Here we go again, changes in IT are comming, now all goes to cloud and parnership, those dais of "x hacker made new system" are in past, now the team development and management computing its the future..
The problem with the free software are that have too little spectre/range of vision.. industry take the message and now know that cannot compete agains free software, so then.. privative companies evolutione the clients to arrange new targets... again free software are steps behind...
In the web: https://db-engines.com/en/system/Oracle%3BPostgreSQL here we have good comparison, Oracle vs Postgres are very close, in fact, for any company with general purposes, theres no sense choose between those two, so obviously will choose Postgres due cheap and using the "ora additions" we have a very close Oracle+postgres implementation. BUT! already former and running enterprices have yet Oracle! so if i have already working, running and knowed DBMS with well trusted manpowers.. please bitch, i don't migrate, for what? if are working dont touch it!
Lest see the most important for 2018:
The new DBMS Postgres 10 brings new things, in November 2017, but they are not so "big things" at all, only three of these are really "new great features" of new value (or old? We should say):
Gentlemen, the world of free software is always steps behind and wrong because they only work with what their poor eyes see ... let's analyze it
Parallel calculation:
the new postgres 10 has a great improvement (the reall best feture) that almost nobody takes into account: support of parallel calculations, that is to say, it can use multiple different indexes without effecting locking of the tables .. this if scalability, here enters postgres in the big leagues .. 4 users (two in fact but logged twice each one) use two indices doubling that same for each connection to multiple tables in a procedure (function, puff does not exist sp in pg),an in one of that columns the data type are a mixed based on a precal math formula!?!? the problem its not the column, its the parallel access to the amount of computed indexes over that column!
Async Replication:
On the other hand we have a lesson learned, asynchronous replication, what?! this its a old century feature, when slow connections are available! why? good because today the enterprises are decentraliced, and not uses lan connections, here the cloud is what commands and not all companies pay wan, in fact none of the readers here has money to have connections that support synchronous replication, ie the successful currentl replication of postgres << 9.99 was a failure for years .. now the development team implements an already made feature present in MSSQL, Sybase and Oracle DBMS's since acient times! please bitch! gimme a break!
Range partitioning:
Unfortunately, partitioning pro comming late, here oracle has years, another lesson learned .. this is only useful in big leagues .. i mean, little users and Pymes don't need use that due currently we have big data disck in normal stores! I will not give a clustering class because no one will understand, so where does this come from? well as can see, big data companies dont change the already orking system (that use oracle) by newer not well knowed, Oracle already made that and obviously with better experience!
That free code are downloading and using to made privative code! in privative systems! since years!
Ok!, good for postgres: is seeing what others in free software see coming, is dying with Stallman and today that the free software really need is many lawyers .. since programmers there are only windoseros throwing code!
None of the 3 new great features are necessary for the environments that are present in small-medium companies .. or companies/users that use postgres .. since they will only have lan networks ... and big data enterprices will not change their system only for few new features.. so these features must be years ago from make some sense.. sorry postgres guys, you must made those features years ago! so postgres developers try to support bad and unstable connection with async replication (as already made other DBMS since years!!)
...as police man said: "please retire, here is nothing to see, go home"
The problem with the free software are that have too little spectre/range of vision.. industry take the message and now know that cannot compete agains free software, so then.. privative companies evolutione the clients to arrange new targets... again free software are steps behind...
Some differences in PG vs Ora ? lest see quickly
Oracle it the mayor vendor and widely used DBMS of the world, a comparison and point of reference are ineviable!In the web: https://db-engines.com/en/system/Oracle%3BPostgreSQL here we have good comparison, Oracle vs Postgres are very close, in fact, for any company with general purposes, theres no sense choose between those two, so obviously will choose Postgres due cheap and using the "ora additions" we have a very close Oracle+postgres implementation. BUT! already former and running enterprices have yet Oracle! so if i have already working, running and knowed DBMS with well trusted manpowers.. please bitch, i don't migrate, for what? if are working dont touch it!
Lest see the most important for 2018:
Name feature
|
Oracle
|
PostgreSQL
|
Description
|
Widely used RDBMS
|
Based on the object
relational DBMS "Postgres" |
Primary database
model |
Only relational DBMS
|
Relational AND also
OBJECT DBMS |
Current release
|
12.2, March 2017
|
10.1, November 2017
|
License
|
commercial
|
Open Source
|
Cloud-based
|
no
|
no
|
Implementation
language |
C and C++
|
C
|
Server-side scripts
|
PL/SQL
|
user defined
functions capable! intoo manycluding PL/SQL |
Partitioning methods
|
horizontal
partitioning since years |
declarative
partitioning since v10 only |
Replication methods
|
Master-master
replication Master-slave replication |
Master-slave
replication only |
Supported programin languajes
|
too many
|
few only including .NET, Java and C/C++
|
Involution: new in Postgres 10
The new DBMS Postgres 10 brings new things, in November 2017, but they are not so "big things" at all, only three of these are really "new great features" of new value (or old? We should say):
- partitioning by rank, increasing scalability
- logical and asynchronous replication (at this time? very later)
- Support to parallel compiles (great, but the least mentioned)
Gentlemen, the world of free software is always steps behind and wrong because they only work with what their poor eyes see ... let's analyze it
Parallel calculation:
the new postgres 10 has a great improvement (the reall best feture) that almost nobody takes into account: support of parallel calculations, that is to say, it can use multiple different indexes without effecting locking of the tables .. this if scalability, here enters postgres in the big leagues .. 4 users (two in fact but logged twice each one) use two indices doubling that same for each connection to multiple tables in a procedure (function, puff does not exist sp in pg),an in one of that columns the data type are a mixed based on a precal math formula!?!? the problem its not the column, its the parallel access to the amount of computed indexes over that column!
Async Replication:
On the other hand we have a lesson learned, asynchronous replication, what?! this its a old century feature, when slow connections are available! why? good because today the enterprises are decentraliced, and not uses lan connections, here the cloud is what commands and not all companies pay wan, in fact none of the readers here has money to have connections that support synchronous replication, ie the successful currentl replication of postgres << 9.99 was a failure for years .. now the development team implements an already made feature present in MSSQL, Sybase and Oracle DBMS's since acient times! please bitch! gimme a break!
Range partitioning:
Unfortunately, partitioning pro comming late, here oracle has years, another lesson learned .. this is only useful in big leagues .. i mean, little users and Pymes don't need use that due currently we have big data disck in normal stores! I will not give a clustering class because no one will understand, so where does this come from? well as can see, big data companies dont change the already orking system (that use oracle) by newer not well knowed, Oracle already made that and obviously with better experience!
Software libre? to make private systems! pufff
That free code are downloading and using to made privative code! in privative systems! since years!
Ok!, good for postgres: is seeing what others in free software see coming, is dying with Stallman and today that the free software really need is many lawyers .. since programmers there are only windoseros throwing code!
CONCLUSIONS:
None of the 3 new great features are necessary for the environments that are present in small-medium companies .. or companies/users that use postgres .. since they will only have lan networks ... and big data enterprices will not change their system only for few new features.. so these features must be years ago from make some sense.. sorry postgres guys, you must made those features years ago! so postgres developers try to support bad and unstable connection with async replication (as already made other DBMS since years!!)
...as police man said: "please retire, here is nothing to see, go home"
Comentarios
Publicar un comentario
no stupid winbuntu users allowed!