10/12/2018 at 10:43 #25598
[This post was originally posted as a reply to this discussion]
Hello Frank, all
Is it correct to say that an EQ is necessarily related to an EIF?
In other words, is it possible to have an EQ that do not read/query permanent data managed by the external application?
And this, no matter how the query is technically made (ws, direct db access, …)?
Sorry if it is a silly question, I am kind of a newbie 🙂
Thx for your help already10/12/2018 at 14:39 #25600
Dear Gram JS,
There is no such thing as a silly question when you ar trying to master FPA, so thanks for the question.
The short answer to the question is NO.
An EO presents information to a user which has either been manipulated and/or the amount of information is unknown.
- For example, when I ask the Nesma website registry for all the users that go by the name of “Gram” the registry does not have to process the data, but I cannot know in advance how many users will be presented. That’s one example of an EO.
- Another example is a personnel system that holds your date of birth. I can ask the system your age. I will always receive one answer, but the answer has been derived from the data of birth, so it has to be manipulated first.”
- It is even conceivable that no ILF or ELF is involved. Take for example a batch interface or a web service that can read Excel or CSV files with multiple records. Processing the batch is one or more EI. A report of how many records have been processed is an EO.
For the EQ only the number of reference logical files (File Types Referenced) is relevant, whether these are internal or external.
(You may have noticed that I used ELF instead of EIF. This is the new Nesma 2.3 terminology.)10/12/2018 at 15:46 #25601
Very clear explanation Frank, thank you very much.
Actually I am trying to make an ‘indicative’ measurement (which is based on logical files count only). And I am a bit lost now because I don’t know when I have count an ELF for the (future) application I am trying to size when it queries (mostly through ws) permanent data from other applications. Also, I have no conceptual data model or whatsoever at my disposal. Do you have a hint?16/12/2018 at 17:25 #25610
It depends a bit on the scale of the size estimate you are making.
The indicative method is suitable for a whole application. You then try to construct a conceptual model of your own application and the data it must query from surrounding applications. In that case you can use the ELF as conceptual element for the data from surrounding applications.
When you are making a size estimate of a limited number of webservices, then the High-level (or estimated as Nesma used to call it) method. Then you don’t bother about ELF, since you will be communicating by means of EI and EO with the surrounding applications.
Hope this helps.19/12/2018 at 10:55 #25613
….why not counting the ELFs when using web-services?
I thought the mean the application uses to consume the data (ws, direct db access, etc) of the external application did not matter?
According to my knowledge, an external entity type meets the following criteria:
•It is used by the application to be counted.
•It is not maintained by the application to be counted.
•It is maintained by a different application.
•It is directly available to the application to be counted..
Used means that the data is also actually made use of in the process of the application.
Maintained indicates that it is possible to add, change, or delete data.
Directly available means that the application concerned always has the current data at its disposal, even though a different application maintains this entity type.
Does the use of ws change anything to that? Do we consider for instance that the data is not “directly” available when accessed through ws?
Thanks already for your help27/12/2018 at 20:29 #25614
When you retrieve data by means of a webservice this is not considered to be directly available. That is why communication with a webservice in a high-level analysis is counted as a pair of EI and EO, rather than an ELF. In an indicative count you don’t bother about these kind of details and you count a data source outside of your application as ELF.
Hope this helps.03/01/2019 at 12:10 #25615
All the best for this new year. 🙂
Identifying the ILFs and ELFs is indeed what I am trying to do here.
Isn’t there always an ELF behind a ws?
Why is it not considered as directly available?
….I thought the way the application retrieves the information (here via a ws) did not matter
….I don’t see fundamentally the difference between accessing the information through a JDBC or via a ws…
Also, you mention that a ws is rather seen as an EI/EO. Even when the information retrieved through the ws only serves (is meaningful to) a larger transaction?
Because when reading the definition of an “Elementary Process (EP)” here below, it seems that in my case the ws query is not a complete transaction in itself as it only feeds information to a larger transaction, which itself is meaningful to the user.
An EP is the smallest unit of functional user requirement that −
Is meaningful to the user.
Constitutes a complete transaction.
Is self-contained and leaves the business of the application being counted in a consistent state.
Thanks again for sharing your knowledge and experience and for your time.
You must be logged in to reply to this topic.