Why Users Dumped Your Open Source App for Proprietary Software
Open source projects that want more users must learn the ways in which it might "lose to the competition."
View full article »
Esther Schindler
If the comments are ugly, the code is ugly
claird
SVG a graphics format for 21st century
pasmith
Take Chrome OS for a test spin
Sandra Henry-Stocker
Solaris Tip: Have Your Files Changed Since Installation?
jfruh
Android fragments vs. the iPhone monolith
mikelgan
What Gizmodo missed about the Pro WX Wireless USB disk drive
Sidekick: The Good News & the Bad News
Either way you look at it Microsoft Data Center management did not follow standards or best practices in this failure. In which case it makes me wonder more about the outsourcing of corporate data much less personal data.
- mburton325
Join the conversation here
Quick, practical advice for IT pros. Made fresh daily.
Want to cash in on your IT savvy? Send your tip to tips@itworld.com. If we post it, we'll send you a $25 Amazon e-gift card.
Slight clarification of my quote
Esther,I wanted to offer a clarification to my quote you used.
AS2 is not a product I dumped; it is a protocol. I dumped a OSS AS2 product that didn't have the management features that our organization needed, and subsequently the OSS wuftpd site that had duplicated functionality in the later-chosen commercial file transfer framework (but didn't have the management tie-ins).
Thanks,
~ Paul
the "too much tweaking needed" factor
One of the things I learned as the sysadmin to research projects is that programmers make the worst end users. (They can also be bad sysadmins but that's another story :-).)
The computingverse couldn't survive without programmers but they like to fiddle and often don't understand why everyone else doesn't either. Often they don't even realize they do it, too, unless you call it on them. If you point out something doesn't work they'll point out the little tweaks you have to make to get it going (sadly, often while making it sound like you're an idiot for having to ask. You should have read the source code to figure it out!).
Often user interface stuff is slid to the bottom of priorities -- adding features is more fun, and bugs are a required issue. Adding to that a lot of the OSS projects are driven by programmers who also are like this. UI is just "adding pretty" to many of them.
Unless a project has a dedicated team of UI people things are probably going to be klunky. And it doesn't even have to be GUI stuff. You often need easier interface stuff for a completely command-line driven app.
More OSS projects need to realize that UI people are worth their weight in platinum.
Partially agreed
A well designed UI is a huge asset, true.My beef with a lot of Window apps is that the underlying data structure is often poorly thought out or even badly broken. The result is a pretty UI on a program in which either too much of the operating code is in the UI or which produces data which cannot be used in other ways.
Also, (G)UI's have to suit the U part. If U is a business user then that's a different use case from a programmer who might actually want to have fairly raw access to the underlying operations.
Amen a million times. When
Amen a million times.When I was still writing custom software, my interface was the thing that my users were most pleased with, as I spent a lot of time prototyping and figuring out what the interface should look like to be within the experience base of my clients. If you make the interface common to the strategy of the OS as much as possible, many people can just figure out how to use it without a great deal of trouble, though in some fields tailoring the interface to the users is more effective.
If you have an already complicated software that is hindered even more by a poorly designed UI, there is just no chance of average users figuring it out. Of course, pay software isn't immune to this either. If you get the opportunity, compare Blender, Lightwave and Maya. Blender is OSS and may be poweful, but has a learning curve of almost 90 degrees. Lightwave is also powerful and used by many shops, but again the UI is so foreign that its hard to get past it without a lot of time... I'd say its learning curve is like 70 degrees. From my usage of Maya, I'd have to say that even though it is just as powerful or more so than the other two, its interface is in the 45 degree learning curve. Maya is still hard to figure out, but the interface doesn't cause most of the problems. Some of the simple things that Blender and Lightwave could do would be just to add in more mouse controls. For instance, being able to use the scroll wheel in a viewport to zoom in and out or clicking to rotate a view adds an awful lot to an interface, without much extra coding.
The worst problem with bad interfaces is that since the current user base is used to the horrid interface, there is actually a movement against simplifying or adjusting it as the people who use it every day will be confused. I've experienced a large UI software with pay software, such as Satellite Tool Kit, which in version 9 vastly changed the interface. I personally can't tell if it is easier for a new user to figure out, because I've used it for 7 years and sit there confused.
Everyone should take some lessons from this article as it doesn't just apply to OSS. Shareware and even smaller companies might not realize the issues that cause their software to be passed by.
Installion Hell with FOSS
This is too true.My open source app that I recently said "screw it!" to was Red5. This is used as an alternative for Flash Media Server for streaming flash video/audio online and a bunch of other stuff.
So here is the story:
My organization needs a way to record audio online and send it over to our server. The old way we have is with Java and FTP which for some reason works intermittently (fails on some computers for reasons unknown). So we looked at Flash and saw that we needed some back-end server software to get Flash to send recorded audio to our server. Not wanting to pay $4000 for FMS, we tried Red5. First I Googled for tutorials and actually found videos on installing Red5. I thought, "Wow! This is going to be cake!" Nope. After watching about 5 of the 21 videos, I decided I knew enough to get it going on our server. I went to their webpage and found it different from the video. Not a big deal, but during installation, none of the options came up that were in the video either. The files downloaded were also in different folders and I couldn't find the ones the video pointed out. Turns out all those 21 videos are for pretty old versions and everything has changed. So I tried reading the ReadMe and Install files downloaded and got annoyed by the fact that I needed three additional programs (including the entire Java Development Kit) just to run it. First I went after the required program, Apache-Ant. They actually want you to modify environment variables and place the files in system paths that don't contain spaces! No one should have to modify anything that requires them to view properties the average users doesn't know exist, then click on the 'advanced' tab, then navigate to the advanced section of the resulting dialogue box. That's a little too advanced for just installing something (especially when most programs are just a double-click and your done). I did all that and still couldn't get the thing to work. So I tried it on my machine thinking it was security restrictions. After having to configure Windows to allow these programs (ant, jdk, and tomcat), I finally got some progress. Although following the instructions exactly how they have it on the webpage doesn't work and you have to know not only how to follow their instructions, but how to catch their mistakes as well. Anyway, ant still doesn't work completely but Tomcat seems too. So I tried the Red5 examples and... nothing could connect. Oh and when you download the latest Red5, you don't get all the files you need! To get these files it seems you have to have svn and run a checkout command (you heard right, to even RUN the thing you have to checkout the project. Again I want to say this is not if you want the source code, this is just to RUN the app). I have svn but I use Tortoise to give me a nice GUI menu option for svn commands. But of course they don't tell you how to make it work in the most popular svn gui client. Nope just good-old command line. The stuff everyone uses right?
Anyway, to everyone with an open source project, if you are going to offer instructions in Linux using command line (which less than 2% of the population use), then at least offer it for Windows users (if there are Windows versions) using the built-in gui. And don't make us hack our system to get everything to work right. At least make a .bat or .cmd file to do most of the work for us. Or ask some easy questions during installation. Plenty of apps do it and it isn't terribly difficult and is definitely worth the time.
Oh and please don't write the bare minimum in your installation files. Writing steps like "add bin to your path" is way too short. Bin what? What is my path? How do I do that? Sure some people (including me) know the answers to these questions but not every programmer and especially end-user know what environment variables are. Try saying something like "If in Windows Vista, right-click on 'my computer', go to 'properties', click 'advanced system settings', go to 'advanced', click on 'environment variables', find path, and add to the end of it the path to the bin folder you downloaded". Was that really so hard? Then just add the same for XP and maybe for Mac if applies. Even so you shouldn't have to modify these values but if you absolutely require it and can't write a file to do it for the user, then at least explain how to do it.
Installation Hell + Setup Purgatory
Installation and setup is part of the user interface. When I can't get a working setup configured because I haven't read the developer's personal diary to know which parameters to set, or I can't even figure out how to install something because the documentation (if it exists) doesn't mention installing a dozen dependencies, I'm not going to trust the thing with my data. Sure, I can find paeans of praise about how wonderful it is "once you get it set up to your liking", but at some basic level I expect the damn thing to *work* without rebuilding my kernel. (Oh, BTW, to all the Linux evangelists: The fact that I *can* rebuild my kernel does not necessarily mean that I *want* to, any more than I rebuild my car engine for entertainment.)When I rent a car, I expect the gas pedal on the right and the brake in the middle and maybe the clutch on the left. I don't expect to have to re-learn the concept "how to drive a car".
What made us dump an open source app we were using?
We haven't dumped an open source app, but we have dumped a proprietary app and the reason was lack of support.We had not been content with the vendor's standard contract and our lawyers had spent months negotiating what they considered to be watertight terms with provision for liquidated damages in the event of software failure.
Unfortunately the the vendor sold up and the new owner didn't want to know us. Although the contract was still valid, having a "throat to choke" proved pointless because what we needed was a working app - not money.
Fortunately my teenage son was able to come to the rescue with an open source app (TeXlive) with plenty of support from the open source community - not that we have found much need for it (support that is).
Open source apps are just that: OPEN!! If there is a bug, you have some chance of figuring out what is wrong and fixing it for yourself ... and everyone else. If you don't know how, you can complain publicly and someone else can have a go. You don't have to wait for proprietary owner who will more likely be interested in adding some new feature than fixing a bug that might only benefit a small percentage of his market.
The OPEN nature of FOSS is
The OPEN nature of FOSS is on one hand its greatest asset and on the other its greatest weakness. Yes I can open up the source and fix the bug or even add a missing feature, I done both. But what happens when the bug or the fix is beyond the sysads skill level? What happens when MY bug isn't viewed as a priority in the community? Now I'm stuck maintaining MY patch set and maintaining MY folk of said project. Which ultimately limits what I can do as an admin. In some cases it easier to explain to the powers that be, that our vendor doesn't support that, or we've submitted a bug report. Then it is to explain the hacks, patches, and duct tape used to fix FOSS.Re: "But what happens when..."
In that case, you're no worse off than if you were relying on some proprietary software.Had this problem with Proprietary Apps
Going all the way back to DBase 4. I called, reported a bug, and the response was along the line of "What do you expect us to do about it?" Since Ashton Tate wouldn't support me, I stopped supporting Ashton Tate. Guess where Ashton Tate is now?
At least with a Free Software/Open Source application you have the option to try and fix the problem yourself (or hire someone to fix it). When Ashton Tate, MicroPro, Borland, Digital Research, Microsoft, etc. decide not to fix something, you are stuck.
You're right about speed
My particular FOSS bete-noir was Evince. I'm a generally happy Ubuntu user. With one exception: reading PDFs. I have found that reading downloaded magazines is so painfully SLOW with this reader that it's a real trial of patience. I frequently, as in several times each session have to wait 10+ seconds for a page to render. Then I find that page is often an advertisement that I just want to page over anyway.It's so bad that I have (re)installed the version of XP that came with my PC into a VirtualBox VM (Shining example of how good FOSS can be) and use Adobe Reader for all my PDFs now.
I did an unscientific benchmark and came to the conclusion that even though it's running in a VM and accesses files on the host across a virtual network connection, Acrobat is still 3 - 4 times faster at fully rendering the whole page than Evince - and even quicker than that to get to a "first impression" of the page, which allows me to skip ahead if it turns out to be unwanted.
Hi Pete! Which ubuntu
Hi Pete! Which ubuntu version were you using? I remember having such issues quite a time ago, but they were already adressed and evince works pretty well here (debian lenny).You should give Okular
You should give Okular (http://okular.kde.org/) a try.pdf
I have a issue with Pete and adobe. I use document viewer and it is lightning fast... I have found no pdf's it cannot open. Adobe also directly supports pdf in linux.As for giving up OSS for proprietary... I do the opposite. I have given up MS Office for Open Office. It is faster and does a better job in my opinion. I use firefox, thunderbird, openoffice, and a bunch of OSS games... In Windows 7!!! In fact, i the only MS software i use is windows. I also use pidgin instead of the ad infested windows live crap.
As for support, I get much better support and chances are the problem already has been addressed using OSS software. I have found the OSS software generally work better in most things.
I think you need to revist your article.
I'm happy for you.
But why do you assume that your experience is the same as everyone else's?Ad-infested ? Wow....
Can you please ellaborate ? I have windows mail open. Also, photo gallery. And others. I see no ads, or, if the app is a web application, there are no more ads than what you can find in google or others - way less than on this page, for instance. Case dismissed.I find just the opposite
I have been running Gentoo Linux for years now, and the only problem I have is finding the right application for the job. Cd-Burners, Office suites are on par, and even have more useful features than their proprietary counterparts. Also, I feel that open source software is more secure than proprietary software, especially when it comes to fixing security bugs.Take Firefox for example. When a bug is found in Firefox, it is fixed that day or the day after. When a bug is found in IE or Windows, it takes MS weeks to months to release a patch.
Now, lets look into Windows7 shall we? I installed avast virus scanner, it asked me to reboot and I said no, because I just needed to install notepad++ for some web development stuff. Figured notepad would not be a big deal, installed it, rebooted and Windows7 crashed, and I had to do a system restore. Windows7 is faster than Vista, but all it is, is just another buggy version of a typical Microsoft Operating System.
Blame the right party
Why is Avast's bug a problem with Microsoft's operating system?I could write a really buggy Linux program that could screw up the machine, woud that make Linux just another buggy version of a FOSS product?
Not quite
Firefox's verified or assigned bugs list starts from October 2001. There are as many open bugs as days have passed this year. There are 511 bugs to which Firefox developers have assigned WONTFIX as the resolution.https://bugzilla.mozilla.org/buglist.cgi?bug_status=ASSIGNED;bug_status=VERIFIED;bugidtype=include;chfieldto=Now;classification=Client%20Software;columnlist=opendate%2Cchangeddate%2Cbug_severity%2Cpriority%2Cassigned_to%2Cbug_status%2Cresolution%2Cop_sys%2Cshort_desc;emailassigned_to1=1;emailassigned_to2=1;emailqa_contact2=1;emailreporter2=1;emailtype1=exact;emailtype2=exact;field-1-0-0=bug_status;field-1-1-0=classification;field-1-2-0=product;field-1-3-0=resolution;field0-0-0=noop;keywords=;keywords_type=allwords;long_desc_type=substring;product=Firefox;query_format=advanced;resolution=---;
Meaningless by itself
This datum is meaningless by itself - how many of those WONTFIX bugs are from earlier versions that are unsupported now - I'm sure that a bug that causes a problem on a current version would never be marked WONTFIX. I don't think its even possible to analyze some closed source apps in this way, as their bug reporting & bug status are not made public.@Pete - Why on earth did you use XP in a VM for Adobe Reader? - the exact same app is available for Linux, same version,same functionality. Also my own experience is Adobe reader is slower by far than Evince. Your post has a slightly astroturfy odour
Linux
A few years ago, I tried a distro of linux for the very first time. At first, it was ok, but then I kept getting very weird bugs in evolution(email app equivalent to outlook). Off and on, for no apparent reason, when I would type in text for an email, it would start to go backwards, from right to left. I checked everything I could, but couldn't fix it. Then I couldn't see a website properly because flash was not installed in the firefox browser. After 2 hours of messing with it, I only got it partially installed & it still didn't work. Then I had to learn how to navigate around a totally foreign software environment. Then I had an issue with playing cd's & dvd's. I did a lot of web searching & never did get that one resolved, so I dumped linux & went back to micrsoft. Now, my win xp machine keeps getting hit with malware, (despite keeping it patched, updating antivirus, firewalls, etc), so I took the plunge and got a new hdd and installed Ubuntu linux. Just about everything is automated, yet I can still use the command line if I want to. Installing flash took just a few mouse clicks & I'm done. The security & performance is so much better than windows. My wife's computer (with xp on it), keeps getting malware too, so I reformatted the hdd last night & put Ubuntu on it as well (and guilt free I might add. I've always felt a bit nervous putting xp on more than 1 machine since I know that is a violation of the EULA). Now I'm considering getting a Linux + certification and keeping linux for good this time. Ubuntu's free upgrade every 6 mo is great! It comes with a built-in firewall that passes grc's test in perfect solid green. The more I learn about linux, the more I like it.Overall, I think price is actually secondary. Performance, features, & security are the important elements here for me, but being free of the EULA, and being able to get the OS or app you want for free (or ridiculously cheap) are very nice benefits nowadays with the tough job situations & poor economy.
There are pros & cons to just about everything, so you have to decide what your priorities are and then find the software that best meets them in your particular situation.
A couple of other points
There are two things that have caused me to drop a F/OSS app. The first is either abysmal (or completely missing) documentation: how the author things the program should work isn't necessarily how a _user_ thinks it should -- if an app is THAT different from anything else on the "market", they it had better include at least a basic instruction manual; if I can't fire up a program and start using it because the author decided to do something in what seems like a less-than-intuitive way, I'm not going to waste time trying to figure out what they're trying to do. Similarly, comments in a config file (or the source code) does not constitute "documentation".Second, I've run into cases where an application claims to be able to do X -- and it can, but does so in just a (very) limited way, or ONLY after adding in a metric buttload of supplemental files, utility programs, and relatively obscure (or not commonly used) libraries.
Programmers don't make bad end users....
Regarding the comment that programmers make bad end users... that's just not true. The problem is more complex than that.The reason FOSS projects typically have so many "knobs and buttons" is that FOSS projects usually don't have a firmly established user community in mind at the beginning that they're trying to satisfy (sometimes they're just hacks that start out with the author as the first end user). In other words, if you're trying to make money selling custom proprietary software, you have a laser focus on who you're trying to satisfy right from the start, and you build as close to exactly what they ask for as you can (perhaps with a little give and take around the edges). With a FOSS project (and I've contributed to many), at the beginning you end up with all these gray areas. "Well, if someone is going to use my project for THIS, then I'd set it up THIS way... if they're going to use it for THAT, then I'd set it up THAT way." Successful FOSS projects eventually get past that infancy phase and they settle into an established user community that they're trying to satisfy, and usually those projects take a marked turn toward less "configurability" and more stability. This is also why FOSS projects often have complex installation needs. Again, if you don't know how your project is going to be used, you can't make a lot of assumptions in how it should be installed, so you end up with a million buttons and knobs.
What would be refreshing to see would be a class of developers emerging who specialize in categorizing all the design / installation choices and then turning them into installation and configuration wizards that are in plain English and follow consistent patterns. Folks who enjoy specializing in that area could make real progress toward dealing with the "installation/configuration" complaint that is kind of pervasive with FOSS.
Crushed between Comercial and Open Source
I am neck deep in a number of CMS projects using Sharepoint, Drupal and Alfresco [all different projects]. There are pluses and minuses to each. I find that basic capabilities are the best with Sharepoint, but to do anything moderately advanced seems to require additional purchases that dig the project deeper into the Microsoft development stack. And even then things seem to unravel into an unmanageable mess. Alfresco has some great base capabilities, but is not very mature beyond its core document management features. Drupal is incredibly powerful so long as you are willing cobble together modules that seem like they should be a part of the base system.Regarding support, Alfresco online support is great, but limited to its core feature set. Drupal has great support, but there are too many newbies leading people astray. Sharepoint support seems geared to competing and incompatible work-arounds and feels appallingly ridiculous--on many occasions the Microsoft recommended method paints you into a corner a few steps down the line.
Personally I think there are enough warts across the entire software landscape.
Lack of Support/Documentation
One thing I always hear about open source is the community support. But, with several applications I have used lately, the community is made up of new users. We can't help each other. We need experts to respond to our queries and problems and sometimes they just aren't available.For instance, Concrete5 has a nice community forum, but the C5 team won't help with install problems. Now, I understand their reasoning. They'd spend all day helping people install it instead of adding features and fixing bugs. But, many of the people that come to try it out are not programmers or system admins or linux gurus and they can't make it work.
Moreover, C5 is supposed to have a very cool API, but they'll be the first to tell you that the API documentation is pretty bad.
If I can't figure out how to install it and I can't use the features it has and there's no one to help me, "free" starts to get expensive, but time is money.
Open Source vs Intuitive
I am a developer and have written my share of open source apps. A big example of the problems with most open source is OpenOffice/NeoOffice. These do most of what you want, but consider:For example, I never could get table names in Writer to save and reload reliably; you can rename them, but sometimes they all vanish. Why? Who can tell me?
It's that stuff that pushed me to buying more copies of MS Office and reduce my use of Linux (so I could run MS Office.)
Open Source issues
I think open software suffers most from the boredom of its volunteer programmers who often focus on interesting challenges rather than basic documentation and compatibility (as mentioned). It is free of course and worth every penny, but it often really wouldn't take much to vastly improve the users' experience. Ironically less-qualified power users working closely with the programmers could help a lot. The Audacity project and audacity forum do a good job of bringing the right people together to get critical bugs fixed.
As a different example, PHP with Apache in Windows should be a mature project that's easy to use at this point. Yet the extensive documentation never mentions that the latest version of PHP (5.3) calls a function that only exists in win XP and after. Not all users are up for re-writing the source code, which is the only way a windows 2000 user might make it work. Doubtlessly most of PHP's developers are linux users.
Yet at the moment I also haven't found anyone who has succeeded at getting a recent version of PHP working with the current version of Apache on Windows. The Windows installer version doesn't work... help forums directed me to the zip version. I couldn't get it to work either after many hours of research and fiddling. At least the config files are heavily documented... but they are dozens of pages long.
I think the best way to increase adoption of open source would be (as mentioned) simplifying and organizing the user interface. Nearly everything should be doable from a GUI. For better or worse, Microsoft has defined the standard interface format for most apps, and open source should generally follow that.
Virtualization is great, as is Wine, OpenOffice and firefox. These coupled with Ubuntu are slowly making open source more comfortable and convenient than closed software. The more mature such projects become, the wider and deeper the user base for open source will be, and the more successful.
VirtualBox and Wine are about 70% complete I think, while OpenOffice, firefox (both a bit slow) and Ubuntu (still buggy) are nearing perfection, but still have a ways to go. These are all huge efforts that are critical to open source's success into the future. It's amazing how far they've come. Full compatibility with Windows software is the most likely to succeed way to destroying windows dominance.
So I think it's all about a familiar interface, compatibility, documentation, forums and marketing of quality support... these are the keys to mainstream adoption. Even speed and bugs are secondary I think.
Are users allowed to comment on and choose proprietary apps?
I think I can see what the problem is here. Are users invited to comment on and choose or not choose proprietary software? The answer is no. With proprietary software, you have paid your money and your staff have to use it - end of story.The solution to the problem you have encountered is simple - treat open source software in exactly the same way as proprietary software. By all means take great care to check out and trial the applications for suitability, but once you decide on it, simply tell your staff that you have acquired the software and that it is mandatory to use it period, the same as you would with proprietary software. Any software deployment where you give individual staff the opportunity to opt out will lead to a shambles which will guarantee the deployment will fail. It is just that with proprietary software staff are not given this choice, because money has been committed to the deployment.
Wow that's a fallicy. Many
Wow that's a fallicy. Many companies that I use software from actually work rather hard to incorporate things for the user. If you are saying your comments on Windows and MS Office have been ignored, then you might be correct. Look up a piece of software called Deep Exploration from Right Hemisphere sometime. Their customer support team is pretty good at responding to problems and requests... I'm now stuck on a version that I can't get my company to upgrade, but it doesn't mean that prior to being left behind, that I wasn't well supported. Is it expensive? Yes. Is it worth the hand-holding and support? I have to say absolutely when I'm stuck doing twenty jobs and none of them involves rewriting someone's code.Why Users Dumped Your Open Source App for Proprietary Software
Why Users Dumped Your Open Source App for Proprietary Software is due to the fact that they are used to the prior proprietary application and find change is hard to due. But look what happens when a proprietary application has amajor overhaul and changes things. There is that learning curve to find out how to use the application again, well try the open source again and keep a open mind to use that application. I have been teaching students the benifits of using open source applications and I teach them how to use them. When they go to use the proprietary application they find it to be hard to use, go figure. They usually stay with the open app due to easy of use and cost (free and legal).Professional Open Source
I completely agree with you if you are talking about situations 5+ yrs ago, now we have open source companies with better support ever before. Fundamentally not all open source companies gives support, now more and more open source companies have realized the need for it and evolved into a nice support business models. I would blame the people who makes the choice not taking into consideration of future...The point isn't who to blame
It's what you (in a FOSS project) can do.I had a bad experience with a restaurant when it first opened, a year ago. Let's say it was slow service, or not enough vegetarian options.
The restaurant has fixed all those problems since then, but it hasn't occurred to me to return because, hey, I had one bad experience and that colored my perception. I have hundreds of other places to go, and no particular reason to think about that one again.
Do you want to blame me? Or do you want to get the message to me that things have improved and perhaps I should consider another visit?
Configuration files are Evil
Take Apache (and its gigantic and cryptic configuration file).
Almost all other free web servers feel necessary to copy it, creating the same hurdle again and again.
I am not fan of unsafe SSL remote administration consoles but others have devised an alternative to this mess: removing the configuration file alltogether.
The excuse for configuration files is often preformances and fine tuning. But everybody ends trying to get the same top performances and denial of service barriers so why not implement the desirable settings as default values?
TrustLeap G-WAN has no configuration file -and it puts all others to shame on the performances department.
Proof that by working a bit harder, good things can happen.
Great article, but we have
Great article, but we have to keep in mind that the "community" doesn't end at the software developers. As more users some in, some have marketing skills and these might contribute by addressing some of the marketing needs. Of course, will they be able to have sway within a project run by sw devs? [If not (eg, if the project leaders don't care about addressing some particular issue or about growing users under X or Y conditions), those that think they are on to a better marketing approach can find some sympathizers and fork or else otherwise show results in some way.]Same concept applies to all other skills and contributions.
You CAN be productive with open Souce software.
Hmm... Well..I am using open source software for some time now. I know it's not perfect, but it is adequate for my needs. I think the biggest problem is not a real bad design, but most times the impatience and unfamiliarity of new users.
First of all - a lot of people compare open source with their commercial counterparts. In most cases this is not a fair comparison, because open source software -as the name implies- is free to build and/or to be used. This also implies a lot of open source projects are on a very tight budget - or no budget at all. In that case it is not fair to demand the same functionality as a version that is payed for, and also has a well payed group of developers behind it. You get what you pay for. If you have a bit patience and accept some little work-around sometimes, most open source software is more than adequate to fit your needs.
Secondly - Most new users of open source software do not see the applications as a whole new design. They want a exact clone of (sometimes very expensive) closed software, but that's not They get impatient or angry when the software does not exactly behave the same. Most times the same end result can be obtained, but the road to this result can sometimes be very different and sometimes be a bit longer an a bit more complicated. Well - to get something for free (what otherwise in some cases would take a massive investment if you choose for a closed source solution) you have to accept it is a bit less perfect and takes a little more time. Now - opponents of open source wold say this is a major setback to productivity. Just for the fun of it I took some rough time measurements and put a closed source application against a open source application. The latest was more complicated and took more time to reach the same result (what application it was is not important - it was the time difference that counted). To my surprise the outcome was just a very little drop in productivity. We are talking for a minute or 12 over a whole day. So yes - open source sometimes can give a drop in productivity, but the amount of it is in almost any case greatly exaggerated.
At this moment I am using open source software and I am really are productive with it. I use Blender for 3-d design, Inkscape and XaraLX for vector graphics, The Gimp and Krita for pixel graphics, XSane for scanning, Picasa and RawTerapee for photowork, Scribus for publishing and OpenOffice.org for some paperwork to name a few. For the sound FX and music parts I use a Linux distribution with rt-kernel (this distro is used for the above programs as well). To produce music and sound I use the Jack sound server in real-time modus and programs like Ardour, SpiralLoops, Rosegarden and Hydrogen. I also can use wine(ASIO) to use some windows VST plugins and windows applications like Reaper, but also can use the native DSP's and softsynths like ZynAddSubFX, Qsynth, Qsampler, Freebirth, Horgand, SpiralSynth to name a few. The hardware is a XBoard49 MIDI keyboard and controller and a E-MU 0404 sound card.
Yes - it takes some time to decently set up a open source environment (but you only have to do that once, after that updates are automatically). You have to accept some work-around in some cases, but if you take that little time investment you can become be very productive with a open source-only environment and produce some high-level output for a very, very, very low level price compared to the closed source counterparts.
Its about support
1)Any IT project will fail if it does not have buy in and support from all parties concerned whether it be on a Microsoft platform or Linux.2) A lot of opensource alternatives have everything a general business needs to operate with no compromise on functionality - this would be a different story with a corporate with specialist functions and needs.
The companies my company supports, which are in the small business sector have mostly migrated to opensource alternatives and have not felt any difference in functionality. This is mostly about support. My technicians are trained to ask the users how they can help any further to automate a task, how to do something and make the use of the computer a pleasure. The other is that I have won over the owners of the companies support - the next was the users.
So my point is opensource works as well - as long as the above is take into account.
disability access
I've recently transitioned back to Windows 7 because there is little or no support for disabled users like myself (speech recognition user). The only two solutions are NaturallySpeaking or Windows speech recognition. Overall cost is about the same as is support from the vendor (i.e. none) I can't wait for an OSS version because I need to work now, not 20 years from now.There is some effort to make NaturallySpeaking work under wine but that effort is starving for lack of talent/resources. This problem highlights one of the issues I have with OSS which is that just because the source is open doesn't mean you have the ability to do anything about it. That goes doubly true for those of us who are disabled and can't manipulate it due to damage hands.
What I'm trying to solve now is how to speak on Windows and live in Linux.
I agree - The Foss community is ignorant to its own detrament
Most of the points of your article about the FOSS community I agree with. That said, those that see themselves as community memmbers, the term FOSS, I would argue, is the communities biggest enemy, or at least the approach to 'FOSS' that most community members have.In fact, I would go as far to say that 'Free' by the majority of this community is read as 'free' of charge. This is a huge hurdle for mainstream adoption, even if there is in fact 'no charge' for the software. Free as in freedom is almost always seen as secondary.
Even the majority of the communities approach to 'Open Source' counter-productive to main stream adoption. In an effort to remain on the opposite side of the scale from 'lock-in' software strategies, most notably from Microsoft, there is an almost 'repulsive' reaction to any 'closed source' software. The most commonly sighted argument being, well, if there is a problem with the software, because it is open source, I can go in and fix it, which is true, but how often does that apply and to what fraction of the community has the ability to just go in and change something? Probably less than 10% of FOSS users, I would guess.
Free of charge, is good. Free as in freedom is good. Software that is open source, is good. But none of these is good if enforced to the exclusion of all else like a zelot.
The only people that understand these points are the market leading firms that need to make profits (not an evil activity) and drive the FOSS market place, and a small fraction of the community.
An effort to re-educate the community is required, for mass adoption of both new users and commercial software and hardware companies to join the fold.
I would like to point out
I would like to point out another part of "marketing" that most FOSS engineers fail to pay attention to -- user interface.If it is not easy for THE USER to use the package and if it does not look 'friendly', then users won't use it.
Engineers usually have the smarts to put something together that works, but generally don't think to make it desirable to the average user. And what makes it desirable to a user is what he/she SEES as well as what it does.
Think of it this way, if the average user had a choice of going into a dirty-dingy liquor store or a nice clean 7-Eleven next door, they will choose the 7-Eleven every time. The engineer on the other hand... :-)
misleading article
The author is ignorant of the fact that most "Software" isn't what you install on your Mac and click and type in to get your work done but something that literally that runs our civilization.Most of the Web-related infrastructure software, for example, is in fact FOSS because it simply works reliably and has great support and value in it.
Any complex LAN infrastructure is bound to have predominantly FOSS software. And have you seen the supercomputers? Most of them run Linux kernel indeed.
It is true that desktop user applications haven't been accomodated by FOSS yet, but some new commercial FOSS applications (that aren't necessarily free of charge) promise to do that.
So the title of the article should have reflected that it is about "Desktop" applications and by far does not apply to software in general...
Your comment is verified by
Your comment is verified by the facts that almost every stock exchange runs Linux (http://blogs.computerworld.com/14637/linux_powers_worlds_fastest_stock_exchange) all Google servers run Linux (http://en.wikipedia.org/wiki/Google_platform) Linksys routers (and many more) run Linux, etc., etc., etc..Are you so anxious for enemies that you want to create em?
I've seen a few variations on this "argument," here and on slashdot, and it rather astonishes me.These people argue that since I point out that some people look at a FOSS app and choose a proprietary alternative, that I am (a) unaware that plenty of people do choose an open source option and (b) suggesting that any proprietary app is better. I don't see how they reach that conclusion. Are you so anxious to confront an enemy that you will work to create one?
Yes, open source is wonderful stuff. I wouldn't be writing a blog about the topic unless I thought so. (And I'm certainly qualified, as I've been writing about open source for most of a decade, and about software development for far longer.) I could write the obvious articles about "why people choose open source instead of proprietary apps," but hey, that's been done, and it's preaching to the choir.
Like anything else, FOSS is not perfect, and it still has barriers to adoption. Unless you acknowledge the barriers, you cannot deal with them. My aim was, as I said, to identify the reasons that people might try out your app and fail to adopt it; I also had a few suggestions about how you might address it for your specific project.
Yes, open source is used all over the place. Great. Do you want to see it used more? Then let's talk about what has to happen to make that goal achievable. Hint: it is not generally useful to accuse the people who do not buy your product (or your vision) of ignorance or malice. Especially when they are on your side.
RE: Are you so anxious for enemies that you want to create em?
My point wasn't about avoiding criticism, but about lack of broad perspective on software industry in the article. For example most of the IT-related software industry has already adopted FOSS standard for licensing so discussing FOSS vs. proprietary in that context doesn't make sense.One of the major problems is interoperability
You missed a very important point why people ditch Open Source, and it is interoperability (I think you might have touched on the subject a bit).Some major companies work with Microsoft Office (for sake of the argument) and if someone told them to move over to Open Office.org, they would laugh about it.
Firstly, it's because MS Office documents don't read/write well with Open Office.
Secondly, because it's free, it means less feature rich (so they think).
Thirdly, they only know about Microsoft and are scared to try anything else, they either don't have the time or don't want to take the time to do so.
Fourthly and most importantly, Open Source doesn't integrate well with existing systems. Especially Microsoft systems. I.e. with SharePoint.
Some Open Source applications do a best effort to do so, and some of them succeeded to some extend. But this, in my opinion, is where Open Source falls short. Although, this might be the case, it might not be Open Source's fault, it could be that proprietary software prevents Open Source to inter-operate.
Some companies have valid reasons why they won't go the Open Source route, whether its support, interoperability and familarity (why change processes which people know so well now, don't fix what isn't broken); others are just plainly stubborn.
Just wanted to add my 2c.
Your point: "Thirdly, they
Your point:"Thirdly, they only know about Microsoft and are scared to try anything else..."
is actually a bit mis-leading.
It is not that they are 'scared' as much as it will cost large amounts of money to switch not only users, but systems to another platform. (Unless the organizations of which you speak are small in size.)
In fact, I would go as far
In fact, I would go as far to say that 'Free' by the majority of this community is read as 'free' of charge. This is a huge hurdle for mainstream adoption, even if there is in fact 'no charge' for Hydraulic valves the software. Free as in freedom is almost always seen as secondary.I'm surprised that you
I'm surprised that you didn't mention planning - most commercial vendors will expose their plans either publicly or under NDA, and its very trying to deal with a community where the planning process is 'it depends what patches people send us'. This applies as much to fixes as to new features. If the product has a poor or non-existent planning process, its hard to plan its deployment and support.I'm also amused by the idea of 'qualified consultants'. Qualified how, precisely? By a desire to promote themselves as qualified? Even if they are competant and experienced in the product concerned, unless they have an ability to force fixes upstream and make them stick, their effectiveness will be limited to helping you fork the product and/or general handholding. Clearly a consultant can help you deploy effectively and avoid pitfalls, but that's not really the same as someone providing defect support.
Open Source Vs. Proprietary
It seems most of the comments are from developers or IT professionals of some sort.I am a college student, studying computer programming, but am here to say I can't right a single line of code. I know a little PHP, but not enough to write a single page without referencing tutorials, and I know XHTML.
I am deeply involved in an Open source project that has grown considerably over the past 4-5 months, and there have been many bugs as well as features that break when a new release is made. I can honestly say I have helped out a great deal in fixing these bugs because I have taken the time to get to know the software, the developers(to the extent I can anyway) and the setup. Its every day someone asks a question about a bug or a problem setting the application up. I do what I can to help, try to reproduce the problem and try to fix it. For my own knowledge, and to benefit the other users. Sure there is a community of people that use the app and discuss it, but there are far more "newbies" who don't know anything about it, and we help.
Every open source project has three groups, the developer(s), the users who just want to use the application, and the users who want to use the application and help out however they can. Proprietary software has the two groups in one, they have the developers and people to test and possibly isolate bugs. But when your paying people to make some form of software, they need to make money, which sometimes means more features and less focus on bugs that are affecting x% of the user base.
Open source projects are focused on the software, but they have to support their families, or themselves, by some form of income, so they can't spend as much time debugging and testing, thats where the community comes in.
I can say I've given up Open Office for MS Office, but because I need it for school. Open Office has problems rendering the documents, often cutting off the right side, I've done my part there to, but its out of my hands now. MS Office is supported by the school, so that is where I had to go.
And sorry for the novel.
There's no such a thing as a free lunch...
There's no such a thing as a free lunch... I know it's a cliché, but it's also true for Open Source Software (OSS). The product may be free, but support isn't, the learning curve isn't. There are many reasons why OSS isn't always the best choice.The price doesn't say all, companies look at the whole picture.
Or it could just be...
Image - like just about every other product in existence what you have says a lot about who you are. In a certain context I might use software X because it's edgy and cool, in a different context I might use software Y because its reliable and conservative.I am reminded of a company that forbade job seekers sending in their resumes in a certain format. This was a statement about how the company saw itself and the people it was looking to hire. Later on I believe, it relaxed this policy, making a statement of an entirely different nature.