IBM Exec Bemoans Lack of Industry-Specific Linux Apps 302
Ian Lamont writes "Bob Sutor, IBM's vice president of open source and standards, used his keynote appearance at LinuxWorld to complain about the lack of industry-specific open source apps. Despite some encouraging signs in the educational field with Sakai, Sutor said that he was 'tired of waiting' for specialized applications to appear in other sectors, adding that the proliferation of different licenses — and changing legal requirements for using the same software over time — is holding some businesses back from using open source applications."
Just a thought... (Score:5, Insightful)
Maybe IBM might wish to consider throwing some money at someone developing those apps.
Re:Just a thought... (Score:4, Insightful)
Ah.. I was about to say that.
In the world of open and free software, people develop software only becuase they are motivated to make one usually for personal gain.
Unless someone motivates them, you wont expect them to make softwares that they do not plan to use. There is no lack of talent.
We could have seen a GTA clone on the Linux front by now, but we don't for two reasons, no one is sponsoring any project and no one really is motivated enough to start one.
If someone throws in enough money, they will get what they want.
Re:Awesome! (Score:3, Insightful)
So, he's 'tired of waiting', eh? I find that waiting for code to appear in front of me is a very poor method of production.
Very true (Score:5, Insightful)
This is extremely true. We are a 95% Linux site- servers and clients, and it is nearly impossible to find industry specific Linux applications. The kicker is- we would be happy with even CLOSED source, commercial applications. There are needs for specific types of software for just about every industry: medical, dental, car repair, musician, distribution, auditing, education, scientific, etc, you name it.
But what is worse is the lack of applications for NON industry-specific business applications. EVERY business needs payroll, GL, AP, and AR. At least there are some choices with those, but surprisingly few. When you are trying to use Linux on the server AND client side, it is quite a challenge. I know. We are trying to get payroll, GL, AP, and AR *right now*. Fortunately, we found Southware- impressive stuff, although it is expensive, and text-based only.
And no, WINE doesn't cut it- we need real support. And no, we are not able to develop such apps ourselves. And no, we don't have the time or resources to start such development projects. But we have money and want to spend it on something that allows us to use the platform of our choice.
What is needed is a COMPELLING, modern, cross-platform, open-source, GUI, business application development environment. It is 2008. It shouldn't be difficult anymore for existing vendors to port their software to all major platforms and support them. Eventually this will happen, at least with a few major vendors. But in the meanwhile, Linux is suffering horribly in business outside of being web/file/print/compute servers.
Notes? (Score:4, Insightful)
Can you run notes without wine?
Myopic execs (Score:2, Insightful)
Why Sakai? (Score:5, Insightful)
I wonder why the example mentioned is Sakai? Anyone working with Open Source Software in the education knows very well that the real success story is Moodle [moodle.org]. Unfortunately the article doesn't go into details in this regard, so I'm left just thinking that it's another case of Big Organization Blindness.
Re:GPL not strong enough. (Score:5, Insightful)
Anything for which they don't distribute binaries they don't have to distribute source. Improving the code only for in-house use means not needing to give any improvements back.
Lack isn't a big surprise... (Score:4, Insightful)
It's quite understandable that a business that needs some industry-spesific application and comes to the conclusion that they must pay to have it made, won't make it open source. It's kind of against traditional business sense to pay for something, and then give it away so that your competitors get access to it for free... And even the usual open source arguments of getting "community development" benefit won't fly here. It's unlikely the competitors would "give back" their modifications/fixes with GPL, if they are just using the software internally, as they probably would with a lot of "industry specific" software. And even if they're distributing it, they probably wouldn't put any effort in "giving back", they'd do it only if somebody found out and actively requested they comply with the license.
I think a yet different kind of licence is needed... It could for example require at least read-only online version control repository for anybody who compiles the software from source and uses it in any way, and also require that the internal version is compiled unpatched from the public repository, and that if some closed code is used with it, the open code must not depend on the closed code (only the other way around). Or something like that, making sure that whatever competitors do with the open software, they can't hide their improvements easily.
changing legal requirements... (Score:5, Insightful)
This "changing legal requirements for running the same software" bit is obviously a crack at the GPLv2 to GPLv3 transition and similar license changes some projects go through.
Anyone speaking on behalf of IBM should realize that the EULAs for proprietary software change between versions all the time. Hell, many of them allow the publisher to retroactively change the terms of the EULA by posting the new wording to a website.
It isn't really fair to call one group out as being inferior for having clearer licenses that just happen to change more noticeably. It's the changes you don't notice until you're being sued that are the really painful ones, after all.
It's a chicken and egg problem (Score:5, Insightful)
Forget about licenses for a minute. There's no law that says "Applications running on Linux must be GPL'd".
The thing is, nobody's going to develop a Linux equivalent of, say, your favourite accounting/payroll application for free. It's as boring as hell to write and doesn't really directly benefit the person writing it so it is (and is likely to remain) the enclave of commercial software houses. GnuCash and the like don't really count here - I'm talking about the kind of application used by business accounts departments with at least a couple of people working fulltime.
And seeing as porting the software to Linux costs money - you've got to pay those developers somehow - you're not going to do it until such time as you've got a reasonable number of customers saying "Do you have a Linux version available? Because if not we're going elsewhere." (And that threat needs to be credible. Migrating your entire accounting department to another application isn't something you do lightly.)
Similarly, the businesses using these applications are more concerned about being able to run them than the OS they run on. "I can save you a bunch of money on Windows licenses" doesn't sound so attractive when you add "...but you won't be able to use products X, Y and Z on which you've built your business, and right now there is no credible alternative."
It's a simple case of economics (Score:4, Insightful)
I know it's very much a chicken and egg situation, but a small developer writing software for specific smaller market sectors will probably only realistically be able to afford to develop for the one platform, and that platform is going to have to be the one used by 95% of businesses - Windows.
There's the potential for this to change with the (relatively) recent arrival a number of good RIA solutions, but these are essentially platform neutral and so aren't likely to help Linux adoption.
Any smaller business trying to find a nearby support company who are fluent in and can advise them on Linux systems will often struggle to do so, and the majority of others will of course advise them (often with generous helpings of FUD) against going the Linux route, as it's not in their (the support company's) best interests.
I'm a user and huge fan of Linux and FLOSS in general, but as a software developer who's pretty much always worked in niche markets, I'm also realistic about these things. I have to be - I need to earn a living.
Re:Just a thought... (Score:2, Insightful)
Thing is even if they did (are?)
a) we're talking about a lot of money to fill even a small proportion of the missing apps
b) for the health of the Linux software eco-system there needs to be many developers of such apps, not just IBM and a few other big corporations
Thing to realise is this is a keynote speech at a conference. Despite the tone of "I'm tired of waiting" he's not really talking about what applications he personally, or IBM as an organisation want/need, he's trying to throw out a challenge to the community in general. I'm sure if there are specific apps that IBM need on Linux they are willing and able to develop them.
Indeed "I'm tired of waiting" sounds to me like an attention-grabbing phrase rather than anything else.
Legal requirements for using software? (Score:3, Insightful)
This is an interesting new kind of FUD about free software: that there are legal requirements for 'using it'. The GPL is quite explicit: running the program is not restricted; unless you are modifying or distributing the software, you can do absolutely what you like. Every other free software licence is the same. This message needs to be spread more widely.
Re:Just a thought... (Score:5, Insightful)
I guess my point is, this is one point where FOSS seems to break down. A lot of industry specific apps are (at least viewed as) highly proprietary and simply won't be prolific in FOSS. This isn't simply a case of no one being interested, or a lack of financial support. More of a NDA / confidentiality / trade secret realm issue perhaps.
Re:Just a thought... (Score:3, Insightful)
Most package managers aren't going to be happy with handling distribution packages for COTS software that is mostly pre-compiled binaries.
That's a burden for COTS distribution unless you're Java-based.
Plus, there's that whole Gtk/Qt mess (which one should I use), testing in multiple languages, and the COTS package having to support multiple platforms.
In the COTS space, for industry-specific apps... Mac is finally getting some attention. However, plain jane dumb users (the kind in OfficeSpace) aren't going to be dropping to the shell and running sh configure, make, make install as root.
COTS developers need to be convinced that it is worth their while to go through the pain of multiple distributions.
Maybe in the OSS world, someone could write a better autoconfigure wrapper (maybe at the dist level?) that can let users know whether their systems meet the minimum requirements to run a COTS app, and COTS developers can "trust" implicitly that requirements checker is always going to be there; and to take care of shortfalls [user is missing some commercial or OSS libs, or needs to upgrade].
A secretary can't understand the output from sh configure and then run to rpmfind or her disto to get libfsckinguselessblahblah.1.4.so
She needs the kintergarden interface that holds her hand and gently walks her through installing missing libs that a COTS developer linked into his/her project.
Throwing money at a problem isn't enough (Score:3, Insightful)
For example, if someone wanted to build an Open Source CAD package, it would be an expensive and time-consuming chore and the odds of other people picking it up and building on it themselves would be very low. For every successful growing F/OSS project out there, another 100+ out there have crashed and burned for a variety of reasons.
What needs to happen is a bunch of people, preferably businesses, with similar interests should hire programmers and developers to work together with a common goal in mind... and that goal shouldn't be to "sell software" either. It should be to build the tool that works the way they need it to work. They can give that tool to other people in their fields so they can read and write their files and grow on from there.
Once upon a time, when a business needed software, they hired programmers to write and maintain them. And an interesting reality is that it's quite likely that programmers are cheaper than software licenses. I could be wrong, but the price of software seems to be ridiculously high. I know my own office spends a ridiculous amount of money for an AutoCAD seat... every year. Maybe it's not as much as the cost of a drafter, but still. After enough seats are purchased and renewed annually in a large enough setting, you could keep two or three good programmers on staff. And if industry related businesses could team their smaller programming staff together, a larger team would emerge and could actually put some interesting things together for a price that is less than or equal to what they are paying now for software that doesn't quite ever meet their needs.
So if industry specific software is needed, the industry needs to get off its ass and stop wasting money on commercial software and put it into programmers.
Mod parent up (Score:3, Insightful)
It's the lack of non industry specific apps - HR, Finance/Accounting. You mentioned that sort of stuff. Stuff that every fair sized company needs.
Then there's also stuff like calendaring/scheduling.
Lots of people are resorting to MS Outlook and Exchange because the OSS options just aren't compelling.
You need an OSS client AND server that work well together for email, calendaring, etc AND the client must run on Windows/OSX - because that's what Microsoft Office runs on.
Because OpenOffice sucks. It does - it takes ages to launch, and there are plenty of WTF bugs - just having some normal character at the end of a line messed up formatting on my OpenOffice Impress presentation - I had to _downgrade_ to an _earlier_ version to have that work right - but that means accepting OTHER bugs.
Say what you like about MSOffice, but the 97 and 2003 versions are still better than OOo. The latest? Well let's say the latest MS Office is an opportunity for OOo to gain significant market share if they get their act together.
And Vista should be a dream come true for Desktop Linux. But..
Re:Holding me back (Score:3, Insightful)
Either this is going to be modded up or its going to be modded down.
Re:GPL not strong enough. (Score:1, Insightful)
Re:How about porting? (Score:4, Insightful)
Games do much better on the console.
Some games do better on the console. Some games are unplayable on a console.
Adobe's been receiving requests for years and years about porting its apps to Linux. There's a market, but somehow the normal rules supply and demand appear to be interfered with here...
No, your ability to reason has been interfered with. There is no market for Linux versions of Adobe apps. There cannot be a market for what doesn't exist. There can be a demand. Demand does not create a market. Only the combination of supply and demand create a market.
If you wish to say that Adobe does not feel the demand is great enough to bother creating the market, than say so. Don't try to insinuate that Adobe hates open systems or that they are in bed with MS.
Re:Just a thought... (Score:2, Insightful)
I'm you are so fucking smart why don't you invest your life savings in an industry-specific open source app? No, I didn't think so either.
Re:Throwing money at a problem isn't enough (Score:2, Insightful)
Why are competitors going to want to help each other, exactly?
Re:Just a thought... (Score:2, Insightful)
I realize IBM is a big company but some of those apps are not cheap to develop. Example, Autodesk invested 10's of millions of dollars in developing the core of Inventor, I'm not talking about all the UI, spit and polish, along with all the changes that have happened over the years.
There are a lot of apps missing for Linux:
Mid-range CAD (AutoCAD, Inventor, SolidWorks, etc)
POS (Point of Sale) for the small/medium business. IBM does have Linux based POS solutions, however, they are exorbitantly expensive for small and medium sized businesses. For small, ma-n-pa type shops they want to have a "PC" that has POS software on it, but also word processing, spreadsheet, internet access, probably even accounting, etc. Not just a POS system and nothing else.
These are just two really, really big examples I can think of.
Re:Just a thought... (Score:3, Insightful)
They already throw a lot of money at Linux.
What's really needed is for those businesses stuck on proprietary windows software, spending thousands or tens of thousands of dollars every year for their niche software to get together and hire some coders to put together a free solution for the benefit of the whole industry. It's boring stuff, no one is going to write this stuff for free. But since they're paying someone already, they might as well get source code they can modify to their own needs. And yes, the competition would get to use it free too, but for an essential piece of business software everyone in the industry will be using it, so there's no competitive advantage there, everyone comes out ahead.
Re:Very true (Score:3, Insightful)
The genius of Visual Basic was that it packaged up everything business types needed to write a GUI-based app for Windows -- a language (with simple syntax), a GUI toolkit, an interface designer, an IDE, and a compiler -- in one package.
This meant you could go from start to finish without having to stop and make decisions all the time about how to fit the next piece into your puzzle. When you wanted to design your GUI, you just used VB's tools for that. When you wanted to write your business logic, you used VB's tools for that too. You never had to step out of the VB environment to figure out how to accomplish something.
Qt isn't really comparable -- you can give someone Qt, but then they have a whole chain of other decisions to make before they have a working app. Visual Basic made all those decisions for you, so your average business analyst-type could concentrate on their app and not on debating the merits of Eclipse vs. Netbeans or C++ vs. Python or what have you.
(You might object that most of the applications that this "all in one" approach made possible are crap, and from a technical perspective you'd be right. But they worked, mostly, and businesses came to depend on them.)
To the best of my knowledge there's nothing really comparable to this for Linux. You can build a stack similar to the stack VB gave you, but you have to build it, and that's a barrier to entry that's too high for the VB audience to scale.
The license complaint is 100% bullshit (Score:3, Insightful)
And then, on top of all that, to buy or use the Free Software in question, it usually isn't even necessary to become bound by the license, so the license becomes irrelevant anyway. If you're going to create a derived work of one of these products (and the need of customization and maintenance isn't all that rare), then the license starts to matter, but again, even the most restrictive (GPL3?) is vastly easier to comply with, than anything in the proprietary world. Where's your customized version of MS Windows or Adobe Photoshop? Oh, that's right, you don't have one, because they haven't offered you a license that lets you maintain those products.
I can't find the tiniest amount of merit in that complaint. There's absolutely not a single shred of truth in it.
Re:Just a thought... (Score:3, Insightful)
I'm not interested in winning support on the basis of my friendly demeanor. I don't have one. I'm interested in supporting individuals by providing them practical infrastructures that allow them to walk away from their feudal masters, and I'm good at it. Personally, I think capitalists should be executed for crimes against humanity, but I'm pragmatic enough to recognize that to truly win, the social organization they provide must be replaced. Once that's done, they'll be rendered irrelevant.
Yes, I'm participating in all the ugly shit in the world, because I'm trapped like everyone else. But there's a difference between making compromises and actively enriching myself on the backs of billions by removing their choices. It's hard to live by ideals in a world so corrupt, but you do what you can.
Re:Just a thought... (Score:3, Insightful)
I'm tired of waiting.... so I'm going to try to goad you developers into starting a few new projects.
I presume that Sutor meant "industry software" as narrow specialized tools used by many companies internally.
But the Sutor seems to not to get the problem with the industry software. (Or he is intentionally pitching at lack of backward compatibility in Vista). Such software has few buyers and often developed with very sophisticated model called "snow ball." 20 years ago, somebody written little tool for DOS to do some work. Then somebody added another function. Then somebody sold it to another company. Then somebody added GUI. Then - Windows 1/2/3 support (because it is backward compatible). Then - Win95 support (because it is backward compatible). Then - support to run in backward compatible mode on Win2k.
As you can guess, most of industrial software look and work horribly. But they are used mostly by engineers who only care about work being done - not how it is done.
The point now, is that the industrial software also has to be rewritten for Vista. It is not all that rosy now in Windows land too. But I heard already that some backward compatible toolkits are made now to convert code developed for Win95/2k/XP to run on Vista. So I guess the ball would keep rolling for some long time.