "4294967295"
>> > at
>> java.lang.NumberFormatException.forInputString(NumberFormatException.
>> > java:48)
>> > at java.lang.Integer.parseInt(Integer.java:459)
>> > at java.lang.Integer.valueOf(Integer.java:553)
>> > at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>> > at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>> >
>> > Thanks,
>> > Dan
>> >
>> > On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary <
>> taylorsinglet...@twitter.com> wrote:
>> > Understandable, Dan.
>> >
>> > Can you tell me the last time an event like this happened?
>> >
>> > Taylor
>> >
>> > On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway 
>> wrote:
>> >> I know this is the weenie answer, but I haven't been able to track a
>> >> specific offending JSON object down yet, since it only seems to happen
>> on
>> >> the firehose, and we're using twitter4j to process that.
>> >>
>> >> If we were able to connect to the firehose more than once at a time, I
>> could
>> >> easily write a tool to detect and highlight the issue.  Short of that,
>> I'll
>> >> try watching the sample stream for a while to see if the same issue
>> pops up
>> >> there.  Will report any findings...
>> >>
>> >> Thanks,
>> >> Dan
>> >>
>> >> On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
>> >>  wrote:
>> >>>
>> >>> Hi Dan,
>> >>>
>> >>> Do you continue to see events like this happening? Can you provide a
>> >>> recent example in as-provided JSON or XML?
>> >>>
>> >>> Thanks,
>> >>> Taylor
>> >>>
>> >>> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
>> >>> wrote:
>> >>> > Anybody else seeing user.listed_count occasionally coming back as
>> >>> > 4294967295?  That value just happens to equate to:  1 + (2 *
>> >>> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to
>> me...
>> >>> >
>> >>> > Anyway

Re: [twitter-dev] 4294967295

[Dan Checkoway]

Check this out...today sometime between 4:01:43 AM PST and 4:01:53 AM PST
(sorry for the ambiguity, those are our "every 10 sec" logging timestamps
for other stuff), we saw the unsigned equivalent of -2 (4294967294) being
sent by twitter in user.listed_count...

Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
java.lang
.NumberFormatException: For input string: "4294967294"
at
java.lang.NumberFormatException.forInputString(NumberFormatException.
java:48)
at java.lang.Integer.parseInt(Integer.java:459)
at java.lang.Integer.valueOf(Integer.java:553)
at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
at twitter4j.UserJSONImpl.(UserJSONImpl.java:86)
at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:101)
at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
at twitter4j.StatusJSONImpl.init(StatusJSONImpl.java:118)
at twitter4j.StatusJSONImpl.(StatusJSONImpl.java:84)
at
twitter4j.StatusStreamImpl.handleNextElement(StatusStreamImpl.java:116)
at twitter4j.StatusStreamImpl.next(StatusStreamImpl.java:89)
at
twitter4j.TwitterStream$StreamHandlingThread.run(TwitterStream.java:529)

Any idea what's going on and/or when it might be fixed?

Thanks,
Dan

On Tue, Dec 14, 2010 at 8:10 PM, Taylor Singletary <
taylorsinglet...@twitter.com> wrote:

> Thanks! This is being looked into. I'll update when I have news.
>
> Taylor
>
> On Tuesday, December 14, 2010, Dan Checkoway  wrote:
> > Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens, but
> here are a handful of timestamps for unrelated stuff that got logged no more
> than 10 seconds *prior* to the 4294967295 error popping out...so they're
> fairly close:
> >
> > Dec 14, 2010 12:34:11 PM PST
> > Dec 14, 2010 1:13:07 PM PST
> > Dec 14, 2010 1:22:48 PM PST
> > Dec 14, 2010 1:27:22 PM PST
> > Dec 14, 2010 1:29:48 PM PST
> > Dec 14, 2010 1:33:36 PM PST
> >
> > Based on the twitter4j stack trace, I can tell you that it was *always*
> user.listed_count that had the funky value:
> >
> > Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
> java.lang
> > .NumberFormatException: For input string: "4294967295"
> > at
> java.lang.NumberFormatException.forInputString(NumberFormatException.
> > java:48)
> > at java.lang.Integer.parseInt(Integer.java:459)
> > at java.lang.Integer.valueOf(Integer.java:553)
> > at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
> > at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
> >
> > Thanks,
> > Dan
> >
> > On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary <
> taylorsinglet...@twitter.com> wrote:
> > Understandable, Dan.
> >
> > Can you tell me the last time an event like this happened?
> >
> > Taylor
> >
> > On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway 
> wrote:
> >> I know this is the weenie answer, but I haven't been able to track a
> >> specific offending JSON object down yet, since it only seems to happen
> on
> >> the firehose, and we're using twitter4j to process that.
> >>
> >> If we were able to connect to the firehose more than once at a time, I
> could
> >> easily write a tool to detect and highlight the issue.  Short of that,
> I'll
> >> try watching the sample stream for a while to see if the same issue pops
> up
> >> there.  Will report any findings...
> >>
> >> Thanks,
> >> Dan
> >>
> >> On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
> >>  wrote:
> >>>
> >>> Hi Dan,
> >>>
> >>> Do you continue to see events like this happening? Can you provide a
> >>> recent example in as-provided JSON or XML?
> >>>
> >>> Thanks,
> >>> Taylor
> >>>
> >>> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
> >>> wrote:
> >>> > Anybody else seeing user.listed_count occasionally coming back as
> >>> > 4294967295?  That value just happens to equate to:  1 + (2 *
> >>> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to
> me...
> >>> >
> >>> > Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned
> >>> > that
> >>> > separately on the twitter4j list, but I wanted to raise the issue
> here
> >>> > since
> >>> > the root cause is twitter sending the weird value.
> >>> >
> >>> > Thanks,
> >>> > Dan
> >>> >
> >>> > --
> >>> > Twitter developer documentation and resources:
> >>> > http://dev.twitter.com/doc
> >>> > API updates via Twitter: http://twitter.com/twitterapi
> >>> > Issues/Enhancements Tracker:
> >>> > http://code.google.com/p/twitter-api/issues/list
> >>> > Change your membership to this group:
> >>> > http://groups.google.com/group/twitter-development-talk
> >>> >
> >>>
> >>> --
> >>> Twitter developer documentation and resources:
> http://dev.twitter.com/doc
> >>> API updates via Twitter: http://twitter.com/twitterapi
> >>> Issues/Enhancements Tracker:
> >>> http://code.google.com/p/twitter-api/issues/list
> >>> Change your membership t

Re: [twitter-dev] 4294967295

[Yusuke Yamamoto]

Just FYI, John Corwin should be also aware of this issue.
He gave me a pull request which workarounds it.
https://github.com/yusuke/twitter4j/pull/10

Thanks,
-- 
Yusuke Yamamoto
yus...@mac.com

this email is: [x] bloggable/tweetable [ ] private
follow me on : http://twitter.com/yusukeyamamoto
subscribe me at : http://samuraism.jp/

On Dec 15, 2010, at 10:10 , Taylor Singletary wrote:

> Thanks! This is being looked into. I'll update when I have news.
> 
> Taylor
> 
> On Tuesday, December 14, 2010, Dan Checkoway  wrote:
>> Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens, but 
>> here are a handful of timestamps for unrelated stuff that got logged no more 
>> than 10 seconds *prior* to the 4294967295 error popping out...so they're 
>> fairly close:
>> 
>> Dec 14, 2010 12:34:11 PM PST
>> Dec 14, 2010 1:13:07 PM PST
>> Dec 14, 2010 1:22:48 PM PST
>> Dec 14, 2010 1:27:22 PM PST
>> Dec 14, 2010 1:29:48 PM PST
>> Dec 14, 2010 1:33:36 PM PST
>> 
>> Based on the twitter4j stack trace, I can tell you that it was *always* 
>> user.listed_count that had the funky value:
>> 
>> Exception in thread "Twitter Stream Handling Thread[Receiving stream]" 
>> java.lang
>> .NumberFormatException: For input string: "4294967295"
>> at 
>> java.lang.NumberFormatException.forInputString(NumberFormatException.
>> java:48)
>> at java.lang.Integer.parseInt(Integer.java:459)
>> at java.lang.Integer.valueOf(Integer.java:553)
>> at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>> at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>> 
>> Thanks,
>> Dan
>> 
>> On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary 
>>  wrote:
>> Understandable, Dan.
>> 
>> Can you tell me the last time an event like this happened?
>> 
>> Taylor
>> 
>> On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway  wrote:
>>> I know this is the weenie answer, but I haven't been able to track a
>>> specific offending JSON object down yet, since it only seems to happen on
>>> the firehose, and we're using twitter4j to process that.
>>> 
>>> If we were able to connect to the firehose more than once at a time, I could
>>> easily write a tool to detect and highlight the issue.  Short of that, I'll
>>> try watching the sample stream for a while to see if the same issue pops up
>>> there.  Will report any findings...
>>> 
>>> Thanks,
>>> Dan
>>> 
>>> On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
>>>  wrote:
 
 Hi Dan,
 
 Do you continue to see events like this happening? Can you provide a
 recent example in as-provided JSON or XML?
 
 Thanks,
 Taylor
 
 On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
 wrote:
> Anybody else seeing user.listed_count occasionally coming back as
> 4294967295?  That value just happens to equate to:  1 + (2 *
> Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
> 
> Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned
> that
> separately on the twitter4j list, but I wanted to raise the issue here
> since
> the root cause is twitter sending the weird value.
> 
> Thanks,
> Dan
> 
> --
> Twitter developer documentation and resources:
> http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:
> http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group:
> http://groups.google.com/group/twitter-development-talk
> 
 
 --
 Twitter developer documentation and resources: http://dev.twitter.com/doc
 API updates via Twitter: http://twitter.com/twitterapi
 Issues/Enhancements Tracker:
 http://code.google.com/p/twitter-api/issues/list
 Change your membership to this group:
 http://groups.google.com/group/twitter-development-talk
>>> 
>>> --
>>> Twitter developer documentation and resources: http://dev.twitter.com/doc
>>> API updates via Twitter: http://twitter.com/twitterapi
>>> Issues/Enhancements Tracker:
>>> http://code.google.com/p/twitter-api/issues/list
>>> Change your membership to this group:
>>> http://groups.google.com/group/twitter-development-talk
>>> 
>> 
>> --
>> Twitter developer documentation and resources: http://dev.twitter.com/doc
>> API updates via Twitter: http://twitter.com/twitterapi
>> Issues/Enhancements Tracker:  
>> 
> 
> -- 
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group: 
> http://groups.google.com/group/twitter-development-talk

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/i

Re: [twitter-dev] 4294967295

[Taylor Singletary]

Thanks! This is being looked into. I'll update when I have news.

Taylor

On Tuesday, December 14, 2010, Dan Checkoway  wrote:
> Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens, but here 
> are a handful of timestamps for unrelated stuff that got logged no more than 
> 10 seconds *prior* to the 4294967295 error popping out...so they're fairly 
> close:
>
> Dec 14, 2010 12:34:11 PM PST
> Dec 14, 2010 1:13:07 PM PST
> Dec 14, 2010 1:22:48 PM PST
> Dec 14, 2010 1:27:22 PM PST
> Dec 14, 2010 1:29:48 PM PST
> Dec 14, 2010 1:33:36 PM PST
>
> Based on the twitter4j stack trace, I can tell you that it was *always* 
> user.listed_count that had the funky value:
>
> Exception in thread "Twitter Stream Handling Thread[Receiving stream]" 
> java.lang
> .NumberFormatException: For input string: "4294967295"
>     at 
> java.lang.NumberFormatException.forInputString(NumberFormatException.
> java:48)
>     at java.lang.Integer.parseInt(Integer.java:459)
>     at java.lang.Integer.valueOf(Integer.java:553)
>     at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
>     at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)
>
> Thanks,
> Dan
>
> On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary 
>  wrote:
> Understandable, Dan.
>
> Can you tell me the last time an event like this happened?
>
> Taylor
>
> On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway  wrote:
>> I know this is the weenie answer, but I haven't been able to track a
>> specific offending JSON object down yet, since it only seems to happen on
>> the firehose, and we're using twitter4j to process that.
>>
>> If we were able to connect to the firehose more than once at a time, I could
>> easily write a tool to detect and highlight the issue.  Short of that, I'll
>> try watching the sample stream for a while to see if the same issue pops up
>> there.  Will report any findings...
>>
>> Thanks,
>> Dan
>>
>> On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
>>  wrote:
>>>
>>> Hi Dan,
>>>
>>> Do you continue to see events like this happening? Can you provide a
>>> recent example in as-provided JSON or XML?
>>>
>>> Thanks,
>>> Taylor
>>>
>>> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
>>> wrote:
>>> > Anybody else seeing user.listed_count occasionally coming back as
>>> > 4294967295?  That value just happens to equate to:  1 + (2 *
>>> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
>>> >
>>> > Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned
>>> > that
>>> > separately on the twitter4j list, but I wanted to raise the issue here
>>> > since
>>> > the root cause is twitter sending the weird value.
>>> >
>>> > Thanks,
>>> > Dan
>>> >
>>> > --
>>> > Twitter developer documentation and resources:
>>> > http://dev.twitter.com/doc
>>> > API updates via Twitter: http://twitter.com/twitterapi
>>> > Issues/Enhancements Tracker:
>>> > http://code.google.com/p/twitter-api/issues/list
>>> > Change your membership to this group:
>>> > http://groups.google.com/group/twitter-development-talk
>>> >
>>>
>>> --
>>> Twitter developer documentation and resources: http://dev.twitter.com/doc
>>> API updates via Twitter: http://twitter.com/twitterapi
>>> Issues/Enhancements Tracker:
>>> http://code.google.com/p/twitter-api/issues/list
>>> Change your membership to this group:
>>> http://groups.google.com/group/twitter-development-talk
>>
>> --
>> Twitter developer documentation and resources: http://dev.twitter.com/doc
>> API updates via Twitter: http://twitter.com/twitterapi
>> Issues/Enhancements Tracker:
>> http://code.google.com/p/twitter-api/issues/list
>> Change your membership to this group:
>> http://groups.google.com/group/twitter-development-talk
>>
>
> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:  
> 

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk

Re: [twitter-dev] 4294967295

[Dan Checkoway]

I know this is the weenie answer, but I haven't been able to track a
specific offending JSON object down yet, since it only seems to happen on
the firehose, and we're using twitter4j to process that.

If we were able to connect to the firehose more than once at a time, I could
easily write a tool to detect and highlight the issue.  Short of that, I'll
try watching the sample stream for a while to see if the same issue pops up
there.  Will report any findings...

Thanks,
Dan

On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary <
taylorsinglet...@twitter.com> wrote:

> Hi Dan,
>
> Do you continue to see events like this happening? Can you provide a
> recent example in as-provided JSON or XML?
>
> Thanks,
> Taylor
>
> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
> wrote:
> > Anybody else seeing user.listed_count occasionally coming back as
> > 4294967295?  That value just happens to equate to:  1 + (2 *
> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
> >
> > Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned that
> > separately on the twitter4j list, but I wanted to raise the issue here
> since
> > the root cause is twitter sending the weird value.
> >
> > Thanks,
> > Dan
> >
> > --
> > Twitter developer documentation and resources:
> http://dev.twitter.com/doc
> > API updates via Twitter: http://twitter.com/twitterapi
> > Issues/Enhancements Tracker:
> > http://code.google.com/p/twitter-api/issues/list
> > Change your membership to this group:
> > http://groups.google.com/group/twitter-development-talk
> >
>
> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:
> http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group:
> http://groups.google.com/group/twitter-development-talk
>

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk

Re: [twitter-dev] 4294967295

[Dan Checkoway]

Yeah, you bet.  Twitter4j isn't logging a timestamp when it happens, but
here are a handful of timestamps for unrelated stuff that got logged no more
than 10 seconds *prior* to the 4294967295 error popping out...so they're
fairly close:

Dec 14, 2010 12:34:11 PM PST
Dec 14, 2010 1:13:07 PM PST
Dec 14, 2010 1:22:48 PM PST
Dec 14, 2010 1:27:22 PM PST
Dec 14, 2010 1:29:48 PM PST
Dec 14, 2010 1:33:36 PM PST

Based on the twitter4j stack trace, I can tell you that it was *always*
user.listed_count that had the funky value:

Exception in thread "Twitter Stream Handling Thread[Receiving stream]"
java.lang
.NumberFormatException: For input string: "4294967295"
at
java.lang.NumberFormatException.forInputString(NumberFormatException.
java:48)
at java.lang.Integer.parseInt(Integer.java:459)
at java.lang.Integer.valueOf(Integer.java:553)
at twitter4j.internal.util.ParseUtil.getInt(ParseUtil.java:120)
at twitter4j.UserJSONImpl.init(UserJSONImpl.java:103)

Thanks,
Dan

On Tue, Dec 14, 2010 at 6:42 PM, Taylor Singletary <
taylorsinglet...@twitter.com> wrote:

> Understandable, Dan.
>
> Can you tell me the last time an event like this happened?
>
> Taylor
>
> On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway 
> wrote:
> > I know this is the weenie answer, but I haven't been able to track a
> > specific offending JSON object down yet, since it only seems to happen on
> > the firehose, and we're using twitter4j to process that.
> >
> > If we were able to connect to the firehose more than once at a time, I
> could
> > easily write a tool to detect and highlight the issue.  Short of that,
> I'll
> > try watching the sample stream for a while to see if the same issue pops
> up
> > there.  Will report any findings...
> >
> > Thanks,
> > Dan
> >
> > On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
> >  wrote:
> >>
> >> Hi Dan,
> >>
> >> Do you continue to see events like this happening? Can you provide a
> >> recent example in as-provided JSON or XML?
> >>
> >> Thanks,
> >> Taylor
> >>
> >> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
> >> wrote:
> >> > Anybody else seeing user.listed_count occasionally coming back as
> >> > 4294967295?  That value just happens to equate to:  1 + (2 *
> >> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
> >> >
> >> > Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned
> >> > that
> >> > separately on the twitter4j list, but I wanted to raise the issue here
> >> > since
> >> > the root cause is twitter sending the weird value.
> >> >
> >> > Thanks,
> >> > Dan
> >> >
> >> > --
> >> > Twitter developer documentation and resources:
> >> > http://dev.twitter.com/doc
> >> > API updates via Twitter: http://twitter.com/twitterapi
> >> > Issues/Enhancements Tracker:
> >> > http://code.google.com/p/twitter-api/issues/list
> >> > Change your membership to this group:
> >> > http://groups.google.com/group/twitter-development-talk
> >> >
> >>
> >> --
> >> Twitter developer documentation and resources:
> http://dev.twitter.com/doc
> >> API updates via Twitter: http://twitter.com/twitterapi
> >> Issues/Enhancements Tracker:
> >> http://code.google.com/p/twitter-api/issues/list
> >> Change your membership to this group:
> >> http://groups.google.com/group/twitter-development-talk
> >
> > --
> > Twitter developer documentation and resources:
> http://dev.twitter.com/doc
> > API updates via Twitter: http://twitter.com/twitterapi
> > Issues/Enhancements Tracker:
> > http://code.google.com/p/twitter-api/issues/list
> > Change your membership to this group:
> > http://groups.google.com/group/twitter-development-talk
> >
>
> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:
> http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group:
> http://groups.google.com/group/twitter-development-talk
>

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk

Re: [twitter-dev] 4294967295

[Taylor Singletary]

Understandable, Dan.

Can you tell me the last time an event like this happened?

Taylor

On Tue, Dec 14, 2010 at 3:41 PM, Dan Checkoway  wrote:
> I know this is the weenie answer, but I haven't been able to track a
> specific offending JSON object down yet, since it only seems to happen on
> the firehose, and we're using twitter4j to process that.
>
> If we were able to connect to the firehose more than once at a time, I could
> easily write a tool to detect and highlight the issue.  Short of that, I'll
> try watching the sample stream for a while to see if the same issue pops up
> there.  Will report any findings...
>
> Thanks,
> Dan
>
> On Tue, Dec 14, 2010 at 6:19 PM, Taylor Singletary
>  wrote:
>>
>> Hi Dan,
>>
>> Do you continue to see events like this happening? Can you provide a
>> recent example in as-provided JSON or XML?
>>
>> Thanks,
>> Taylor
>>
>> On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway 
>> wrote:
>> > Anybody else seeing user.listed_count occasionally coming back as
>> > 4294967295?  That value just happens to equate to:  1 + (2 *
>> > Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
>> >
>> > Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned
>> > that
>> > separately on the twitter4j list, but I wanted to raise the issue here
>> > since
>> > the root cause is twitter sending the weird value.
>> >
>> > Thanks,
>> > Dan
>> >
>> > --
>> > Twitter developer documentation and resources:
>> > http://dev.twitter.com/doc
>> > API updates via Twitter: http://twitter.com/twitterapi
>> > Issues/Enhancements Tracker:
>> > http://code.google.com/p/twitter-api/issues/list
>> > Change your membership to this group:
>> > http://groups.google.com/group/twitter-development-talk
>> >
>>
>> --
>> Twitter developer documentation and resources: http://dev.twitter.com/doc
>> API updates via Twitter: http://twitter.com/twitterapi
>> Issues/Enhancements Tracker:
>> http://code.google.com/p/twitter-api/issues/list
>> Change your membership to this group:
>> http://groups.google.com/group/twitter-development-talk
>
> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:
> http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group:
> http://groups.google.com/group/twitter-development-talk
>

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk

Re: [twitter-dev] 4294967295

[Taylor Singletary]

Hi Dan,

Do you continue to see events like this happening? Can you provide a
recent example in as-provided JSON or XML?

Thanks,
Taylor

On Tue, Dec 14, 2010 at 2:13 PM, Dan Checkoway  wrote:
> Anybody else seeing user.listed_count occasionally coming back as
> 4294967295?  That value just happens to equate to:  1 + (2 *
> Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...
>
> Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned that
> separately on the twitter4j list, but I wanted to raise the issue here since
> the root cause is twitter sending the weird value.
>
> Thanks,
> Dan
>
> --
> Twitter developer documentation and resources: http://dev.twitter.com/doc
> API updates via Twitter: http://twitter.com/twitterapi
> Issues/Enhancements Tracker:
> http://code.google.com/p/twitter-api/issues/list
> Change your membership to this group:
> http://groups.google.com/group/twitter-development-talk
>

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk

[twitter-dev] 4294967295

[Dan Checkoway]

Anybody else seeing user.listed_count occasionally coming back as
4294967295?  That value just happens to equate to:  1 + (2 *
Integer.MAX_VALUE)  Sure looks like an unsigned version of -1 to me...

Anyway, it's breaking twitter4j.TwitterStream stuff.  I've mentioned that
separately on the twitter4j list, but I wanted to raise the issue here since
the root cause is twitter sending the weird value.

Thanks,
Dan

-- 
Twitter developer documentation and resources: http://dev.twitter.com/doc
API updates via Twitter: http://twitter.com/twitterapi
Issues/Enhancements Tracker: http://code.google.com/p/twitter-api/issues/list
Change your membership to this group: 
http://groups.google.com/group/twitter-development-talk