Linux Needs Critics

An anonymous reader writes "Keir Thomas berates the fact that the world of Linux almost entirely lacks critics. In fact, he says, Linux people tend to see genuine critical evaluation as a bad thing. FTA: 'The problem with this anti-criticism approach is that it's damning Linux to an eternity of navel gazing. Nothing can ever get any better. The best hope we have are the instances where a few bright sparks, with their heads screwed on the right way, get together and make something cool (as happened with, say, Firefox back in the day). But that's rare and can't be relied upon.'"

Comments

You should look into linuxhaters

[Useful Wheat]

You should look into the Linux Haters Blog published here: http://linuxhaters.blogspot.com/ The author is a former Linux contributor, and he offers many valuable insights into some of the issues with the code, for example:

Let me attempt to summarize. A) PulseAudio needs to work with existing applications, so it implements an ALSA emulation layer, except, it's not complete. Only 70% of ALSA applications work. So it's like, totally ready. B) So, in the true open source fashion, you should port your app to be a native PulseAudio client. Except that you can't. There's this yet-another-audio-library called libsidney, but it's not ready yet. (Hmm, this sounds familiar...) C) Fedora led the way in incorporating PulseAudio before it was ready, breaking audio for thousands of users. Then because open source is about copying good ideas and bad ones, a ton of other distros adopted it as well. Amazing guys. In a way, you've spread bad code that breaks audio on thousands of computers faster than a virus could have. And it's immune to antivirus! D) so now that we're in this "mess" (as the lead developer of PulseAudio calls it*), LSB comes along and says "we're going to standardize how your write audio apps!" Oh, but wait, ALSA's now "old" (we hardly knew ye), and I can't directly program PulseAudio. Hmm... So the article's brilliant solution? Standardize on the PulseAudio-safe subset of ALSA. WHAT THE FUCK. I can just imagine the future alsa man page. A big listing of functions, with a nice little asterisk next to those functions that you shouldn't use unless you want your app to totally FAIL on a system which has been sodomized by Pulse Audio. I can just see the developers of commercial Linux sound apps (all three of them) jumping for joy. And thus unfolds another chapter in long history of failed sound systems on Linux. Can they make it much worse? I, for one, am excited to see how much worse they can make it until we all go back to listening to square waves on our PC speakers. * BTW, also notice that it's the PulseAudio guy calling Linux audio a mess. Did he forget that it was his project that took the existing mess, and unloaded a giant steaming turd on it? Congratufuckinglations. You've just made it worse. You're a truly a worthy OSS contributor.

He's pretty harsh, but he always has a point behind it.

Re:Nonsense

[DJRumpy]

I wouldn't necessarily say it has a lack of critics. I think it is more pointed to the Linux communities response to said criticism. They tend to be very dismissive, defensive, and unresponsive to any criticism whether it's deserved or not.

From TFA:
"Most of the time the world of Linux tends to be anti-critical. If anybody in the community dares be critical, they get stomped upon."

There are plenty of critics

[kimvette]

I've been complaining about the dumbing down of Gnome (they think gnome users are idiots - just look at the file dialog for one example), the crappy Flash player Adobe puts out for Firefox (why can't DHTML float over flash like it can in MSIE? Is the problem Flash or Firefox? Either way, it's been broken since day one and needs fixing), OpenOffice is spaghetti code and I/O is very slow, *something* needs to be done so more preconfigured systems can be shipped (NVidia & GPL "license" incompatibility creates legal issues when it comes to shipping preinstalled systems), X11 and VNC are horribly inefficient over a WAN, whereas Windows' Remote Desktop Protocol works great even over dial-up connections, oh, and yeah, developers still suck when it comes with users who bother to submit bug reports - especially the OpenOffice folks. They just don't want to fix horrid architectural issues or bugs, because developing new buggy features is more interesting than fixing their previous garbage.

Having said that, I do recommend Linux whenever and wherever it makes sense. I've slowly been convincing the Rabbi at my congregation to go F/OSS at home, the congregation's infrastructure is going to be 90% linux, my business is >90% Linux, and some of my customers run Linux. However, there are many cases where Linux just is not a good fit. It's not the one-size-fits-all BFH. Sometimes a a screwdriver or wrench is a more appropriate tool.

Where is AutoCAD?

Where is the Adobe Creative Suite? (I personally get by with inkscape + gimp + pdfedit + Krita, but my art director NEEDS the Adobe CS (So it's Windows at work and OS X at home for him). It takes me ~3 hours to do a task that takes him under a half hour in Illustrator or Gimp, because to get the same final product requires a lot more manual steps in Gimp and Inkscape; no layer effects, no droplets, Macro recording and playback doesn't exist in any user-friendly way (and no I am NOT about to get into scripting gimp. I'll stick to shell scripting server maintenance and monitoring, and writing installers. thanks anyway!)

Where is Quickbooks for Linux? They have a server component that runs on Linux, but where is the Quickbooks Pro desktop app?

Where are Linux-based embroidery apps? Windows XP is going to be on my new Dell Precision notebook so I can design embroidery patterns. I draw them in Inkscape but I need them to be converted to an embroidery format my machine can understand. So, I do the design AND conversion in Windows, then I don't have to reboot to run the embroidery machine.

Also, more specific to Linux itself (meaning the kernel, not the integrated distro end users refer to as Linux): Where are the merges from RedHat, Ubuntu, Novell, and so forth? Each vendor has incredible extensions to the kernel which makes automounting, user-space drivers, WiFi, and various other features work better than the vanilla kernel. Why can't LSB become a reality, and along with that, a more stable-yet-almost-bleeding-edge kernel come from kernel.org? That would make it much easier for users of $foo and $bar distros to run new hardware without losing fixes and enhancements added by the various vendors? Ubuntu works extremely well with WiFi (but I hate their standard desktops, and I hate ubuntu's administrative GUI) and with 11.x OpenSuse works almost-but-not-quite as well as Ubuntu. DeadRat, er, RedHat/Centos, not so much. Fedora? Every time I've tried it, it's been on bleeding-edge motherboards and would kernel panic or simply not boot, whereas (K)Ubuntu and OpenSuSE would always Just Work(TM). Centos/RedHat? I run it on servers, but hate it for desktops.

I love running Linux, but it is not a one-size-fits-all solution. I can't even use it as my sole OS at home any more because embroidery software I need doesn't exist. :(

Lots of us users are plenty critical of Linux, even though we are Linux evangelists. It's just that while many/most developers take feedback readily (the KDE team is particularly good in this regard!) others