
When someone criticizes our favorite operating system on the Linux blogs, they typically get called Windows fanbois and are sent packing.
But when someone from the right side of fence takes the trouble to put together a video and accompanying slide show explaining *why* they think Linux sucks — and, moreover, what they think needs to be done to fix it — it’s hard not to pay attention.
Sure enough, that’s just what came up on the Linux blogs last week thanks to the efforts of Bryan Lunduke, cohost of Jupiter Broadcasting’s Linux Action Show. Lunduke’s presentation made quite a splash at LinuxFest Northwest a little over a week ago, and the waves are still crashing on the blogosphere’s shores.
‘Anti-Freedom All Around’
Some bloggers were indignant: “Wow. I’m surprised so many liked this slanted and JUST WRONG presentation,” wrote Chris Cox among the many comments on the site. “It’s anti-FOSS, anti-Linux… and pretty much anti-Freedom all around.
“If this is what the Ubuntu community is … leave me out. I can get this kind of garbage from Redmond,” Cox concluded.
Many, however, appreciated Lunduke’s perspective: “Excellent presentation Bryan,” wrote Johnny bonelessmonkey. “So glad to see some pragmatism seeping into the Linux community.”
Similarly: “It’s always refreshing to see someone ‘stirring the calm waters’ a little,” concurred Michael.
‘I Can’t Imagine How That Got Posted’
Bloggers on Digg got riled up too — to the tune of more than 1,100 Diggs and 300 comments in a few short days — so we here at LinuxInsider knew we had to dig deeper. Specifically, we took a small poll of bloggers’ opinions — and then talked to Lunduke himself.
“I can’t even imagine how that got posted,” Montreal consultant and Slashdot blogger Gerhard Mack told LinuxInsider. “That’s 45 minutes I wish I had spent doing something else.”
The first part of Lunduke’s presentation “was either whining about things already being worked on or factually inaccurate, and it doesn’t help that he is arguing against updating to new kernels and versions of Xorg,” Mack asserted. “Where does he think driver problems get fixed?”
‘That 10-Year-Old Myth’
Package management “isn’t an issue,” Mack added, and “I really wish he hadn’t repeated that 10-year-old myth that distro makers need to make packages for each Linux distro.”
In fact, “application makers can just come up with distro-agnostic auto-install packages of their own, and the Linux distros can repackage as needed,” he explained. “That’s what the Linux standards base is for.”
Otherwise, “they can create packages for the larger distros — Red Hat and Debian for server, or Red Hat, Suse and Ubuntu for the desktop — and get 90 percent of Linux installs that way,” Mack added.
“I think we all want commercial-quality apps, and it seems like we all want other people to pay for their development,” he charged.
“Either as time goes on, Linux will end up better-supported or it won’t; it’s all down to business sense,” Mack concluded. “In the meantime, we have OSDL and the distros spending real money paying for development of things they see lacking.”
‘Valid, but Not New’
Many viewed the matter differently, however.
To wit: Does Linux suck? “Yes,” Slashdot blogger yagu told LinuxInsider via email.
“So does XP, Vista, (your favorite Windows version here), Mac OSX, and just about any other system. Systems really suck at things they don’t do well,” he said.
“I think for Linux it becomes a real problem when users are on their own fixing their computer,” yagu added. “Linux elite always say, ‘RTFM’ (read the fine manual) and learn something in the meantime!’ Problem is, real people want to do lots of things, and learning how to install new Linux kernels tends to not be one of them.”
Lunduke’s points “are valid, but not new,” yagu added. “Yeah, we fight to get wireless and sound working, and we learn more in the process. And in more mainstream distributions, we roll out pretty non-sucky Linux (most notably, Ubuntu).”
‘An Excellent Presentation’
It’s important that the Linux technical elite “continue the work and fix the Linux issues, and it’s reassuring to see they’re still working toward that goal,” yagu said. “And Bryan gives an excellent presentation … anyone serious about understanding work to be done around Linux needs to take a look at his video.”
It’s all part of the “getting Linux ready for the desktop” movement, yagu asserted. “It’s an unforgiving demographic: Users need their sound and video to work, and they need their Photoshop work-alike tools.”
In the meantime, however, “there are a lot of things Linux continues to *not* suck at,” he added:
- “infrastructure for the greatest search engine in the world (Google)
- embedded devices (TomTom)
- PVR/DVRs (TiVo)
- smartphones (Android)”
No Single ‘Desktop’
“That was an interesting presentation,” Chris Travers, a Slashdot blogger who works on the LedgerSMB project, told LinuxInsider. “However, I think that the author’s presentation might be missing the immediate focus at the moment.”
Specifically, “the big problem is that there isn’t a single homogeneous environment called ‘the desktop,'” Travers asserted. “Instead, we see really three fundamentally different markets in these areas: the corporate desktop, the technical workstation, and the consumer desktop.”
Lunduke’s presentation “focused exclusively on consumer systems, although some of the problems were consistent across the desktops,” Travers noted.
In many cases, corporate desktops are viable today with Linux, as are technical workstations — with some exceptions, he argued.
“The consumer desktop, however, is quite a different environment, and while I have no doubt that Linux will get there, I don’t think the current distros can get us there,” he said Travers. “For Linux to succeed in this market, it will be necessary for hardware vendors to come together and collaboratively build a distro for their environments. The availability of Linux on netbooks may be a sign that these developments are on the horizon.”
‘Many of the Problems Are Not Real’
“Linux Rocks!” exclaimed blogger Robert Pogson.
“The presentation raises many concerns that should be addressed, but many of them are niches and not really holding back GNU/Linux on the desktop,” he told LinuxInsider.
Rather, “most of the issues that affect a consumer who buys a PC with GNU/Linux are minimal,” Pogson added. “The OEM and the distro have done the configuration. If the consumer sticks with supported peripherals, they have no problem.”
GNU/Linux has reached its present level of popularity “by being better than that other OS at a lot of things” — such as stability, malware and performance — “but it is still evolving,” concluded Pogson. “The consumer does not need to change his software just because there is a new release, so many of the problems mentioned are not real.”
‘A Bit of a Power Struggle’
What should be done to “fix” Linux?
“It really depends on which side of the Linux divide you are on,” Slashdot blogger hairyfeet told LinuxInsider. “There seems to be a bit of a power struggle going on between the hacker types, who like Linux being heavy on CLI and like things the way they are, and the usability types, who realize that to gain market share Linux will have to move away from being so CLI-heavy, as have OS X and Windows.”
Speaking from “the usability side of the fence,” hairyfeet ticked off a few requirements:
- 1. Easy privilege management. “There needs to be a simple, GUI-based way to elevate privileges, similar to the ‘Runas’ that you have in Windows,” he said. “Also, if you can’t actually use it, don’t allow it!”
2. A stable framework to target. “There needs to be a stable core framework that is supported across ALL distros and that anyone can write their drivers and apps to,” hairyfeet explained. “Companies like to write once and forget about it. After all, who wants to support last year’s widget when you need to be writing drivers for this cycle?”
3. An NDIS style wrapper for all-in-one printers and webcams — “Lexmark in particular,” he said. “Ask any Wal-Mart which are their No. 1 sellers in printers and all-in-ones, and you will have more Lexmark in the top 5 than the other companies combined. Yet they still have virtually zero support in Linux. Since these ‘Winprinters’ are simply calling on the Windows GDI to do the heavy lifting, there shouldn’t be any reason why a wrapper style driver couldn’t be written. If the Winmodem could be defeated, so can the Winprinter.”
4. The death of CLI. “For the home user, the CLI is simply unacceptable in ANY form,” he asserted. “It is too foreign, requires the memorization of too many strange Unix commands, and is frankly simply too powerful. It is a lot harder to break something with checkboxes and radio buttons than it is with a mistyped CLI command.”
‘5 percent Think I’m Wrong’
Finally, with such a diversity of opinions on the topic, we couldn’t resist checking in with Lunduke himself to get his own impressions of the reactions.
Some 30,000 people watched the video in the first 24 hours, Lunduke said, “which is funny, as there were only 65 people at LinuxFest who saw it in person” since the room was only supposed to hold 40.
Aside from all the comments in forums and blogs, Lunduke had also received 100 or so emails by last Friday.
Breaking down the responses into percentages, Lunduke offered this:
- “5 percent think I’m wrong and hate freedom and want to kill puppies;
- 25 percent think I have a point, but am presenting the wrong solutions;
- 30 percent think I am partially right;
- 35 percent are happy [that] people are saying the things that were in the video;
- 5 percent think I may have a point, but they’re unhappy that I wasn’t kind to a particular project they like or are involved with.
‘A Large Amount of Support’
“The response has been, surprisingly, fairly positive,” Lunduke told LinuxInsider. “Sure, there has been the normal amount of angry and upset reaction.
“But — and I am thrilled about this — people have been voicing a large amount of support, and it seems to have sparked off a great deal of constructive discussions about how to solve some of these issues in a practical way,” he added. “I think it has resonated with a lot of people who have been noticing the same things.”
For some reason simply trying to lay out obvious issues that don’t even get to the level of the OS nuts and bolts in actual use and are more of a marketing dynamic brings down a rain of napalm on your head.
Something like "most consumers who want the OS to work like an appliance, will take a look at the CLI and walk away. You can talk about how the CLI isn’t needed, but most users will do a Google search to find a solution because that’s what Windows users do. And they will see thousands of hits with CLI instructions. I can’t tell you how often a Linux noob has showed me a pooched install they screwed up from busted dependencies and were totally unaware of repositories, the GUI front end for APT or Synaptics, and typed in some old CLI terminal commands or .deb or .rpm package."
I’ve had people ask me how much Micro$oft was paying me and worse from a comment like that. Or God help you if you point out OSX tripled their marketshare with a GUI front end on a BSD bedrock and a couple of grand cost of entry during the Vista disaster, while Linux barely moved the needle.
The sad part is being 50+ with 30 years of IT as the way I’ve payed my bills and a solid background in ‘nixes, mostly doing training and instruction, and getting blasted by a kid who just installed Ubuntu and has a bad case of Microsoft derangement syndrome.
You only get the mindless blasting in the consumer desktop discussions. Server or embedded threads have a tiny amount of this.
"while Linux barely moved the needle."
Not at all the case.
In April, 2009, NetApplications showed GNU/Linux at 1.02%
In April, 2006, they showed 0.38%. In that period PCs increased in number from about 1000 million to 1300 million, a 30% increase. GNU/Linux showed a 268% increase in share and about 350% increase in numbers. By any measure, GNU/Linux is soaring.
If 1% were true, would M$ be making layoffs? Nope. The number is closer to 10% in the real world, not NetApplications’ world.
The reasons for GNU/Linux success are obvious:
– no one in their right mind will upgrade a 5 year old PC with XP to Vista but they will go to GNU/Linux and see a good improvement.
– the old PCs make decent thin clients with GNU/Linux, too
– GNU/Linux has a good chunk of netbooks
– Vista sucks on lots of PCs especially the first batch when it was released
– 7 is Vista II. Few want it.
– even a 10 year old PC that shows 1024×768 is still useful with GNU/Linux. Guess what? Folks are tired of chucking 3-4 year old PCs just because that other OS does not work on them.