Re: Proposal for an errata management

> On 23 Nov 2015, at 14:31, Yakov Shafranovich <yakov@noom.com> wrote:
> 
> How do other groups in the W3C handle errata?
> 
> I know that on the IETF side, the RFC Editor has errata tools in their
> system used for publishing RFCs and the errata get automatically
> displayed along side the RFCs. Example here:
> 
> https://www.rfc-editor.org/errata_search.php?rfc=2616

This looks handy… but we do not have anything like that. Each working group sets up an errata page the way they want. For a long time, it was simply a web page maintained by the staff contact (with errors coming in by email), then there was the Wiki age, when a public wiki page was set up that people could use, but we are now in the github age, so a bit of an unchartered territory (there aren't that many groups that pushed all its work onto github like we did).

Ivan

> 
> On Sun, Nov 22, 2015 at 10:12 AM, Ivan Herman <ivan@w3.org> wrote:
>> Dear all,
>> 
>> I played with Javascript and the Github API the past few days, and I have a
>> proposal improving what I originally set up.
>> 
>> There is a mock_up file (that I used for the testing) at:
>> 
>> http://w3c.github.io/csvw/errata/mock_up
>> 
>> The important point is:
>> 
>> - The HTML file runs a script that retrieves issues with a specific label
>> ('Errata') from a github repo (I used my old development repo which is
>> fairly dormant right now in that test mock_up file)
>> - The errata themselves are displayed among sections based on a specific
>> label assigned to that section. The same erratum can appear several times if
>> it is labelled accordingly. Finally, if an erratum does not use any of those
>> section specific labels, it is displayed in a separate section
>> - Each erratum is displayed with some of its important characteristics,
>> including other possible labels. There is also a possibility to add a
>> comment to the discussion (if any) starting with the word 'Summary:', which
>> is then displayed separately. This may be a good practice when raising an
>> erratum is followed by some discussions
>> 
>> The file above shows what it does; the file below
>> 
>> http://w3c.github.io/csvw/errata/errata
>> 
>> shows how this would translate to our case, with a description of the
>> process.
>> 
>> That means that report is done automatically, the average handling of an
>> erratum is entirely on the issue list and nowhere else. Keeping the report
>> on github also allows for an easier change on the text/workflow, etc, so it
>> may be worth keeping there (instead of hosting the report on W3C).
>> 
>> One thought that I did not implement: we do have a number of issues that we
>> postponed for a possible future release. We could label these as errata and
>> add a separate section in the report for 'postponed issues'. Although these
>> are, technically, not errata, but it would be still o.k. in the W3C
>> terminology (W3C errata often include future improvement things.). But we
>> may decide to keep those apart.
>> 
>> WDYT? Should I proceed and use the errata report page as above as the
>> 'official' one?
>> 
>> Cheers
>> 
>> Ivan
>> 
>> 
>> 
>> On 18 Nov 2015, at 14:47, Ivan Herman <ivan@w3.org> wrote:
>> 
>> Actually… I may have found some ways of including the issues automatically
>> into the errata page. It will require some javascripting, but I do not mind
>> playing with this. This means that the content of the errata page will be
>> filled automatically. I would still keep that page on github, if anybody
>> wants to make some change at some point later…
>> 
>> Ivan
>> 
>> 
>> 
>> 
>> On 18 Nov 2015, at 14:23, Ivan Herman <ivan@w3.org> wrote:
>> 
>> … while I was at it, I have made the necessary changes in the document
>> config for the REC, with the provisional date of the 17th of December.
>> 
>> Additionally, I have set up an errata page. Although it has a W3C URI
>> 
>> http://www.w3.org/2013/csvw/errata.html
>> 
>> it redirects to
>> 
>> http://w3c.github.io/csvw/errata/
>> 
>> This is just a proposal on how we can proceed with errata; because we always
>> used GitHub, it seemed logical to use github for that purpose, too.
>> 
>> It is a bit of a pain that the accepted errors have to be recorded on the
>> errata page manually, although it may help in separating the real errata
>> from the fake ones. Nevertheless… if somebody knows some good javascript
>> tricks to get the data directly from github on the fly rather than doing it
>> manually, that would be better. I believe there is Github API, so it may not
>> be impossible. I will have a look, but, in the meantime, what is there may
>> be fine.
>> 
>> If you prefer another approach, let us discuss it now, while we still have
>> time…
>> 
>> Cheers
>> 
>> Ivan
>> 
>> ----
>> Ivan Herman, W3C
>> Digital Publishing Lead
>> Home: http://www.w3.org/People/Ivan/
>> mobile: +31-641044153
>> ORCID ID: http://orcid.org/0000-0003-0782-2704
>> 
>> 
>> 
>> 
>> 
>> 
>> ----
>> Ivan Herman, W3C
>> Digital Publishing Lead
>> Home: http://www.w3.org/People/Ivan/
>> mobile: +31-641044153
>> ORCID ID: http://orcid.org/0000-0003-0782-2704
>> 
>> 
>> 
>> 
>> 
>> 
>> ----
>> Ivan Herman, W3C
>> Digital Publishing Lead
>> Home: http://www.w3.org/People/Ivan/
>> mobile: +31-641044153
>> ORCID ID: http://orcid.org/0000-0003-0782-2704
>> 
>> 
>> 
>> 


----
Ivan Herman, W3C
Digital Publishing Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
ORCID ID: http://orcid.org/0000-0003-0782-2704