Hello. I am a strong free software supporter, trying to help in any possible way free software to move forward. I'm involved in the French Trisquel community mostly as a translator and in Replicant, the 100% free Android version as a developer. Now that I introduced myself, here's the subject of this mail: the psview software included in the http://www.pdfreaders.org/ list.
My big problem with this software being in that list is that the way you present it "psview (online)" suggests that this software is to be used online on the http://view.samurajdata.se/ website. Although this software is indeed free software, using it online on the server the user doesn't own makes it Software as a Service (SaaS) : this is bad because the user has no control over the computing that is done on this server while it handles the user data to return back only the output.
Even if the software that runs on the server is apparently free software and that the source code is given, the following issues remain: * We have no way to be sure that the source code that is distributed is actually what's used on this software: we have to trust the good will of the person who owns the server. * Even if the code is free, we have no guarantee that the person who owns the server won't keep the files we submit and do a nasty use of these. * We don't know on the top of what it runs : the system running on the server may contain some non-free parts that could do a nasty use of the data we send to the server.
etc. These are not individuals problems in themselves but consequences of Software as a Service.
Promoting such uses is not, in my opinion, promoting freedom. And this opinion seems shared with Richard M. Stallman, who wrote an article about this called "Who does that server really serve" (I guess you already know about it): http://www.gnu.org/philosophy/who-does-that-server-really-serve.html.
Perhaps promoting the use of this server the user has no control over wasn't your intention: maybe your intention was to mention that such a software exist and since it's free, can be used by anyone on a server they own. This makes the use of this software respect the user freedom and can indeed be encouraged.
Anyway, when I went on this page and saw this, my impression was that you recommend using this on the server the link points to.
I suggest you to add some information along with the link to this software on the http://www.pdfreaders.org/ website or to remove the link to it (but this is perhaps a bit too much, even if I'm convinced that almost all the users who'll click on this link won't download the software to use it on their own server but us it directly on this server, thinking they're taking good care of their freedom).
Thanks for the attention to my message, I'm looking forward to hearing from you.
He has a good point, we should discuss this...
Am Dienstag, 16. August 2011, 23:05:58 schrieb PaulK:
Hello. I am a strong free software supporter, trying to help in any possible way free software to move forward. I'm involved in the French Trisquel community mostly as a translator and in Replicant, the 100% free Android version as a developer. Now that I introduced myself, here's the subject of this mail: the psview software included in the http://www.pdfreaders.org/ list.
My big problem with this software being in that list is that the way you present it "psview (online)" suggests that this software is to be used online on the http://view.samurajdata.se/ website. Although this software is indeed free software, using it online on the server the user doesn't own makes it Software as a Service (SaaS) : this is bad because the user has no control over the computing that is done on this server while it handles the user data to return back only the output.
Even if the software that runs on the server is apparently free software and that the source code is given, the following issues remain: * We have no way to be sure that the source code that is distributed is actually what's used on this software: we have to trust the good will of the person who owns the server.
- Even if the code is free, we have no guarantee that the person who
owns the server won't keep the files we submit and do a nasty use of these.
- We don't know on the top of what it runs : the system running on the
server may contain some non-free parts that could do a nasty use of the data we send to the server.
etc. These are not individuals problems in themselves but consequences of Software as a Service.
Promoting such uses is not, in my opinion, promoting freedom. And this opinion seems shared with Richard M. Stallman, who wrote an article about this called "Who does that server really serve" (I guess you already know about it): http://www.gnu.org/philosophy/who-does-that-server-really-serve.html .
Perhaps promoting the use of this server the user has no control over wasn't your intention: maybe your intention was to mention that such a software exist and since it's free, can be used by anyone on a server they own. This makes the use of this software respect the user freedom and can indeed be encouraged.
Anyway, when I went on this page and saw this, my impression was that you recommend using this on the server the link points to.
I suggest you to add some information along with the link to this software on the http://www.pdfreaders.org/ website or to remove the link to it (but this is perhaps a bit too much, even if I'm convinced that almost all the users who'll click on this link won't download the software to use it on their own server but us it directly on this server, thinking they're taking good care of their freedom).
Thanks for the attention to my message, I'm looking forward to hearing from you.
* Hannes Hauswedell h2@fsfe.org [2011-08-17 03:59:09 +0000]:
He has a good point, we should discuss this...
On the website we link there is the source code of the software. So I would propose we just remove the "online" and write him, that everybody can set it up. What do you think?
Regards, Matthias
Am Dienstag, 23. August 2011, 13:56:32 schrieb Matthias Kirschner:
- Hannes Hauswedell h2@fsfe.org [2011-08-17 03:59:09 +0000]:
He has a good point, we should discuss this...
On the website we link there is the source code of the software. So I would propose we just remove the "online" and write him, that everybody can set it up. What do you think?
His point was a different one: That the source code is available makes the software better from a global point of view (because people can make copies and improve on them), but using a gpled web-service provides no benefit over using a proprietary one from personal point of view (we actually encourage people to send their data/files to another web-site for processing!) and is definitely a lot worse than using Free Software locally. The only thing remotely equal to using Free Software locally is combining AGPL and Zero-Knowledge infrastructure, as previously described by Richard[1].
If we recommend a web-service, for something that can aswell be done without a web-service, we should make this very clear. Otherwise people will call us hypocrites: "You say Adobe does not respect my privacy, because they *could* snoop on my files, but recommend sending my files directly to some dubious homepage, because of some license? And you can't even prove that the software running on the site is actually Free Software?"
Hi Hannes,
ok, I got your point, and I think it is valid. Just to make sure the criteria to be listed on pdfreaders.org would than be:
1. Free Software 2. running on a desktop computer
so they are competing with Acrobat Readers, Foxit, and other non-free software which is promoted on the websites.
Is the above correct? Than we should put that on the website and remove the online reader.
Regards, Matthias
Am Mittwoch, 31. August 2011, 09:57:18 schrieb Matthias Kirschner:
Hi Hannes,
ok, I got your point, and I think it is valid. Just to make sure the criteria to be listed on pdfreaders.org would than be:
- Free Software
- running on a desktop computer
so they are competing with Acrobat Readers, Foxit, and other non-free software which is promoted on the websites.
Is the above correct? Than we should put that on the website and remove the online reader.
Regards, Matthias
I will take care of answering Paul. Maybe we should use this opportunity to also state that the pdf-reading component has to be free and that this condition rules out skim. What do you think?
Maybe an extra section on the About page:
---
"Conditions for PDF-Readers"
To be listed on this page a PDF reading software must fulfil certain conditions. First and foremost it has to be Free Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
Currently we only cover regular desktop PDF-Readers. Software for editing, converting or creating PDFs is not listed; Software for mobile platforms is not listed and web-applications ("Software as a Service") are not either.
If you come across software that fulfils these conditions and is not listed, please let us know!
---
What do you think? Maybe have translators look over it? [its late right now ;) ]
* Hannes Hauswedell h2@fsfe.org [2011-08-31 22:49:18 +0200]:
I will take care of answering Paul.
Thanks.
Maybe we should use this opportunity to also state that the pdf-reading component has to be free and that this condition rules out skim. What do you think?
I think it is good. Also add Bernhard in CC:
Maybe an extra section on the About page:
"Conditions for PDF-Readers"
To be listed on this page a PDF reading software must fulfil certain conditions. First and foremost it has to be Free Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
Currently we only cover regular desktop PDF-Readers. Software for editing, converting or creating PDFs is not listed; Software for mobile platforms is not listed and web-applications ("Software as a Service") are not either.
If you come across software that fulfils these conditions and is not listed, please let us know!
What do you think? Maybe have translators look over it? [its late right now ;) ]
I like it. Bernhard, do you agree this is a good definition of our criteria?
Regards, Matthias
Am Freitag, 2. September 2011, 13:58:52 schrieb Matthias Kirschner:
- Hannes Hauswedell h2@fsfe.org [2011-08-31 22:49:18 +0200]:
I will take care of answering Paul.
Thanks.
Maybe we should use this opportunity to also state that the pdf-reading component has to be free and that this condition rules out skim. What do you think?
I think it is good. Also add Bernhard in CC:
Maybe an extra section on the About page:
"Conditions for PDF-Readers"
To be listed on this page a PDF reading software must fulfil certain conditions. First and foremost it has to be Free Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
Currently we only cover regular desktop PDF-Readers. Software for editing, converting or creating PDFs is not listed; Software for mobile platforms is not listed and web-applications ("Software as a Service") are not either.
If you come across software that fulfils these conditions and is not listed, please let us know!
What do you think? Maybe have translators look over it? [its late right now ;) ]
I like it. Bernhard, do you agree this is a good definition of our criteria?
Maybe "Criteria for PDF-Readers" is a nicer heading then "Conditions for PDF-Readers".
* Hannes Hauswedell h2@fsfe.org [2011-09-02 13:32:35 +0200]:
Maybe "Criteria for PDF-Readers" is a nicer heading then "Conditions for PDF-Readers".
Yes, I agree.
On Friday 02 September 2011 13:32:35 Hannes Hauswedell wrote:
Am Freitag, 2. September 2011, 13:58:52 schrieb Matthias Kirschner:
- Hannes Hauswedell h2@fsfe.org [2011-08-31 22:49:18 +0200]:
"Conditions for PDF-Readers"
To be listed on this page a PDF reading software must fulfil certain conditions. First and foremost it has to be Free Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
In our technical definition we should make clear that we mean "additional" libraries for proprieary systems. Check out the Gnu GPL line on this., The components that normally come with _all_ versions of a proprietary operating systems can be used by Free Software on it, even with the GNU GPL. We should not get stronger than this.
I don't know what library Skim uses so I cannot say, but the criteria should be understandable, otherwise people will rule out all pdf readers on proprietary platforms. They all use proprietary libraries at least some.
Currently we only cover regular desktop PDF-Readers. Software for editing, converting or creating PDFs is not listed; Software for mobile platforms is not listed and web-applications ("Software as a Service") are not either.
Note that most of my laptops are quite "mobile" and can deal with networks connections being there or not and changing, so "mobile" does not seem to be a fitting term. If you mean small handheld devices or touch screen usage, please say so. :) (My company did an application called Kontact Touch, which can be used on mobile and immobile machines.)
Thanks for progressing here, this is hard, but important work! I hope my comments are useful!
Best, Bernhard
* Bernhard Reiter reiter@fsfeurope.org [2011-09-05 11:14:57 +0200]:
Thanks for progressing here, this is hard, but important work! I hope my comments are useful!
Hannes, are you able to update the text?
Thanks, Matthias
Hi Bernhard,
Zitat von "Bernhard Reiter" reiter@fsfeurope.org:
On Friday 02 September 2011 13:32:35 Hannes Hauswedell wrote:
Am Freitag, 2. September 2011, 13:58:52 schrieb Matthias Kirschner:
- Hannes Hauswedell h2@fsfe.org [2011-08-31 22:49:18 +0200]:
"Conditions for PDF-Readers"
To be listed on this page a PDF reading software must fulfil certain conditions. First and foremost it has to be Free Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
In our technical definition we should make clear that we mean "additional" libraries for proprieary systems. Check out the Gnu GPL line on this., The components that normally come with _all_ versions of a proprietary operating systems can be used by Free Software on it, even with the GNU GPL. We should not get stronger than this.
We are stronger than this, because we are about Free PDF-Readers.
I don't know what library Skim uses so I cannot say, but the criteria should be understandable, otherwise people will rule out all pdf readers on proprietary platforms. They all use proprietary libraries at least some.
Yes, but these libraries can be replaced. E.G. Okular on Windows might link against some Windows libs, but the same piece of source-code will produce the same piece of software on a completely free system.
With Skim this is not the case. Skim uses PDFKit which is installed on every MacOSX and cannot be removed, but there is no gain in Freedom for the user, because *all* functionality depends on this proprietary lib. Needless to say you cannot port Skim to another (free) OS.
So I think our policy is sound. We have discussed this multiple times, but of course I open for any new input on the matter :)
Currently we only cover regular desktop PDF-Readers. Software for editing, converting or creating PDFs is not listed; Software for mobile platforms is not listed and web-applications ("Software as a Service") are not either.
Note that most of my laptops are quite "mobile" and can deal with networks connections being there or not and changing, so "mobile" does not seem to be a fitting term. If you mean small handheld devices or touch screen usage, please say so. :) (My company did an application called Kontact Touch, which can be used on mobile and immobile machines.)
Thanks for progressing here, this is hard, but important work! I hope my comments are useful!
Thanks for your input! I am in Israel now and only in an internet cafe once in a while so decision-making will be slow, if it involves me. I will be back in Berlin on October, 1st.
Regards, Hannes
Hi Hannes,
On Tuesday 06 September 2011 13:09:31 Hannes Hauswedell wrote:
First and foremost it has to be Free
Software[link to our definition]. This includes -- but is not limited to -- the software component responsible for rendering the PDF (e.g. Skim[link] does not qualify, because it uses a proprietary library for PDF rendering).
In our technical definition we should make clear that we mean "additional" libraries for proprieary systems. Check out the Gnu GPL line on this., The components that normally come with _all_ versions of a proprietary operating systems can be used by Free Software on it, even with the GNU GPL. We should not get stronger than this.
We are stronger than this, because we are about Free PDF-Readers.
..then the above explanation of the criteria should be improved.
A software using a proprietary pdf-rendering library which comes with all versions of the operating system is Free Software to begin with, so it matches our Free Software definition I guess. Also requiring the pdf-render library to be Free Software means is another requirement and you need a definition what is the render library. Does it for instance include support library for the rendering, e.g. like an XML library that might come from the system? If so, why?
I don't know what library Skim uses so I cannot say, but the criteria should be understandable, otherwise people will rule out all pdf readers on proprietary platforms. They all use proprietary libraries at least some.
Yes, but these libraries can be replaced.
All proprietary libraries can be replaced. If you apply this thinking you'll end up without any Free Software reader on proprietary platforms. Which is not what I believe is sensible.
E.G. Okular on Windows might link against some Windows libs, but the same piece of source-code will produce the same piece of software on a completely free system.
Every library can change behavioural details, so it will not behave similiar on GNU systems. Almost all GNU systems I have heard of are not entirely "free" either, many use proprietary BIOS, firmware or microcode on their chips. Any of these components will change the behaviour of the software. If you come up with such a criteria it should be sound.
With Skim this is not the case. Skim uses PDFKit which is installed on every MacOSX and cannot be removed, but there is no gain in Freedom for the user, because *all* functionality depends on this proprietary lib.
That is a good reason to demand that Skim is not listed, I agree.
Needless to say you cannot port Skim to another (free) OS.
Sure you can, just "port" or use a different library. :P
So I think our policy is sound. We have discussed this multiple times, but of course I open for any new input on the matter :)
The proposed explanation of the policy does not explain the borders of this decision. I suggest to explain this in more detail saying that the frontend and the backend of the PDF software needs to be Free Software. And then the "backend" needs to be defined a little bit. Maybe containing the core logic of the PDF format or something.
Thanks again for working on this, I am merely trying to check if the argument on the explanation makes good sense so you are most effective in communicating this great campaign.
Best Regards, Bernhard
Hi Paul,
your points are vary valid. We discussed this internally and decided to keep the site's focus on Desktop Software and the "traiditional" case of Free Software vs Proprietary Software. We will state this policy more clearly and remove the psview web service.