This is pretty incredible. It really feels as if a bunch of people were asleep at the wheel, mumbling in their dreams about how Linux on the desktop was totally ready to go, and then Linus and Luke threw a glass of cold water on their faces and went “No it’s not, I can’t sign a PDF goddammit”
My only gripe so far with their series was when Linus complained that right-clicking and doing “Save as” on a GitHub link saves an HTML file instead of the file itself. But here’s the thing: If you’re a non-developer and you have to go to GitHub to get everyday software running, then the ecosystem has failed you. We can’t both say “Linux is an acceptable Windows substitute for most people” and also say “Well of course you need to know how to run a shell script, it’s Linux!”
> My only gripe so far with their series was when Linus complained that right-clicking and doing “Save as” on a GitHub link saves an HTML file instead of the file itself.
However, that's a function of the browser, and not of the OS itself. If I were to do that using Firefox on Windows or Linux, it would save as a file regardless. So, my gripe would be that he apparently can't distinguish problems with the OS itself, and those of software running on the platform.
Yeah exactly. I was kinda like "Well dude, if you click a link and it takes you to an HTML page, certainly clicking Save As will save an HTML page". I kind of feel like Linus should know that. But the higher point stands: A lot of my friends and relatives should not need to be on GitHub pulling down scripts so they can enter a video call or do a stream or screencast. The only exception being if there is a direct link to a GitHub releases page with download links for executables.
Why though? Isn't most windows software downloaded from <insert random developer's website here>?
I think having software more centralized not only makes things easier but also reduces the "sketch factor" of downloading random executables from random websites
The point is that a regular user is unlikely to ever need to go to GitHub and run a script on Windows or MacOS.
However, it’s still surprisingly common on Linux.
The truth is, though, Linux today already has the capabilities to make these things far more seamless than Windows or MacOS. The problem is that since the user base is so technically inclined already, they don’t see these as problems. Or if they do, they don’t see them as being important enough.
And that’s why I think the LTT challenge has been so successful in getting distributions, DEs, etc to make changes because it does a good job of showing the developers which of these issues, that may seem minor to someone with a technical background can be a complete deal breaker for a regular user.
> The problem is that since the user base is so technically inclined already, they don’t see these as problems. Or if they do, they don’t see them as being important enough.
They may not even realise that it is a potential issue. I'm a technically inclined user, and it wouldn't even occur to me to consider someone doing what Linus did there because I know because I "just know" it won't work and would gloss over it completely.
Fortunately (for everyone else) I'm not in the habit of designing UIs.
It was painful to watch. I often sign PDF documents on Linux. I see three ways of doing it:
- using annotations in the PDF viewer. Okular supports this, I think Evince too. So you don't have to install anything normally
- using Xournal. That's was I do.
- Pasting an image of your signature using LibreOffice Draw. I know someone doing this.
It's pretty simple and you don't need to install Wine + a Windows software to do this. But you have to know it I guess. I'm sure I would be in a similar situation as them, should I have to sign a PDF document on macOS or Windows. I know the PDF viewer on macOS does that from observing former colleagues. On Windows I would just install Xournal I guess.
There could be a "Sign this PDF" button in PDF viewers to make this common action obvious and solve this particular use case.
None of the three things you describe affixes a digital signature, just an image of a signature. If you need to sign a PDF for some legal purpose, you need the former, not the latter.
When we talk about signatures on PDF, it's almost always explicitly about eSignatures and certificates. I doubt a photo of your physical signature is legally enforceable anywhere without a proper certificate and an audit trail.
Sometimes I just have a feeling Linus is playing dumb. I know he knows better but for example when you get a timed task to compress some files into a ZIP you sure are not foolish enough to make it harder for yourself and select files gigabytes in size if this is not a requirement of the task. And then he even wonders why compression is taking so long.
>Sometimes I just have a feeling Linus is playing dumb.
I don't think he was playing dumb in that case. I think he was frustrated by the Linux challenge and didn't enjoy it one bit so he did all the tasks in a sloppy, "let's just get this over with" manner, which caused him to make silly mistakes that could have been prevented if he would have done 60 seconds of googling "how do I x on Linux" before plowing head first into everything.
But I can completely understand him. He's the CEO and the face of a multi million dollar media company and a family man with kids, hardly the person with the spare time and inclination to be learning the quirks of a new OS between running a company and spending time with a family, when he just wanted to use that little spare time to play the vidja games he liked on Linux instead of Windows, not study the new OS beforehand.
And I think his attitude can be found in many average joes who don't find the experience of learning a new OS enticing and just expect a seamless transition from Windows/MacOS to Linux.
13 comments
[ 3.2 ms ] story [ 35.0 ms ] threadMy only gripe so far with their series was when Linus complained that right-clicking and doing “Save as” on a GitHub link saves an HTML file instead of the file itself. But here’s the thing: If you’re a non-developer and you have to go to GitHub to get everyday software running, then the ecosystem has failed you. We can’t both say “Linux is an acceptable Windows substitute for most people” and also say “Well of course you need to know how to run a shell script, it’s Linux!”
However, that's a function of the browser, and not of the OS itself. If I were to do that using Firefox on Windows or Linux, it would save as a file regardless. So, my gripe would be that he apparently can't distinguish problems with the OS itself, and those of software running on the platform.
I think having software more centralized not only makes things easier but also reduces the "sketch factor" of downloading random executables from random websites
However, it’s still surprisingly common on Linux.
The truth is, though, Linux today already has the capabilities to make these things far more seamless than Windows or MacOS. The problem is that since the user base is so technically inclined already, they don’t see these as problems. Or if they do, they don’t see them as being important enough.
And that’s why I think the LTT challenge has been so successful in getting distributions, DEs, etc to make changes because it does a good job of showing the developers which of these issues, that may seem minor to someone with a technical background can be a complete deal breaker for a regular user.
They may not even realise that it is a potential issue. I'm a technically inclined user, and it wouldn't even occur to me to consider someone doing what Linus did there because I know because I "just know" it won't work and would gloss over it completely.
Fortunately (for everyone else) I'm not in the habit of designing UIs.
It was painful to watch. I often sign PDF documents on Linux. I see three ways of doing it:
- using annotations in the PDF viewer. Okular supports this, I think Evince too. So you don't have to install anything normally
- using Xournal. That's was I do.
- Pasting an image of your signature using LibreOffice Draw. I know someone doing this.
It's pretty simple and you don't need to install Wine + a Windows software to do this. But you have to know it I guess. I'm sure I would be in a similar situation as them, should I have to sign a PDF document on macOS or Windows. I know the PDF viewer on macOS does that from observing former colleagues. On Windows I would just install Xournal I guess.
There could be a "Sign this PDF" button in PDF viewers to make this common action obvious and solve this particular use case.
I don't think he was playing dumb in that case. I think he was frustrated by the Linux challenge and didn't enjoy it one bit so he did all the tasks in a sloppy, "let's just get this over with" manner, which caused him to make silly mistakes that could have been prevented if he would have done 60 seconds of googling "how do I x on Linux" before plowing head first into everything.
But I can completely understand him. He's the CEO and the face of a multi million dollar media company and a family man with kids, hardly the person with the spare time and inclination to be learning the quirks of a new OS between running a company and spending time with a family, when he just wanted to use that little spare time to play the vidja games he liked on Linux instead of Windows, not study the new OS beforehand.
And I think his attitude can be found in many average joes who don't find the experience of learning a new OS enticing and just expect a seamless transition from Windows/MacOS to Linux.