archive request limit?

[Gerard van Enk]

Hi,

I only get 20 statuses per archive request. Has this been changed
temporarily or does the documentation needs to be updated? It used to
be 80.

Gerard

[Alex Payne]

It's still 80 in the code...

--
Alex Payne
http://twitter.com/al3x

[Gerard van Enk]

On 27 jun, 20:57, "Alex Payne" <a...@twitter.com> wrote:
> It's still 80 in the code...

and why am I only getting 20 statuses at once? I just use the
http://twitter.com/account/archive.xml request. A few days ago (or
weeks, can't remember the last time I checked), it was 80.

----
Gerard
http://twitter.com/gvenk

[Cuyler]

I am writing a script to retrieve and save all my Tweets. A few days
ago I was getting 80 posts per request and now it is only sending 20
posts just as Gerard described.

In addition, there is strange behavior as when I try to access earlier
posts with the "page=n" option. After a point, the request does not
return the proper page, but instead returns the 20 latest posts (the
same result as if http://twitter.com/account/archive.xml&page=1 was
specified). A few days ago, if I tried to access a page with a page
number that was too high, I would get an XML containing no posts
(which makes more sense to me). But now a high page number always
returns the 20 latest posts.

Even worse, I cannot access all of my earliest statuses. I have
something like 273 tweets, but when I try to access page=10, I receive
just the 20 latest tweets. With 20 tweets per page, that means I
could only collect about 200 of the 273 tweets.

I had assumed this was some kind of experiment to reduce traffic.

On a separate issue, I wonder why each status returned by the account/
archive request contains user information. The user information is
the same for each status isn't it? It seems wasteful to repeat it
over and over.

Also the one piece of user information that would be useful when
retrieving the archived statuses is the total number of statuses. That
can be retrieved with users/show request, but as long as you are
sending user information in the account/archive response, I wonder why
you don't include the statuses_count.

Actually what would be nice is an indication of how many pages can be
retrieved. The number of statuses per page is variable and there is
no indication of when the last page is retrieved other than the page=1
posts are returned.

Cuyler

[Ahmed Qurashi]

Having the EXACT same problems as Cuyler above:

1. Calls to the 'archive' method return only 20 items. Not 80 as
stated in the documentation.
2. After passing in page arguments higher than 10, response retrieves
the first page in the sequence.
3. Messages are overly verbose, and contain redundant user info.

I'd love to have one grand archive call that retrieves every message
ever created in one lump sequence. A true archiving utility, that
could be limited to the public API in a reasonable timeframe, say one
archive call per authenticated user per day.

As it stands with the current limits I can only archive 400 tweets per
hour maximum, up to the last 200 tweets written.

[Evan Weaver]

We can give you this grand archive call.

In the meantime, the 20 per page limit will remain. I've updated the
documentation. For disallowed pages, would you prefer an HTTP error
code, an empty API response as before, or what?

Thanks,

Evan

--
Evan Weaver

[Cuyler]

"Evan Weaver" <ewea...@twitter.com> wrote:
> For disallowed pages, would you prefer an HTTP error
> code, an empty API response as before, or what?

I think either would be fine as long as it was documented. Is there a
limit to how far back you can go? Right now it seems like you cannot
retrieve beyond 200 posts (in addition to the 20 requests per hour
limitation).

Cuyler

[Evan Weaver]

Yes, 200. We will increase this to 800 within a few weeks.

Evan

--
Evan Weaver

[Sewcrates]

I'm having problems retrieving the oldest tweets. I have an account
wiht 1,269 tweets, and it only allows me to request up to page 10
before returning the redirection. Is this by design? I noticed
twitter.com does the same thing: it only goes back to page 10 when
reviewing old tweets.

On Jul 4, 12:46 am, "Evan Weaver" <ewea...@twitter.com> wrote:
> We can give you this grandarchivecall.
>
> In the meantime, the 20 per page limit will remain. I've updated the
> documentation. For disallowed pages, would you prefer an HTTP error
> code, an empty API response as before, or what?
>
> Thanks,
>
> Evan
>
>
>
>
>
> On Wed, Jul 2, 2008 at 12:23 PM, Ahmed Qurashi <ahmedqura...@gmail.com> wrote:
>
> > Having the EXACT same problems as Cuyler above:
>
> > 1. Calls to the 'archive' method return only 20 items. Not 80 as
> > stated in the documentation.
> > 2. After passing in page arguments higher than 10, response retrieves
> > the first page in the sequence.
> > 3. Messages are overly verbose, and contain redundant user info.
>

> > I'd love to have one grandarchivecall that retrieves every message

> > ever created in one lump sequence. A true archiving utility, that
> > could be limited to the public API in a reasonable timeframe, say one

> >archivecall per authenticated user per day.
>
> > As it stands with the current limits I can onlyarchive400 tweets per

> > hour maximum, up to the last 200 tweets written.
>
> --

> Evan Weaver- Hide quoted text -
>
> - Show quoted text -

[Evan Weaver]

Yes, by design. It will be changed eventually.

Evan

--
Evan Weaver

[Brian Layman]

I've not delved into the API yet.  I'm in the designing phase of what I want my app to do.  But it seems that even archival call limit of 1 per day would be more than needed.

Does the API allow you to send "Most recent recieved message" as an optional parameter in these calls? Then the API would only need to retrieve a small amount of data. I'm constantly annoyed that Twhirl retrieves all of the tweets with each API call upon its restart.  Really, most of the time, it should only return a few new messages.

On an archival call, "newer than this ID or "older than this ID"  parameters might be useful.