Hi Michael,
thanks for your reply. 
Simplifying and exaggerating a bit, I'd try to sum up the described
problem as "There's not enough revenue for ecosystem companies, but
those are essential for LibreOffice." and the described solution as
"Let's discourage enterprises/organizations from using LibreOffice from
TDF, and hope they'll use paid versions from ecosystem companies instead."
Right; it -could- be seen as a simple "developers for users" trade off.
I'm not sure it is a trade-off though: I think we'll win more users by
having happy enterprise users and more investment in feature / function
and a richer product myself.
I fully agree with the latter part, let's describe it as:
more contributors => richer product => more users
My assumption is still that having more users will also result in more
contribution:
more users => more contributors => richer product
The nice thing is that if you combine both of them, that results in:
... => more users => more contributors => richer product => more users
=> more contributors => richer product => ...
(which could probably be better depicted by some nice graph showing a
circle with continual growth, but I think the idea is clear...)
The interesting question then is how/where to start.
And my concern is that the "Personal Edition approach" will lead to
fewer users and I'm not so sure this will ever be compensated either in
the medium or the long run.
There is risk in any change, but also risks in stasis - particularly
when we know the status quo doesn't work well.
The "Personal Edition approach" *might* work of course; it's not what
I'd personally expect, though.
From what I have heard, there's also a tendency in (particularly in
large) organizations to only use products backed by some kind of SLA, so
there is some contractor to contact (or blame) in case of problems.
I've met a few organizations like this - but they seem to be extremely
rare. Can they even get an SLA for Firefox or Chrome ?
Ilmari has answered the explicit question regarding Firefox support
already. (I wouldn't have known myself.)
In any case, the wording "only use products backed by some kind of SLA"
probably was a bit too strong and there are other factors limiting what
software that applies to.
While browsers are certainly mission-critical these days, the facts that
they are available for free (i.e. gratis) and the existence of web
standards make it easier to have multiple browsers in parallel or switch
between them (at least these days, was certainly different in the past),
making that a somewhat less "critical" component in my eyes regarding
professional support.
In theory, document standards should allow switching between office
suites or using them in parallel as well, but that is known to be much
more difficult in practice, due to interoperability issues and
additional components on top, like macros or all kinds of third-party
software, so the office suite becomes some kind of "platform" that is
mission-critical and not easily replaceable.
I think organizations with such an approach (use software with SLA for
mission-critical tasks, in slight variations of where this applies) are
not too uncommon among larger organizations; maybe too few of them are
using LibreOffice for various reasons (never heard of it, never heard
there is professional support available, doesn't fit their needs,...).
I fully agree that it's unfortunate if migrations to LO (and FLOSS in
general) are/were done/encouraged only because it's "free as in free
beer", "no cost at all", which certainly isn't key to success for either
the enterprise nor the LibreOffice ecosystem.
So - lets turn this around - can anyone thing of more than
five enterprises that paid for support or instead (just as good)
contributed meaningfully to LibreOffice instead ? Munich, and ...
At least those 3 quickly came to my mind
Sure; there's a reason I picked five 
Adding the two from the follow-up email, those that came to my mind
(without checking git log):
* NISZ
* SIL
* TU Dresden
* BaseAlt
* BSI (German "Bundesamt für Sicherheit in der Informationstechnik")
And of course, there's RedHat, if all enterprises except those listed in
the "Ecosystem partners" section on the website [1] count...
Regarding paid support, I've at least heard from two or three
organizations, but don't know what amounts of money were/are involved
there; that's certainly something the involved ecosystem companies (so
basically you and Thorsten) know better...
So - I was talking of new contributors; how many can we think of that
are new since 2018 ? =)
Regarding all contributions, NISZ falls into that category; but I don't
know details regarding paid support, those are presumably not publicly
available.
To be balanced, when talking only about new contributors, I think it'd
also be good to consider how many new (enterprise) users there actually
are since 2018. Do you know whether such numbers are easily available?
=> It is the norm to deploy LibreOffice from TDF in
enterprises, and pay nothing for support &
maintenance that can go into development.
+ its that good.
Might one (main) problem be that LibreOffice (from TDF as well as its
enterprise derivatives) just is not widely used by companies whose IT
strategy involves paying for their office suites (yet)?
We're really quite widely used; our 200+m users includes many large
government and business deployments.
IMHO, it'd be ideal to try to get more organizations switch to
LibreOffice editions from whatever they're using now which I'd expect to
increase demand for professional support as well.
I think this was one of the headings in my mail. With the current %age
up-take of professional support we run out of world population before we
get enough developers to make LibreOffice fly.
I tend to have a more optimistic view here, given what happened e.g. in
Munich and the contributors mentioned above.
I think one/the key point where we have different expectations is
whether we believe (enough) large enterprises and governments tend to
ask for professional support when using LibreOffice as mission-critical
software. - And we have obviously made different experiences there as
well, which may be one reason for this.
Besides the existing examples already relying on support and/or
contributing back to LibreOffice themselves, there are also those that
you mention as "bad examples" in your initial email:
Another pathology is that there are companies who ship
LibreOffice, often claiming support, but then file all their tickets
up-stream and hope they are fixed for free. Naturally they are cheaper
in government tenders, they use our brand, they leave the customer
with hundreds of un-fixed bugs, and all of the users with a terrible
experience.
In those cases, a demand for support clearly exists (but is met in a
non-optimal way; I think Sophie's email [2] describes it well).
Maybe trying to steer that into another direction would be a huge step
forward?
In case the customers are actually making bad experiences, is it
possible they'll be more willing to pay more when they do the next tender?
While public tenders have a tendency to select the provider with the
lowest price, criteria like expertise in the area - proved e.g. by
having LibreOffice-certified staff - could possibly of help here.
So a key question here might be where those companies look for (and
easily find!) information how to do things better next time (or right
from the start)?
As written in my previous email [2], I agree that many larger
deployments involving "professional use" will probably want to use an
edition with some kind of professional support (e.g. due to Service
Level Agreements, long-term support, more stability, new features) and
the TDF-provided version won't fit their needs, regardless of whether it
has a "Personal" tag attached or not.
Therefore, also from the experiences that the City of Munich made, I
tend to expect that affected organizations will find this out
Well - it took Munich a long time to find this out I think; furthermore
our marketing tends not to make people effectively aware of the
existence of, nevermind the benefits of, support / migration / training
- even in the abstract. It also tends to make people believe the
software is created by Volunteers + TDF at many points. I guess
enterprises think that TDF is sustained by donations from end-users, and
volunteers just train themselves & contribute - so ... no need to
support the ecosystem ? =)
The more I read and think about the topic, the more it seems to me that
this may be a/the key problem.
In general, I'd be fine about making it more obvious/prominent that
professional support via ecosystem companies is available, and it's
recommended to organizations to actually make use of it, and think this
is actually crucial to improve the situation.
The primary place where I'd expect such notes is on the LibreOffice website.
I just visited libreoffice.org, and the way where I'd look first is
under "Get Help" -> "Professional Support", which mentions this after
introducing the different kinds of certification:
"If you are interested in their services, the list below shows our
certified developers and professionals - with their affiliation - in
randomized order."
I wouldn't read any explicit recommendation into that, maybe that'd be a
good place to encourage/emphasize that further?
You may notice the other discussions here arguing for a replacement of
the explicit recommendation to get support & services from the download
page (which we know doesn't work) with a suggestion.
As mentioned, I myself have no general objections against recommending
professional support and services at appropriate places and even
clarify/strengthen that where appropriate, thus "positively encouraging"
making use of those.
However, the "Personal Edition approach" with the suggested label is
rather discouraging the use of TDF LibreOffice, which I see as a
different approach, one I don't like.
While those companies may not contribute meaningfully to LibreOffice
upstream, I tend to think that they will probably manage to do their own
branded build of LibreOffice ("MyOffice" or whatever, without a
"Personal" tag attached), and then offer that with the same
(nonexisting) support for basically the same price.
Sure - but they will have to put their own brands on something and
associate the terrible reputation for support with their brand not the
LibreOffice project =)
I'm wondering, however, whether that is actually perceived much
differently from what actual ecosystem companies provide. Both are "some
vendor-branded office suite based on LibreOffice", or is there some
additional obvious distinction for those who don't know more about the
background?
Related question: Would LibreOffice's current Trademark policy [1]
(currently or in adapted form) actually require to not use the name
"LibreOffice" when disabling the "Personal" branding, e.g. via autogen
option?
I'm wondering what Linux distros will do if so, e.g. I can hardly
imagine some enterprise Linux distro like RedHat's or SUSE's shipping
"LibreOffice Personal" - and don't think it would be helpful to have
something similar to what the Firefox/Iceweasel situation was like in
Debian.
If there are large numbers of users who refuse to contribute anything
in enterprises - and they will switch away if we ask them to: it seems
to me we're unlikely to get much from them anyway.
I pretty much agree here with what Justin wrote in [3].
I even think it's fine if there are organizations using LibreOffice for
free if it actually fits their needs. As long as there are enough others
that contribute in some way, that should not be a problem, so I don't
see a need to "force" them to use something different instead.
(And I do not think that an unsupported LibreOffice fits the needs of
most, in particular large, organizations, so there should be "enough
need" for professional support.)
In the end, I think that in general, even "just" using LibreOffice is a
contribution by "spreading the word" and making the software known to
more people, as happens e.g. in universities that might otherwise fall
back to providing just other (proprietary?) software to their
students, who will later become the people to decide in enterprises, etc.
(That of course doesn't really apply where LibreOffice isn't suitable
and results in a horrible experience...)
At least among the (non-IT) people that I talk to, many initially have
no idea what I'm talking about when I mention LibreOffice, about half of
them know OpenOffice, though, and ~everybody knows MS Office.
I try to think well of people, so I do think there are a large pool of
well-meaning people who like our brand, and product and who - if
effectively steered - and can provide a clear and plausible rational to
their management: "We can't deploy the Personal edition - we need to use
the Enterprise edition" - will seek out and support the project in this
way. Every LibreOffice deployment must have an enthusiast behind it: but
(apparently) none of them can effectively encourage anyone to pay; what
can we do to help them ? =)
This assumption/case (there are already people who want to contribute,
but their management doesn't want to) wasn't clear to me earlier.
In fact, my assumption of a wider use of LibreOffice leading to more
contributors mostly works where IT management does a good job without
having to be forced to do so. And I fully agree that the "We want to use
LibreOffice just because it doesn't cost anything" approach is generally
not helpful, but may even cause more harm then good.
(The same is true for FLOSS in general, not just LibreOffice, IMHO.)
In any case, as others have already said, I personally don't like the
idea of "actively discouraging" the use of TDF's LibreOffice, but it'd
be great to have an approach to more positively encourage the use of
enterprise editions.
Differentiation is like that. Somewhere we have to have a page which
says: "Who should use what version" - and/or the enterprise people have
to have a way to say: use XYZ version for ABC reason. I hear lots of
good-ideas about adding proprietary features & value for ABC - but
they're not attractive to me. A simple marketing message would be great
that does this differentiation in one place outside of the software.
I fully agree.
Ultimately, the goal should be to somehow convince organizations
currently using other office suites to migrate to LibreOffice
(Enterprise), and I think that the popularity of TDF's LibreOffice plays
a vital role there as well.
Unless someone tells them that TDF's LibreOffice is not suitable for
their enterprise - their first (and final) stop is to deploy
GratisOffice I'm afraid, ~all the data points in that direction. They
see GratisOffice as more genuine, and legitimate and authentic than
Collabora Office eg. which is pretty depressing given how much we put in.
Yep, the question is just what the best way to tell them is.
Regarding the main "target audience":
Is the assumption that (most of) those people who need to be told
actually don't know that there is and they might want professional
support or that they know, but still ignore it?
In the former case, adding the relevant information (professional
support available and encouraged for enterprises) more explicitly to the
website or maybe using existing mechanisms to inform the user (like
mentioning support in the donation/contribution infobars or add a "Tip
of the Day" with that info) might be a good way of dealing with this, in
my opinion.
For the latter case (they know, but don't want to), this wouldn't help,
of course, but I think there are legitimate use cases in that area as well.
One scenario e.g. is also if an organization uses TDF's LibreOffice, and
does code contributions or pays an ecosystem company to fix bugs or
develop new features, and is happy to get them with the next TDF (minor)
release.
While I understand that you prefer a license-based approach to this one,
this is still contribution we shouldn't discourage (but rather
encourage) in my opinion.
(I know of at least one university that went with this approach.)
Then the other problem I have is a soft heart & over-enthusiastic
programmers. Customers tend to put these puppy-dog eyes on and say: "I
can't afford it, but surely you'll do it anyway at ~below cost; it's all
for the good of the project / product" - and I get suckered (again), and
then on top our staff do an extra specially good job that goes beyond
the minimum the customer asks for - and consume a ton of time doing that
- and in any circumstance estimating time for projects is a nightmare -
so when you look at the economics, it ends up loss-making. It is far
from trivial running a business.
I really appreciate and highly respect that way of not doing things just
for business sake but with a deep interest in LibreOffice itself. =)
But I see the difficulties and that actually sounds related/similar to
the topic being discussed here...
Obviously, it's much easier for me to make assumptions on how things
may/should work without actually running an ecosystem company, so thanks
for your engagement.
Best regards,
Michael
[1] https://www.libreoffice.org/download/libreoffice-in-business/
[2]
https://listarchives.documentfoundation.org/www/board-discuss/msg04720.html
[3]
https://listarchives.documentfoundation.org/www/board-discuss/msg04681.html
