Hi Ryan, Happy to help.
Yeah, I can't cite a lot of adoption on DSLs among F500. I know the tech companies and their closely-tied partners doing R&D/commercial proving have worked with them fairly heavily, but I can't cite any outright commercial/client-facing instances. ∞ Andy Badera ∞ +1 518-641-1280 ∞ This email is: [ ] bloggable [x] ask first [ ] private ∞ Google me: http://www.google.com/search?q=andrew%20badera On Mon, Oct 5, 2009 at 1:05 PM, rbr <[email protected]> wrote: > > Thank you Andrew. After a little research I am going with your > suggestion of the rules/template engine. (Actually, we already have > one that is rarely used that I am extending.) > > I like the idea of the DSL. However, I believe it would be difficult > to get buyoff for such a solution for a client facing application such > as this. I asked around in my network of consultants and none of them > have seen DSLs getting corporate buyoff in major (International > Fortune 500) corporations. they often site worries about long-term > sustainability. "If the original developers of the DSL leave, > transition could be more time-consuming, thus costly." Have you seen > DSL catching traction in major organizations and/or heavily utilized > client-facing applications? I'd be interested in hearing a little > about it. We have numerous applications that I have considered this > approach but, have not been able to find examples where it has been > used in comparable-sized (20k+ employees/1.5 million client users) > organizations. this would help greatly in creating executive buyoff. > > Thanks again. > > rbr > > On Oct 4, 5:34 am, Andrew Badera <[email protected]> wrote: >> Try as I might, I just cannot see where operator overloading has a >> darn thing to do with this. Nor what operator overloading and >> constructors notionally have to do with each other. Constructors >> can/do set parameters in OO, period, whether you're defining operator >> overloads or not. >> >> Could you explain why or how overloading an operator would help >> someone define an email address default pattern, Charles? >> >> ∞ Andy Badera >> ∞ +1 518-641-1280 >> ∞ This email is: [ ] bloggable [x] ask first [ ] private >> ∞ Google me:http://www.google.com/search?q=andrew%20badera >> >> On Sat, Oct 3, 2009 at 9:20 PM, Charles A. Lopez >> >> >> >> <[email protected]> wrote: >> > this reminds me of operator overloading. The function runs code depending >> > on >> > what parameters are received. In your case, elements are the parameters. >> > You >> > Add or modify data with constructors. >> >> > just my $.02. >> > 2009/10/3 rbr <[email protected]> >> >> >> Hello, >> >> >> I have a problem to solve and have come up with a few solutions. >> >> however, none are as clean as I would like them. I was hoping that >> >> someone would have done something similar and have good suggestions >> >> for solving the following problem: >> >> >> Problem: >> >> >> We receive data via a webservice from clients. They come in as XML and >> >> we map these elements to one of our Domain objects. We do not receive >> >> a few required elements currently and build them using the included >> >> elements via a hard-coded pattern. We now have the request from our >> >> clients to allow them to define this pattern. >> >> >> Example: >> >> >> We receive user credentials in mass uploads to our system. Currently >> >> we do not recieve an email. Our system requires an email so we build >> >> an email for each user by contatenating elements we do receive. In our >> >> case we use firstnamelastn...@[client.com]. Our clients are requesting >> >> the ability to define this pattern. So, for example, one client wants >> >> the following pattern. >> >> >> [firstname.firstletter][lastnam...@[client.com] >> >> >> (I just made up a faux pattern definition language for illustration >> >> purposes. We have not defined this aspect yet.) >> >> >> So, we would like the ability to define this patter per client in a >> >> config file, read it in, and build it appropriately. >> >> >> Does anybody have experience with this type of problem? It does not >> >> seem like it would be uncommon and we have come up with a few >> >> solutions. However, I would like to get outside input to make sure we >> >> are not missing something. >> >> >> We have certainly considered the strategy pattern. And that is >> >> porbably our "best" solution thus far. However, I would prefer a >> >> solution that would not require new coding when a new (not previously >> >> defined) pattern is introduced. >> >> >> Thanks in advance! >> >> >> rbr >> >> > -- >> > Charles A. Lopez >> > [email protected] >> >> > Registered Microsoft Partner >> >> > New York City, NY >> >> > I'm running on Windows 7 Build 7100 >> >> > Quality Software Works- Hide quoted text - >> >> - Show quoted text - >
