PYLON GA 1.5 Release Notes



The main theme of GA 1.5 is incorporating usability feedback gathered from customers in trials during Q2. GA 1.5 is targeted for release on August 17.

New features

Common target for stories & engagements. In GA 1.3 & earlier, customers need to use 2 separate CSDL targets to return stories and their engagements: fb.content for stories, and fb.parent.content for engagements. So for example, two conditions are required to return stories that mention “NFL” and their associated comments, likes and reshares:

fb.content contains “NFL” or fb.parent.content contains “NFL”

Based on customer feedback, for ease of use we will introduce a new target, fb.all.content, which will return both stories and their engagements based on story text with a single condition such as:

fb.all.content contains “NFL”

Timezone support for daily time series analysis. When customers perform time series analysis at daily intervals, DataSift calculates each daily result on a UTC-based clock. This means that when defining a time range for an analysis that covers a 24-hour period on a US-based time range, users who generate two interactions inside the same US-timezone day, but different UTC-based days, will be incorrectly counted for two dates. In GA 1.5, DataSift will allow users to pass an optional UTC offset parameter on time series analysis queries which defines a timezone for daily interval analysis.

Capacity Notifications. PYLON users will now receive email alerts when their account-level daily capacity reaches 50%, 90% and 100%. By default, users will receive one email per threshold reached. To turn these off or receive a single daily or weekly email about these events instead, go to

New endpoint for tag discovery. To make it easier to use VEDO tags applied to interactions in your recordings, a new PYLON API endpoint has been introduced. When you perform a GET request to, specifying a CSDL hash as a parameter, you’ll be able to see a list of tag trees applied in your recording. The list is subject to all the same privacy-based audience size gating rules which govern use the /analyze endpoint. If you use the /tags endpoint to identify tag trees which are available for analysis, you can optimize your analysis queries by only using tag trees in query filters which are present in your recordings, rather than trying all tag trees included in your interaction filter.
Here’s an example response from the /tags endpoint:


Platform Protection Limits

Time range limits on time series analysis. For platform protection, DataSift will restrict analysis time ranges for time series queries based on the customer’s chosen interval. The limits will be:
minute interval <= 60 minutes
hour interval <= 24 hours
daily interval <= 7 days
weekly interval <= 4 weeks
monthly interval <= 1 month


SPU Billing introduced. With the release of GA 1.5 the DataSift platform will begin billing SPUs for all PYLON customers. SPU usage will be counted against the customers purchased allowances. No SPUs will have been counted prior to the GA 1.5 deployment date.

Bug Fixes

Report remaining capacity in PST. Because daily data volume limiting is performed on a PST-based clock, the remaining_index_capacity and remaining_account_capacity fields returned by the /get endpoint need to report based on data indexed in the current PST day. Currently they report based on UTC days, and this will be corrected in GA 1.5.

Analysis filters will support 30 logical operators. Analysis filters, sometimes referred to as “secondary CSDL”, have a non-deterministic limit on the number of logical operators (and/or/not) which may be used. In GA 1.5 this limit will be set at 30 for all queries.


In response to customer feedback, we will be making changes to the new time series limits introduced in PYLON GA 1.5:

  • change limit for daily interval from 7 days to 32 day
  • change limit for hourly interval from 24 to 336 (2 weeks)

The changes will be released on Tuesday, August 25th at approximately 02:00 - 03:00 PDT. There is no expected downtime during this window.