Monday, November 23, 2009

Has the market rejected The Linux Desktop?

That seems to be the premise for the following article:
Here is my point by point response:

It’s still too much of a pain:
Compatibility should be the vendor's realm. To pass the buck to the OS then the device spec should be open. For some vendors this would be unconscionable.

The divide and fail strategy:
First: Linux is not an OS, it's a kernel. Second: The problem is trying to measure success based on proprietary goals. The goal of FLOSS is to be Free. There are actually many other contexts in life where being free is more of a hassle but we still prefer it. e.g. uniforms eliminate the "need" to think what you are going to wear. Does that make them better for everyone?

Not enough innovation:
"The primary value proposition for Linux is that it’s just as good as Windows"
Wrong premise. The primary value proposition of a FLOSS operating system is its Freedom.

There is innovation although it happens in bits and pieces and components that you don't really see. They are transparent to the user. Here are some examples
- Software repositories
- The free desktop specification
- The compiz desktop
- Desktop folders on KDE (easy to access most used folders)
- Print to PDF from any application

I agree that there are many people doing the Windows copycat. Why do they do it? because they can. It's their freedom and their right. There is no right answer. Sometimes true innovation causes confusion as people tend to resist change.

Businesses want someone to blame:
With Linux they can, there is RedHat, Novel, Mandriva, Canonical and many other vendors that will stand up for the FLOSS OS they sell. Many FLOSS software has professional support, actually that's usually the business model for free software. There is a real drive to provide support as no one is ever "captive client" as it happens with proprietary software.

Summary: The market hasn't "rejected it" it doesn't know the benefits as people keep trying to put the benefits in terms of proprietary software. I guess the problem is not with the product but with the "marketing".

In the end "FLOSS" is a counterbalance to proprietary software and formats. A way to free users from vendor and format lock in. That's what, in my view, makes it better. Once users understand that the current lock-ins don't have to be "normal" they'll realize that there are alternatives.

Wednesday, November 04, 2009

Karmic Koala: Who's to blame for the issues found by early adopters?

Someone at my LUG's email list pointed to this article:

This is not a response or rebuttal to that article as it seems to be quite objective reporting on facts. This is instead a reflection on why being early adopters and expecting perfection is an utopia.

I read the article and it gave a clue of why these issues are popping up:

"Early adopters tend to be the most technical of the technical."

As part of the community and early adopters we should share part of the blame:

There was time to test the alphas, betas and RCs. I'm sure the developers do not have access to all the different combinations of hardware out there. Linux is a community effort. It is only by contributing with testing that we can ensure the release will come out cleaner.

Even other companies with huge resources have extensive pre-release beta programs and beta testers take it upon themselves to improve those companies software.

It is not common sense, but it is common professional practice that with ANY software for a production environment (This would include your primary desktop) you either:
- test the pre-release versions and felt comfortable before upgrading or
- you wait until other people found the issues and hope you don't have a unique one.

In both scenarios you make every effort to have a way to roll back the change (e.g. data backups or even OS partition image)

Another common practice is to have your production systems at version n-1 and a test system at the level you are planning to move next.

Being early adopters the "most technical of the technical", you'd think that they either contributed testing the pre-release versions or at least tried it in a separate partition before rendering their main system unusable.

So, if you are technical: don't wait to test and report issues, you may uncover something the developers don't know or you may have a clue on something which is baffling the developers.

If you are not comfortable with this either because you are not technical enough or don't have time to test, then keep using version n-1 until most of the issues of the latest version have been resolved.

I do think that Canonical should be clearer on this and, on the downloads page, advise non technical users to use the LTS version. After all, non technical users don't know about the good technical practices I mentioned before.

Canonical's part of the blame is removing the link to download the LTS.