I was asked this question in an interview and had no idea. And I still don't. I thought they were created for this purpose, not to replace something that already exists. Any thoughts?
In oversimplified terms web services 'replaced' older RPC technologies like COM+ and CORBA
CORBA might be what the interviewer had expected you to answer. Or maybe RMI for java2java communication.
Probably the interviewer wanted to name some particular term but they improved the way different applications and systems communicate with each other.
Related
I was asked this in an interview and have not been able to find an answer. I do not know whether it refers to SOAP or REST, or whether it refers to the development or the execution lifecycle. Given this ambiguity, can you give answers for any of these possibilities?
Thank you and sorry for not being more specific.
It is not really something I can straight up answer easily; you would be better reading this document to get the full understanding.
Web Service Management: Service Life Cycle
I've heard a lot of buzz through the years about "Self Describing" web services and I'm curious what it means. I've flipped through the W3C standard and it doesn't really help.
I understand you can say "Give me this argument and call this functionality and I'll return something like this", but how is that actually helpful? How can software know what is relevant to a client given some context?
Can anyone give real world examples of this concept, and explain how it's better than some other alternative? Or maybe how useful/useless it is?
It's really designed as a standard for describing in a cross-platform cross-language manner an interface that a developer can use to develop a SOAP based way to exchange information with a web service.
Another alternative would be providing a library that provides a local interface to a blackbox communcation scheme, which is fraught with compatability/security issues.
Or providing documentation, which may be difficult to find, have compatibility issues, be out of date, or incomplete.
In short, it's very useful.
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 7 years ago.
Improve this question
On our embedded device, we currently use PHP for its web interface, and unfortunately it's quite slow. We've been experimenting with Python, but is seems (at least on FPU-less ARM architecture) to be as slow as PHP.
Therefore we're thinking about implementing web interface in some compiled language like C++, but so far the only thing we've found is Wt, which looks more like desktop than web framework and it's documentation is rather complicated for a beginner.
So my question is: do you know about any good web frameworks for C/C++? What would make me totally happy would be something like C++ Django, but I doubt such thing exists :-)
What is the reason your site is slow in PHP? Have you profiled the code to figure out the bottle necks? Replacing the entire codebase with another language is likely to be too large a tool for the job.
Perhaps if some portion is computationally intensive, then maybe you can implement a PHP module in C++ to do the job.
If I were you, I would give Wt a try. I don't think you will find another solution as complete and easy to use as Wt with similar performance. The mailinglist are active, and has regular posts of people who use it on embedded devices. The Wiki (here) of the project mentions some numbers for embedded deployment and performance on ARM.
You can take a look at CppCMS. I don't know how mature or good it is, but it sounds like what you are looking for.
Another option is to write the main application in Python/PHP and then rewrite the critical parts in C, which seems more sane to me.
I had just seen this link as a referrer to CppCMS. Several questions/answers about it.
Q. How is it mature?
A. There is a Wiki 100% based on it. Developers blog runs on CppCMS as well. So you can see it quite works.
well. At least for author.
Q. "What would make me totally happy would be something like C++ Django"
A. Actually it is "C++ Django". Many Ideas had been taken directly from Django, like
forms, template inheritance and some more. So you would probably feel at home
Q. Nanogear
A. I had taken a look on it... It has many plans but not-too-much code.
Q. How many projects using it.
A. Not many. It is quite new project.
Q. Who am I?
A. To be honest I'm the developer of CppCMS.
Do you really need a web framework or just an html templating library?
See here for template options: C++ HTML template framework, templatizing library, HTML generator library
One thing to look at here is using C++ for some sort of web service instead of an actual web app development environment. On one of my current projects we generate JSON(and parse for requests) then let the client(web browser) deal with the rest. You have a great deal of computing power available to you on the client machine, mix in some javascripty magic and boom, now your embedded device can go back to doing what its supposed to do. I personally kind of like Wt and think it is your best bet if you want to dev the whole "web app" thing in C++.
As to the specifics of how you get your C++ code to start serving up web services, you can use the amazing boost.asio library(they have an http server example) and json_spirit(I think its on codeproject)
How about Nanogear. I haven't used it, so I'm not sure if it is a good fit. Here is a quote from the website.
C++ is a great, general purpose
programming language but no modern,
easy-to-use and powerful C++ web
frameworks were made during these
years. We try to fill the gap with
Nanogear. Inspired by the Zend
Framework and Rails we want to bring
C++ to the web, easily.
Here is a link to the older googlecode site Google Code Nanogear Looks like there is more info there.
Try Lua. There is Xavante embeded web server somewhere on LuaForge.
It's very fast, very small and very extensible - as all about lua is.
You have power to use Lua where scripting fits betteer, and use C++ functions where in need of more speed and conection to existing native code.
Since no one else has said it, I will. Try Java. Java is a compiled language with a good web framework. It is also very scalable. Every coperation I have worked for uses java as its main web-technology, and this is mainly due to this scalability factor. Not sure if it will work with your architecture, but it answers the question:
Are there any web frameworks for compiled languages like C++?
You may give a try to embedded version of CppCMS.
It's quite new and exprerimental development but It looks like it is
much more suitable then Wt because it is ore web oriented (more closer to Django)
and not GUI like.
Check out the ffead c++ application framework (http://code.google.com/p/ffead-cpp/), its developed on the lines of Spring for JAVA.
You can have a look at Tntnet. It uses its own template language which looks a bit like PHP, only with much more additional tags.
I'm implementing an SOA at a large company, and I'm not sure which web service specifications (WS-*) actually make sense to implement. At a minimum, I'm looking at WS-Addressing, WS-Security, WS-Eventing, and WS-ReliableMessaging. However, there are several other standards that look interesting, but I don't know which ones are widely adapted. I don't want to implement a standard (and force all the developers to follow them) if they're not mature or necessary.
EDIT:
I'm asking this question not about a specific situation, but in general. There are quite a few WS-* standards that don't seem to have a lot of practical use (at least to me), so I'm really curious about which ones are widely used.
Thanks for your help!
KA
WS-Adressing is widely used, and quite useful. For WS-Security, consider the set of mechanisms you'll need (based on your usage scenarios).
Only SOAP is widely adopted. If you care about reach, going beyond WS-Security and WS-Addressing is asking for trouble (even WS-Security can be hard for a lot of people). If you are creating services for internal use in a large company, then I wouldn't worry as much. Something like WCF would allow you to provide endpoints with different bindings for a wide range of consumers without writing any additional code.
There are two types of web services : REST and SOAP. They represent different protocols of sending data over the internet.
SOA is an acronym standing for Service Oriented Architecture. It is a way of architecturing your system using multiplet tiers (applications) one atop the other. Web services, mostly soap based are used to implement this archtecture, but they are not the only way.
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
I've been given the task of laying the groundwork of a SOA for my client. The goal is to open up various processes in an end-client independent way and also to make data available offline e.g. for reps visiting customers.
I do have extensive experience with J2EE (Websphere) and web services but I would appreciate advice on how to build up such an SOA.
Where are the pitfalls? What about security? How finely granulated should services be? etc.
Links to tutorials and book recommendations would also be useful.
Thanks!
Pitfalls
Versioning/backwards compatibility: it gets really hard to change a contract once you have loads of clients. I have seen many sites version the APIs by introducing the version in the URL
Granularity
Each service should be reasonly self-contained (don't expect people to do 3 calls before they get what they need)
Platform Independence
Try to give more than one way of accessing your APIs (WS, JSON, REST...)
People can't agree on what SOA actually means.
http://martinfowler.com/bliki/ServiceOrientedAmbiguity.html
(although consensus may have grown since that was written)
I suggest quizzing your client to find out exactly what they mean - if anything. Then give them something that actually provides business value, while ticking any SOA boxes that might coincide with that effort.
Call me a SOA-skeptic. Fowler's lament still seems right on.
I would focus on the more general problem: your client has 2 or more applications that have to collaborate together. Look at old school integration patterns.
(source: amazon.com)
Found this IBM Redbook (#sg246303) which is quite a good introduction to the basics of SOA.
As Alan said, I'd start reading the Enterprise Integration Patterns book. There are a number of ways to implement them either using a messaging system directly such as JMS or using open source projects like Apache Camel, for example see the pattern catalogue.
I'd also look at understanding how to build good RESTful services using JAX-RS with Jersey as a simple way to expose resources for your systems to anyone on the web from any language/platform easily without falling into the SOAP/WS-* deathstar :)
Get an ESB (enterprise service bus): Mulesource is a good choice (Opensource, Mature, yet bleeding edge) . Once you understand it, you will understand SOA.
The goal is to open up various
processes in an end-client independent
way and also to make data available
offline e.g. for reps visiting
customers.
The second half of that isn't really an SOA topic, it's more of a replication to mobile devices problem. I would stay far, far away from trying implement a buzzword and focus on the problems that you are stating. Web services are good way to open up process to client independent ways.
So far the best book I found is SOA Compass also available on Amazon