[steering-discuss] LibO registration page gone?

Hi guys,

As not everyone might be on steering-discuss yet, forwarding it here
We can discuss it on sc-discuss then

  So let me try to do that [ I believe I'm subscribed now ]

From: Anthony Papillion <papillion@gmail.com>
> I had posted this in the wrong mailist and intended it here.
>
> I have just been informed that the registration page has been removed
> from LibO. All it took was on bug report to have it removed.

  It was removed before the first Beta; correct.

> Did I miss something? I don't remember a good and deep debate about
> this. Removing the registration page, in my opinion, is short-sighted.

  Personally; I prefer to have the debate framed in terms of whether we
should add something like that back.

While it's true that it is a nuisance

  In the commenter's words - it is indeed a nuisance. Here is at least my
rational:

* Cons of registration page
  + It makes OO.o look like some shareware / marketing-ware thing
  + It clutters our startup, with extra confusing clicks
  + It has some annoying "annoy me again later" option you have
    to carefully avoid.
  + There are real privacy, data collection and storage issues
    around the data that is entered
  + LibreOffice currently has no infrastructure to register and
    store that data at all, never mind in a compliant way
  + Very many people will simply ignore it, thus giving us only a
    subset of users, and presumably those who care little for
    their privacy
  + Ubuntu had a -massive- flamewar before removing a somewhat
    more defensible dialog about something very similar, and
    currently do not ship with this dialog cf.:
    http://news.cnet.com/8301-1001_3-10044054-92.html

* Pros of registration page
  + It allows us to collect data on people that have no problems
    entering their personal information into this sort of thing
    + we could (hypothetically) use that data for something

Maybe even just a name and state/country collector? Removing the
data collection point altogether will create a huge blind spot
between us and our users.

  So - I really don't think this creates a huge blind-spot. I would be
amazed if Mirrorbrain cannot give us state/country information and
breakdown from its list of IP addresses that it re-directs for
downloads. Peter is some kind of wizard - so I suspect we can get this
now. Failing that we know our approximate geographical penetration from
the font analysis work.

  We can also get a similar set of data we can analyse geographically by
looking at hits on our link to the extensions repository.

  So - in summary ... putting this 'back' is not a free operation - it
takes real work, it also annoys real users (including some of the
commenters, and me too FWIW ;-).

  Now - the migration page - that is different, we have a plan to fix the
issues Mechtilde carefully identified there, and I need to execute on it
(working ... :wink:

  HTH,

    Michael.

Hi guys,

> As not everyone might be on steering-discuss yet, forwarding it here
> We can discuss it on sc-discuss then

  So let me try to do that [ I believe I'm subscribed now ]

Me too now.

* Cons of registration page

Agree with all of that. Another way of viewing this IMO, is that if it
didn't already exist noone would suggest adding a dialog shown during
startup in order to collect what I believe is very dubious and skewed
information. e.g. I don't think a single Linux distribution that shipped
OpenOffice.org left the dialog enabled, so noone using the OOo shipped
with Debian/Ubuntu/Fedora/RHEL/Suse/etc was counted.

C.

Hi Caolan, all!

[...]

> * Cons of registration page

Agree with all of that. Another way of viewing this IMO, is that if it
didn't already exist noone would suggest adding a dialog shown during
startup in order to collect what I believe is very dubious and skewed
information. e.g. I don't think a single Linux distribution that shipped
OpenOffice.org left the dialog enabled, so noone using the OOo shipped
with Debian/Ubuntu/Fedora/RHEL/Suse/etc was counted.

Since distributions like Ubuntu do also provide to collect usage
statistics of installed packages, there is already some similar
information source. Ubuntu does that to improve the collection of
software, to better suit their user's needs. So the real point is: What
is required for us to improve LibO and to make it more successful?

Example: If a project gets a certain size, then there will be the need
to base decisions (both for the project and the product) on valid
information (number of users, there likings, common work flows,
essential templates, ...). In my opinion, gut feeling (only) doesn't
serve some of our needs in the long run.

Personally, I don't like the "register" approach, but I do like good
ways of gathering helpful information that isn't available elsewhere. We
can be sure that the structure of people using OOo/LibO isn't identical
to those we get direct feedback from. So the question is (as you stated
it), if there is nothing in the product ... what would be our approach
to let our users help us to better shape LibO?

The idea "User Feedback" / "Improvement Program" / "Usage Tracking" that
is available in OOo is one puzzle piece. There are some rough edges, and
the data currently needs some manual pre-processing, and there is still
missing information - but it already helped to convert "just features"
into something that "just works for the majority".

So my proposal would be: There might be no need for a registration
dialog, but there is a need by development, marketing, UX to get some
information. Maybe not today, but tomorrow (literally speaking). But, I
also second Florians thoughts that there should have some discussion in
advance before removing it.

By the way, some people worked on (limited) data collection for The
Gimp. Just fun to look at (serious data): http://ingimp.org/

Cheers,
Christoph

All these are good things, but not provided by the infamous registration
dialogs IMO. If someone wants to stick a menu entry in to enable an
"improvement program" which tracks what menus, dialogs, buttons, etc.
are most used in order to get useful data as to what gets used most then
that's a cunning idea. I was greatly impressed by the frequency-use data
you were able to present for, say, the printer dialog work.

C.

Hi Caolan!

Thanks for your answer - it seems that we think into the same
direction :slight_smile:

> Personally, I don't like the "register" approach, but I do like good
> ways of gathering helpful information that isn't available elsewhere. [...]

All these are good things, but not provided by the infamous registration
dialogs IMO. If someone wants to stick a menu entry in to enable an
"improvement program" which tracks what menus, dialogs, buttons, etc.
are most used in order to get useful data as to what gets used most then
that's a cunning idea.

Yep. Concerning the registration dialog and some requirements/wishes by
the marketing team - my mail was more meant how we can make the user
feedback functionality more versatile to suit their needs, too (e.g.
what platform, language settings, screen/window size, ...).

[...]

Cheers,
Christoph

+ some corporate or Gov users/customers HATE this, as they are touchy on data privacy, newtork traffic, proxy setting, user assistance, help desk costs to explain why, bla bla bla....

In big corporations / Gov, 100.000 small individual annoyances is a huge annoyance.

Regards