Standardization in the libs supplied with the basic distributions to allow a comun base for installing new programs, use of yhe embedded approch and a lot of flash memory to make the switch On/Off of the PC pratical for every day application and reduction of the passwords required are all needed to make Linux acceptable to the masses.

Naah, I reckon the AJAX vendors are fighting a losing game. Check out http://www.cosmopod.com. They've got the right plan.

Guys,

First: in what way is OOo insufficient?

Second: There are solutions for accounting on Linux in the SME market - SQL Ledger immediately comes to mind but Quasar Acounting is a good fit for any business - from a small one-man piano tuning outfit to a 250 store retail concern.

The basic version of Quasar is GPL, with graphical clients for Linux and Windows and server parts for Linux. It contains all functionality except point of sale, which is provided by a commercial verion.

http://www.linuxcanada.com

Not an employee, just a satisfied user.

Good article! I'm a consulting engineer that has waited since I installed Red Hat 5.0 for a Linux desktop. I dunno how you can convince folks like Adobe and AutoDesk to port their software, but until they do, this open-minded, technically literate user won't be asking for a Linux desktop.

Regarding your second question about usability, Linux assumes that every user can function as a sysadmin. That's even the basis of the documentation. I can't afford to use a lot of non-billable time configuring and maintaining a desktop. When I can maintain my own Linux desktop, without support, and not lose productivity, I'll reconsider.

Proprietary ISVs can't "port to Linux".

Linux has an incredibly stable userspace/kernel API and ABI. But everything in userspace is up to the distribution. The way that proprietary ISVs work depends on porting to one distribution.

Even if all the distributions agree on support for a binary standard, LSB, no ISV in his or her right mind is going to support an application on "Joe's Random Linux that Happens to Pass LSB Today".

Expensive apps, those for which it had been worthwhile to do a UNIX port, can port to Red Hat and/or Novell. EDA and DCC apps are already doing this.

Less expensive apps are going to have to make way for open-source replacements that get built and integrated by the distribution, or go browser-based, or both. (I can see a "free financials" package that lets you subscribe to proprietary tax calculation service engines, for example.)

I've been running Linux on the desktop since 1995, and moved everything over in 1996. And I think that in the long run binary diversity is a good thing, since monoculture is a security risk and the ability to rebuild allows for innovation in areas from power management to global automated testing.

Good points Don.

I think for clarification, what I meant by "port to Linux" is that I want a major distro to be supported. It would be great if that distro was one that already had a large base of desktop users.

Financial support from the companies that are making their living off of Linux (IBM, Red Hat, Novell, etc.) would certainly be a boon for game developers, especially mainstream studios who simply cannot devote precious time and resources for niche version of their games. It's hard enough just to make a splash (necessary to pay bills and keep the studio going) on Windows and consoles let alone Macs or Linux boxes.

I work for a large studio nad have a great rapport with those who operate it. I could likely forge a deal to port some of their games to Linux had I the financial might to set-up a proper studio (which would run on open source software whenever possible).

Unfortunately, most of the companies I mentioned above don't care about games, they care about corporate desktops used for business. That the potential for innovation useful to their businesses is likely to result from Linux game development (such as contributions to various open source projects that help the dev team do their work more efficiently) appears to be something they have no intention of investing in.

Either its an unknown quantity full of risk or they simply don't care. Linux NEEDS games. Games will drive innovation like gangbusters. But the sad truth (at this moment in time) is that I could be sitting on the greatest game franchise in the history of games and it wouldn't do any of us one darn bit of good... no money = no deals = no development.

Distributions should leverage Linux's strength in being customisable by being perfect for a specific area of application. For example an office desktop with less emphasis on games and multi-media. An consumer desktop with emphasis on games and multi-media. A consumer appliance desktop as for the consumer desktop but for use as a no maintenance appliance. A kiosk/call center/school/instutional type distro which is stable and securely locked down so as to be tamper proof by the user.

People will change if as is easily possible, Linux offers a perfect fit for an application, particularly if maintenance and support overhead is reduced by increasing security, stability and freedom from viruses, worms and spyware. The fact that it may be more difficult to install common programs such as spyware, viruses, and for example media players for accessing porn or music at work or at school, can be sold as an advantage rather than a disadvantage.

For lots of users linux needs to support colour profiles.
Photoshop and so on cannot function without that and therefore a lot of people cannot move.
Why don't I ever read solutions for that?

The mistake often made is that if there were apps for Linux, they'd be on the shelves at the local office suppliers.
As one who has used Linux as my sole platform for private and business use over near a decade, I haven't experienced the lack of applications that I needed to interoperate in a sea of Windows. Critical stuff like Lotus Notes under wine, Citrix Linux client, Cisco VPN client for Linux, StarOffice/OpenOffice.org for all the stuff including teaching using powerpoint presentations, X3270 connections to Mainframes, build/admin/troubleshoot and fix large SPARC servers (Sun up to E10K and Fujitsu Primepower up to 2500 systems), that was a laptop with only Linux installed, free of viruses and malware, more efficient and versatile than any Windows setup other colleagues used. It was the consumate desktop. The only problems encountered was getting basic information from people who ran the Windows servers, they only knew how to point and click without any real understanding of what they were doing, so were of little or no help.
Imagine, you install Lotus Notes under wine and you have a problem, they give you server IP addresses, your ID and password, but couldn't help further. Next you discover they have escalated the problem to their Director and he says if you need to go back to accessing Notes via a browser, it would incur a cost which would have to be sanctioned by your Director. Instantly POLITICAL by asking a simple question. With delight, you figure it out yourself in 10 minutes and tell them what the problem was, I was able to politically tell them they knew very little of Lotus Notes, it was a very simple problem they should have been able to answer directly.
I also escaped the annual software audits and saved the company license fees which were seriously squeezing our bottom line, but were falsely considered to be necessary for us to conduct business.
Linux can do it, it will just take time to turn the herd that is in full gallop.

The solutions are out there, they are apparently well-kept secrets.

Appgen software (makers of MyBooks Pro) offers total application source to Appgen 4GL developers. And MyBooks is so inexpensive that you could justify the price paid as going to initial support and call the software free.

cpaasp.com (e-accounting) provides Internet-based hosting services, giving ANY platform user the ability to run their Windows platform apps. In this model - the desktop becomes irrelevant and the user can select what they want.

Server-based computing models have demonstrated their value in the past, and it's obvious that this is a trend that is facilitated with the Internet and Web technologies. The pendulum begins to swing back...

m

People believe that there is no choice other than windows, and even if they were aware of Linux, they dont want to spend their time and effort re-learning how to learn their computer. There is also a real perception that 'free' software must be poor quality / unsellable software.

We have a local wireless group that gives free computers (500Mhz, 256Mb,10GB etc)running Mepis Linux to the community. They set them up on their wireless network without charge and let them surf unrestricted also without charge. They have been operating for 18 months and, in a densely populate European city have only managed to give away 50 of these machines. Many of the recipients of these computers have subsequently installed Window$, knocking them off the network.
The reason that they have uninstalled Linux is always the same, they know how to complete the jobs they want to do with M$ and can't be bothered to learn new skills under Linux.

I've used both windows and Linux(and Unix) since their beginnings, but I do most of my computing in windows for two reasons:

1) Several applications that I use are not available for Linux and the "equivalent" isn't anywhere near being equivalent, let alone compelling enough to get me to convert packages!

2) I see no compelling reason to go through the headaches of moving completely to Linux when I have all the applications that I need (including windows ports of open source packages) to run at my fingertips.

It is my opinion that you win the war by winning over developers. Microsoft certainly understood that, especially in the early days of Windows. They still focus a huge amount of their resources and energy on developers (we have all seen the Steve Balmer "developers, developers, developers" dance).

If Linux can win over ISVs by making it easy to develop applications that run on many distros we will start to see a healthy ecosystem around Linux.

I don't think the big ISVs with existing franchises on Windows are going to be the first to move. The first to come are going to be new/small ISVs looking for a nich opportunity. The big ISVs will follow once the ecosystem has started to gain just a little momentum.

One problem I see is that large portions of the open source movement have philosophies that are antithetical to small companies making money off of the Linux platform with software. There are only three possible outcomes from this that I can see. No ecosystem will ever develope and open source will continue the way it is. Delivering decent solutions to a highly technical skilled crowd and reverse engineering Windows applications that will always be a step behind. Another outcome would be that rich client software dies altogether and web based applications totally dominate making the OS insignificant. The third and the one I am hoping for is that there is a small change in the world of Linux and open source software that recognizes the need for a rich ecosystem around the platform and makes changes to encourage that ecosystem. What I saw at the Desktop Architects meeting was a start.