|
Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks. |
On February 10 2017 12:18 WarSame wrote:What is an elegant way to set a bunch of div colours in SASS? I would ideally like to do something like $brown: #955436
$lightblue: #AAE0FA
$pink: #D93A96
$orange: #F7941D
$red: #ED1B24
$yellow: #FEF224
$green: #1FB25A
$darkblue: #0072BB
.block-(colorexample)
background-color:$colorexample
where I could have divs with classes like block-brown and have it automatically sub in the colour. Should I just not bother and make a block-colour for every single colour I have?
Please don't.
Name your CSS classes after what they represent, not after what they look like. Otherwise you'll either have to search and replace all the instances when your Boss tells you that all brown buttons should now be green, which negates all the advantages of CSS, or you have a class "brown" that is now green.
|
Last questions:
About JOIN, if I got a ManyToMany Relationshipt between say, Students and lecturers:
ListL result em.createQuery("SELECT s, l FROM Studens s join s.lectures")
for(Iterator i = resultL.iterator(); i.hasNext();)
{
Object [] element = (Object[]) i.hasNext()//each Object[] containing a Student at [0] and a lecture at[1]
//using get-Methods to get Student/Lecture Name and printing it out
My Output would be equivalent to the joinedtable, where for every Student, all lectures get Listet:
_Student_ _Lecture_
Hans Lecture 1
Hans Lecture 2
Hans Lecture 3
Max Lecture 2
Max Lecture 3
Anna Lecture 3
. .
. .
. .
. .
Correct?
I could also acess the name of the Students directly, via using "SELECT s.name, l.name" instead.
An implicit join would also work like:
Select s, l,
FROM Student s, lesson l
Or to only get Name+all courses of one Students:
Select s, l,
FROM Student s, lesson l
WHERE s.name = "Hans"
Last question regarding general understanding:
All Querys in the end will be translated into SQL and send to the Database, this happens either after commiting an transaktion or calling flush() (which can be dangerous since it leaves the transaktionscope open, If the transaction fails afterwards, changes made before flush() are still in the database unless handled properly). However, how do you acess objekts then, that arent in the persitence context, but in the DB.
EntityManager em = emf.createEntityManager();//Session created
EntityTransaction tx = null;
try{
tx = em.getTransaction(); //transaction context created
tx.begin() //start transaciton();
// Customer001 previously not in persistence context/Cache
customer myCustomer = em.find(Customer.class, 001);//Now in persistence context
/*
Interaction with the object I pulled from the database - How can this work,
if the select-statement only get send after the commit?
Or get everything that you need for the tx get pulled from the DB before/em.find()
circumvents the commit? What about Entities pulled with JPQL-Querys?
*/
String name = myCustomer.getName()//
myCustomer.setName = "MC";
tx.commit();//End of transaction, transactions get send to DB
}catch (RuntimeException e){
if(tx != null) && tx.isActive()) tx.rollback();//Rollback transaction in case something went wrong, and tx got created
throw e;}
finally{em.close();} //closing the session
|
Poland17522 Posts
Just fetch the objects you need before starting the transaction. You should really only use transactions for insert/update statements, not fetching.
|
On February 10 2017 18:49 Manit0u wrote:
Just fetch the objects you need before starting the transaction. You should really only use transactions for insert/update statements, not fetching.
So it doesn't work, or it is just really clunky? In case of it working, is my assumption correct, that fetching always happens immediately and only inserts/update/delete statements get hold back until .commit() is called?
|
On February 10 2017 18:49 Manit0u wrote:
Just fetch the objects you need before starting the transaction. You should really only use transactions for insert/update statements, not fetching.
Not true if you care about isolation and atomicity. It can lead to serious issues specially in financial applications.
|
Poland17522 Posts
On February 10 2017 19:20 Hanh wrote:Show nested quote +On February 10 2017 18:49 Manit0u wrote:
Just fetch the objects you need before starting the transaction. You should really only use transactions for insert/update statements, not fetching. Not true if you care about isolation and atomicity. It can lead to serious issues specially in financial applications.
It really depends on what you need and what you're working with. Also, different SQL servers handle transactions differently (MS SQL for example will not allow for any other queries - even reads - until the transaction is commited and closed). Sure, if data integrity is critical, you want to wrap everything in transactions (and possibly add other mechanisms of ensuring that) but for the majority of use cases you won't need that level of paranoia.
|
This isn't paranoia. It is about a program behaving correctly or not. Saying that I don't need to do a transaction because I am only reading is the same as saying that I don't need to use a lock because I'm reading data (except that you won't crash because the database is always acquiring integrity locks). In some cases it is fine because there is an implicit transaction or auto commit in place. Yet if someone is learning about transactions, he should do the right thing.
|
But I still struggle with the fact that the DB gets the data on commit, while I use fetches inside the transaction without having commited first. Does fetch always immediately get passed on to the DB and only write/delete/inserts/everything that manipulates data get send only on commit?
(In other words, persistence context stuff like adding/changing/removing data in/from it gets applied immediately, the Database only gets changed on commit. Fetching inside a transaction doesnt change anything on the DB but loads aditional entities in the persistence context? On commit, the persistence context will be synchronized with the DB)
|
I don't know what the JPA (?) layer does but speaking from the database point of view, your transaction delimits a unit of execution that is Atomic and Isolated (A and I of ACID). Atomicity means that: Everything succeeds or everything fails. For instance, if you do 2 inserts and the 2nd one fails because of a constraint violation both inserts roll back. It doesn't mean that these inserts are sent at the end of the transaction. If you fetch something, then update it, and fetch again, your second fetch will return the updated value - even if the transaction isn't committed.
Isolation is more complicated but ideally, while your transaction is underway you don't see the actions of other transactions that are run concurrently. There are subtleties because isolation impacts performance and database implementations typically offer several options of isolation level.
|
On February 10 2017 17:38 Morfildur wrote:Show nested quote +On February 10 2017 12:18 WarSame wrote:What is an elegant way to set a bunch of div colours in SASS? I would ideally like to do something like $brown: #955436
$lightblue: #AAE0FA
$pink: #D93A96
$orange: #F7941D
$red: #ED1B24
$yellow: #FEF224
$green: #1FB25A
$darkblue: #0072BB
.block-(colorexample)
background-color:$colorexample
where I could have divs with classes like block-brown and have it automatically sub in the colour. Should I just not bother and make a block-colour for every single colour I have? Please don't. Name your CSS classes after what they represent, not after what they look like. Otherwise you'll either have to search and replace all the instances when your Boss tells you that all brown buttons should now be green, which negates all the advantages of CSS, or you have a class "brown" that is now green.
In this case they are going to be blocks to build a monopoly board, though, so they need to be named by colour.
|
On February 10 2017 20:31 Artesimo wrote:
But I still struggle with the fact that the DB gets the data on commit, while I use fetches inside the transaction without having commited first. Does fetch always immediately get passed on to the DB and only write/delete/inserts/everything that manipulates data get send only on commit?
(In other words, persistence context stuff like adding/changing/removing data in/from it gets applied immediately, the Database only gets changed on commit. Fetching inside a transaction doesnt change anything on the DB but loads aditional entities in the persistence context? On commit, the persistence context will be synchronized with the DB)
The DB definitely does not only get the data on commit. JPA is just built on top of JDBC, which gets a connection and starts a transaction if you are using your transaction manager. As others have tried to explain, and probably better than I will, you are actively participating in a database transaction at that time.
If you are trying to look at the database while your code has a transaction going on, you won't be able to see it -- that's the part of the isolation levels etc. It might feel like everything happens on commit, but, that's just when everything is actually committed, not when it is transferred.
If you are in the scope of a transaction, you can read the changes that you have made but others can't (most of the time).
Here is some reading on JDBC Transactions
Underneath the hood, when you start the transaction, it takes the connection that you have and turns off auto commit. Then every time you run something with JPA, it takes the jdbc connection and prepares a statement and then executes it -- all of that is done until you call commit.
|
So, C question.
I want to create a program that takes in input of the total number of variables that will have to be put into an array
then create an array that has size based on that input
but im reading that arrays can't be created with a variable as the size
wtf is that about, that seems really annoying
|
On February 12 2017 04:57 travis wrote:
So, C question.
I want to create a program that takes in input of the total number of variables that will have to be put into an array
then create an array that has size based on that input
but im reading that arrays can't be created with a variable as the size
wtf is that about, that seems really annoying
Are you compiling C89 standard or C99 standard? The rules are different for each. The short answer is that the compiler wants to know exactly how much memory will be needed on the stack for your array.
The simple solution is to allocate on the heap using malloc, but that might not be what your class wants you to do.
|
I think they want me to just initialize each array to size 50, because they told us in specifications that they won't need to hold more than 50 elements
but I don't like the huge lack in knowledge that comes with that hehe
I am finding I am not a big fan of C.... too much stuff that takes too long to do
|
|
|
On February 12 2017 05:29 travis wrote:
I think they want me to just initialize each array to size 50, because they told us in specifications that they won't need to hold more than 50 elements
but I don't like the huge lack in knowledge that comes with that hehe
I am finding I am not a big fan of C.... too much stuff that takes too long to do
If you know the max size will be 50 its as simple as:
const int MAX_SIZE = 50;
When you compile your code the compiler knows that the arrays size is 50. It knows exactly how much stack space to allocate and it is happy.
extern int MAX_SIZE;
Contrast with the above where you're getting the max size from somewhere else. The compiler doesn't know what the value will be and it won't let you compile.
|
If you do this:
const int MAX_SIZE = 50;
T arr[MAX_SIZE];
then arr is a VLA. Don't use VLAs.
|
what about #define ARRAY_LENGTH 50 ?
|
|
|
I was under the impression that a constant didn't change value, but when I wrote C89 was the worst months of my life so I'll just refrain from answering anymore about it since apparently I don't remember anything.
|
|
|
|
|
|
EPT
