Обсуждение: Re: [GENERAL] Re: [HACKERS] [Fwd: SGVLLUG Oracle and Informix on Linux]
Re: [GENERAL] Re: [HACKERS] [Fwd: SGVLLUG Oracle and Informix on Linux]
От
lynch@lscorp.com (Richard Lynch)
Дата:
At 10:56 PM 7/23/98, The Hermit Hacker wrote: >On Thu, 23 Jul 1998, Maarten Boekhold wrote: >> In fact, they are handled by SQL: CREATE DATABASE and DROP DATABASE. The >> createdb and destroydb tools just call these SQL statements.... > > Here's an odd thought: > > Let's remove the "I don't want to think" utilities like >{create,destroy}{db,user} and force DBA's to actually use the *proper* >functions. :-) Actually... While the man pages indicate that these invoke psql, and that a postmaster must be running, and somebody really smart could infer that that means that there is SQL to do the action, it would be much, much better if the man pages explicitly stated that it was merely a shortcut to using the sql. -- -- -- "TANSTAAFL" Rich lynch@lscorp.com
> >> In fact, they are handled by SQL: CREATE DATABASE and DROP DATABASE. The > >> createdb and destroydb tools just call these SQL statements.... > > Let's remove the "I don't want to think" utilities like > >{create,destroy}{db,user} and force DBA's to actually use the *proper* > >functions. > While the man pages indicate that these invoke psql, and that a postmaster > must be running, and somebody really smart could infer that that means that > there is SQL to do the action, it would be much, much better if the man > pages explicitly stated that it was merely a shortcut to using the sql. I think only doing it the SQL way would be fine. Documentation would, of course, have to cover it. I want, no need, to know what functionality belongs to SQL and what belongs to PostgreSQL. I've certainly not got any qualms about dropping into psql to do things. I like psql. Bruce Tong | Got me an office; I'm there late at night. Systems Programmer | Just send me e-mail, maybe I'll write. Electronic Vision / FITNE | zztong@laxmi.ev.net | -- Joe Walsh for the 21st Century
Re: [DOCS] Re: [GENERAL] Re: [HACKERS] [Fwd: SGVLLUG Oracle and Informix on Linux]
От
Bruce Momjian
Дата:
> > >> In fact, they are handled by SQL: CREATE DATABASE and DROP DATABASE. The > > >> createdb and destroydb tools just call these SQL statements.... > > > > Let's remove the "I don't want to think" utilities like > > >{create,destroy}{db,user} and force DBA's to actually use the *proper* > > >functions. > > > While the man pages indicate that these invoke psql, and that a postmaster > > must be running, and somebody really smart could infer that that means that > > there is SQL to do the action, it would be much, much better if the man > > pages explicitly stated that it was merely a shortcut to using the sql. > > I think only doing it the SQL way would be fine. Documentation would, of > course, have to cover it. I want, no need, to know what functionality > belongs to SQL and what belongs to PostgreSQL. I've certainly not got any > qualms about dropping into psql to do things. I like psql. They have to connect to template1 to do the work. Currently, they don't need to know template1 even exists, so it seems like an added burden. I will add a mention to the createdb, destroydb man pages. createuser does psql too. -- Bruce Momjian | 830 Blythe Avenue maillist@candle.pha.pa.us | Drexel Hill, Pennsylvania 19026 + If your life is a hard drive, | (610) 353-9879(w) + Christ can be your backup. | (610) 853-3000(h)
> They have to connect to template1 to do the work. Currently, they don't > need to know template1 even exists, so it seems like an added burden. I > will add a mention to the createdb, destroydb man pages. createuser > does psql too. And as a result, I didn't know what template1 was for until now, and I fear there's more to it than just this. Up until this point, I assumed "template1" was an example database which could be copied, or something. At least that's what a template is to me. Okay, I've suspected there was more to "template1" for a little while now, but I'd not gotten around to looking into it more. Still, my first impression was it was a sample database. ;) Maybe a name like "master" would be clearer, or maybe that means something else to someone. Bruce Tong | Got me an office; I'm there late at night. Systems Programmer | Just send me e-mail, maybe I'll write. Electronic Vision / FITNE | zztong@laxmi.ev.net | -- Joe Walsh for the 21st Century
Re: [DOCS] Re: [GENERAL] Re: [HACKERS] [Fwd: SGVLLUG Oracle and Informix on Linux]
От
The Hermit Hacker
Дата:
On Fri, 24 Jul 1998, Bruce Tong wrote: > > They have to connect to template1 to do the work. Currently, they don't > > need to know template1 even exists, so it seems like an added burden. I > > will add a mention to the createdb, destroydb man pages. createuser > > does psql too. > > And as a result, I didn't know what template1 was for until now, and I > fear there's more to it than just this. Up until this point, I assumed > "template1" was an example database which could be copied, or something. > At least that's what a template is to me. In a sense, that is exactly what it is. When you do a 'createdb', it uses template1 as the "template" for the new database, and then buildds from there...
> I think only doing it the SQL way would be fine. Documentation would, of > course, have to cover it That last sentence says it all..."Documentations would, of course, have to cofver it." The reason I used createdb to generate the my database, is that is what the man page said to do. Unfortunately it is hard to search in the man pages for something like: How do I create a database?. I think I came accross the createdb man page via the postgres man page's see also section...james
> > > Let's remove the "I don't want to think" utilities like > > >{create,destroy}{db,user} and force DBA's to actually use the *proper* > > >functions. IMHO (actually make that IMVeryHO) This is probably a bad idea... We should just update the man pages to detail the SQL code that can be used instead of the command. It doesn't hurt anything/anyone to leave the programs as they are, and can even be helpful to people writing scripts to automate management of their servers. Chris
> > > > Let's remove the "I don't want to think" utilities like > > > >{create,destroy}{db,user} and force DBA's to actually use the *proper* > > > >functions. > > IMHO (actually make that IMVeryHO) This is probably a bad idea... We > should just update the man pages to detail the SQL code that can be used > instead of the command. It doesn't hurt anything/anyone to leave the > programs as they are... True. > ... and can even be helpful to people writing scripts to > automate management of their servers. This is already possible with psql. Most of my psql work is done via a makefile, infact. I tend to put my SQL into a file such as create.sql and destroy.sql then my project makefile can handle the rest just by having psql read those files when needed. Sure, I do that for development, but it would work for maintenance. Bruce Tong | Got me an office; I'm there late at night. Systems Programmer | Just send me e-mail, maybe I'll write. Electronic Vision / FITNE | zztong@laxmi.ev.net | -- Joe Walsh for the 21st Century
> > > > Let's remove the "I don't want to think" utilities like > > > >{create,destroy}{db,user} and force DBA's to actually use the *proper* > > > >functions. > > IMHO (actually make that IMVeryHO) This is probably a bad idea... We > should just update the man pages to detail the SQL code that can be used > instead of the command. It doesn't hurt anything/anyone to leave the > programs as they are, and can even be helpful to people writing scripts to > automate management of their servers. man pages already updated. -- Bruce Momjian | 830 Blythe Avenue maillist@candle.pha.pa.us | Drexel Hill, Pennsylvania 19026 + If your life is a hard drive, | (610) 353-9879(w) + Christ can be your backup. | (610) 853-3000(h)