There isn't a .mak file for pg_dumpall, nor for any client program
except psql as far as I can see. In fact, the nightmare difficulties of
keeping MS type makefiles in sync with our GNU type makefiles is one of
the big reasons that we chose to adopt MinGW instead as the supported
build environment for the native Windows port.
With the Win32 port there will be native Windows versions of all
programs. It would certainly make sense to have a "clients only" Windows
install package, as well as a "server+clients" package - although I am
not working on the packaging. I have no knowledge about Mac clients,
although I believe the full range of programs work on OS-X.
You might be able to get someone (e.g. me :-) ) to work for $$$ making
packaged Windows binaries of other 7.4.1 client apps, assuming it is
possible, if they are required before the next release. I doubt anyone
will work on it otherwise - those of us interested in Windows apps are
concentrating on the full Windows port.
My own solution to the problem, if I am using a non-Unix workstation, is
usually to connect to a Unix box (via ssh and/or VNC) and run the
required client from there. That plus Samba and/or NFS allow me to put
things where I want and run them where I can.
My own opinion is that Cygwin is usually fine for client use such as
this - I agree that I wouldn't run it for production server use.
Robert Treat wrote:
>So the counter to this of course is "Hey Paul, we'll give you a complete
>database server for free to replace your existing databases and save you
>tons and tons of $$$, and all we ask in return is that you write a small
>terminal program for your OS and donate it back." That sure sounds
>like a good trade to me, don't you think?
>In reality though I'm pretty sure that psql and pg_dumpall do compile
>under Win32 and MacOS, if not under 7.4 then certainly under 7.5, so
>really these options are already available to you, they are just not in
>a pretty package. (hey, let me CC one of the developers in case not...)
>Now, if you want to help by compiling the client programs and wrapping
>them up into a nice one click install for windows users (or mac users)
>we have project space available on gborg.postgresql.org you're more than
>welcome to use. Otherwise you'll have to wait for someone else to decide
>they want to maintain such a thing, and no one has stepped up yet
>(though it has been discussed before)
>On Wed, 2004-02-11 at 08:58, Paul Smith wrote:
>>I am evaluating Postgres for an existing Java application. I already have
>>the app running with Oracle ($$$) and Sybase (oooold!) and I am trying to
>>convince my management that it is worth going fully open source. (We already
>>use Java, JBoss, Eclipse etc.) But the problem I am having, and it seems I
>>am not alone, is the absence of Client support in postgres. I can fully
>>understand why the db server is Linux/Unix as this covers the majority of
>>servers used for enterprise applications, but I can not understand the
>>abscence of platform support for the client software. Where is the Mac psql,
>>or the Win XP version of pg_dumpall? Clients do not use Linux boxes, they
>>use Wintel or Mac boxes. You do not have to like it, but ignoring them
>>doesn't make them go away.
>>I do not need the fancy GUI interfaces like PGExplorer, they are nice, and
>>handy to port my scripts and test my stored procedures, but I can't use them
>>to automate the creation of my database.
>>Sybase had iSQL for scripts and bcp for data loading. Oracle had sqlplus and
>>sqlldr. These are the basic client side tools required to automate the setup
>>and maintence of a non-trival database, and they are available on almost any
>>platform you care to name.
>>And before anybody suggests I use Cygwin, I have used Cygwin, I do not like
>>it (mild understatement) and nor am I prepared to learn Tcl just to run a
>>script that already works. Run a script to load a telnet session to run a
>>script to run my script? Be serious.
>>Please, guys. Open source is the way to go. We need to convince accountants
>>that it is the sensible, safe and cheap way to develop software. But that
>>means we have to do our jobs! When we write client-server systems, write the
>>clients as well as the servers.
>>Can it really be that difficult to write a native mode psql? Have you
>>considered doing it in Java?
In response to
pgsql-advocacy by date
|Next:||From: David Costa||Date: 2004-02-12 22:33:44|
|Subject: PHP/Postgresql Advocacy|
|Previous:||From: Magnus Hagander||Date: 2004-02-12 21:02:07|
|Subject: Re: Rant to the guiding lights of Postgres|