I have researched the vendor for the D&I Survey and present the
following information and vote at the bottom.  The goal of this change
is technical to limit spamming as well as improve the deliverability of
the survey and therefore the response rate.

-KAM

Operator: LimeSurvey GmbH https://www.limesurvey.org/about-us/imprint

"The worldwide leading open source survey software
as a professional SaaS solution or as a self-hosted Community Edition."

Licensed: GPL v2 or later (https://www.limesurvey.org/stable-release)



Due to the operator being German, the data protection Terms of Service
are excellent and follow BDSG, TKG and GDPR.  See
https://www.limesurvey.org/policies/terms-conditions, Section 10: Data
Protection.

As is typical of the strong German data protection laws, the privacy
policy is excellent as well:
https://www.limesurvey.org/policies/privacy-policy

The only nit is that technically the terms of service point to the
privacy policy in German:
https://www.limesurvey.org/de/richtlinien/datenschutzrichtlinie so a
minor thing they should fix.

Otherwise, I think it's an excellent vendor providing no concerns for
the ASF to use them as a service provider for the survey.

My only key recommendation is that we make sure the survey is set to
"Turn on the Anonymized responses- option" which will "...mark
participants who complete the survey only with a 'Y' instead of
date/time to ensure the anonymity of your participants."

Therefore, I call a vote and +1 to use limesurvey, request a list of
committer addresses, load them into the SaaS offering and use this to
send to all committers rather than use committers@ for the survey for 1
use only. 

We should also still allow anonymous entries, ask PMCs to post about the
survey and spread the word on our social media.

We should also ask Infra to join in a small test of the survey and to
whitelist as appropriate the surveys on our system as well as to provide
a current CSV file export to KAM to load into the survey software.

If this vote passes, various Jira like DI-30 should be updated to
reflect this approach.

On 11/2/2019 3:12 PM, Kevin A. McGrail wrote:
> Bitergia isn't the actual sender, it would be limesurvey.  I will look
> into how it sends on behalf of but the idea is not to use a mailing
> list software but to have the survey software send each individually.
>
> I doubt di30 talks about this as I have been suggesting offlist how to
> improve the deliverability and response rate of the survey.
>
> On Sat, Nov 2, 2019, 12:35 Sam Ruby <[email protected]
> <mailto:[email protected]>> wrote:
>
>     On Sat, Nov 2, 2019 at 10:26 AM Kevin A. McGrail
>     <[email protected] <mailto:[email protected]>> wrote:
>     >
>     > The Apache.org email addresses are easily harvested from our mailing
>     > list archives.
>     >
>     > This would be an export from LDAP or similar of all @apache.org
>     <http://apache.org>
>     > addresses which is the same as committers@ but will be sent directly
>     > instead of routed through a mailing list.
>     >
>     > There are significant deliverability and response rate concerns with
>     > using a mailing list.
>
>     I may have misunderstood the intent of
>     https://issues.apache.org/jira/browse/DI-30.
>
>     If there is a need to create an alias for all committers, that could
>     be easily constructed.  Bitergia would send a single email to our
>     infrastructure, and our infrastructure would be forwarded to each id
>     on the list.
>
>     If such an alias were created, it should either be set up to only
>     allow emails from known Bitergia emails, and the alias should be taken
>     down when not in use, as it would be a vector for spam.
>
>     - Sam Ruby
>
>     > Regards,
>     > KAM
>     >
>     > On 11/2/2019 5:53 AM, Justin Mclean wrote:
>     > > Hi,
>     > >
>     > > I would also be uncomfortable in creating a list of people to
>     email and making that available even internally. Pervious
>     experience with surveys (non D&I) at the ASF have shown several
>     times that mistake are made and/or emails addresses harvested
>     without permission. If we do go down that path I would also like
>     to know how we are creating this list e.g what would be the
>     criteria to be on it.
>     > >
>     > > committers@ has a wide distribution and with correct messaging
>     we can use it very little effort and risk.
>     > >
>     > > Thanks,
>     > > Justin
>     >
>     > --
>     > Kevin A. McGrail
>     > [email protected]
>     >
>     > Member, Apache Software Foundation
>     > Chair Emeritus Apache SpamAssassin Project
>     > https://www.linkedin.com/in/kmcgrail - 703.798.0171
>     >
>
-- 
Kevin A. McGrail
[email protected]

Member, Apache Software Foundation
Chair Emeritus Apache SpamAssassin Project
https://www.linkedin.com/in/kmcgrail - 703.798.0171

Reply via email to