Re: Web versions

"Schanzenbach, Martin" <[email protected]> Sun, 14 Mar 2021 13:36:28 +0100
Newsgroups	gmane.org.gnu.discuss
Message-ID	<[email protected]>

> On 14. Mar 2021, at 12:57, Schanzenbach, Martin <[email protected]> wrote:
> 
> 
> 
>> On 13. Mar 2021, at 16:48, Alfred M. Szmidt <[email protected]> wrote:
>> 
>>  Sounds like multi-user UNIX-like system, or modern GNU/Linux
>>  multi-user system.
>> 
>> On a multi-user system you can keep your own files in our home
>> directory.  You can decide to copy a program you like from one
>> location to your home directory.
>> 
> 
> The same is true for JS/Webassembly. In fact, one could argue that this
> is a significant part of the value offering (offline use of the web application).
> You can copy the whole site offline and continue using it.
> Yes, there MAY be interaction with a REST API, but that is a completely different
> story not directly related to webassembly at all.
> 
>> With Webassembly / Javascript (specifically in the form of SaSS) you
>> are at the mercy of whoever is hosting the program to run it.
> 
> This issue is completely unrelated to the technology.
> You are at the mercy of the (HTTP!) hoster of the software repos you use in your
> distro. It is simply by the fact that the distro project (may!) state that you
> receive free software. The same can be and is done for web applications.
> Reproducible builds may help a savvy user to confirm if trust is not enough.
> But the same could be done for WA.
> The application it self may rely on communication with a remote systems. But I though this
> is what AGPL is for.
> 

Maybe here is where we can actually untie the knot. The issue is the software architecture.
It can be abused by having the local application heavily depend on the backend.
Sometimes this is necessary (see email). But I guess you could formulate guidelines in this regard.
But I think it is important to look at such technology without prejudice.
The current reality includes much more than SaaS. And implementing a GNU package probably
would not result in such an architecture anyway. It (Webassembly) would be an organic compilation target.
As mentioned having GNU tools available on machines you do not have control over (i.e. your friends machine)
makes this infinitely valuable IMO.

>> Maybe
>> even depending on that site for storing your data.  The issue is
>> _intent_ of how these things are to be used -- depriving users of how
>> they can run their programs.
> 
> Have you tried running emacs on a C64 recently? You will have to modify the
> source to compile it to the platform. You can do the same even if the source also happens to
> compile and run as webassembly or ELF.
> 
>> 
>> The suggestion in this thread was to make GNU port to webassembly, and
>> then be run in a web browser, from someone elses machine.
>> 
> 
> No, it was not:
> 
> "
> ince WebAssembly is now a reality, maybe you guys should get to making the browser versions of AAAAAAAALL your software?
> "
> 
> A browser version can be a webextension. It can also be provided through a webserver (see distribution repository).
> As it happens, if it were to be provided as a web extension it would likely be possible to provide
> it through the distribution repositories as well. Which shows how bizzare the argument is.
> 
> In fact, a "browser version" can even be a full browser including the application! (e.g. mattermost)
> Yes, ugly, but that is not the point.
> 
>>  In my opinion question is if that all is free software. Not if it runs
>>  remotely.
>> 
>> That it is free software is a side issue.  Sometimes the issues of
>> software freedom are not just about the four freedoms.
> 
> Yes, but it is completely unclear from this thread what the issue actually is.
> "You are not sufficiently in control" is not it.
> 
> BR
>
signature.asc (application/pgp-signature, 833 B)