Open-source software may not make major inroads into industry-specific enterprise applications, according to an IBM open-source guru.
The next 10 years will be "do or die" for this type of application, said Bob Sutor, vice president of open source and standards at IBM, in a LinuxWorld keynote address Wednesday. That was one of a series of 10-year predictions he made to mark the tenth anniversary of IBM embracing Linux.
So far, he said, there is little open-source software written for use in specific industries.
"I'm getting tired of waiting," Sutor said. "Either it's going to happen or it's not going to happen."
Many enterprises use general-purpose applications such as Mozilla Firefox, but few have industry-specific Linux applications, Sutor said. The public sector, especially education, offers glimmers of hope with software such as the Sakai collaboration and learning environment, he noted. For other industries, open source may take a long time to gain traction or may never gain it, Sutor said.
"You may believe that there will come a day where all software is free software, or open-source software ... [but] it's not tomorrow, and it's probably not next year, and it's probably not 10 years from now," he said.
One thing that has kept some enterprises from embracing open-source software is the proliferation of different licenses, Sutor said.
"When customers say 'I'm ready to use open source,' [they] don't want to see the license du jour," Sutor said. They won't tolerate a lot of ongoing change in the legal aspects of using a piece of software, he said.
Fortunately, out of approximately 60 open-source licenses approved by the OSI (Open Source Initiative), a handful of licenses are used for roughly 90 percent of open-source projects, Sutor said. Among them are Apache, Eclipse, Mozilla and versions of the GPL (General Public License) and Lesser GPL. Over the next 10 years, these will be refined and there will be less pressure to craft different types of licenses, he believes.
Meanwhile, Linux will be used in a wide variety of devices and various Internet-based services, such as cloud computing and SaaS (software as a service) and judged less as an operating system on desktops or even the x86 hardware platform, Sutor predicted.
"Linux may become much more widely used, but you won't know it. ... It's just there," Sutor said.
However, Linux will stay at the top of the heap for open-source platforms, with no new operating systems coming along by surprise to supersede it, Sutor said. He laid this to Linux's proven adaptability to different needs over time.
More news, commentary, and predictions from The Industry Standard:
- Prediction: Apple hits 18% share of US computer market by end of year
- Analysis: Neither Fortify nor Network World understand open source
- Analysis: Microsoft on Symbian's open-source move: Good luck with that
- News: Not all e-Bullion customers are worried by outage
- Special Feature: Where are they now? The Industry Standard tracks down 10 dot-coms from the Web bubble of the late 1990s
Reprinted with permission from IDG News Service. Story copyright 2008 IDG News Service Inc. All rights reserved.
Comments
I would suggest that Bob visits his optician, for there are many open-source industry-specific enterprise applications out there, should he care to look.
In Healthcare IT two high-quality FOSS enterprise applications that I've come across recently are Mirth ( http://www.mirthproject.org/ ) a cross-platform standards-compliant interface engine, and Tolven ( http://www.tolvenhealth.com ) an EHR solution.
I can't comment about other verticals, but I suggest that the parable of 'look, and ye shall find' is appropriate here.
IBM doesnt help...
Example, where I work we have hired some IBM consultants to interface two of our systems together. These are industry standard systems and there are 100's of other companies out there that have exactly the same interface problem as we do.
This interface application is an ideal open source product, as it has nothing to do with our core business.
IBM has at no point suggested implementing this as an open source solution. OK our mgt might still say no, but they haven't even had the conversation.
It would only take a few examples of this in my Industry for a snowball effect to start happening. IBM consultants are in the IDEAL situation to get those snowballs started.
I work in a very typical company in my sector, I do suggest lots of good ideas to Mgt, but its only when an external consultant comes in that mgt listen. Which is why IBM is in an ideal position.
my 2 cents.
If Mr. Sutor is so tired of waiting for those applications, why doesn't he move his employer to take the lead? The Tivoli products would be the ideal candidate. But this would probably hurt IBM's bottom line, so it can't be done.
I think the things that he really wants to see are AutoCAD, AutoDesk, and the Adobe suite. Many design houses would switch out their 100s of high cost OSes if only they could run the applications that are central too their business.
I don't know why AutoCAD hasn't moved. I can only imagine how many engineers are now getting into the market who run Linux desktops but must run VMWare and Windows to use 'the tool' they need to get their work done.
Come on Adobe, get your stuff together!
I don't think he is talking about having proprietary 3rd party developers like AutoDesk and Adobe release their products on Linux, I think he's tired of waiting for legitimate open source alternatives to their products. There is no decent open source CAD application, although GIMP is a marvelous alternative to Photoshop. Anyone who thinks the GIMP is quite there yet, obviously doesn't use it. The GIMP does at least 90-95% of what Photoshop can do, plus a handful of things Photoshop can't. The problem is that nobody is marketing the GIMP to the corporate world. This goes to show that even if we have good open source alternatives, getting people to start using them is harder than one would think. We NEED marketing.
OK, here's a LONG one:
IBM & Mr Sutor could do what for years I have been advocating (to and outside of IBM): MAKE LOTUS SMARTSUITE AN OPEN-SOURCE OPTION.
How hard can it BE? Don't TELL me about cannibalizing Notes/Domino. Don't TELL me about compilers and patents. DON'T TELL ME ABOUT MARKET SHARE! IBM seems all to happy to claim having 10 or 33 million corporate users of SmartSuite, yet totally short-shrifts/screws those of us who are NOT corporate users of SmartSuite.
To me, it's as simple as this:
1. Find and offer pay to worthy developers (inside of or outside of IBM/Lotus)who are former and agitated/longingly-waiting users of SmartSuite and tell them what is expected.
2. Make plans for 3 teams of 5 to 10 developers each: Each is given 4-6 weeks to play with the SmartSuite features, ONLY IN USER MODE, and tele-link them with assignments expected of office users and features asked for non-corporate end-users
2. Go back through your library of patents in SmartSuite owned/co-owned by people IBM's patent/development attorneys have been unable to identify for patent liability releases.
3. Make each team use SmartSuite on monitored/tamper-evident laptops (this is NOT the hack/decompile stage) to streamline the schedule of what they will need to attack and then write up reports of weaknesses vs OpenOffice.org, Corel, ABI Suite, ms office, etc
4. Plan to take the best people from each of the 3 teams and reduce to two teams which will be propositioned for full-time work which will demand they be sequestered for 6 months SOLID; TAKE CARE OF THEIR FAMILIES!
5. SHOW the Open Source developers the actual code stripped of the non-IBM code
6. Let them suggest dev tools to either re-write or encapsulate the code in OS-agnostic languages (your GOAL is to supplant ms office simply by making the suite accessible to any OS not engaging in anti-competitive tactics)
7. Carry out the sequester plan, and in mid-stream, let the Unix/Linux/Apple/windows users get a peak at it, and for HEAVEN'S SAKE, don't let it be Symphony. Toss that out and start updating SmartSuite! No more "maintenance fixes...
8. Price it for $100 for corporate and support-seeking users; $0-25 for end-users. SmartSuite IS WORTH PAYING FOR, so it need NOT be totally free, unless you want academic uptake to roar ahead.
IBM has for YEARS sat on its butt, squirming and hawing about patents, market size, and so on. Instead of mulling openly, OPENLY DO SOMETHING that is not an insult to SmartSuite's features.
One reason I am so irritated is because:
-- SmartSuite needs for Lotus Approach a stand-alone executable for the database & apps, similar to MS Access
-- Approach needs sliders on its detail table
-- SmartSuite needs better web intelligence so that its WYSIWYG heritage extends outside the desktop
-- WordPro needs updating to handle more file formats
-- SmartSuite has a vibrant end-user community that needs a reinvigorating surge of love for IBM
-- SmartSuite (particularly Lotus Word Pro and Lotus Approach) hands-down beat OO.o in probably 60% of mission-critical areas that make document creation, editing, and usage much easier, and OO.o devs trapped in Not-Invented-Here syndrom blithely and arrogantly ignore the SmartSuite features that could have softened the pain of SmartSuite users continually INSULTED by half-baked features in OO.o that CLAIM to address issues but do NOT (master/detail documents in text editing; WYSIWYG database forms, templates and applications)
Most of all, my personal stake in SmartSuite is my screenplay/dialog-tracking database and front end, and until SmartSuite is rejuvenated without destroying its lineage and appearance, I cannot realistically deploy for user testing the stuff I want to see thrive in SmartSuite.
IBM, I am NOT a programmer, yet I feel that i'll have to learn to trust (and overcome my fear of being screwed over by unscrupulous) developers/hackers to help me shunt my Approach-based apps to other dev tools such as Kdevelop or QTDesigner, which ask far more of me than Lotus Approach does.
Not, if QT/Trolltech or Open Source had/has a clandestine merit-based group that looks for paranoid people like me to submit apps for dual-license development, I'd like to know. If I release to the various testing/submission repositories what I'm doing (before it's updated to enter the real world), i'll just be preempted by other commercial apps. Personally, I'd like to see Celtx and StoryLines benefit from my work, but I'd prefer to continue on my own, too, in Lotus Approach if it ever is resurrected with a long-term support path.
Personally, I think (and, call me grousing or paranoid) that IBM is polluted by a number pro-ms-stock-holding employees who will do or say whatever it takes to keep SmartSuite from regaining a market position it used to have. I also suspect that IBM and ms probably entered into some agreement that would promise that ms would not assail IBM in some way so long as IBM lets SmartSuite languish. I also fear that IBM is acting so OpenSource-friendly that it is cowering out of fear of PR wrath by Open Source types who sensually gyrate at night over OO.o, when most of them haven't objectively USED SmartSuite to know what real weaknesses OO.o has relative to SmartSuite.
Now, if my Karma gets shot to hell, so be it. I'm sick and tired of seeing SmartSuite sit only in a languishing corporate-product position. If IBM wants to help us SmartSuite users and simultaneously NOT kill off OO.o, then dammit, put away the gauntlet and clubs and maces and collaborate with SUN to merge the best of both Suites, (OO.o's cleaner code base portions) with SmartSuites cleaner code portions PLUS SmartSuite's user interface. Or, give users the choice of interface flavor, especially access to WordPro's "Sections" and "Divisions" tabbed-interface features with a SANER view of multiple docs instead of the ms word-mimicking ruled line crap and lame print-preview, WP's multiple document views (no, not tiling and screen splitting you OO.o users), and to Lotus Approach's WYSIWYG, drag-n-drop forms creation, more-than-adequate WYSIWYG charts and reports creations.
But, PLEASE, extend and enhance Approach's database capabilities, with native support for Postgres and MySQL (yeh, SUN again), and beef up the already-pretty decent (but vastly superior to Base's templates) offerings of templates and applications.
Mr. Sutor's assistants, how many more passionate, painful pleas must you hear before acting? No, Symphony is NOT non-insulting. It should have been released as an updated, recognizable extension of SmartSuite, not a half-baked clone of the pseudo-underwhelming OO.o. As OO.o is, it will NEVER, EVER be able to 1-for-1 be a drop-in replacement of SmartSuite for me, and i suspect not for maybe 60% of creative users of SmartSuite. Besides, you could give Sue Sloan and her colleagues a whole new world to swim and bask in.
I have heard of OSI. But, I'd prefer IBM sponsor Smart-Suite-based hobbyists and developers who make apps that have great potential to attract existing and new users to redouble their use of SmartSuite. Heck, Anthemion Software writes Storylines, and they support OpenOffice.org. Their app is wonderfully simple, elegant and EASY to use. Mine is thick, somewhat difficult to use, but still is powerful in ways of reporting and linked tables manipulation of data presentation.
Also, I forgot to mention that Vista does a debilitating number on Lotus Approach.
Please, please do something about that. 1-2-3 STILL has no built-in support for mouse wheel scrolling.
I would like to be able to re-develop my app so it can be considered as an app and template (in free and for-pay versions) in SmartSuite. My apps in L/A ran FINE in Win98. Vista brings my dozens of linked tables and dozens of forms and searches to a freaking CRAWL. Unfortunately for me, I have not been able to get neither VirtualBox nor other emulators (other than Win4Lin of old, as in the 2002/3-ish versions) to run win98. Win98 is leaner, faster, and friendlier to SmartSuite, and it seems IBM has dropped the ball and let ms get away with a release SmartSuite-crippling code.
"5. SHOW the Open Source developers the actual code stripped of the non-IBM code"
Actually, I meant
5. SHOW the Open Source developers the actual code stripped of the non-IBM code (for patent infringement suit avoidance) and have them use modern tools not conceived of back when SmartSuite was patented, and therefore enable the new devs to restore the lost functionality
In my opinion the necessary conditions for "grassroots" industry-vertical open-source efforts are: non-trivial software needs, and critical mass of software developers as users. We here in finance have QuickFIX (and /J), QuantLib, Marketcetera (that's us) and others. Other industries like medicine, education and others seem like fertile ground as well. Of course there are other ways in which successful open-source projects get started, the open-sourcing of an existing proprietary product, or by multi-lateral agreement of established industry players. I think it is these mechanisms that have yet to come up to speed in many verticals.
It goes far, far beyond Adobe and AutoCAD. Just about every industry and profession has some kind of highly specialized software. Take my wife's industry, clothing print design. There are special (some extremely expensive) apps for designing knit fabrics; for designing print fabrics; for doing color separations, etc. (Yes, they use Photoshop some, but they use other much more specialized stuff as well.) The entire printing industry has their own set of apps. The accounting world alone has tons of special and custom apps, it ain't all Quicken and Quickbooks. Tax accountants have special apps, inventory accounting has special apps. It goes on and on. And guess what? 99% of this stuff is on Windows.
Yeah, it would be nice to have a port of CS3 and AutCAD, that would get people's attention. But as popular as those two packages are, that's just a drop in the bucket, really. Kind of gives you a new appreciation for the importance of Wine and virtualization, doesn't it?
Many of these specialized programs are hideously expensive, and that presents an opening to FOSS. What is really needed is for ISV's to start targeting these specialized industries, using the tried and true philosophy of Open Source: give away the app, remove barriers to entry, allow everyone to work on it, grab a foothold in the market by being the no-cost alternative, then charge for support, training, service. Works for Red Hat and a ton of others. It can work in specialized industries too.
One thing I don't understand about remarks in this article: why are open source licenses more scary than proprietary licenses, where you pay by the seat, have to worry about keeping licenses up to date, can get sued by the software police for having unauthorized copies, have to insure your hardware dongle, without which the app won't work, etc etc etc?
The history of open source is a textbook lesson in economics and politics. The apps that go open, and the way they do so, are exactly the ones you'd expect. Open source may have been around for a while now, but the human condition has been around far longer.
The source code being open is the least of most user's desires. It's only the start of a very long set of needs, such as documentation, bug fixes, security patches, ports, updates, support, trusted material repository, etc. The average developer doesn't find any of those fun, especially for free. As much as I respect and admire Stallman's principles, a solution for typical end users has to be more encompassing. "Open source" isn't even the right name for it -- more like "Open maintenance".
I'd give vendors obseletion warnings. Take one central piece of software, round up some capital, and inform the vendor with an open letter that if they don't make a linux version of their software to meet your needs, you will, and yours will be GPLv2.
But then, I don't know if CEO's of large companies are used to the idea of using FOSS to play hardball..
the problem with linux and dont get me wrong i use it and love it is the mentality round it. it stuns me that just because the o/s it self is open source that *everything* that goes with it like games apps atc must be open source as well.
why not do a port popular programs like what ibm refers to linux and keep it closes source and charge a fee for customers to buy it... theres an idea. i wonder where i have seen this idea before????
All I can say is that the guy should do three things:
1) see what's really out there in OpenSource... you'd be surprised at what's out there. Unfortunately, that doesn't include a MCAD app I want to use. (though the ECAD stuff actually looks pretty good)
2) Offer IBM money to subsidize the development and porting of industry-specific applications and help with marketing exposure afterwards.. A great many niche developers aren't porting because they are uncertain about market / profit from Linux users, if there are only a few customers in Win/OSX, why develop for Linux?
3) Create a site that indexes all the industry-specific Linux / Open Source apps out there (starting with a team to scour google for them) and buy ads for it.
Note that whining is NOT on the list of things I recommend. If IBM wants to be able to recommend OpenSource for their customers for everything, it's a problem they can fix by throwing a few million dollars at it.
It's mind-boggling how some people just can't be satisfied. The stuff that's already free makes verticals that much easier to produce. But noooo... he doesn't want to do even that work. Laziness knows no bounds. Tell me, when you say open source is nice because it's free, are you actually saying it's nice because someone else did all the work? Let's be honest and call a spade a spade.
Why aren't people voluntarily working on industry specific applications? What an fning stupid question. Giving away software that took thousands of man-hours is just plain dumb. Oh but Red Hat does it! Yes but Red Hat is in an industry where support is critical. How often have you felt the need to contact Adobe by phone? Would you pay an annual fee to do this? Of course not. Adobe would go bankrupt if they open-sourced their software.
Time to face a harsh reality: Open-source software is like socialism, it works better in theory than practice.
Anyone who thinks the GIMP is quite there yet, obviously doesn't use it. The GIMP does at least 90-95% of what Photoshop can do, plus a handful of things Photoshop can't. The problem is that nobody is marketing the GIMP to the corporate world.
This is wrong and shows a marked lack of knowledge about digital image editing and the history of the gimp. A couple of VFX houses in the late 90's took the gimp and added HDR support to it so it would be useful in the motion picture industry as a dust-brushing tool, etc. The gimp devs liked the idea but decided to implement their own solution instead of using the existing code. Ten years later and the Gimp still doesn't support HDR, barely supports CMYK, and don't even get me started about blending modes... Hollywood and graphics professionals know all about the gimp, they just _can't_ use it because of its technical failings.
Post new comment