Slashdot Mirror
The Death of Licensed Enterprise Software?
tfsm writes "Andy Singleton wrote a short, interesting article about the looming death of traditionally licensed, proprietary, enterprise software over at The IT Manager's Journal. In it, he talks about the declining revenues of software giants such as Siebel. There are several causes, but one, he suggests, is erosion from Open Source offerings."
Great articles. No sales numbers. No real explination given. Just lots of guesses and assumptions my some guy. I'm impressed. About all I've learned is that Siebel's licensing revenue is down. That, and it's written by a guy whose job is to sell software to big companies. Wow.
I don't respond to AC's.
In addition, the traditional "support" argument is pretty much out of the picture these days.
If you're talking about boxed software then support is limited to a "knowledge base" database and rudimentary and usually dire scripted phone support.
Support for customised applications is expensive and no different to the support contract you are likely to get from a good consultancy that has built your enterprise application on an open source platform.
Do it Yourself. This is the keystone for future business success.
You don't run a business, do you? Any business person can tell you that this is 100% wrong. You should only design your own software if 1. Your needs can't be filled by off-the-shelf stuff 2. There's some kind of value or competitive advantage to doing it yourself 3. You can afford it.
Unless you're a software company, software is just a tool like any others. You may as well have said that the only way to success in any business is to build your own trucks instead of simply bying them.
I don't respond to AC's.
Rather than a product. The problem with software (any information) as a product is that there is no scarcity, it's easy to copy and make more.
Markets require a supply and a demand, to make any information a product rather than a service you have to find a way to limit supply of something which isn't naturally scarce, licenses, keys, dongles etc. Without these, limitations the supply increases to infinity and the price therefore tends towards zero.
You may not like him, but Richard Stallman is a bloody clever bloke. The GPL and similar basically eliminate the artificial scarcity limitations imposed by most commercial software vendors.
Deleted
This trend is definitely true in smaller companies. Why spend thousands of dollars on proprietary software when you can get an open source project for free that you can modify to your heart's content? Granted, you're going to spend time and money to make those modifications, but it can be worth it when you get exactly what you want/need.
Compared to 5 to 10 years ago, the number of open source software apps available now is mind-blowing. So much so that whenever we are researching and deploying a new application, we immediately go looking for the open source one. The proprietary version is a last resort.
Whoever Has the Most Toys Wins!
The commoditisation of software that open source represents is a rising tide. There are two ways to handle a rising tide: float or sink.
Siebel, like many big software firms, are unable to float. They don't use open source for their processes, so don't benefit from it. They are stuck in a niche, so are basically anchored to the sea floor while the water rises around them. Their customers have the choice of remaining anchored with them, and drowing as well, or cutting free and floating.
It's a bit sad if you're in the position of the drowing man. But it's been the same in Big Auto, Big Steel, Big Textile, Big Science, Big Pharma, Big Business... competition is a tough game.
The smart money is on those firms that learn to float. IBM, CA, Novell, Apple. Maybe Sun and SAP. Apparently not Siebel, definitely not Microsoft.
My blog
I couldn't agree more. If you're not in the business of building software, you shouldn't be building software. Same goes for trucks, warehouses, cranes, desks, and anything else your business may require.
Hell, even if you *are* in the business of building software, you may well be better off buying it in instead. My company is a software house (we do web-apps), there's no way we'd write our own web server or RDBMs, we'll get one from a third party (be that mySQL, MS SQL Server or Oracle, etc).
It's official. Most of you are morons.
it shouldn't be more expensive (even though, in the past, it has been). and the point of this article is that, in fact, software is getting easier, or else there wouldn't be so much open source software supplanting 'old-school solutions' in the first place ...
...
OK, I run a retail store. Are you suggesting that, instead of buying a working POS solution for about $800/workstation, that I should make my own? You can't really be serious...
if your business is selling pigs, and you need software to control the sale of those pigs, it is far better for you to have your own, purpose-built, customized software for the selling of pigs, than it is to 'copy someone elses model'.
Why?
it shouldn't be more expensive (even though, in the past, it has been). and the point of this article is that, in fact, software is getting easier, or else there wouldn't be so much open source software supplanting 'old-school solutions' in the first place
Fine. Who do I contact to write a point of sale system that handles inventory, purchase orders, vouchers, printing tags (multiple formats), printing receipts (multiple formats), supports all major POS software, has integrated credit card processing, tracks customer purchases, and seamlessly transfers data to/from Quickbooks for $2400 (we have 3 workstations)?
I don't respond to AC's.
That isn't quite correct. If someone comes up with a nifty utility to base their business on, then yeah, the OSS community will duplicate it in no time, but then, so will commercial vendors.
It seems most of these complaints come from companies who charge money for the most trivial of crap, such as this password generator. Nevermind that it has one of the worst user interfaces ever designed (look at that screenshot), it's a freaking PASSWORD GENERATOR! Trivial software is trivial to reimplement. It's offensive that someone would even charge money for this.
Now if a company develops something non-trivial, for which there aren't already a thousand similar products, this shouldn't be a problem. For example, just try finding an F/OSS product that can compete with 3D Studio Max or Maya. Blender isn't even in the same league. Photoshop? The Gimp is neat for web logos or hobbiest graphics, but doesn't even fully support the most fundamental Photoshop features such as native CMYK color.
Siebel Systems makes non-trivial software, but it is only non-trivial in that it is large. It isn't innovative; it's just a lot of work. I don't know of any OSS products that compete with, for example, their customer management software, but if there are, I would not doubt that it is because Siebel's stuff sucks (I've used it), and some smart developer got fed up and decided to show Siebel how it's done. If they do a better job, should we feel sorry for poor Siebel for losing revenue to the F/OSS guy, or should we root for the OSS project because any multi-billion dollar company which can't make a better project than a handful of F/OSS programmers needs to die?
Another example is the game market. There are neat OSS technologies such as the Irrlicht engine, but Itari and Blizzard aren't exactly concerned about F/OSS games taking over their market. When's the last time you played an open-source game which was even comparable to Farcry, Starcraft, or Alpha Centauri in terms of refinement, scale, and fun factor?
With all that said, I don't see how F/OSS is any different than another commercial competitor. An intelligently run business targets their product to account for competitors' weaknesses and tries to downplay its strengths. Seems to be working for Microsoft, and every single one of their core products have powerful and mature F/OSS competitors, yet their revenue has grown every year.
Specifically, the F/OSS community may be great at making low-level technical stuff, such as libraries, web servers, and DBMS software, but it isn't very good at polishing user interfaces (compare Visual C++ to KDevelop or Anjuta, though this being Slashdot will probably prefer the latter two regardless), at making high-end enterprise software (MySQL is neat but it can't even touch Teradata), or making the absolute highest-performance software (Apache is sort of fast, but Zeus and even recent versions of IIS can blow it away, especially in static page serving [That said, most corporations are even worse at making performance software, using bloatware tools such as MFC to make bloatware apps such as Norton Utilities]).
In short, the reasons given sound like the kind of reasons given by the kind of companies that make password generators or horribly poor quality customer management software and then complain that the F/OSS community is stealing your marketshare. Hell,
Computer Science is no more about computers than astronomy is about telescopes. --E. W. Dijkstra
Read the author's bio:
Andy Singleton is president of Needham, Mass.-based Assembla, which brings "inspired by open source" applications and development processes to enterprise software.
His business is implementing enterprise open source software. Of course he's going to say that that's the future.
I think the key point of your argument was this:
Pay for a contract and you will get highly knowlegable engineers to solve whatever problem that crops up within the confines of the contract.
This is true whether the software in question is a based on closed-source or open-source. You need to find a vendor who will meet your business needs first; whether that solution is closed- or open-source should be a secondary consideration
Open-source software allows more businesses to get into the "Enterprise Software" business by using existing (Free) products that they're free to tailor to customer requirements, and the customer is free to find a new consultancy if the old one goes under / is bought out / starts to suck.
One thing I noticed about a lot of these packages is that they often have terrible underlying data models, or at least data models that do not fit a particular business well. They tend to throw indirection and duplication at the problem. Perhaps this would not be a big problem if the package hid the underlying mess from the users, but if you ever want to extract and use data from such monsters, which is a common request, you have my sympethy.
If you want to make a better tool, allow one to model the particular company, sticking with certain conventions for hooks into the package.
Or better yet, sell development and expertise to help companies build one to fit their own company. In other words, become a domain expert company instead of software box company, and market that expertise. "We know how to build sales-force systems" instead of selling a pre-packaged blob of software. Such a company could still sell software, but in bits and pieces or as part of a bigger semi-custom-built package. Build a Lego kit that fits sales-force software (for instance) instead of the whole deal itself.
Domain specialists seems to be a missing software-related niche.
Table-ized A.I.