Hello,
Last summer, there was a discussion on this list about "Strategy and serendipity" to which I contributed the following remarks:
Maybe the role of the FSFE is to go beyond advocacy and help knock some stubborn heads together, to eliminate people's parochial and needlessly competitive attitudes, to actually persuade people to commit to realising an actual vision in a genuinely serious way, perhaps to secure resources to allow this to happen. Urging people to use Free Software products that don't exist just won't do the job.
Recently, in the "suggestions/request for fsfe" discussion, it was pointed out that people want to use Free Software solutions but have no clear course of action they can take. And more recently still, I encountered another kind of situation that perhaps highlights the deficiencies of merely advocating for something without anything being organised to make that advocacy meaningful.
As luck would have it, it has something to do with arguably one of the more successful and engaging campaigns that the FSFE has run: the PDFreaders campaign [1] requesting that public institutions cease advertising proprietary PDF-reading software (particularly that of a single vendor), emphasise the standardised nature of the PDF format, and acknowledge the existence of Free Software solutions.
(One can imagine taking elements of such a campaign in other directions. For instance, public broadcasters like the BBC systematically promote proprietary social networking platforms whose harmful social effects go far beyond those caused by the promotion of proprietary PDF reader applications. Just as one might wonder whether Adobe had somehow procured favours from public institutions, so might we wonder what bargains have been made between taxpayer-funded media organisations and multi-billion-dollar corporations.)
But it is the aspect of promoting Free Software solutions in a campaign that becomes troublesome, particularly if those solutions do not manage to address end-user needs. It was precisely this problem that I encountered recently. Having been sent a PDF form that needed completing, I rediscovered the apparent lack of support in Free Software applications for performing this task.
Now, I may have mentioned this before in another context, that being a discussion about an alternative microkernel-based foundation for GNU Hurd, which I think most people would agree is quite a different area from PDF application software. But ignoring the rather out-of-place nature of the encouragement from an influential figure of the Free Software movement [2], a need for improved PDF document handling in Free Software is clearly recognised.
Yet the situation is still confused and rather unsatisfactory:
https://askubuntu.com/questions/29230/is-there-software-that-can-fill-pdf-fo...
While there may have been other factors involved in my own recent experiences, such as whether such documents would even be communicated securely, I ended up with the feeling that it would not be possible to perform an important task using Free Software. No amount of advocacy would remedy this situation.
I imagine that some people would rather let "the market" figure out how Free Software products can be improved, that people motivated by a "business case" would see to it that such improvements are made. Unfortunately, the accompanying "business model" usually ends up taking the software proprietary, pursuing dubious monetisation strategies, or looking after niche groups of customers.
What really needs to happen is that a coherent vision be articulated and then all necessary action pursued to realise it. If people are supposed to be able to use Free Software for their needs - interacting with PDF documents, in this particular case - then organisations must go beyond advocacy and actually facilitate the development of the features that are part of that vision. That may well involve the financing or sponsorship of developers, rather than "inspiring volunteers" or whatever the mantra - ostensibly empowering creativity, actually perpetuating exploitation - tends to be.
So, while the PDFreaders campaign was helpful in raising awareness of Free Software and competition issues, as well as having useful side-effects in terms of liberating some PDF handling software, it also needed to be accompanied by initiatives to sustain Free PDF software development. Without such accompanying initiatives, we find ourselves confronting the same situations over and over again, always on the defensive and never really having something compelling to offer to those willing to try something new (or, for that matter, those of us who have supported Free Software all along).
Well, that was probably far too many words stating the obvious, but I suppose the message will eventually sink in.
Paul
[1] https://fsfe.org/campaigns/pdfreaders/pdfreaders.en.html
[2] https://lists.gnu.org/archive/html/l4-hurd/2018-05/msg00001.html
I like your line of thinking.
The Reuse software is a great example I think, rather than starting a full-blown European GNU project.
I've heard the issues with PDF documents before in the Netherlands. Perhaps it makes sense to pinpiont the software requirements, and work to development a generic extension that can be included in one or more pdf editors?
Nico
On Sunday 9. February 2020 19.57.50 Nico Rikken wrote:
I like your line of thinking.
Nice to hear from you again, Nico! :-)
Well, I think we cannot just make assumptions about things magically happening all by themselves. Instead, we have to consider the economic models involved, and that requires us to consider the bigger picture.
The Reuse software is a great example I think, rather than starting a full-blown European GNU project.
Reuse [1] only really quality-assures existing Free Software licensing, however. Finding myself back in an academic environment, with licensing applied unclearly to software or even an absence of proper licensing in some cases, I do see the need for things like Reuse. But its role in getting new Free Software written is going to be somewhat less directly significant compared to other initiatives, in my opinion.
Now, the GNU project [2] is actually relevant here, since it aims to offer a complete system. But interactions with the wider world mean that it might not address every possible user need. Thus, amongst the guidance about helping the GNU project and Free Software in general [3], we find the familiar FSF high- priority project listing [4].
But PDF software improvements were taken off that list in 2011 [5] (whose link to GNU PDF now leads to non-free software), with it then having been suggested to reinstate them in 2016 [6]. The response is arguably disappointing:
"We're still debating whether to make the addition or not, as PDF is in long term decline -- not nearly as steep as Flash, but still becoming ever less central."
I guess it depends on perspectives. I look at quite a few datasheets: all of them are PDF documents. My bank produces PDF documents for invoices, receipts and other records. The suggestion to use PDF.js is a good one, which is presumably why it is bundled with things like Firefox. I imagine that a lot of people get a lot of benefit from it. So much for a format in decline, especially since as a kind of archive format it will be with us for a long time to come.
I've heard the issues with PDF documents before in the Netherlands. Perhaps it makes sense to pinpiont the software requirements, and work to development a generic extension that can be included in one or more pdf editors?
It would certainly be worthwhile. What we apparently see is that familiar phenomenon where "mission accomplished" is declared but where the mission is open-ended; where "we got this" is announced but years later the software is abandoned (in other cases) or not covering the needs of the users (in this case).
Given that the problem area is likely to be one of the proprietary forms technologies from Adobe [7], only some of which have been pushed into ISO 32000 (which being an ISO standard isn't genuinely open, anyway), it might be unfair to expect Free Software support to track these technologies, particularly completely proprietary ones [8], and still maintain a usable level of support. Indeed, it is creditable that support does exist and appears to be maintained [9].
I can see a few areas of work, however...
1. Advocacy of, and insistence on, genuinely open standards that meet the needs of users. People may need to exchange documents (including forms), but the technologies involved have to be genuinely open and accessible. If insisting on a properly standardised version of PDF means that certain necessary features (such as forms) are missing, then other means of providing such features need to be advocated and provided.
2. Monitoring of organisations - particularly public agencies - to ensure that people stick to the rules. It is too easy for institutions to procure proprietary software (various Adobe "suites" in this case, perhaps) and to impose proprietary technologies on random people. There should be mechanisms to prevent and correct non-compliance, and such policing should not have to fall on random volunteers (as it did with the PDF Readers campaign).
3. Assistance for developers implementing necessary and advocated functionality. Too often, such work is framed as something for volunteers to do, but such work is also fairly thankless and tedious, with other people seemingly content to wait for the hard work to be done and then to use it for their own personal gain. I would argue that people who are intellectually curious and who engage themselves in such work without asking for reward are actually being exploited. Is the Free Software movement comfortable with perpetuating the trends of exploitation that now pervade wider society?
4. Cultivation of expertise and knowledge sharing amongst developers. Another phenomenon is where people are willing to undertake development tasks, but then they are more or less driven away by those with the expertise and knowledge. Reasons for this include plain old-fashioned selfishness (keeping the opportunities within a clique), but it can also include a lack of time due to other commitments (they already do such things at work, or that they are happy keeping their Free Software activities at "hobby" level), or even a lack of enthusiasm ("I'm done with this, so why should anyone else be interested in it now?"). I've seen all of these.
Making (3) happen is seemingly straightforward: just pay someone! But it has complications all by itself, as we see with all the different schemes that are concocted to find money and then to give it out fairly to people who supposedly did the work. But I also feel that (4) has a complicating influence as well. If the person who could do the work doesn't want to (or cannot), and yet cannot empower someone else so that they may do it, either, then the difficulties suddenly become much more severe.
Then again, this is a discussion list and I don't have all the answers ready, so maybe other people can bring their insights to bear on these problems.
Paul
[4] https://www.fsf.org/campaigns/priority-projects/
[5] https://www.fsf.org/blogs/community/gnu-pdf-project-leaves-high-priority-pro...
[6] https://www.fsf.org/blogs/community/a-preliminary-analysis-of-high-priority-...
[7] https://en.wikipedia.org/wiki/PDF#Interactive_elements
[8] https://en.wikipedia.org/wiki/XFA
[9] https://gitlab.freedesktop.org/poppler/poppler/issues?scope=all&utf8=%E2...
By putting the Reuse project and the GNU project against each other I tried to favor a small targeted project compared to a large, seemingly all-encompassing effort. But as you showed me, even GNU has its points of focus.
With regards to funding, there might be other organizations willing to sponsor an effort, as long as the goal is clear. So that is where to begin.
In regards to the PDF issue, it makes sense to first get an idea of the scope of the problem. Like I mentioned, I heard some reports about this issue, but am not aware of the current problems on this topic. Collecting theses reports can help to scope the issue: what organizations are not using open standards, what proprietary extension are we talking about, and what use-cases does it relate to? Then we could move forward by requesting the organizations to adopt open standards, by suing them into compliance, or by developing software that provides a workaround.
The pdfreaders.org website currently lacks information on typically used extensions. Also there is no way to file a complaint on a specific case that could help paint a better picture of the current situation.
Kind regards, Nico
On Tuesday 11. February 2020 07.46.27 Nico Rikken wrote:
By putting the Reuse project and the GNU project against each other I tried to favor a small targeted project compared to a large, seemingly all-encompassing effort. But as you showed me, even GNU has its points of focus.
The aim of something like GNU now should be a computing environment that allows people to conduct their lives, to interact with other individuals and organisations, and to focus on making their lives better, whether it is to help them organise their lives, pursue their interests, be creative, or simply to be educated and entertained.
Of course, to counter those who claim that their phones offer such an environment, it should be noted that such an environment should naturally be Free Software from top to bottom and uphold the user's need for things like privacy and safety. Otherwise, we wouldn't be invoking the GNU name in talking about it.
With regards to funding, there might be other organizations willing to sponsor an effort, as long as the goal is clear. So that is where to begin.
This is one area where the FSFE should be central, in my opinion. It should be a venue where things get done by people being able to bring different skills and resources to bear on problems that need solving. Where else would sponsors and those with financial resources go to get such things done otherwise? And where else would developers look for meaningful areas of endeavour that advances society in some way?
I know we live in an age where entrepreneurism is encouraged and people are meant to have great ideas, seek capital through "the market", and so on, but we all know how that usually ends up. The skill-sets for success as an entrepreneur and for producing decent, lasting, non-exploitative systems can be rather different, and we shouldn't expect people to master both realms. Otherwise, we really are minimising the chances of anything constructive getting done.
In regards to the PDF issue, it makes sense to first get an idea of the scope of the problem. Like I mentioned, I heard some reports about this issue, but am not aware of the current problems on this topic. Collecting theses reports can help to scope the issue: what organizations are not using open standards, what proprietary extension are we talking about, and what use-cases does it relate to? Then we could move forward by requesting the organizations to adopt open standards, by suing them into compliance, or by developing software that provides a workaround.
I agree. For all I know, my experience was based on a misunderstanding (perhaps there wasn't an actual form involved), or involved me not having the right software (despite trying different versions of different tools), or involved me not choosing the right settings (in the frustratingly opaque and poorly-designed interfaces, thinking specifically of anything that seems to be developed for GNOME these days).
We need to understand the capabilities of Free Software today, what the missing capabilities are, how those capabilities are used, what people's experiences are with the technology, and whether (and in which ways) Free Software fell short. We need to assess which projects would form the basis of supporting what is missing, whether they are capable of adding support, and whether they are interested in doing so.
We also need to establish patterns of behaviour in organisations that have power over us so that the software we use is not rendered inadequate due to unnecessary technological change. This leads to another area of endeavour entirely: that of organisations focusing on sustainable, enduring, undemanding solutions instead of pursuing needless change and causing confusion, frustration and waste.
The pdfreaders.org website currently lacks information on typically used extensions. Also there is no way to file a complaint on a specific case that could help paint a better picture of the current situation.
The PDF Readers campaign focused on anticompetitive and coercive practices, which was an important thing to do. But as I have noted before, we also need to be able to answer the follow-up questions of what people should otherwise be using after telling them that a document format and a software package are not the same thing.
At the time of writing, it is just you and I having this discussion, Nico, because our mails have not made it to the list, but I hope that others have useful insights that they might eventually be able to share with us.
Paul
Hi,
to focus on the PDF form handling:
Am Freitag 31 Januar 2020 17:29:18 schrieb Paul Boddie:
Having been sent a PDF form that needed completing, I rediscovered the apparent lack of support in Free Software applications for performing this task.
I ended up with the feeling that it would not be possible to perform an important task using Free Software. No amount of advocacy would remedy this situation.
An example to the contrary: The tax authority of the German state "Lower Saxony" (almost 8 Million inhabitants) is using GNU/Linux desktops in about 12,000 workplaces. Now the current government coalition plans to switch to Windows, this is political and can be influenced by advocay. As documented on our public German speaking FSFE list, it is unclear how this change came to be, probably advocay from the different direction, see the threat (in German): https://lists.fsfe.org/pipermail/fsfe-de/2018-August/010390.html
This public tax authority contracted my company Intevation a few years ago to improve form handling in a few aspects in okular (and the libraries it is using), which we did. The connection is: If they were planning to use okular even longer, they would contract more improvements to PDF handling. They don't because of the political process of being forced to go to Windows.
Am Dienstag 11 Februar 2020 23:22:59 schrieb Paul Boddie:
For all I know, my experience was based on a misunderstanding
PDF is not a good format, in my view. There are several different ways of handling form elements and the ability to embed some active javascript elements is not a good choice for an open standard for simple forms. PDF also does not adhere to the "minimal principle" (https://fsfe.org/activities/os/minimalisticstandards.en.html).
So "misunderstandings" are likely, computer usage is complicated and just getting to understand what would be a good next step for which user group requires a lot of effort to begin with.
It is hard work to "track" what the proprietary developers of the PDF format do and it needs people that do this for years, which means they need to be financed. We at Intevation were toying with how this can be done a few time, but haven't found something promissing yet. Maybe a crowd-funding for some features? But the work can probably not be estimated well enough. A pay-as-you-want windows build of okular? Maybe, but it needs serious time invest and expertise.
The problem is: Getting somewhere with better form handling in PDF, we are not looking at a 20 k€ project, but it would need years with funding around 200 k€ per year or more to get somewhere.
Best Regards, Bernhard Disclaimer: As co-owner of Intevation I have a "commercial" interest to help making more Free Software (that is the only software we make) and earn a living and pay my employees as fair as I can. (It should be obvious from this post, but just in case. ;) )
On 2/13/20 9:58 AM, Bernhard E. Reiter wrote:
It is hard work to "track" what the proprietary developers of the PDF format do and it needs people that do this for years, which means they need to be financed. We at Intevation were toying with how this can be done a few time, but haven't found something promissing yet. Maybe a crowd-funding for some features? But the work can probably not be estimated well enough. A pay-as-you-want windows build of okular? Maybe, but it needs serious time invest and expertise.
The problem is: Getting somewhere with better form handling in PDF, we are not looking at a 20 k€ project, but it would need years with funding around 200 k€ per year or more to get somewhere.
I haven't worked with PDF implementations, but €200k/year sound reasonable for an actual program to track the functionality of proprietary developers and hopefully lead the way on some points.
Ironically, or perhaps rather: Tragically, this would be easy to do if public authorities were focused on using free software in their daily work. With pooling of resources, they could get a lot of these things done and end up with software that was free and in the end also technically better as well as cheaper and more flexible for each customer. But they are not, and as a result there's much less resources available for free software development, and hence some core products are perceived to be lacking.
And that's something advocacy *can* do: Change this political understanding and hence make public organisations (among others) dedicate more funds to development of software that is not only good, but also free.
Best Carsten
On Thursday 13. February 2020 09.58.23 Bernhard E. Reiter wrote:
An example to the contrary: The tax authority of the German state "Lower Saxony" (almost 8 Million inhabitants) is using GNU/Linux desktops in about 12,000 workplaces. Now the current government coalition plans to switch to Windows, this is political and can be influenced by advocay. As documented on our public German speaking FSFE list, it is unclear how this change came to be, probably advocay from the different direction, see the threat (in German): https://lists.fsfe.org/pipermail/fsfe-de/2018-August/010390.html
The difference between us as Free Software advocates and those promoting proprietary software is usually a large amount of money being available to the latter group, meaning that people can very easily be influenced by all sorts of "incentives" to switch to particular kinds of proprietary software. It is not limited to discounts and rebates on that software, either, nor on bundling deals with other proprietary software that an organisation might already depend on.
Having a captive customer base provides the revenue stream to introduce shiny new features that entice the kind of executive-level manager who wants to put their mark on things or the easily distracted and vocal users who don't feel that their job is complete without something new to play with. Not having such a revenue stream - partly due to decency in not wanting to exploit or control users, partly due to the perpetuation of the idea that "open source" is some kind of endless free buffet - means that Free Software advocacy is not able to drive the agenda.
This public tax authority contracted my company Intevation a few years ago to improve form handling in a few aspects in okular (and the libraries it is using), which we did.
This is interesting to hear and useful to know.
The connection is: If they were planning to use okular even longer, they would contract more improvements to PDF handling. They don't because of the political process of being forced to go to Windows.
I guess it could be pointed out that local businesses and the local economy would benefit if public agencies exercised their right to control their technological destiny through Free Software, whereas imposing proprietary software is likely to see most of the money leave the local economy. But I imagine that the response involves a combination of highlighting local "solution partners" doing customisation or configuration and nothing more demanding, a demonstration of corporate generosity (such as promises to fund "partnerships" with public institutions), and vague statements about the supposed merits of trickle-down economics.
Am Dienstag 11 Februar 2020 23:22:59 schrieb Paul Boddie:
For all I know, my experience was based on a misunderstanding
PDF is not a good format, in my view. There are several different ways of handling form elements and the ability to embed some active javascript elements is not a good choice for an open standard for simple forms. PDF also does not adhere to the "minimal principle" (https://fsfe.org/activities/os/minimalisticstandards.en.html).
I agree with you about PDF as a format, and I imagine that its incorporation into public sector standards may have involved simplified subsets. I also don't think that a vendor-driven technology should be a standard, either, but even the most broadly accepted of various common open standards have a degree of vendor manipulation.
So "misunderstandings" are likely, computer usage is complicated and just getting to understand what would be a good next step for which user group requires a lot of effort to begin with.
It is hard work to "track" what the proprietary developers of the PDF format do and it needs people that do this for years, which means they need to be financed. We at Intevation were toying with how this can be done a few time, but haven't found something promissing yet.
PDF serves a need for an encapsulated "print-ready" format, because despite printing being rather tangential to many processes these days, it is convenient to have documents that are effectively frozen. Potentially intersecting with those needs are other needs that should be investigated.
It is possible that the other needs (forms being an example) might not require the peculiar characteristics of PDF. If so, it becomes pertinent to offer other technologies as alternatives.
Maybe a crowd-funding for some features? But the work can probably not be estimated well enough. A pay-as-you-want windows build of okular? Maybe, but it needs serious time invest and expertise.
Why would only Windows users be thought of as paying customers?
The problem is: Getting somewhere with better form handling in PDF, we are not looking at a 20 k€ project, but it would need years with funding around 200 k€ per year or more to get somewhere.
I agree. And that brings me onto the topic of sustainable and predictable funding for work, which means that we should be looking beyond occasional and unpredictable deliveries of funding in amounts that are often inadequate or even exploitative (when considering the scope of some projects and the obligations that people feel when accepting funding).
Paul