Hello, I have a need to create a spreadsheet upon demand (use triggered). Because of the volume of records, I cannot have the user wait and was thinking of creating a job queue. My question is if I create the queue within the website (separate running thread) and the last user logged off, does IIS shutdown the website even though a spreadsheet could be in the middle of being created or is it smart enough to realize a thread is "busy" doing something so stay running? If not, is the recommendation to create a windows service that would had the creation of these spreadsheets or? Thanks for your time, Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Sorry - never been gifted in painting a picture :( The pool, connections, and transactions to the mainframe are all created/handled in the windows service. An app creates an instance to a class within the windows service that will execute transaction logic making it's way to the mainframe. This logic checks first if the pool has a link and uses it if it does or creates a link and then passes it to the pool to hold on to.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello Eddy, thank you for your response... Windows service vs a class-library: My apps include a classic asp website, a asp.net wcf service site, a asp.net webforms site, a asp.net mvc site, as well as scheduled tasks. What keeps that library running and accessible from all calling apps? The intent is to use one pool, not one per app to cut down on the number of connections to our mainframe. The 3rd party tool creates two connections (first is to establish connection, second is the actual call to the service) per transaction to be able to call a specific service on our mainframe. The first connection can be "cached" / held on to which is what the pool manager is intended to do. Thanks again for your time..

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello, I want to create a windows service that will act as a connection pool manager for a 3rd party tool to various apps (web, utilities, etc) that are running on same server. I'm struggling with how the other apps call this service without creating additional instances of the exe. I've played with load assembly and references and it seemingly is creating an additional instance. Proof: I have a static class that is populated (initialized) upon load of the service. When an app tries to create an instance of a different class that calls methods within the static class...the static class is empty (uninitialized). So my conclusuion is a second instance is created? How can I design this to do what I'm looking for... Thank you for your time... Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Hello, My project is creating a connection pool for a 3rd party product. I wanted to make a windows service (because of being in the IDE, it's start as a normal exe) that will control the pool. I have a static class in my windows service that is initialized upon startup of the windows service -> this class is the pool manager. Right now my attempts are referencing the windows service project in visual studio from a test project. The startup projects are the windows service and then my test project. I watch the windows service start up seeing the pool manager be initialized. When the test project creates an instance of a class inside the windows service which uses the pool, that instance claims the pool manager class isn't initialized even though it's defined as static and was intialized at windows service startup. The task manager doesn't claim another exe is running so why am I loosing the static content? Hopefully I explained the above clearly. Thank you for your time... nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello, I am working on an mvc site. I understand the concept of the Model, however we are forced to work with the existance of our main frame. The data comes back in the form of a parameter block (fixed structure defining each field and it's length). This transaction to the mainframe would occur outside the controller (obviously). So to move the data from the transaction to something the controller uses can either be done with it's own class (my company calls them entities) or to double purpose the model as the information from the mainframe would basicaly (with same additions) be the fields of the model. Hopefuly all that made sense. So the question is: do I create entity class and transfer all the data(properties) into the model after every tranasction once the fields are returned to the controller or do I double purpose the model by using it as the entity as well with the understanding that these fields that aren't to be used in the view would show up because the model fields and entity fields are mingled together? Any advise would be greatly appreciated... Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Hello, I have a dll that i'm calling 'library' (contains core classes necessary for my app to run). I have a series of dlls that are responsible for various tranactions (security, detailed data, etc) that am doing a loadassembly to when a certain type of transaction is requested. So my website has a reference to the library. The library does loadassembly to transaction dll(s). But there are classes inside the library that the transaction dll(s) could/should use to cut down on duplicate code. So the question...is this type of circular reference acceptable (1 reference is done early and 1 is done late) or how should this be designed? Thank you for your time. Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Thank you for your reply. Yes, a person can have 1-N authorizations. I guess i'm still confused with 1 to many in this context. From the user record, I wanted to include the auths table to get at the auths relating to that person. In order to have a 1 to many, the 2nd table must have an additional field(s) defined as the primary key for uniqeness. But EF seems to want a matching of 1-1 of primary keys. Not sure how that's can be. Unless there's something about the mapping I completely am overlooking. Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello, Forgive if this is in the wrong forum. I didn't see a better one to use. i'm confused on the mapping of a 1 to many. I have 1 table with a primary key of userid. I have another table with the primary key of userid and authid. I created an association of a 1 to many (because one user record to many occurrances of security authorizations for that user). When I assign userid to userid - that's fine but obviously I don't have a match for authid -> and the compiler isn't happy about that. So how does 1 to many work or am I missing something :omg: - in order to have a many, the primary key can't match because it 2nd tables record's primary key has to be unique? Any help would be appreciated. Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Thank you Eddy, I have a grid that supports drill-down so that's not a problem. I was just hoping to use entities instead of having to convert but from what you said, doesn't look like that's possible. Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello, When loading grids with data tables, if you want to have nested tables (grid shows a + to expand and show the 2nd table) you create a datarelation - method available in the dataset object. I have a list(of ???) that I feed the grid for it's datasource where the ??? is an object(entity) that has fields (just like a table). Is there an equivalent to creating a datarelation so I nest the entity objects or would I have to convert the entities to datatables to do this? Any help would be appreciated... Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

thanks Eddy

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

self-contained. ok good to know.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Thanks Dave, Do you object to a design that seems to be flooded with "shared" use?

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Thanks Luc, So if he has a routine that accepts 1-n arguments that could be different per call and does a search and update from/to a database - not a good candidate for a shared routine - correct or? Is there a limit to how many current calls can be made to a shared routine?

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

Hello, I saw this code of developer we have doing something at my company and he's using the word "shared" in front of routines everywhere. For example, he has a dll that he's referencing from a website and lots of routines in the dll are coded "shared". These routines aren't doing simple formatting or soemthing that, they are doing DB work, etc. Is there a concern in how many times a shared routine can be called currently? I personally do not like his design at all and i'm curious what everyone else's opinions are about the usage. Is using shared in this way a lazy thing or ? Any opinions are appreciated (within this topic - ha ha). Thanks Nathan

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Others might disagree but in my experience you are waiting because of the whole start up of .net framework/etc besides the effort of the website to get going. You'll experience it again if the application onend fires and it has to startup again. Like you said, you would need a polling app to keep the website alive so you don't incur the startup time.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Very helpful. Thank you for the explanation.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

Erik, Re-looking it, it obviously has something to do with nullable types but do you have any idea why declaring it as nullable allowed a constructor? Thanks, Nathan.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous

I guess that was simple. thanks.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous