dcsimg

The Importance of Fitting In

Is the success of Linux directly proportional to its ability to integrate with existing proprietary systems like Windows? If so, should free software developers be spending more time integrating with it instead of building better software for free platforms?

Last week we took a look at proprietary software and how its integration might affect the Linux desktop. Is it good, bad, or somewhere in between?

There are several reasons why proprietary software might be useful to the Linux desktop. People are used to the way it works, it could support proprietary data formats better, it might simply work better.

Why is this an issue? Because if we want users to make the move to free software like the Linux desktop, then we should try and make it as easy as possible.

Linux is gaining in popularity every day. So is solid, open software its strength, or the ability to integrate into the existing Windows monopoly? Perhaps it’s a bit of both.

The importance of fitting in

This whole issue raises a valid point about Linux and free software in general. Should we try harder to fit into the Windows world?

On the one hand, we create free software to scratch an itch. We use it because we like it, not because we’re trying to take over the world (that will be a “completely unintentional side effect,” remember?).

That quote from Linus suggests that displacing Microsoft and gaining market share should not be a primary focus of open source software, although of course he doesn’t speak for everyone. Instead, we should focus on creating great software. As a result of great software and a better way of doing things, users will naturally migrate to Linux over time. However, does that mean that we should ignore the integration issue altogether?

Truly at the end of the day, the Linux desktop doesn’t have to do anything that its developers don’t want it to do. So your games don’t run on Linux. Too bad, it’s not Windows. If you want Windows, then run Windows! If you want a different, reliable operating system, then use Linux – you just might have to leave some of your software behind. We aren’t locked into a proprietary model and don’t have to be.

On the other hand, if Linux can inter-operate with Windows, then it will gain acceptance and adoption at a much faster rate. Like it or not, the world is mostly locked into proprietary software. Take Microsoft Office, for example. How can we expect businesses to switch to free software if they can’t read and write Microsoft formats? How can we expect users to switch to Linux if they cannot connect their iPods and listen to their music?

Various major computer manufacturers have started selling computers with Linux as an option, instead of Windows. If users who purchase these machines cannot use their software, can’t access their data and cannot use their devices, then does Linux really win?

Linux is king of the Internet and has conquered the super computer realm, yet it has hardly made a dent on the desktop. So perhaps in this arena fitting in is important, if we want to make a dent on the desktop market any time soon.

Kicking the habit

Some degree of interoperability is useful, but at what point does it become dangerous? Should Linux be able to play every file format people might use, should it synchronise with every mobile device, media player and palm pilot out there? Should it just work?

Today, we understand how much damage the prevailing vendor lock-in model has done to the computer industry. We know that proprietary data formats are bad for interoperability and ultimately hurt end users. Security expert Bruce Schneier agrees, writing:

“With enough lock-in, a company can protect its market share even as it reduces customer service, raises prices, refuses to innovate and otherwise abuses its customer base. It should be no surprise that this sounds like pretty much every experience you’ve had with IT companies: Once the industry discovered lock-in, everyone started figuring out how to get as much of it as they can.”

How then do we balance the support of the proprietary model with helping end users enjoy free computing. Is it even our call?

Perhaps we shouldn’t be dictating what formats people use at all and just let them make their own choices. Then by the same token, they shouldn’t be dictating what data formats we support. Should we be working to support patent encumbered proprietary data formats to make life easier for those who choose (or didn’t choose) to use them? Or should we aim for liberation through imitation? How many users, after successfully migrating to Linux, converted their entire music collection to a free file format such as Ogg Vorbis? Probably not a whole lot (after all, Linux can play MP3s so why bother, right?).

Personally, whether the Linux desktop integrates perfectly with the Windows world doesn’t make much difference to me. I only use free data formats and avoid products which would lock me into one particular vendor (like an iPhone). I realise however, that this is a luxury I have because I made a choice to be that way. I do recognise however, that it might make a big difference to others.

Where would Linux and free software be today, had we shunned existing markets and completely done our own thing? If Samba didn’t exist because Andrew Tridgell decided that he didn’t care about interoperability with Windows, would Linux have such a great market share on servers and network attached storage? Where will we be tomorrow if we don’t continue to integrate ourselves today?

The Samba project is a useful case study on the benefits of integration. It was written to scratch an itch, that is, to get Unix computers talking with Windows machines over a network. That’s a prime example of free software at work, the result of which is excellent communication between Unix and Windows, and indeed a robust replacement for Windows servers. In line with those words from Linus, Samba didn’t set out to destroy the Windows file server market, that would be a completely unintentional side effect. Then again, the very reason that Samba is successful is because it does integrate.

The Mono project has brought support for Microsoft’s .NET framework (which they are using to try to kill Java and dominate the programming space) to Linux. As a result, we are now seeing more free software applications written in various .NET languages like C#. How do we reconcile using and integrating proprietary platforms with pushing freedom from vendor lock-in? How do we make sure that we don’t shoot ourselves in the foot, long term?

How much should free and open source development focus on integration with proprietary software and platforms? Should Linux distros try harder to integrate into the world of Windows and how much should it do its own thing? On one level, free software must continue to follow its own path – scratch its own itches and innovate in its own right. On a higher level, if it can integrate with the proprietary world, then it will be much more attractive to those coming from that space.

Looking down from on high

Free software has come this far by just being itself. We use it because we like it. Should we be focusing more on integrating with Windows in order to increase market share, or should we just let nature take its course and those who want to switch will? In the end, a majority will “see the light.”

Then again, it’s easy to sit in our current position and ponder whether we should aim for greater integration with platforms like Windows. In truth, we already integrate quite well. We take for granted the fact that we buy a memory stick can actually access the data. Likewise, the ability to write to NTFS volumes, open Word documents, browse networked Windows shares, use calendars on Exchange and even authenticate to Active Directory.

Perhaps it’s time we took a page out of Microsoft’s book and started to “embrace, extend and extinguish” proprietary data formats and software. Integrate ourselves with them, and once we have the power, break them. Aside from the fact that this approach is horrible, nasty and completely unethical, the danger is whether in the end we would actually switch to totally free solutions, or continue to use those old familiar encumbered ones. We might have the market share, but then we’d have a whole new battle on our hands.

One thing is certain, however. Integration with existing systems has helped to rocket the popularity of Linux and make it much more attractive to business, education, government and even end users. If you’re a developer of free software, keep doing what you’re doing. Make great products and the users will come. Consider integrating with existing systems to make your software more attractive, but just don’t lose yourself along the way.

Comments on "The Importance of Fitting In"

tjrob

Interoperating with Windows is a non-issue. The REAL reason Linux is not used very much on desktops is the fact that it is much more difficult to administer a Linux desktop than either Windows or Mac OS. For instance, the latter two OSs just work on a laptop; not so for Linux — it takes hard-core Linux expertise to get Linux working on many/most laptops. While experts can do that, it\’s hopeless to expect most users to figure out the arcana of ACPI, ndiswrapper, and finding codecs for movie and MP3 players. Unfortunately, most of these problems cannot be solved by technical means.

Reply
k_short@lanl.gov

Integration is a slippery slope. Once you get started down that path, we\’d end up trying to be everything for everyone. That simply isn\’t achievable, nor would it be good for Linux.

It\’s also easy for us to sit and ask these questions, being more knowledgeable than probably 80% of the computer user base in the world. However, when you take it at the level of \’Ol Lady Parker sitting down to a Linux box to check out her grandkids on the family website, I\’m just not sure I can see her researching, downloading, and installing plugins and codecs to get the job done. That\’s what the desktop realm is all about, and frankly, I could care less if Linux is easy for Mrs. Parker or not. I like it, I know how to use it, and I can make it work for me. If its the business world, there will be people paid to make it easy for everyone to get on board with Linux and smooth out the bumps before everyone runs across them. Not so for \’Ol Lady Parker, but if she really wants to use Linux, she\’s gonna have to learn up like the rest of us. Is it our job to take away the flexibility and complexity that we love so much just to get Grandma on board?

Reply
tomaz

Does Linux even want to be used by the kind of people who are content to use Windows, the iPhone, and, increasingly Mac OS? I think Linux serves well a different market niche, i.e. technically literate users. What happened to NextStep/OpenStep after it was acquired by Apple and Apple focussed on compatibility with Windows and friendliness to non-technical users is a case in point. It has (from a technical user\’s perspective) become completely unsatisfactory and unusable. There is a market and a niche for every kind of product, there is more than enough offering for consumers, technical users need a technical platform of our own, which meets our requirements and is not contaminated by (entirely different) ideas, approaches and compromises which need to be made to keep the non-technical consumer happy.

Reply
csmart

@tjrob, so are you saying that we should ignore integration and just build a better product? The reason that laptops don\’t always work well is because manufacturers build them specifically for Windows. OS X works on custom hardware only which Apple controls. By making sure that Linux runs on these \”designed for Windows\” laptops, aren\’t we integrating anyway?

@k_short, I agree with you that integration is a slippery slope. One concern is that making everything \”easier\” will make Linux more bloated and buggy. However, so long as that doesn\’t become the norm and I can still have my technical distribution without the junk that\’s fine for me. A bloated, \”easier to use\” distro might have its place, so long as it doesn\’t affect all the other distros too.

@tomaz, that\’s _the_ question. Do we care about market share, or are we just in the business of making great software. I think it\’s the latter and I believe that because of it, Linux market share will _eventually_ rise. I disagree with the idea of deliberately going out to increase Linux market share, because I think that you just end up introducing junk which clogs up Linux. I guess at its heart, Linux is still a hacker OS, but that doesn\’t mean that it can\’t make it work.

At the end of the day, so long as the system remains open source, we are free to build both \”easier\” and more technical distros. That way we can have best of both worlds, right? Or can we..

-c

Reply
hesperidae

\”Interoperating with Windows is a non-issue. The REAL reason Linux is not used very much on desktops is the fact that it is much more difficult to administer a Linux desktop than either Windows or Mac OS.\”

Funny, but in my experience, the exact opposite is true. I set Granny types up with Linux all the time and you\’d be hard-pressed to pry it from thier hands. I\’ve also yet to run into a machine (that I build for a client for Linux install) that didn\’t load appropriate drivers for all hardware at first boot. While I do have \’hard-core Linux expertise\’ I rarely get the opportunity to use it because I do my homework before spending my clients\’ money to ensure that all works as intended. This shaves HOURS off an install as compared to any version of Windows. (I can\’t speak for MacOS). The drivers issues and ndiswrapper crap has either been un-necessary or automated for a couple of years at least… That is if you spend a few minutes on the interet researching your purchase. I for one evaluate everything thoroughly before spending my or anyone else\’s money. This does not just apply to computers.

Reply
dbayer

@tjrob, interoperating with Windows is exactly the issue. From the standpoint of a desktop system, Linux is not any more difficult to administer than Windows or MacOS, just different. In the enterprise, you will rarely find anyone holding purse strings willing to replace the entire client/server infrastructure with something they aren\’t familiar with. So long as Linux (and MacOS too) can happily work alongside it\’s Windows counterparts and access the existing Windows servers, there is the possibility of replacing Windows with \”something better.\” The enterprise must also factor in training costs, not just for the admins, but for the users as well. There are Linux apps available to fill the space of most Windows apps, but there is a very large percentage of users who get frustrated or confused when they don\’t \”just know\” how to do what they need. Training would be required for many users to move even to something as easy to use as OpenOffice.

These days, the same thing applies to many home users. If I want to run Linux but still want to share files with my wife\’s Windows computer, they need to talk to each other.

I am not advocating the pursuit of market share. I am advocating the pursuit of \”something better.\” Linux and Linux apps have matured a great deal as time has passed. That is due in part to increasing popularity and visibility. At the risk of getting flamed, I see more buggy Linux apps than I do Windows apps. The number of buggy apps has decreased as the programmer base has grown, and that only happens through visibility, curiosity, and adoption.

Reply
ruilapa

Happy Windows user is a dumb user!

I\’m currently replacing Windows with Linux on an insurance company. Easy? Not really! Here\’s why it is being done.

Windows administrators don\’t care about security. Some try to get there… but they are still ages away from encrypted filesystems, top notch firewall, stack protection, True functional anti virus, worms and fingerprinting and log analysis.
They disaggree obvious! But how do they know that their windows network driver is not hackable?
Scripts? Automated sentrys? SMART monitoring? Realtime firewall logs? IDS?
Un-Differentiated routing? Static Arp per network? IP stack optimization? Kernel scheduler?

Its all standard… unfortunately a standard slow crap of a biological nightmare
. Their solution… Destroy and Clone, centralized filesystems, no laptop backups, Roaming Profiles, destroying every machine the user logs in, carrying Desktop nightmare files like PST\’s, crappying IO, Bandwidth, etcetera

IMHO Windows networks are just installed and deployed! Active Directory controls permissions, groups, users, VPNs… Its a centrallized nightmare of outdated info, crappy records, garbage info available to everyone who wants to look at! No real security is applied and no real logging is ever looked at.

Print Servers are another nice place to steal info… they are point and click… Raw senders. Protected printing? Authenticated printing? Compressed transmissions? What is that? They are the suckers of bandwidth, unsecured and used by ghost campaigners against Corporate Administration.

File Servers… Have you tried to increase an NTFS volume? FEAR! Linux as volume managers for long time now. Windows rests on hardware solutions. Either that or

VmWare! There are more Windows licenses on VMware than on real machines! Virtual Desktops, Virtual Appliances, Development Servers, Testing and never deleted Servers. You name it… They clone it!

Do I think Linux solutions can do better? Without a doubt!
It\’s it easy? NO!
Why?
1. You know more than they ever WILL!
2. Your stealing their jobs replacing them with automations.
3. They have top notch databases… they like to call Excel files.
4. M$ Access, Project? First think if there should even be alternatives.
5. Linux uses a lot of hardware… specially if you make your structure full redundant… which they never needed until Linux came.

But Linux is not easy! After spending 1 month customizing 1 pretty darn secure and fast laptop, it hurts basically due to:
- Centralized and OFFLINE authentication mechanisms
- Hibernation and suspend
- lm_sensors
- Centrallized VPN routing control
- Multi Display Support (projectors, 2nd screen, …)
- Native Extended laptop Keymap support
- Lack of integration and full compliance with the desktop with things like Adobe Reader, Flash
- Codecs
- Flash sound
- F***** \”exec\” on USB plugged devices! ;)
- SSH Client depends on SSH Server
- ALSA Volume 0 and Mute
- Remote Visual Control and Helpdesk with Vino crapped up idea of authentication without and SSH tunnel
- And the thousands applications that do not even exist on Linux

Do I love Linux? Still using it for 14 years… and still loving it!

Reply
tomaz

@csmart: \”Do we care about market share, or are we just in the business of making great software. I think it\’s the latter and I believe that because of it, Linux market share will _eventually_ rise. I disagree with the idea of deliberately going out to increase Linux market share, because I think that you just end up introducing junk which clogs up Linux.\”

Absolutely, I agree.

With one possible proviso: although the market share will eventually rise for the reasons you give, I don\’t think the ultimately achievable market share is huge nevertheless. Most users simply don\’t have the capacity, the will to learn, or any interest, in using a quality, properly designed, powerful O.S./computing platform. The average user likes to get pissed every Friday night, watches football all weekend, and loves being able to control his iPhone and his Apple TV with a couple of simple grunts and his two thumbs. This sort of user will never have any interest in a quality technical platform – and the vast majority of the market is like this. Microsoft and Apple cater for these markets very well – and increasingly they neglect all other markets.

No operating system will – in my opinion – ever be able to cater well to both this market and the technical user market. And the technical user market will always be (relatively) small, and probably getting smaller with decreasing standards of education worldwide.

\”I agree with you that integration is a slippery slope. One concern is that making everything \’easier\’ will make Linux more bloated and buggy. However, so long as that doesn\’t become the norm and I can still have my technical distribution without the junk that\’s fine for me. A bloated, \’easier to use\’ distro might have its place, so long as it doesn\’t affect all the other distros too.\”

I don\’t think in the long run this is possible. The bloated distro ultimately impacts the mainstream distros, who have to start making compromises to accomodate the bloatware distro, and before long, dealing with the ensuing issues becomes the centre of development effort.

I think NextStep / Apple Mac OS X is a fairly good case study of this phenomenon. By way of example, there is a whole raft of technologies Apple has to support in the mainstream O.S. to support what started out as a \”legacy\” iTunes application (I believe one of the last Apple apps still to be written using Carbon).

Apple being able to support Spotlight (a great feature for the average user, but a total nightmare for a serious user in my opinion) requires modifying the kernel in an essential way. You can\’t have Darwin without the kernel hooks necessary to support Spotlight.

\”so long as the system remains open source, we are free to build both \’easier\’ and more technical distros. That way we can have best of both worlds, right? Or can we..\”

Well, I think you can already tell my opinion is we cannot.

@dbayer: \”I see more buggy Linux apps than I do Windows apps. The number of buggy apps has decreased as the programmer base has grown …\”

I\’m not sure I agree with the latter statement. My own experience is that as the programmer base grows, the quality of the programmer base inevitably falls.

There are only so many people good at anything, be it software development, design, programming. As you expand the base, you increasingly end up attracting more and more of less and less good people.

The biggest \”programmer\” base is probably in web \”programming\” … which results in the lowest quality \”programs\” around.

Microsoft employs more programmers than does Apple than did many existing and former high quality software companies like DEC, NeXT, Sun – and the quality of the software they make is inversely proportional to the number of people/designers/programmers they employ.

Reply
andrewwebsdale

k_short – You COULD care less, which means the opposite of what you\’re trying to say. I hope you don\’t get booleans mixed up in the same way!
tjrob – if you\’re not a shill you sound like one. Where did you get all this codec rubbish from? Use a (cross-platform) app like VLC (VideoLanClient). This will even play DVDs without CSS – I\’ve never had to look for a codec for linux – EVER.
The ONLY problem Linux has & has ever had is the MS \”monopoly\” which means that hardware compatibility is always an issue, especially with laptops. If there was a level playing field, MS would become a minority player in the general purpose computer market.

Reply
tomkersey

I run ubuntu on about 5 machines in a network. some are desktops, some are laptops.
they all work well and do everything i want to do on a computer. I have a fine recording studio running ubuntu-studio. the graphics and office apps are fine. I am publishing a book with Open Office and Scribus. ubuntu plays music and videos in every format i\’ve found so far. I write music using the free composition software. why would i ever want it act like windows? Ridiculous notion.(sic?)
(don\’t know if i spelled that right.)

Reply
tomkersey

Oh I forgot to mention that all the machines in my network are a different hardware architecture. We don\’t need no stinking proprietary hardware!

Reply

We stumbled over here different web address and thought I might check things out. I like what I see so now i am following you. Look forward to going over your web page yet again.

Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>