WS-RetailIndustry
It talks about the benefits to retailers of outsourcing IT solutions (particularly fulfillment systems), and I think web services is something that enables organizations to easily outsource entire IT systems and processes.
I'd suggest everyone who needs some real-world, applicable evidence of the benefits of web services to read it, since most of the time you just hear the generic "we've used web services and they're just great.... .... ....?"
more links
Harry Fuecks has PHP predictions for 2005, and even though I agree with his premise to date, I don't like his characterization of SOA. I think PHP developers should seriously consider at least learning about SOA so they know when to apply it, even though it doesn't apply to every app or script, it's the best tool that exists for its purpose.
Web Services are being used to load live data into Excel spreadsheets, DUH!
that's really all I can report now. I did catch up on some other web services and xml news like XLink and such, but none of it really hit my area.
</div>
great OS read
"It [Open Source] is only good for nondifferentiating software, just as buying from Microsoft is only good for nondifferentiating software, because everybody can get either one of those things. Your competitor can have the same stuff as you."
This one hit like a revelation. I've been focused on an open-source project that will primarly be used by developers, and so I've found it hard to see benefits in opening up/giving away the labor (or lack thereof) I put into making development tools so that other developers can make money. But when looking at it from the proper perspective - the customer's, Perens has hit on what I now see has been the biggest power behind the open-source movement. He sums it up nicely with:
"...they [companies] can spend less in a cost center for nondifferentiating software than they otherwise would, and then they can take some of their software budget and move it over to the differentiators..."
I've also heard the phrase, "commoditization of the software stack." and that struck with me while reading this article. it really does make all the sense in the world for customers to use open-source software where they can. and the most valuable services will be those of analyzing a companies systems and advising when to apply OS, when to do in-house, and when to buy proprietary.
in developing for developers, as I'd like to do, it's hard to be compensated for your work by developers who are used to getting their tools for free. from what I can tell, the common approaches to this are either dual-licensing strategies (MySQL), where you make the development tool, offer it under GPL, and under a more commercial-friendly paid license that will be paid by commercial production-level developers, or keeping it completely GPL and getting the OS community to support the development of the tool, making it much less expensive to develop.
I'd prefer the latter approach, but no-one else seems all that interested in php web services. if you or someone you love is suffering from lack of things to do, contact me immediately.
</div>
catching up
Web Services AND open-source are positioned to drive packaged software prices down for the first time in a decade, according to META Group.
this article really applies to all projects, not just open-source ones, and I may at some point write a post describing my disagreements with the first benefit it lists, bypassing WSDL, but it does a good job discussing XML benefits for developers.
more evidence that web services are expected to grow.
XInclude is recommended to be a new standard. new standards are almost always good in my eyes, and especially ones published by the W3C.
that's all for now. I'll try to get back to managing my time well enough to make posts every day or two. it might all depend on how long it takes me to beat Halo 2 campaign on legendary difficulty.
</div>
real world "Open Services" battles
Jeffrey McManus, eBay Web Services evangelist said, in response to an accusation that eBay's Web Services were not 'Linux-friendly': "If our API somehow didn’t work with Linux, that would be one thing, but, hello, this is XML we’re talking about."
to which Jeff Licquia responded quite reasonably: "I suppose we in the free software world aren’t used to the idea that we have to pay money for the privilege to access an otherwise free site."
at the risk, or inevitability, of rousing the slumbering OS debating giant that is my brother, this particular example demonstrates some of the misunderstanding of (or unacceptance toward) non-OS business models that I think the majority in the OS community has. Both of the contrasting points are true and to me just signify the importance of getting the Holy Grail of an Open Source Web Services licensing, business, and development model.
Since I completely agree with Licquia's response to McManus, and the debate went in a different direction afterwards, I'll have to go ahead an extend that original topic on my own.
Although Licquia ultimately concludes that it should at least be easier to sign up users into the eBay API program, and I agree with that, I do not agree with a conclusion that eBay is now charging a fee for information which is otherwise free.
The data may be the same, but as we all learned in Information Systems 101, data != information. Data is just raw data, like this:
(0,1,2,1,0,0,0)
and information is more like this:
Readers of Luke's blog last week, by day, in linear order are (0,1,2,1,0,0,0)
Now, the reason that's relevant here is that the way data is presented is what makes it information. So the same data could be used to convey different information to different people, and that is exactly what eBay is doing with their Web Services vs. their website.
When they use their data on their website, they massage the data all up with their site's features, including a listing of advertisers' logos and links, who no doubt paid money for it. Whereas their Web Services is a more raw form of the data, expressed in a much more generic information set as XML. And that's great for others to work with, but it also means they lose the features of their site, like the paying advertisement links. And eBay is very much different from Amazon in the respect that they do not actually sell products, but rather they sell the services of their site. If they offer a means by which these services could be replicated by anyone else, they are essentially removing the need for themselves. Their value then is only in their data, and to stay solvent, they must recover their costs+ (economic costs, which include profit)
Whether their products, services, and/or prices are reasonable is up to the market to decide, including all the open-source and proprietary developers that will use their product. In this sense, eBay has made their model and is asking the market to figure out how to make it work. If the market can't, then their web services API is doomed. If only the proprietary market can make it work, it will survive in that market.
To make it work for open-sourcerors, a reseller license, or an end-user registration API, or many other things could be tried.
For open-source and web services, it is imperative to find a correct balance of the ease of distribution that open-source carries as its main strength, while still accurately compensating service providers to keep their services active. The freedom of distribution makes it extremely hard to comine the two requirements. There will be many attempts, and it may just be that the models will be as unique as the applications they are applied to. Everyone may need their own.
</div>
