Also On This Page →

Nearby

This tutorial is outdated. For a more up-to-date version, see the Quick Start Guide.

Some of you who may not be familiar with the way that minutes are recorded for W3C meetings might benefit from a bit of a primer/FAQ for scribes.

Use of IRC

Use of IRC is preferred for scribing since the clean-up and preparation of the final minutes can be fairly automated using the various IRC bots, etc. Please let the chairs know if you will be unable to scribe from IRC before the meeting in which you are likely to be selected (e.g. you are atop the scribes list).

At the beginning of the call

Ensure that the call is setup by issuing the following command:

Trackbot will invite the other bots for you and setup the access controls. You then need to identify yourself to Zakim as the scribe:

or

You should also identify the chair:

It is also good practice to add a link to the agenda:

During the call

Identify the agenda topic being discussed

What should be recorded?

W3C minutes should record the sense of all of the discussion held. This may be different than what some are accustomed to (e.g. some other organizations simply record action items and resolutions). Thus, when a topic is being discussed, the scribe should capture the sense (no need to try to get every word)

Note that the person's name is recorded as <name>:

If the scribe is unsure of whom is speaking, then they should record in the minutes:

Someone will hopefully fix this in IRC using a sed command:

otherwise, when reviewing the minutes, the speaker should recognize that it was they who made the point and they should ask the chairs to amend the minutes to reflect that fact. If they do not, then at least the point is recorded (even though no one will ever know who made the point).

Note that if someone makes a specific proposal, then every effort should be made by the scribe to capture it fully and accurately in the minutes. It would be unfortunate if all that were recorded was: joe: makes a proposal for resolving issue-NNNN

Action Items

When recording action items, the form should be as follows:

When recording a comment on an existing action item, the form should be as follows:

You can also close action action items:

As a good practice, always use the form "ACTION-NNNN" when mentioning an existing action item in the record. This syntax will be picked up by the email tracker later.

Issues

As for actions, always use the form "ISSUE-NNNN" when mentioning an existing issue in the record. This syntax will be picked up by the email tracker later. Note that you cannot create or close issues using IRC commands.

Resolutions

When recording a resolution, the form should be as follows:

or, if the resolution isn't related to an issue:

Making/fixing mistakes

When you make a mistake/typo in recording something, you can use sed commands to add corrections to the log:

However, this doesn't work when the typo is with either of the following keywords: ACTION or RESOLUTION.

If the problem was that you mistyped the keyword ACTION or RESOLUTION, then simply enter the ACTION or RESOLUTION again, fixing the mistake.

If, however, the ACTION or RESOLUTION was incorrectly recorded, then these need to be fixed using the form:

The two syntaxes above are unfortunately not being picked up by trackbot, thus one has to go to the Web interface later on and fix the title of the action item.

After the call

Ask rrsagent to generate the minutes:

RRSAgent will reply by providing the link to the generated record.

If you find any mistake, fix them using Making/fixing mistakes. You can then ask RRSAgent to regenerate the record as many times as you need, as long as it's within the same day.

Once the record has been approved by the Group, send a text version of the record to the Group list. trackbot will pick them up and associate actions and issues with the record. You can use the comma tool ,text to generate a text version of the generated record:

 http://www.w3.org/2008/04/02-sml-minutes.html,text

See also the XML Security WG's Scribe-Instructions Page, thanks to Frederick Hirsch.


plh@w3.org