Trapped on Windows
I was recently approached by a friend of my mothers who was absolutely done with Microsoft, she just couldn't stand Windows any longer. I didn't ask what the specific issues were, but imagine they were the standard problems.
She had heard about Mac, but the problem was that she used some software for her business which was not supported on Mac, which they used to generate labels for products they sold. She was going to check if the software was supported on Ubuntu. I'm not holding my breath.
This is sort of a fascinating niche... people who desperately WANT to change, and most likely would be totally happy with Ubuntu, but they're trapped on windows, locked in by a simple business app.
I don't think this is a very small niche, either. Think of the many small companies built around some pencil ordering app, or that use some wonky windows app to operate their machine tools, or etc.
I suppose every one of us has run up against someone in this niche. Uncle Harry at last year's Christmas party, or that guy you sat next to on the train, or your best friend from college.
This niche seems much like the gamer niche, stuck in a chicken-and-egg situation. The app is developed for Windows because that's what people use. People stay on Windows because that's what the app was developed to run on.
How do we break that cycle and enable these people to gain their freedom?
I wonder if the solution might not lie in setting up a system whereby SMB's can donate towards a bounty on specific port/app functionality, having some kind of feedback/voting system for the release of the bounty to the appropriate devs.
Just a thought...
I wonder if the solution might not lie in setting up a system whereby SMB's can donate towards a bounty...
I've written before about my experience with bounties in FOSS, and remain skeptical that it is something we can rely on. There have been some notable successes (Blender's open sourcing for instance, but in that case it was more of a ransom model than a bounty model), so it may be suitable in certain situations.
I suppose a flowchart would work best for situations like this:
Does the program work under wine? Yes - DONE.
No - Run the program in a VM such as Virtual Box.
The later one requires a more difficult setup and use because then you have to deal with transferring the file back into Ubuntu, so it's not always an ideal solution. I wish there were a better solution, but I can't think of any. I know a few years ago I remember seeing (I think Xen) a program that made Windows programs run as if they were native in Ubuntu, but were being run through Windows. It worked liked Parallels for Mac, but I'm not sure what happened to it. So to give you an example, click on the 3rd link "Switch. Run Windows." at http://media.parallels.com/video/STM/demo/us/ and you'll see how they address issues like this.
My thoughts are that we (Ubuntu) should band together and take a holistic approach. Fragmentation/complexity is the enemy of small business, so how about a suite of tools that every small business owner needs, designed in a way that supports workflow? Large(r) companies already have things like SAP to manage their businesses, so there is some precedent (and success) with this integrated approach.
How about a new project in Launchpad with some use cases for small business? If each small business task is implemented as a module (e.g. Labels, Invoices, ...) and all those modules fit together nicely, then we have something "the competition" doesn't.
Comments?
how about a suite of tools that every small business owner needs
This is an excellent idea. Basically, don't worry about the *really* niche stuff, but cover the things that 80% of business owners would need. Invoicing, contact management, label printing, inventory, yada yada, and integrate it into Ubuntu very cleanly.
Maybe a lot of these tools already exist out there, just not installed on ubuntu by default and kind of hard to locate. If this is true, it suggests perhaps it would make sense to package it as a derivative, like Mythbuntu or Ubuntu Studio.
What SMB-oriented applications beyond the usual Ubuntu set would make sense to have integrated in something like this?
I have a thought - why not look at this as an open-source challenge rather than a Linux problem? We all know that when the software is FOSS the Windows v. Linux problem generally disappears. Remember, the reason (most) of us got into this was freedom of choice.
I say this thinking specifically of a piece of hardware I bought a few years ago- a bike computer made by Polar. The interface was ancient - some sort of acoustic coupling to transfer data to an analysis program. The software from the company was one step above vapourware - and I can't imagine they were protecting any IP by keeping the software closed-source. In the hands of a fan, the software could have flourished - the hardware was fine, after all.
When these small hardware companies learn that they can get their software written and debugged FOR THEM - with very little expenditure on their part, I think the ball at least starts rolling.
In work and with parents etc, I have successfully converted people from Windows to Ubuntu by careful management of expectations and also a good bit of 'wine'.
It's amazing how many of those "I cannot live without " apps turn out to be quite dispensible when they finally have a go on Ubuntu.
Also, wine is a fantastic *fantastic* tool. Using it (and an instance of dosemu for some particularly skanky old code) I have successfully moved industrial control systems et al. Indeed wine is sufficiently good these days that with (admittedly quite complex) hand-holding it can play huge 3D-heavy games like Eve-Online quite effectively.
Both Virtualbox and WINE are free, Open Source, and work really well. If it won't work under wine, it will work in a VM. You can now use some kind of seamless mode in most VM's, where the application inside the VM gets and icon on the Ubuntu desktop, silently lunches the VM when you run it and just displays it's own window. The experience is more of less like Wine, but without the compatibility issues, because it's really running Windows. If you've got reasonable hardware, less than a few years old, this works well - especially if you can load the Windows VM with Windows 2000 - a much lighter and faster to boot Windows than XP.
This is, in reality, the only plausible solution for these problems. There are, as someone else pointed out, thousands of these little applications that people are relying on. They're only supported on Windows, will never get ported anywhere else and they're too niche to get cloned.
The best way to solve these issues is to change the way the person works or runs things - there's very often a simpler, better way to do it that doesn't rely on running some dinky proprietary piece of software. This is generally referred to as 'Business process re-engineering', if you're paying for it. This is not always possible, unfortunately, and sadly, even when it is, almost never accepted as an alternative - most people just don't want to change, even to make things better.
I second the VirtualBox solution. I'm rather dependent on QuickBooks for my mostly-failed business. I know there are some good open-source accounting applications, but frankly none of them even come close to competing with QuickBooks.
So, instead of constantly having to reboot into Vista to do accounting and then back to Ubuntu to get Real Work done, I just run WinXP in a VirtualBox VM.
Sure, it may not be the sort of solution that my grandma(TM) could set up, but once it's set up, it really is pretty easy to use.
Obviously the more ideal solution would be if these software companies would get on board with Linux, even if they stayed closed-source. But I just don't see that happening, not for quite a while anyway.
We all appreciate the fact that it *is* possible, but it's just not a real option for someone who owns a business, because if it breaks we keep both pieces.
This is a great example because it's probably dealing with very specific software, possibly for very specific hardware.
It's more frustrating when the software isn't as specific. For example, I've worked with companies that a decade ago invested into proprietary database applications that still refuse to move to an Free solution. It gets even worse when they pony-up $10,000 for a new license to that inferior product, since you know they've just delayed themselves at least another 10 years.
My overall advice to anyone involved in FOSS is to present the option and assist clients to move away from proprietary software. Give them the information to make an informed decision. Supporting integration with non-free software is overall a waste of time. Most companies already have a lot invested into proprietary software. Free and open software isn't new, and they are suffering from having made a bad decision in the past. All too often companies turn the mistake of supporting proprietary software into a repetitious error because they are never willing to cut their losses.
Hmm, that's interesting. I imagine the main roadblock here is that "run it under a virtual machine" sounds like technobabble and probably would scare off the faint hearted.
But I've seen a lot of work go into making VMs easier to set up and run. I wonder if perhaps the real solution here would be to integrate some simple to use feature for running windows apps in a VM sandbox. Give it some friendly name ("Window Pane" springs to mind) and make a wizard for setting it up automatically trigger when putting a windows software CD in the machine, or when executing an EXE installer.
Don't forget that for running Windows in a VM you will need a Windows licence.
Don't forget that for running Windows in a VM you will need a Windows licence.
Yeah, I thought of that and am not sure how this would address this in a clean way. Maybe after base system installation prompt the user, "Would you like to also run Windows programs (Y/n)?" and then "Please insert your Windows CD." It would generate a base VM at that point, that the system could use for later installed windows apps.
1. Find the app
2. Build a replacement
3. Use the name of the app it replaces in the description in the deb file.
Now, I'm dying to put my python concepts to the test, and may even try quickly if I can find an app I need in Windows that I think I could move over to Linux.
The other problem in this is that if the developer doesn't need the software it may be less of an incentive, but getting the basics up somewhere where it can be taken on as a project would be good.
Being more Ubuntu specific, maybe we need a 'new projects' section on the Ubuntu Forums? Subcategorized into 'Original' and 'Based on a Windows App' where each title has to show the new name, ' based on Windows app' $app?
We definitely need some more guides out there for programming in Ubuntu too... (All Languages)
This would probably be a good spot to plug the new 'Quickly' tool being developed for Ubuntu:
https://wiki.ubuntu.com/Quickly
It's aimed at becoming a tool for developing and packaging small applications for quickly. It still has a way to go before it'd solve problems for these types of needs, but it's definitely aiming for this niche.
I'd encourage anyone with an interest in doing python development towards a framework to make making these niche apps easier, to join the quickly project and help make it more powerful.
I was stuck on MS access. Then i learned PHP and it solved my problem :D
glabels is surprisingly good for what it is. Show it to her and see if it could work?
I gather the problem is that she needs to be able to create the labels and then send them to someone else in the company to print.
So she creates them and outputs them to a pdf. Then she sends the pdf to the other person who prints them. Still seems trivial *if* glabels would work for her.
This is the big drawback for Linux which needs to be addressed. I hope things like the Software Centre will make it easier to have apps written and make it available. I was hoping the Linux Foundation would step and make some serious changes so develpers can make their software easily available for Linux but I'm not holding my breath. Hence, I now hope that Ubuntu can show the way with the use of the Software Centre. I would like to see the Software Centre pushed to the max, it can do this.
My wife has dozens of thousand of records locked up in a proprietary genealogy program's data files. There have been a few programs that have reverse engineered the formats and there is the ability to export or an open format with fewer features. I can't be certain I would loose or alter data using either path as a migration route to something like Gramps. That is really the only thing keeping Windows in the house.
Web based apps such as
-Google Docs
-Gmail
-Wikis
-redmine
-drupal
The more stuff people do in the cloud, the less important their Operating System becomes.
People use Windows at work, go home to their mac, which they love, until everyone uses a Mac at home, and demands support for more and more job-related functionality for it.
The same must happen with Ubuntu. Leave your Windows problems where you're paid to deal with them.
1. First we make equivalent apps that do the same thing.
2. Then we make them able to import the proprietry apps files so people can switch.
3. Then we make noise wherever we can to make sure people who could benefit from our app know about it.
4. Then we make it more open, free and useful in ways that the proprietry app maker wouldn't dare (for example there's a reason Microsoft isn't duplicating Google's free-as-in-beer Docs... it kills their cash cow!)
5. Profit!
1. First we make equivalent apps that do the same thing.
That would be awesome, but I see two pretty insurmountable problems. The first being that since these are niche business applications, the chance that someone with the time and talent to create them for Ubuntu is vanishingly small. But let's say there are an infinite number of developers to write all of these niche apps; this leads to the second problem... how does the potential Ubuntu convert *find* them?
I'm not posing these as rhetorical questions - I actually think they *can* be solved. I just am not sure *how*.
Lots of applications are sold as parts of bigger services. For example people will generally use label printer software sold bundled with the label printer.
So for lots of these things it is not a big deal. Lots of the software is fundamentally not difficult to make.. most are just little more then cobbled together VB scripts.
This is why we have Wine and we have Mono. That is the whole point behind those. You can't depend on Mono and Wine to just run Windows binaries. They can do that but the main usefulness of these products is for people porting software. Porting from .NET to Mono or Win32 to Winelib is a hell of a lot easier then to port C#/Win32 C++ to Python/C + GTK.
But the fundamental decider of all of this is customer demand. If customers request Linux support then those applications WILL get ported to Linux, or they WILL be replaced with alternatives.
That is all there really is to it. Customer demand.
Its a very simple thing. It is not a easy thing, but unfortunately unlike popular opinion simple things and simple solutions are often very difficult. People can wave their arms around about widget libraries, languages, and unified user interfaces and other slashdot fodder, but that is fundamentally irrelevant.
The easier to make Linux applications/port windows apps and have Linux applications programmed by amateurs* to "just work" across Redhat/Centos/Debian/Suse/Mandriva/etc the lower the threshold for people to start porting software en mass, but that is of secondary importance.
------------------
Our society is geared towards Windows and how Microsoft does business. They won the fight for the desktop so they set the rules for the desktop. You have huge industries created that are just setup to deal with Windows problems. Mom and pop computer stores were people take their PCs to get updated and remove viruses, Geek Squad, etc etc. Look in any newspaper, in any phone book, on craigslist... people with Windows problems making people who solve Windows problems wealthy.
And that is just the tip of the iceberg. Education, software salesmen, and all sorts of other groups of people make huge amounts of money on Microsoft.
You know why Windows Active Directory was able to defeat Novell Netware? Because migrating from older Netware to newer Netware was expensive and difficult, while software salesmen could sell Windows 2000 to end users. People who sold Windows made MILLIONS from people buying Windows. Microsoft targeted salesmen with their software, not the end users.
Goto any community college or technical institute. People are paying hundreds of dollars per credit hour because they have no fucking clue on how to use Windows and use MS Office, but they believe they need those skills to get a job. So they struggle through it and learn to use the software. And you know what? They are right.
Nothing like that exists for Linux.
These, including your software examples, are the barriers that Linux must overcome.
It's not enough that Linux is better or cheaper then windows. Linux must _DESTROY_ Windows with how kick-ass it is and it simply is not at that level, yet,
But the fundamental decider of all of this is customer demand. If customers request Linux support then those applications WILL get ported to Linux, or they WILL be replaced with alternatives.
I think that's a key point - people need to *request* Linux support. I think a lot of us Linux users tend to fend for ourselves pretty well by now. And we all pretty much know how the story ends if we try contacting the company asking for Linux support, so why bother?
Does this suggest something like organizing letter writing campaigns for key niches would have an effect? Especially if it were coupled with some sort of outreach/education effort to people in that niche. Hmm...
It isn't a small niche, you're right. It's thousands of "small" niches. These vertical markets are going to be the hardest to break into. My father is an example of this. He owns an auto repair shop. He hates windows. However, the only apps that will help him run his business and are affordable only work on Windows. Period. Note that by "affordable" I mean "Costs less than the tens-of-thousands of dollars that the big systems cost, but still cost a few grand."
So the way this is fixed is when someone who is stuck in one of these niches is willing to splash out the time, effort, and/or $$ to create an alternative. There is a lot of momentum to fight against here.
OR
Convince a company that currently develops one of these tools to support more platforms. That would be an even bigger uphill battle...
So the way this is fixed is when someone who is stuck in one of these niches is willing to splash out the time, effort, and/or $$ to create an alternative. There is a lot of momentum to fight against here.
OR
Convince a company that currently develops one of these tools to support more platforms. That would be an even bigger uphill battle...
Yes. And you're right that both are tough - in the first case, by definition, the people in these niches tend to have skillsets that match those niches, as opposed to having software development skills. And in the second case, they'll be rationalizing from a business perspective whether it is worth the investment.
The basic answers in both cases seem to lead you to think, "It can't be done", but I don't buy that. I mean, long ago these guys all used typewriters and filing cabinets, and I'm sure switching to using computers was an even bigger challenge, especially given that a lot of the people in the niches at that time didn't even know how to *use* computers. (I remember 'cause I was doing summer jobs at companies like these as a kid.) I think it can be done, we just have to think outside the box as to how.
In other words, are those really our only choices? For example, in the first case what about people who are going *into* those niches fresh, with or without Ubuntu experience, but who aren't yet tied to any particular platform and don't have a plethora of files in proprietary formats we'd have to worry about converting?
Or in the second case, instead of trying to win over software developers who target that niche and support only Windows, what about finding companies that are already sold on Ubuntu and trying to convince them to support the niche?
It'll only be possible when the community completely embraces Wine and other inter-platform technologies like Java to being integrated in the Linux desktop.
A lot of people claim we don't _need_ Wine out of the box, but hey, I'm sure they haven't tried teaching their dads how to install Wine so they could use [some random Windows app].
"...the problem was that she used some software for her business..."
http://appdb.winehq.org and search
Or dual-boot, or Wubi (google it), or if her hardware supports it, run MS in a Virtual Machine under linux, or run linux in a VM under MS. Or, with the thousands of apps in the repositories, it's entirely possible there's an app in there that can do the job. Or even go with 2 systems. It isn't a matter of either/or.
In fact I did mention WINE as an option, although my gut tells me whatever program she uses might be too specialized. A VM might work.
if you can get it to run on wine, that is a great first step. But it depends on how the total solution goes. If they need other apps to use the data from the one important app, then it becomes a lock-ed in situation. Getting one app to run, by itself, might be doable.
Wine won't run Quickbooks. The data files aren't convertable, and exporting transactions are a joke. Because of this, migration can only occur at the end of the year -- the busiest time of year for a bookkeeper -- and you still need access to past transactions. Yes, there are ways to do this; no one has any howto's approved by a CPA though.
And the open source equivalents are not as simple to get going with, nor can you find anyone trained in them.
So I'm committed to running it in a VM, along with the other things that make a complete bookkeeping environment: Adobe Reader (QB doesn't do Foxit), Excel, and (optionally) MS Access.
The situation is terrible. Fixing it isn't simple, but if someone did, you'd see free OS's in a lot more small businesses.
I'm not one of those you speak of. I don't have a problem with Windows and never have. And I don't feel trapped at all. I started out as a die hard Mac user until I introduced Windows to my bag of tools and don't have a problem with Mac except for the fact that they cost way too much. And now I am getting into Linux which I don't have a problem with either except for the lust for Windows bashing that some Linux users have. I'm getting to the point when I see the term "Welcome to freedom." or anything similar maybe I will just dismiss it and move on and hunt for comments from Linux user who are going to really help me learn. Yes there is freedom to get a distro for nothing but I have found the freedom of time is restricted when it comes to working with Linux. I have spent more time trying to figure things out and get things to work in some of the Linux distros let alone trying to figure out which one will suit my needs best. Don't get me wrong I think Linux is great and I'm not ready to give up on it yet. I love the concept and I love open source software and say thanks to all who have the programing and genius ability to create these great apps. I will continue to work with it as well as Windows and Mac. I work in Windows on a Mac under VMWare and get the work done just fine thank you. So now I have a little more and having fun on all of them.