LibreOffice: the next five years [LWN.net]

By Jonathan Corbet
July 9, 2020

The LibreOffice project would seem to be on a roll. It produces what is widely seen as the leading free office-productivity suite, and has managed to move out of the shadow of the moribund (but brand-recognized) Apache OpenOffice project. The LibreOffice 7 release is coming within a month, and the tenth anniversary of the founding of the Document Foundation arrives in September. Meanwhile, LibreOffice Online is taking off and, seemingly, seeing some market success. So it is a bit surprising to see the project's core developers in a sort of crisis mode while users worry about a tag that showed up in the project's repository.

LibreOffice was based on firm free-software principles, and is an egalitarian organization overall, so it is not surprising that the appearance of a "Personal Edition" tag in a recent 7.0 release candidate raised some eyebrows. Company-dominated projects will often withhold features for "enterprise" customers, delaying their arrival into the second-class "community edition" or keeping them entirely proprietary. But LibreOffice is not supposed to be such a project; it is owned by an independent foundation and its development is driven by a few companies. LibreOffice is freely shared by everybody, or at least it has been so far.

The fuss quickly reached a level that required the Document Foundation's board to issue a statement about what was going on. The board emphasized that there will be no license change for LibreOffice, and no changes to "the license, the availability, the permitted uses and/or the functionalities". But there is still something going on:

This "Personal Edition" tag line is part of a wider 5 years marketing plan we are preparing and it has the purpose of differentiating the current, free and community supported LibreOffice from a LibreOffice Enterprise set of products and services provided by the members of our ecosystem.

So, while nothing is going to change, there is still a plan to create different versions of LibreOffice, some of which will need to be paid for.

Some problems

The driving force behind the changes is easy enough to understand; it is one that many successful free-software projects face. LibreOffice is a huge program, and developing it takes a lot of work. According to this marketing plan [PDF] put together by the project, nearly 70% of the changes to LibreOffice come from developers paid by "ecosystem companies"; those companies pay about 40 people to work on LibreOffice. That is not a small expense; the companies involved will only be able to sustain that level of development if LibreOffice is bringing in a corresponding amount of revenue.

In a lengthy post titled "Some problems", project co-founder Michael Meeks explained that this revenue is not coming in. Part of the problem is that Microsoft provides "poor to non-existent support to the majority of users" of Office, he said, so nobody thinks in terms of buying support for any office suite:

It is routinely the case that I meet organizations that have deployed free LibreOffice without long term support, with no security updates etc. Try the Cabinet Office in the UK (at the center of UK Government), or a large European Gov't Department I recently visited - 15,000 seats - with some great FLOSS enthusiasm, but simply no conceptual frame that deploying un-supported FLOSS in the enterprise hurts the software that they then rely on.

Companies think of LibreOffice, he said, in the same way that they think about web browsers, which are available for free and are well supported. But work on web browsers is paid for by advertising, which is not a model that works for an office suite.

The problem is compounded by companies that sell inexpensive "support" for LibreOffice, but which are not involved in its development and are not really able to provide that support. Those companies "file all their tickets up-stream and hope they are fixed for free". Companies working in that mode have no problem pricing their offerings below those of the companies doing the actual work (and thus winning much of the business that does exist). In addition, they simply call their offerings "LibreOffice", which actually looks more authentic than services from other companies, which are trying to build their own brands around LibreOffice support.

LibreOffice, he concluded, has tried to do something unique and is finding its path to be difficult:

As we look around the industry we see tons of organizations exploring ways to solve similar problems. LibreOffice's is ‑particularly‑ challenging, because we aspire to being a vendor neutral project. There are reasonably well-known ways to build a company controlled, branded, FLOSS project - we know and love lots of them: openSUSE, Fedora, Nextcloud, ownCloud etc. this is the norm. With TDF we tried to do something far harder - to create an vendor neutral ecosystem that can help retain the community spirit while delivering on our mission. That has proved extraordinarily harder.

The result of all this is that the LibreOffice ecosystem "is under long term stress".

The plan

In response to these problems, members of the LibreOffice community have been working on a five-year marketing plan, the core of which can be seen in the slides linked above. The intent is to create differentiated versions of LibreOffice while avoiding open-core or proprietary business models. Part of that involves getting a better handle on the LibreOffice brand.

The plan starts by creating the concept of the "LibreOffice Engine", which is a term to describe the core LibreOffice code. It is meant to be a way to enable products selling under their own brand to associate themselves with LibreOffice while maintaining their own identity. "LibreOffice Engine" is described in the plan as a sort of equivalent to the highly successful "Intel Inside" branding effort. Presumably this term would be trademarked by the Document Foundation; the plan does not get into what constraints would be put on who could use the trademark (and how).

Then, there is the Personal Edition, which would be "forever free" and only available from the Document Foundation. This release would be tagged, according to the plan, "volunteer supported, not suggested for production environments or strategic documents". The alternative would be "LibreOffice Enterprise", which would only be available from "ecosystem members". This version would come with commercial support and a corresponding price tag.

LibreOffice Online seems to be a place where a lot of tension resides, perhaps unsurprisingly, since that is where the bulk of the money is being made with LibreOffice now. Companies would like to keep parts of LibreOffice Online to themselves, but that threatens to disrupt the volunteer part of the development community. The plan involves the same split between "personal" and "enterprise" offerings, but adds a little note: "There will be an X month gap between the release of the two versions: LibreOffice Online Enterprise and LibreOffice Online Personal".

The hope is that this plan will give the true "ecosystem members" something attractive to sell and, to an extent, free them from the difficult challenge of competing with the free LibreOffice offering. It is, in many ways, reminiscent of the path Red Hat took years ago to differentiate its Enterprise Linux offering, complete with insinuations that the free version might not be fully trustworthy. That approach has clearly worked well for Red Hat; it would be hard to argue that it has not worked well for the wider Linux community too.

Free software is an inherently challenging base upon which to try to build a company. Many in the free-software community are happily indifferent to the fate of companies working with the code, but without successful companies we would not have much of the code that we depend on every day. As Meeks pointed out, LibreOffice without companies would look a lot like the cobweb-strewn OpenOffice project; it is hard to see that as a win for anybody. So one can only wish LibreOffice and the Document Foundation luck as they seek a way to solve this problem while remaining true to the free-software principles that sparked the project's launch in the first place. Ten years of LibreOffice is nowhere near enough.

to post comments

LibreOffice: the next five years

Posted Jul 9, 2020 20:40 UTC (Thu) by atai (subscriber, #10977) [Link] (7 responses)

"The water can carry the boat. The water can sink the boat" a Chinese saying about the power of a community over its governing body or a project in this case.

LibreOffice: the next five years

Posted Jul 9, 2020 22:33 UTC (Thu) by timrichardson (subscriber, #72836) [Link] (6 responses)

The argument that small companies are stealing by charging service fees and then sending tickets upstream is a bit scary. Firstly, there are services beyond bug fixing, such as training, deployment, conversion, which the fees would cover. Maybe small, local businesses just do this better?

Secondly, even if a service company did nothing but file bug reports upstream, a strong open source project can't credibly have a contributor shouting 'theft' because all contributors 'steal' from each other, this is the mutual dependency which sustains an open source project. If a company has such control over a project that they perceive economic harm because they can't monetise their code contributions to the core product, the mutual dependency is broken.

LibreOffice sounds like it is on the way to being a 'noblesse oblige' type of open source.

LibreOffice: the next five years

Posted Jul 9, 2020 23:02 UTC (Thu) by NYKevin (subscriber, #129325) [Link]

I'm of two minds here.

If the tickets are well-written and otherwise "reasonable" (contains steps to reproduce, can actually be reproduced reliably, etc.), then it's obviously none of TDF's business who's actually filing those tickets. But if they are more along the lines of "[User]'s installation is broken. Please fix it.", then I think TDF does have the right to ask those responsible to either file better tickets, or stop filing them altogether ("This is a bug tracker, not a support line. [We/our partners] sell support services for $X per seat per month. Please let us know if you are interested."). The article is unclear about which of those two extremes we're talking about here, so I imagine it's something in between, like this:

"[User] reports that on alternate Sundays, saved spreadsheets get randomly corrupted."
"What do you mean by 'randomly' and 'corrupted?' Can you try X and Y, and send us a copy of the spreadsheet?"
"We're currently discussing that with [User], but they don't want us to send the spreadsheet because they consider it private. Can we send Z instead?"
"No, Z really isn't good enough; please ask them to make a test spreadsheet with fake entries. Also, you still need to tell us what you mean by 'randomly corrupted.' And what happened when you tried X and Y?"
[etc.]
A year later, an entirely unrelated bug report identifies the heap corruption responsible, but the first ticket isn't closed because [User] never told anyone the problem stopped happening.

I could imagine TDF being frustrated by bug reports like that one, and I would not think it too outrageous to say "No filing bugs against us unless the human being sitting behind the keyboard has personally witnessed the bug at least once." But this is pure conjecture; I don't know if that's what's actually happened here.

LibreOffice: the next five years

Posted Jul 11, 2020 23:42 UTC (Sat) by Trelane (guest, #56877) [Link] (4 responses)

care to point at where the "stealing" argument is? Afaict, you're the first to claim such.

LibreOffice: the next five years

Posted Jul 15, 2020 7:12 UTC (Wed) by timrichardson (subscriber, #72836) [Link] (3 responses)

"The problem is compounded by companies that sell inexpensive "support" for LibreOffice, but which are not involved in its development and are not really able to provide that support. Those companies "file all their tickets up-stream and hope they are fixed for free". Companies working in that mode have no problem pricing their offerings below those of the companies doing the actual work (and thus winning much of the business that does exist). In addition, they simply call their offerings "LibreOffice", which actually looks more authentic than services from other companies, which are trying to build their own brands around LibreOffice support. "

This is basically a complaint of charging for work while someone else does it for free, taking revenue that 'belongs' to someone else, and IP abuse. So I summarised it as a complaint of stealing.

LibreOffice: the next five years

Posted Jul 15, 2020 18:51 UTC (Wed) by Trelane (guest, #56877) [Link] (1 responses)

I suspect that it's not that they're "stealing" (whatever that may mean in a Free project) but rather that they're not contributing to development. Which is the overall point at hand -- how do we fund LibreOffice development? Factually, if you don't contribute to development (either funding development or by contributing patches) then you're not contributing to development.

The second half is that by selling support below sustainable price, coupled with not contributing code or paying for coders, makes it harder for shops that charge higher prices because they hire coders.

None of this is stealing, There's no IP issue (unless they don't comply with the mpl/lgpl) there no theft, there's no "ownership" of revenue (if that were ever a thing). It _is_ however, a call to finding a plan which helps fund development, because the current approach isn't working. If you don't care, then don't worry about it. If you do care, you can look through this and figure out what you can do to help given now and maybe help with the planning.

LibreOffice: the next five years

Posted Jul 16, 2020 3:37 UTC (Thu) by timrichardson (subscriber, #72836) [Link]

Firstly, I personally care. I donate on a regular schedule, I advocate, do bug reports, and even once or twice some bug fixes. My donation level alone is a lot more than I pay for Office 365.

As for "stealing", a better word is "takers" (as used by Dries Buytaert), the concept is the same, please look past my exaggeration.

I don't really like this 'Takers' attack, because I think it misses the real problem with LibreOffice. Open source has been very successful and revolutionary, in some cases. But it has also become 'fairy dust' that has been sprinkled over other projects that won't achieve a genuinely self-sustaining project. There is an assumption that such projects prove that open source is broken. I think the signs are that Libre Office is a failed open source project, but I challenge that it means open source is broken.

I think that when you find yourself talking about "takers" being a threat to the project, you are probably looking at a project that is not a good choice for open source. This is because the 'makers' are seeking to make money from their contribution, and the 'takers' are a threat to this. Previously, you were a user. But now we have 'makers' who call others 'takers'. Where are these terms in the licence?

This problem has a solution: traditional IP protection. Some projects may have painted themselves in a corner because they used an open source licence (maybe to use the fairy dust to raise money) and now they miss the IP barriers they need to exploit their work. I think many 'open core' and 'community licence' initiatives are just ways of walking back from that.

Recently, cloud SAAS happened, and LO code-base became an attractive launching pad for SAAS entrepreneurs and investors to exploit, and now they want their return. (I don't think the move to open document standards is much of an economic opportunity, that is, I don't think there is any chance that desktop LO will make much money).

I understand people in the LO project are motivated to protect the resourcing of the project, in some cases their jobs depend on squaring this circle. But to use open source code in this way and then stigmatise people as 'takers' when they complying with the licence terms and in some terms actually bringing users to the project is not something I like.

LibreOffice: the next five years

Posted Jul 18, 2020 3:20 UTC (Sat) by giraffedata (guest, #1954) [Link]

This is basically a complaint of charging for work while someone else does it for free

Not really. There isn't even any complaint here -- just an explanation of a problem the publisher of LibreOffice has.

You can conclude that somebody's doing something wrong, because TDF shouldn't have this problem, but you can't attribute that to the LWN article or its source.

LibreOffice: the next five years

Posted Jul 9, 2020 22:44 UTC (Thu) by NYKevin (subscriber, #129325) [Link] (9 responses)

> moribund (but brand-recognized) Apache OpenOffice project.

I think it is fair, at this point, to ask why AOO still exists. All it is accomplishing is siphoning less-technical (Windows) users away from LibreOffice. I do not understand how that supports the purposes of the Apache Foundation.

Amazingly, their website (https://www.openoffice.org/why/index.html) still says things like this:

> Apache OpenOffice is the leading open-source office software suite for word processing, spreadsheets, presentations, graphics, databases and more.

As recently as May, they made a blog post (https://blogs.apache.org/OOo/entry/apache-openoffice-need...) asking for more contributors, so they're clearly still trying to make a successful project here. I just don't see why.

LibreOffice: the next five years

Posted Jul 10, 2020 2:45 UTC (Fri) by thumperward (guest, #34368) [Link]

Spite. This is literally the only reason.

Note that in 2017 (as per the link in the article) AOO was preparing for its 4.1.4 release. Three years later, it is preparing for its 4.1.8 release, with a pie-in-the-sky 4.2.0 still unscheduled. Of the (laughably titled) core development team, it appears that literally three (3) people are capable of regular builds.

If this were some random incredibly obscure Apache project donated to the org in the usual way (by being thrown over the wall) then this might be excusable. Instead, Jim Jagielski is still intimately involved. This is literally a decision by the Apache Foundation, at the root level, to endanger computer users for the sake of having a few million fewer people using GPL software. Everyone involved in this endeavour should be permanently ostracised from the software development community.

LibreOffice: the next five years

Posted Jul 10, 2020 9:10 UTC (Fri) by andrewsh (subscriber, #71043) [Link]

Why can’t they just stop pretending. It helps absolutely nobody.

LibreOffice: the next five years

Posted Jul 10, 2020 14:00 UTC (Fri) by wazoox (subscriber, #69624) [Link] (2 responses)

AFAICT, many, many Windows users still go the OpenOffice route, silly as this may sound. That's the brand they know, so that's what they're downloading, installing and using. I don't know if that represents less of more users than LibreOffice overall, but it's certainly a large group anyway.

LibreOffice: the next five years

Posted Jul 11, 2020 17:49 UTC (Sat) by NAR (subscriber, #1313) [Link]

I've seen this just in the past few weeks when a Windows user planned to install OpenOffice. Established brands have huge momentum...

LibreOffice: the next five years

Posted Oct 3, 2020 9:50 UTC (Sat) by scientes (guest, #83068) [Link]

To this type of user the products are basically identical. The only difference is the bugs that let the document take over your windows computer (both versions have these, of course).

LibreOffice: the next five years

Posted Jul 11, 2020 10:21 UTC (Sat) by chithanh (guest, #52801) [Link] (3 responses)

Ironically, this move has the potential to push users back to OpenOffice. The purpose of the name change is to make the gratis LibreOffice less palatable to corporate users. If the alternatives being a paid LibreOffice Enterprise version, or the stale and less-featured but still working OpenOffice without the embarrassing "Community Edition" branding?

Combined with the fact that corporate users will be aware that this was a move directly against them, and if that move doesn't bear fruit will expect more actions by TDF to get them to pay for LibreOffice.

LibreOffice: the next five years

Posted Jul 11, 2020 20:50 UTC (Sat) by NYKevin (subscriber, #129325) [Link] (2 responses)

I'm honestly a bit confused about which corporate users will even care here. I can think of three groups:

1. The people who want to roll their own support, with their own IT department (which may or may not be contracted from another company). They will use whatever's free and has the features they want (probably LibreOffice community edition, most of the time). The more technically inclined corps might patch out the "community edition" language from the about box to prevent their users from whining about it. But none of them are going to give any money "back to the community," because they just want an office suite that mostly works and isn't unreasonably expensive to "support" (by reimaging the machine every time it looks at the user funny).
2. The people who want to buy support at the lowest price. They will adopt whichever of LibreOffice, AOO, or various proprietary options costs the least money per seat (including the cost of either first or third-party support contracts). The "community edition" language is unlikely to matter to them; it's all about price.
3. The people who want a product that works, is widely-accepted, and is backed by a major corporation (so that it won't suddenly go bankrupt or something). These people will buy Microsoft Office, G Suite, or another proprietary competitor. LibreOffice is not on the table for them.

I'm really struggling to think of a fourth group of people, who care about the provenance and first-party support of their software, but lack the instinctual urge to buy proprietary. You don't really hear about for-profit companies subscribing to the FSF's philosophy, so the reasoning would have to be money-oriented. The company would need to be large enough to want to buy a first-party support contract, but small enough to find Microsoft etc. products too expensive. That seems like a very niche target audience to me.

LibreOffice: the next five years

Posted Jul 18, 2020 17:21 UTC (Sat) by jmclnx (guest, #72456) [Link] (1 responses)

QUOTE: I'm honestly a bit confused about which corporate users will even care here.

This is one area where they may run into a bit of trouble.

Some companies allow people to use Linux as a Work Desktop knowing they can interact with Office Files using LibreOffice. I am fairly sure if some of these Companies need to pay for an Enterprise edition, they will push their Linux users over to Windows or MACs seeing they probably have a site license for Microsoft Office.

LibreOffice: the next five years

Posted Jul 22, 2020 13:03 UTC (Wed) by film_girl (subscriber, #134009) [Link]

>Some companies allow people to use Linux as a Work Desktop knowing they can interact with Office Files using LibreOffice. I am fairly sure if some of these Companies need to pay for an Enterprise edition, they will push their Linux users over to Windows or MACs seeing they probably have a site license for Microsoft Office.

I *really* don’t think that’s the case if we’re just talking about client software. Compatibility is likely a much bigger concern than having to potentially pay for a LibreOffice license (and that assumes that the company would have the mechanisms in place to audit what software is installed on a user’s Linux desktop — if they are shelling our for an MDM solution for Linux, I can’t see there being a big argument about paying for an LO license). LO is fine for simple documents, but I’ve been burned too many times by it when using a complicated spreadsheet or specially formatted document to ever trust it for important documents that others are going to edit.

I work at a big tech company and there is no objection to desktop Linux, but there is also no IT support (there is a wiki in a git repo with instructions for getting on the corporate VPN), so yes, there might be an objection to the company paying for a license (though I could almost certainly expense it), but there would be no worry about whether the edition used said “community” or “personal” or anything else. Again, the bigger issue would be “will this mangle docs my colleagues send to me?”

Frankly, in a corporate environment, the problem LO has (which is alluded in some of the posts but not explicitly) is that collaboration is increasingly a requirement. Employees who need to be able to edit documents in real-time amongst colleagues and to leave/reply to comments, can’t use the LO client for that. And data protection standards would understandably limit ad-hoc adoption of any sort or LO Online product (regardless of brand). If a business is already paying for Office 365 or G Suite, the Linux desktop users are going to need to use those services in the browser. Collabora, by its own admission, is seeing a lot of success in its online product, and I’m sure that space will increasingly represent its future revenue.

LibreOffice: the next five years

Posted Jul 10, 2020 3:45 UTC (Fri) by pabs (subscriber, #43278) [Link] (4 responses)

I wonder how the Linux/BSD distros redistribution of LibreOffice factors into these plans.

LibreOffice: the next five years

Posted Jul 10, 2020 13:33 UTC (Fri) by giggls (subscriber, #48434) [Link]

Will probably work the Iceweasel/Firefox way.

Looks like we might get CentOffice.

LibreOffice: the next five years

Posted Jul 11, 2020 0:27 UTC (Sat) by swilmet (subscriber, #98424) [Link] (2 responses)

What is missing for Linux distros wrt LibreOffice, in my opinion:
- For a leading/bleeding edge distro like Fedora: easily pay a personal subscription to install a more stable version of LibreOffice.
- Vice-versa for a stable distro version like Debian stable: pay a subscription to install the fresh version of LibreOffice with the latest features, and emphasizing that it has better support for file formats of other office suites.
- Or, in any case, pay a subscription to choose where on the "technology wave" we want to locate ourselves when it comes to install and use software we rely on.

It's something that can be accomplished with container technologies (Flatpak, Snap, …), to install several versions in parallel (the LTS version for writing new documents, the fresh version for opening MS Office documents, for example).

However, for a personal usage of LibreOffice on Linux, I'm not aware of an easy way to install an LTS version (and pay a subscription). I.e. not an IT department wanting to install LibreOffice at scale, but just one person wanting a more stable version than what the Linux distro ships (in my case, Fedora that I install for somebody else, and the somebody else prefers a more stable version of LibreOffice).

LibreOffice: the next five years

Posted Jul 11, 2020 1:10 UTC (Sat) by pabs (subscriber, #43278) [Link] (1 responses)

Probably Flathub is the place to implement that stable version of LibreOffice for Fedora idea. ISTR they support paid apps.

The Debian maintainer of LibreOffice packaging already provides gratis backports of the latest stable releases of LibreOffice to the Debian stable release, no need for a subscription there.

LibreOffice: the next five years

Posted Jul 11, 2020 14:00 UTC (Sat) by swilmet (subscriber, #98424) [Link]

Flathub just supports donation links, not paid apps. But it would be possible to create an app store based on Flatpak, but I think it still requires a significant development and legal effort.

So let's say I'm a simple individual who wants an LTS version of LibreOffice on Linux, to install on 2 or 3 computers for me and someone else, what's the options?

- https://www.collaboraoffice.com/subscriptions/ : there is Collabora for SMBs, Supported Version up-to 99 users, but by clicking "Buy Now" we end up on a Contact Sales form, there is no way to directly choose Linux, pay with PayPal/Stripe/whatever and then install the LTS version (with Flatpak or another way). It's not streamlined for individuals/personal use.

- https://libreoffice.cib.de/ : the "Buy now" is a link to the Microsoft Store, so it's not for Linux.

LibreOffice: the next five years

Posted Jul 10, 2020 4:16 UTC (Fri) by felixfix (subscriber, #242) [Link] (3 responses)

I seldom use any "office productivity" programs, such as fancy document writers, spreadsheets, etc; but when I need them, such as for work documents, LibreOffice is all I have. I'd be glad to send them some money once in a while, but I wonder if that's at all useful. 40 programmers probably cost several million USD a year, and even 10K donors like me wouldn't go very far towards helping that.

Maybe what I need to do is remember to send them $5 or $10 every time I need to use LibreOffice to read some terrible API documentation, or when someone sends out notes written in a spreadsheet, and bill it to the employer on an expense report (which probably needs to be done as a spreadsheet....).

I better check their website for each a donate clicky.

LibreOffice: the next five years

Posted Jul 10, 2020 4:23 UTC (Fri) by felixfix (subscriber, #242) [Link]

That turned out to be relatively painless. Click the donate link, follow that to an SPI page, fill in some info, done. And it's tax-deductible in the US!

LibreOffice: the next five years

Posted Jul 10, 2020 6:34 UTC (Fri) by gfernandes (subscriber, #119910) [Link]

That's not a bad idea. Is something like the annual Wikipedia fund raisers possible here?

LibreOffice: the next five years

Posted Jul 10, 2020 12:32 UTC (Fri) by eru (subscriber, #2753) [Link]

At least Europeans can do it with a direct bank transfer, which hopefully has less overhead (just sent a small contribution).

A Microsoft Office365 subscription for personal use costs 69€ a year in Finland. I guess if all LibreOffice users who can afford it donated a similar amount per year, the project would be fine. I think I'l take that as my donation goal.

LibreOffice: the next five years

Posted Jul 10, 2020 9:54 UTC (Fri) by flussence (guest, #85566) [Link] (10 responses)

I personally don't have any problem with the “identical code under two different names, but with a price tag for corporate parasites” business model. It's worked for SQLite. It probably would've saved Pidgin a lot of Oracle-induced stress over the years too.

LibreOffice: the next five years

Posted Jul 10, 2020 10:48 UTC (Fri) by eru (subscriber, #2753) [Link] (6 responses)

They made a BIG communication blunder by tagging the non-enterprise version "Personal", and having text "intended for individual use" in the About box (https://www.theregister.com/2020/07/07/libreoffice_community_protests_at_introduction/). For me, this gave flashbacks of dealing with intentionally crippled personal editions of this and that software, way back when I used MS-DOS in the dark 1980's.

Calling the non-enterprise version just plain "LibreOffice", and noting it is only volunteer-supported without implying some kind of individual-use-only restriction would achieve much the same distinction, without raising so many hackles.

Please, LibreOffice, fix this!

LibreOffice: the next five years

Posted Jul 10, 2020 14:33 UTC (Fri) by mmeeks (subscriber, #56090) [Link] (5 responses)

> Calling the non-enterprise version just plain "LibreOffice"

ie. the status quo.

> and noting it is only volunteer-supported without implying some kind of
> individual-use-only restriction would achieve much the same distinction,
> without raising so many hackles.

the current download page (which funnels ~95%+ of our web hits) says:

"For business deployments, we strongly recommend support from certified
partners which also offer long-term support versions of LibreOffice."

Sounds like your wish already came true. The only problem with this is
that it is demonstrably utterly ineffective, and has been so for many years
despite being seen by millions of downloaders.

Having something (perhaps not 'Personal') that reminds people of their
(moral) obligation to support the software they rely on was thought to be
potentially more effective - who can say ? but the status quo doesn't work.

LibreOffice: the next five years

Posted Jul 10, 2020 21:09 UTC (Fri) by perennialmind (guest, #45817) [Link] (1 responses)

I thought the Docker/GitLab naming convention was reasonable: Community Edition vs Enterprise Edition. I think that communicates the peer vs customer distinction well enough. Although that doesn't communicate the stewardship angle.

LibreOffice: the next five years

Posted Jul 11, 2020 7:05 UTC (Sat) by edomaur (subscriber, #14520) [Link]

Yes, I'm of the same mind, in my eyes LibreOffice Community is better that LibreOffice Personal.

LibreOffice: the next five years

Posted Jul 11, 2020 2:13 UTC (Sat) by Kluge (subscriber, #2881) [Link] (2 responses)

How about simply "Unsupported Community Edition"?

LibreOffice: the next five years

Posted Jul 13, 2020 11:28 UTC (Mon) by epa (subscriber, #39769) [Link] (1 responses)

Calling one version unsupported implies that a different version is "supported". But actually being supported or not is not an aspect of the software itself, but depends on whether somebody is doing work to support it. You can download Ultimate Premium Enterprise Edition and it still won't be any more supported. You'd need to pay somebody to support it, if you want that -- and the paid support could equally well be offered for the ordinary community edition, which is still free software after all.

LibreOffice: the next five years

Posted Jul 13, 2020 20:19 UTC (Mon) by perennialmind (guest, #45817) [Link]

The kind of support that matters most runs both ways. You can support the project with code contributions, direct monetary contributions, or you can provide monetary support to those who maintain the software with the proviso that they address your particular maintenance needs. Those are all qualitatively different from paying for services that don't positively impact the LibreOffice project as a whole. A varied collection of people and organizations maintain the project: they seem to want a distinction between a relationship with a member of the first-party collective from that with an unaffiliated second-party. If you only care that they support you, that distinction won't sway you; they're hoping it'll sway civic-minded users.

If I've read the situation correctly, Enterprise vs Personal is an attempt to draw a distinction between paying patrons and welcome guests. So the real question is: does LibreOffice have proprietors? Because that metaphor only works if they can claim a degree of ownership.

LibreOffice: the next five years

Posted Jul 14, 2020 11:30 UTC (Tue) by jkingweb (subscriber, #113039) [Link] (2 responses)

> I personally don't have any problem with the “identical code under two different names, but with a price tag for corporate parasites” business model. It's worked for SQLite.

Is this a model SQLite uses, never mind had success with? What is the different name for corporate parasites? I'm not aware of one.

Taking it as a given you are correct, SQLite is not your typical open-source project. The code is tightly controlled by a single for-profit entity which has a pool of provable expertise that backs up the value proposition of paid support. The example of SQLite doesn't seem to me terribly applicable to LibreOffice's problems.

LibreOffice: the next five years

Posted Jul 17, 2020 4:59 UTC (Fri) by NYKevin (subscriber, #129325) [Link] (1 responses)

SQLite (or more accurately, Hwaci) does indeed sell "licenses" which basically say "We promise that the software is in the public domain and nobody will sue you over it." They also clearly state that a license is not required and so the license is little more than an indemnification agreement to satisfy paranoid legal departments.

They charge $6000 USD for this service. I'm not aware of any "different name," however, and it does not appear to include any sort of per-seat support contract. Those are sold separately.

LibreOffice: the next five years

Posted Jul 18, 2020 14:36 UTC (Sat) by jkingweb (subscriber, #113039) [Link]

Yes, I was specifically incredulous about the "under a different name" part (I'm well aware Hwaci sell licenses and support). One might say the distinction is trivial, but the different name for LibreOffice appears to be a source of considerable confusion according to the article itself, so I'm pretty sure the distinction is important here.

LibreOffice: the next five years

Posted Jul 10, 2020 13:35 UTC (Fri) by mmeeks (subscriber, #56090) [Link] (14 responses)

First - thanks for the write-up Jonathan ! One thing: this X month gap thing is/was some random brain-storming that didn't come from the ecosystem and should never have made it into that public deck IMHO. I don't believe the ecosystem needs that - the existing approach of a time-based release, everyone working on master and having regular feature freezes every 6 months gives time to differentiate by contributing: which drives the contribution behavior LibreOffice wants =) We really do need to remind people that it is important that people contribute - as so many volunteers do to improving LibreOffice, and that as of today many enterprises are heavy users while currently contributing nearly nothing - some (arguably) without even knowing they should =) My goal would be to differentiate the Ecosystem version from the Personal version in the marketing - rather than the software per-se. Anyhow - as always, I think you capture the debate well. Thank you !

LibreOffice: the next five years

Posted Jul 10, 2020 23:15 UTC (Fri) by penguin42 (guest, #72294) [Link] (13 responses)

IMHO differentiation can be the thing that kills it.
Most businesses are now using online/cloud based systems that they pay almost nothing for (e.g. G apps) when they're already using them for some other reason. They expect online to be the default - splitting it off as a separate product/name kills the base.

Everyone hates using the online systems for big documents because the web interfaces suck - it's almost like we're back in the early 90's with the split between DTP and word processors, with the online systems at the level of word processors. Those DTP packages died as the word processors got better; the same things is happening to the desktop apps.
I think having the desktop package be a {local instance + enhanced client} would work - for the people doing local editing the structure under-the-hood might not be needed, but then having something that gave you all the power of a desktop client while allowing you to collaboratively edit with people using the web interface or a mobile would be great.
IT orgs in some big companies don't want to do anything themselves; so having someone else host they're happy to pay for; but then there is the opportunity for hosters to integrate and charge more for things like integration with document management systems, or sell supported systems you can run on a cloud of your choice (be it in house or public).

LibreOffice: the next five years

Posted Jul 11, 2020 9:52 UTC (Sat) by mmeeks (subscriber, #56090) [Link] (2 responses)

> I think ... having something that gave you all the power of a desktop client while allowing
> you to collaboratively edit with people using the web interface or a mobile would be great.

Have you tried Collabora Online - (Collabora did 95%+ of the commits to LibreOffice Online ;-)
and it basically targets exactly what you describe. The feature/function comparison vs. Google Docs
and MSOffice Online is really rather impressive =)

LibreOffice: the next five years

Posted Jul 12, 2020 15:50 UTC (Sun) by penguin42 (guest, #72294) [Link] (1 responses)

(shells out for a few hours and tries it)
OK, so C.O does work similar to G.Docs/G.Office - but reread my comment; my point was that doesn't get you to a point where
people will keep L.O. alive - companies are just using G.Office. Either L.O.Online needs to be as feature complete as the desktop edition (and usable easily on a single desktop) or the L.O.Desktop app needs to seemly integrate with the Online version (not just save/load); what you have now is differentiation that's killing the Deskop development but leaving you with an online version that's not as good as the desktop - there needs to be less separation between effectively two independent projects.

[I used the Univention CODE owncloud VM images; it was more than a bit of a fight; I almost gave up a few times;:
a) the required email/registration key
b) the univention users and owncloud users seem independent rather than integrated; so I added users to the wrong one
c) As a test VM it still wants FQDN - most of it works OK if you give it just a hostname in your /etc/hosts, but it borks in some places; it got to trying to bring up CODE itself when editing a odt and hung until I added the FQDN to my /etc/hosts ; I only diagnosed that from firefox console)
d) The docs coule do with suggesting RAM/CPU count for the VM image

LibreOffice: the next five years

Posted Jul 13, 2020 10:26 UTC (Mon) by mmeeks (subscriber, #56090) [Link]

> OK, so C.O does work similar to G.Docs/G.Office
Thanks for trying it =)
> Either L.O.Online needs to be as feature complete as the desktop edition (and usable easily on a single desktop)
It shares all the code under the hood, so lots of feature work on one is complementary to the other.
> what you have now is differentiation that's killing the Deskop development
What harms desktop development is not the lifeboat - but the large economic holes below the waterline I think ;-)

LibreOffice: the next five years

Posted Jul 12, 2020 7:14 UTC (Sun) by epa (subscriber, #39769) [Link] (9 responses)

DTP packages died? So print newspapers are using Microsoft Word for layout? I doubt it.

LibreOffice: the next five years

Posted Jul 12, 2020 12:17 UTC (Sun) by penguin42 (guest, #72294) [Link] (8 responses)

You missed the *D* - Desktop - back in the early 90's it used to be that people used 'word processor' stuff for business letters, etc and small reports; but if you wanted to write a large document you broke out something more powerful; I'm sure the newspaper guys have something specialised - but everyone else is fine with something like Libreoffice or MS Word even for a large complex document.

LibreOffice: the next five years

Posted Jul 12, 2020 18:12 UTC (Sun) by k8to (guest, #15413) [Link] (7 responses)

People will now do newsletters in word, or similar.

Doing a multi-column book in word remains insane and there are certainly desktop publishing packages that are still in use for large complex documents.

LibreOffice: the next five years

Posted Jul 12, 2020 19:34 UTC (Sun) by Wol (subscriber, #4433) [Link] (6 responses)

Don't forget that many authors used (and found it VERY good) WordPerfect to write their books. That's WP4 - the version that I believe was ported to DOS from DG (I don't know the history back that far - I started with 5.1 which, imho, STILL beats Word).

For example, much of the early Disk World books were written with WP4. I don't know what TP used later on ...

Cheers,
Wol

LibreOffice: the next five years

Posted Jul 12, 2020 20:16 UTC (Sun) by mpr22 (subscriber, #60784) [Link] (2 responses)

Writing a book (what kind of book? a prose novel? a technical manual with minimal or no illustrations? a figure-heavy technical manual?) is one thing; taking the text and preparing it in camera-ready form for publication-grade printing is another thing.

LibreOffice: the next five years

Posted Jul 12, 2020 21:54 UTC (Sun) by Wol (subscriber, #4433) [Link] (1 responses)

I was thinking of a book with many chapters, heavy use of footnotes, a long novel, etc etc. I think a fair few people even managed maths-heavy texts in it! You know the stuff - things TeX has a reputation for ...

WordPerfect could EASILY handle a book length novel, Word had a reputation of getting its knickers in a complete twist even before you'd finished the first chapter. Current experience with today's Word makes me wonder if it's actually improved much ...

Cheers,
Wol

LibreOffice: the next five years

Posted Jul 13, 2020 13:50 UTC (Mon) by brian_lindholm (subscriber, #42351) [Link]

Aye. I wrote my Ph.D. dissertation in WordPerfect 6.0c for DOS. Multiple chapters spanning 120+ pages, with a crap-load of equations, figures, footnotes, and cross-references. It worked pretty well.

Using Word for large documents was rougher. I remember writing on another heavily technical document at work using Word 2000, and it crashed regularly once the page count got above about 20. Word 2010 was significantly more stable (though the ribbon-based UI made it harder to find advanced features). I haven't exercised the latest Word enough to form an opinion. [And Word's auto-formatting?!? Gah! Don't get me started.]

LibreOffice: the next five years

Posted Jul 13, 2020 6:10 UTC (Mon) by dtardon (subscriber, #53317) [Link]

That books were written using WordPerfect doesn't mean they were printed using WordPerfect... I bet the first thing a print shop would do with such a book was to import it into QuarkXpress or PageMaker.

LibreOffice: the next five years

Posted Jul 13, 2020 15:58 UTC (Mon) by NYKevin (subscriber, #129325) [Link]

My perception is that nowadays, a lot of writers use specialized proprietary (usually) software such as Scrivener. In fact, when searching for FOSS equivalents, I had to search for "Scrivener alternatives" because there doesn't seem to be a standardized name for this particular niche. It's not really DTP, because it's not easily capable of the advanced layout functionality you would find in a tool like InDesign.

(Basic idea: It lets you break a big document into a hierarchical tree of smaller documents. You can then temporarily "glue" them back together in an ad-hoc fashion as needed for publishing or editing. It also offers a number of other specialized "writer-oriented" tools such as a random name generator, per-session word count targets, etc.)

LibreOffice: the next five years

Posted Jul 14, 2020 7:11 UTC (Tue) by eru (subscriber, #2753) [Link]

George R.R. Martin (the Game of Thrones author) is said to still write with WordStar on MS-DOS. He has a separate computer for this, another for using other software and the net (https://time.com/99432/george-rr-martin-game-of-thrones-computer/).

LibreOffice: the next five years

Posted Jul 10, 2020 13:38 UTC (Fri) by njamba (guest, #140094) [Link] (1 responses)

This is the "software dialectic": how to find a way to ensure economic sustainability without sacrificing the ideals of free/open-source software. Michael Meeks' post on the state of the Libreoffice ecosystem makes for some sobering reading (https://lwn.net/Articles/825602/)

LibreOffice: the next five years

Posted Jul 11, 2020 10:08 UTC (Sat) by chithanh (guest, #52801) [Link]

Eh. Quoting from that post,

* These naming changes seem to suggest some people need to pay ? I won't be able to install it in XYZ enterprise if called Personal

This is in large part a feature. Clearly people can use it for free, just as before - but they would occasionally see the splash and thing: "I wonder why it says Personal" (or another tag) there.

So the purpose of this name change is market segmentation, and more precisely to embarrass certain users in order to push them towards the enterprise version. The side effect of this is that it will create confusion (is LibreOffice now "open core"? No, but the branding makes it appear as if it were) and risk further devaluation of the brand. Joel Spolsky wrote about this more than 15 years ago:

In the world of software, you can just make a version of your product called “Professional” and another version called “Home” with some inconsequential differences, and hope that the corporate purchasers (again, the people who are not spending their own money) will be too embarassed at the thought of using “Windows XP Home Edition” at work and they’ll buy the Pro edition. Home Edition at work? Somehow that feels like coming to work in your pyjamas! Ick!
[...]
So, while segmenting can be a useful tool to “capture consumer surplus,” it can have significant negative implications for the long term image of your product.

Microsoft came to their senses and removed the "Home Edition" label from the Windows XP loading screen with SP1. It will be interesting to see if TDF is going to repeat history here.

LibreOffice: the next five years

Posted Jul 10, 2020 15:19 UTC (Fri) by karim (subscriber, #114) [Link]

I have the utmost respect for the work done by LibreOffice contributors. I depend on it every single day and prefer it to anything else. But I'll admit to this: I've never contributed a penny to any of its development. Worse: I suspect (possibly wrongly) if I did it would make no difference because the bulk of the money comes from corporate support contracts. Worst: everyone, including companies/orgs, expects software to be either free or subscription-based -- that's where the models have landed for better or worse.

I'll claim zero expertise when it comes to open source monetization, but for having been involved in open source for 20+ years one of the things I've concluded is that OSS works best as a by-product, not THE product. Large companies provide swaths of open source code to commoditize functionality or ensure *their* implementation becomes *the* standard. Others contribute to open source projects as marketing. Others, yet, are dedicated to maintain projects as part of specific roles they have in certain organizations that derive value of this maintainership. But making money on the OSS itself, that's tough. The often-cited case of Redhat is misleading in my opinion. What companies/organizations are buying there is a neck to choke for *all* parts of the Linux distro they are using. It so happened that Redhat offered the proper corporate dress-code to be accepted as a valid front, and they've been rewarded for that. There is no example to be emulated here by any single OSS project.

What's unclear, though, is why there isn't some sort of model where organizations such as the "Documentation Foundation" don't have a way to officially recognize the contributions of certain companies as being of exceptional value to the ecosystem. Why isn't there a way for a specific vendor to be recognized officially as a "Platinum" contributor to LibreOffice? Maybe this can then be used by said vendor to differentiate their support offering from fly-by-night cheap support provider who makes zero contributions?

Again, no expert here and no particular agenda being pushed, just throwing some ideas here in case they may be useful to others.

LibreOffice: the next five years

Posted Jul 10, 2020 15:23 UTC (Fri) by ldearquer (guest, #137451) [Link]

Although I agree with the bottom line, I don't feel too comfortable with some of the arguments in the presentation.

In particular, this:
"[In the last 2 years] Around 68% of contributions to the source code are from ecosystem companies (TDF paid developers are just 4%, volunteers are 28%)"
and...
"Without contributions from the ecosystem, we would have issues in keeping up with user expectations"

Basing a branding/licensing decision on this kind of statements leaves me a bad taste, and maybe that's why I always preferred strong copyleft licenses.

I don't know enough of the Libreoffice history, and surely this argument is legitimate here (i.e. it extends for much longer time), but IMHO there are certain basic things that shouldn't change in a project based on who has contributed the most in the last 2 years.

LibreOffice: the next five years

Posted Jul 11, 2020 0:15 UTC (Sat) by grantma (subscriber, #5225) [Link]

Reminds me of PostgresQL. It looks like they are trying to establish a version of the PostgresQL ecosystem. Don't fret, we will have enough if this is what is going on.

LibreOffice: the next five years

Posted Jul 11, 2020 1:05 UTC (Sat) by pabs (subscriber, #43278) [Link]

I wonder if a better solution here would be something like "LibreOffice, provided by Debian". Then the packaging or UI could list what "provided by Debian" means ("volunteer based security updates and bug fixes, contributed upstream") and doesn't mean ("contributes $$ to Collabora"). There could also be donation banner/links in the default UI.

LibreOffice: the next five years

Posted Jul 12, 2020 9:18 UTC (Sun) by mfuzzey (subscriber, #57966) [Link]

I think there are two meaning of "support" that are getting conflated here.

There is "supporting the end users" by fixing their problems so that they can do their work using the software more effectively.
And there is "supporting development" to enable future updates and improvements.

Doing the second indeed requires providing money or resources to those that do most of the development the first not necessarily.

To give the obligatory car analogy. I need to service my car. I could:

1) Do it myself (as I used to do when I was younger with less money, more time and when cars were simpler)
2) Go to a local "generic" auto service shop
3) Go to an official dealership service center.

Now for *complicated* things 3 is very likely the best technical solution. Although their local mechanics may not be any better than the generic service shop they do have access to the full backing organisation of the maker.
But, for simple things like a standard service 2 is probably good enough and likely has other advantages (like being closer, maybe friendlier etc).

It's similar for end user software support. Yes those in the best position to fix a complicated bug are the original developers but most end user support requests are user error or misunderstanding. A small local support company can provide better training, will have people that natively speak the end user's language and are likely more reactive than an organisation on another continent that employs most of the developers.

So, when an organisation is looking at how best to support its users (in the short term with the current software) deciding to do it themselves or to contract with a local support company can make perfect sense, even though it does nothing for the long term sustainability of the development.

I think there *are* situations where the two can be more blurred. Say I don't have a standard car but some custom one off, going to the original builder for service probably makes a lot more sense. Or if I have some special server farm with custom hardware, kernel and everything.

Also end user support is like insurance, you are paying money to mitigate a future risk. That may, or may not, be worth it. It depends on the risk when something goes wrong. On servers where the ability of the business to function may depend on them being up (ecommerce sites, production control systems etc) having support is a no brainer but for an office suite not so much. Sure if all the office suite software "disappeared" the company could be in great difficulty but, realistically, that is not going to happen. We're more talking about bugs in a few special cases affecting a small number of users.

How to fund future development is indeed a legitimate question but one that I think needs to separated from "how should I support the users".
So going to the big company or government agency and saying "look you've got thousands of people using this software, it would be good if you could contribute to future development so it will be better for you tomorrow" seems a very reasonable thing. But saying "you should take a support contract with *us*", not so much because there may very well be other options that make more sense to them.

Bug Bounty perhaps

Posted Jul 12, 2020 18:20 UTC (Sun) by Richard_J_Neill (subscriber, #23093) [Link] (8 responses)

I love LO - and have used it for years - and filed bugs where I can. However, there are a number of small bugs in LO that drive me a bit mad, and which I would gladly pay to get fixed - or at least make a contribution of a few £100. However, there's no bug bounty or any way to commit £ to a bug. For example (and I think I've filed all these as bugs over the years).

* When editing slides, you have to constantly switch between notes view (to type your speech) and slides view (to change the slides). Why can't the notes text-box be pinned below the slide (especially when there's plenty of unused room in the GUI).

* When dragging to auto-fill in the spreadsheet, it correctly continues the sequence "n, n+1" to "n+2, n+3...". However, if the sequence begins with just "n", then it should continue as "n,n,n,n". (I recognise this is a matter of taste).

* Scrolling in the spreadsheet has 2 "gears": dead slow, and way-way-way too fast. It's too easy to overshoot off by miles.

* Can we have a global preference to never auto-"interpret" data entry in a spreadsheet? If I type "1/2", I mean the string "1/2", not 0.5 nor 1st February, nor 2nd January.

* Autocompletion in a spreadsheet column does a "top of column down" scan, not a "nearest the mouse scan". So if I have for example "Application, Apple, Apple, Apple, Apple" and then type "A", it always completes to the rare initial word, not the most common or nearest.

* Saving a presentation "with fonts" can make it balloon to 50MB - and the zipfile includes a lot of fonts that shouldn't even be there.

Bug Bounty perhaps

Posted Jul 13, 2020 0:27 UTC (Mon) by NYKevin (subscriber, #129325) [Link] (4 responses)

> * Can we have a global preference to never auto-"interpret" data entry in a spreadsheet? If I type "1/2", I mean the string "1/2", not 0.5 nor 1st February, nor 2nd January.

Suppose you open a spreadsheet authored by someone who has this preference switched off. Their spreadsheet has "1/2" and they really do want it interpreted in (whatever way Calc happens to interpret that particular string). It would be problematic if your instance of Calc failed to recognize this desired interpretation and displayed "1/2" instead of (whatever the author intended).

So the setting becomes per-file instead of global (to avoid the above problem). But, on the one hand, that's a lot less useful to people like you, since you still have to remember to switch it on every time you make a new file (or, if there's a global preference that changes the default value for new files, it would still have the "wrong" value whenever you're editing a spreadsheet created by someone else). On the other hand, there is already a per-cell setting which does (basically) the same thing. IIRC you can enable such settings for all cells in a worksheet at once. So this would not enable you to do anything you couldn't do already with a few extra clicks. It would also increase the interface complexity of the system, because the two settings would need to interact with each other in some way (one takes precedence, or the per-file preference changes the default value of the per-cell preference).

TL;DR: This feature request only makes sense if you are the only person in the universe who creates spreadsheets.

Bug Bounty perhaps

Posted Jul 13, 2020 1:24 UTC (Mon) by Richard_J_Neill (subscriber, #23093) [Link] (3 responses)

That's a good point. It depends when the spreadsheet interprets data. Which of these is correct?

* Interpretation A, spreadsheet stores string data exactly as typed in.
1. user enters "1/2"
2. spreadsheet stores it as "1/2". This is what is saved in the file format.
3. whenever the spreadsheet is displaying data, it formats it magically as "1/2", 0.5, or "1/Feb"
[default formatting can be overridden by cell-formatting]

* Interpretation B, spreadsheet saves typed-data, making a one-time inference.
1. user enters "1/2"
2. spreadsheet interprets it, and stores it as "1/2", 0.5, or 2020-02-01::date. This is saved in the file format.
3. whenever the spreadsheet is opened, it displays the underlying data literally.

I assumed it must be (B), and that the data-type-inference is done at the point where I actually enter the data. Maybe it isn't?

In which case, A would mean that if I (as a UK user) enter "1/2" and save the spreadsheet and email it to a colleague in the USA, then
the same spreadsheet will show me "1st Feb" and my colleague "2nd Jan", because the magic-formatting is invoked as a display-time, not an entry-time
function.

Anyway, thanks very much for clarifying; if it is as you say, I agree.

Bug Bounty perhaps

Posted Jul 13, 2020 16:09 UTC (Mon) by NYKevin (subscriber, #129325) [Link] (2 responses)

I don't work on Calc, so I could be mistaken. But I can't believe that interpretation B could be correct, because if you click on the cell to edit it, it always manages to reproduce the original text you entered, so that text has to live somewhere...

Bug Bounty perhaps

Posted Jul 13, 2020 18:33 UTC (Mon) by Wol (subscriber, #4433) [Link] (1 responses)

Why?

Excel certainly seems to work in mode B. And if I enter 7/7 as a date in LO, it's quite happy to let me use that in a formula and add 30 days. Or I change the format to number and it becomes a Julian offset.

All LO needs to do for that is attach the *implicit* format to the cell, and it can convert between external and internal format. I know I had the experience (in Excel) of entering the 7th of July ( as 7/7 ) and being surprised when it was displayed as £1 - the cell was defined as currency format :-)

Cheers,
Wol

Bug Bounty perhaps

Posted Jul 14, 2020 6:51 UTC (Tue) by NYKevin (subscriber, #129325) [Link]

> Why?

I refer the honorable commenter to the answer I gave some moments ago:

> because if you click on the cell to edit it, it always manages to reproduce the original text you entered, so that text has to live somewhere...

Mode B was described to me as not preserving the original text, so I assumed that it was incompatible with my observations. If you are going to change the definition of mode B now, I cannot guarantee that my comment will continue to make sense.

> All LO needs to do for that is attach the *implicit* format to the cell, and it can convert between external and internal format.

I don't understand what "implicit format" means in this context. My mental model is this:

- Every cell has a format (enum) and a contents (string). (Cells probably have other data attached to them, such as formatting, but that's not important right now.)
- The default value of contents is the empty string. This is not saved for most cells (because most cells remain empty).
- The default value of format is "general," which means "guess what I mean based on what I typed." (I'm not sure if Calc actually calls it "general.") This is also not saved for most cells (because most cells remain "general").
- Changing the value of contents does not cause format to change from "general" to something else (e.g. try right-clicking on a non-empty cell and see if its format has changed...). It may cause the cell to be rendered differently (guessed-numbers are right-aligned, guessed-dates are formatted based on locale, etc.), but that's a display issue, not a data issue.
- If we change the behavior of "general" in any way, existing files will break, because the same contents will be rendered differently.
- If we change the default value of format (e.g. to "text"), existing files will have the "wrong" default value applied to them, which may or may not cause problems.

If my mental model is wrong, I welcome corrections.

Bug Bounty perhaps

Posted Jul 13, 2020 6:35 UTC (Mon) by dtardon (subscriber, #53317) [Link] (1 responses)

> * Can we have a global preference to never auto-"interpret" data entry in a spreadsheet? If I type "1/2", I mean the string "1/2", not 0.5 nor 1st February, nor 2nd January.

One can always write " '1/2 " (note the leading apostrophe) when a literal string is needed. That works no matter what the cell format is. (But it is better to set format to Text for cells/rows/columns that only contain textual data. That way one doesn't have to think about this when inserting data.)

Bug Bounty perhaps

Posted Jul 13, 2020 18:35 UTC (Mon) by Wol (subscriber, #4433) [Link]

That won't always work in Calc.

Unfortunately, I know from interactions on the dev list, that there are quite a few "gotchas" in cell data entry, and the way to fix them is to - wait for it - COMPLETELY rewrite the module that manages this :-(

Technical debt, unfortunately :-(

Cheers,
Wol

Bug Bounty perhaps

Posted Jul 16, 2020 13:33 UTC (Thu) by bfields (subscriber, #19510) [Link]

"However, there are a number of small bugs in LO that drive me a bit mad, and which I would gladly pay to get fixed - or at least make a contribution of a few £100. However, there's no bug bounty or any way to commit £ to a bug."

Out of curiosity, are there projects like LibreOffice that have been able to raise any significant funds that way?

I'm afraid to most end users a few £100 probably sounds like a lot to spend to get a bug fixed, and yet you'd have to have a lot of people doing that every year to actually get to an income stream that would reliably fund a developer's salary.

LibreOffice: the next five years

Posted Jul 22, 2020 13:07 UTC (Wed) by robbe (guest, #16131) [Link]

> The problem is compounded by companies that sell inexpensive "support" for
> LibreOffice, but which are not involved in its development and are not really able
> to provide that support.

You’re describing a problem that vendors of cars, washing machines, and proprietary software all have as well. The main difference with free software being, that the base price of these products is not zero.

Most of these vendors have some kind of partner program, authorised dealership, or whatever. But none of them can fully control who sells or supports their products (not for lack of trying, mind).

It seems that such a partner program exists already for LibreOffice (https://www.libreoffice.org/get-help/professional-support/). Maybe put more love into it? For example, just clicking on "Get Help" does not steer me there, its not mentioned on the page. I’d also appreciate a way to search for someone in the vicinity, or at least speaking my mother tongue.

> […] nobody thinks in terms of buying support for any office suite:

Sell something called „support and subscription“, then? People *are* accustomed to paying subscription fees for proprietary software, including MS Office.

LibreOffice: the next five years

Posted Jul 23, 2020 12:39 UTC (Thu) by ydebilloez (guest, #140333) [Link]

Changing the name of LibreOffice into LibreOffice Personal and LibreOffice Enterprise is a bad idea. Why do we need to demote LibreOffice into telling it is only for personal use? It seems room needs to be created to market Enterprise versions. LibreOffice is going the path a lot of open source projects are going. e.g. Odoo, SugarCRM, ... at a certain moment, the free version will be impossible to find. Please keep LibreOffice as it is or even promote it. I suggest the following: LibreOffice or LibreOffice professional for the free version LibreOffice Enterprise, LibreOffice online, ... etc for the revenue generating version.

LibreOffice: the next five years

Posted Jul 24, 2020 11:25 UTC (Fri) by Klavs (guest, #10563) [Link]

Why not do the same thing browsers has been doing for a LONG time.. with the SSL CA list.. choose who gets shown in a "support company" list.. and then it can clearly show who users of LO can get support for.
and IMHO sort them by how the LO perceive their value of contributions..
so basicly - do some sort of stats based on people known to work with one of the "support delivering" companies - and sort them based on whatever metrics LO project perceives relevant (top ones will get most calls from customers) - and possibly charge a comission fee - to support LO hiring OWN developers (firefox f.ex. takes many 100k - for including a CA in the browser and also f.ex. for including a search engine in its default list)..