headers
Kevin Walzer | 1 Dec 01:18

Questions re: DarwinPorts API for TkDarwinPorts


I've begun looking at the DarwinPorts API for TkDarwinPorts, and also at
the code for the "port" utility, and I'm getting a bit lost: I'm not
sure where to begin in trying to hook directly into DarwinPorts'
internals. This is, doubtless, a function of my still having some things
to learn about Tcl. Still, it seems to me that calling the DarwinPorts
API directly from my application is re-inventing the wheel; I'm not sure
why I should do this when port already does it so much better than I
probably could.

I understand that there's no real advantage to using Tk if I'm just
going to wrap "port" as if I were calling any a shell command (as
version 0.1 of TkDarwinPorts does). If that were my goal, I'd simply use
AppleScript Studio and have a polished GUI up-and-running pretty
quickly. Still, I'm not sure I want (or need) to re-create "port" for my
own GUI. Would it be feasible to actually use a good deal of the "port"
code in my own script? That seems to me to be a more promising approach.
After all, my goal here is to provide a higher-level interfce to what is
already provided by the DarwinPorts infrastructure (and as Fink
Commander does for Fink), rather than re-implementing a good deal of
that infrastructure on my own.

Any advice is appreciated.

--
Kevin Walzer, PhD
WordTech Software--Open Source Applications and Packages for OS X
http://www.wordtech-software.com
http://www.smallbizmac.com
http://www.kevin-walzer.com
(Continue reading)

Permalink | Reply | 5 comments |
headers
Henry Vogt | 1 Dec 12:05
Picon
Favicon

qt-mac doesn't compile - solved.

Hello again,

just for the records - everything went smoth after i removed the 
(previouls installed) x11/qt Port from my system. So there is 
(obviously?) some sort of conflict between the two versions, being 
installed on the same machine, which i wasn't aware of.

Am 30.11.2004 um 11:18 schrieb Chris Ridd:

> On 30/11/04 10:00 am, Henry Vogt <hv <at> tuebingen.mpg.de> wrote:
>
>> Hello,
>>
>> Trying to build the aqua/qt-mac Port, it stops in
> [...]
>> I'm using Apple's Developer Tools (Xcode 1.5), if that matters:
>> gcc version 3.3 20030304 (Apple Computer, Inc. build 1666)
>>
>> Has anybody an idea what's wrong or mwhat i'm missing ?
>> Thanks in advance.
>
> I built qt3-mac successfully a week or so back, and am using the same 
> build
> of gcc/g++ as you. You've obviously got the same Portfile as me - have 
> you
> got the up-to-date patches in qt3-mac/files?
...

Kind Regards
--

--
(Continue reading)

Permalink | Reply | 1 comment |
headers
Chris Ridd | 1 Dec 12:08
Favicon

Re: qt-mac doesn't compile - solved.

On 1/12/04 11:05 am, Henry Vogt <hv <at> tuebingen.mpg.de> wrote:

> Hello again,
> 
> just for the records - everything went smoth after i removed the
> (previouls installed) x11/qt Port from my system. So there is
> (obviously?) some sort of conflict between the two versions, being
> installed on the same machine, which i wasn't aware of.

Ah. The x11/qt port is an older version, and maybe that's the cause of the
conflict.

Well spotted :-)

Cheers,

Chris
Permalink | Reply | 0 comments |
headers
Ismael | 1 Dec 16:18
Picon

Help

Hi all. I need some help. I'm trying to install gromacs 3.1.5_pre1 on  
MacOSX using darwinports. However, I get this message:

Ismael:~ ismael$ sudo port install gromacs
Password:
--->  Configuring gromacs
Error: Target com.apple.configure returned: configure failure: shell  
command "cd  
"/Users/ismael/darwinports/dports/science/gromacs/work/gromacs 
-3.1.5_pre1" && LDFLAGS="-L/opt/local/lib"  
CPPFLAGS="-I/opt/local/include" ./configure --prefix=/opt/local  
--exec-prefix=/opt/local/gromacs" returned error 1
Command output: checking how to hardcode library paths into programs...  
immediate
checking whether stripping libraries is possible... no
checking dynamic linker characteristics... darwin7.6.0 dyld
checking if libtool supports shared libraries... yes
checking whether to build shared libraries... no
checking whether to build static libraries... yes
configure: creating libtool
checking whether ln -s works... yes
checking for special C compiler options needed for large files... no
checking for _FILE_OFFSET_BITS value needed for large files... no
checking for _LARGE_FILES value needed for large files... no
checking for _LARGEFILE_SOURCE value needed for large files... no
checking for fseeko... yes
checking for sqrt in -lm... yes
checking for sfftw.h... no
checking for fftw.h... no
configure: error: Cannot find any single precision sfftw.h or fftw.h
(Continue reading)

Permalink | Reply | 5 comments |
headers
Dr.Ernie | 1 Dec 17:23

Re: Questions re: DarwinPorts API for TkDarwinPorts

Hi Kevin,

Its been a while since I looked at the code, but what I think really 
needs to be done is to factor out 'port' into more discrete 
subroutines, which you could call directly from your Tk port.   That 
is, right now (if memory serves) port contains both the parsing logic 
for the CLI, as well as the application logic for running the ports.

If so, then the optimal solution would be to factor out port a little 
more cleanly, so it becomes just a shell over the 'port' API, which is 
to say a peer of TkDarwinPorts.

Then again, I could just be totally blowing smoke.  Any of the 'port' 
maintainers care to weigh in?

-enp

On Nov 30, 2004, at 4:18 PM, Kevin Walzer wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> I've begun looking at the DarwinPorts API for TkDarwinPorts, and also 
> at
> the code for the "port" utility, and I'm getting a bit lost: I'm not
> sure where to begin in trying to hook directly into DarwinPorts'
> internals. This is, doubtless, a function of my still having some 
> things
> to learn about Tcl. Still, it seems to me that calling the DarwinPorts
> API directly from my application is re-inventing the wheel; I'm not
(Continue reading)

Permalink | Reply | 4 comments |
headers
Chris Ridd | 1 Dec 17:16
Favicon

Re: Help

On 1/12/04 3:18 pm, Ismael <ismael <at> laguna.fmedic.unam.mx> wrote:

[...]
> That is also necessary to do if your compiler doesn't search
> /usr/local/include and /usr/local/lib by default.
> You can find information at www.gromacs.org, or in the INSTALL file.
> 
> 
> 
> 
> I installed the fftw 3.0.1 by myself (before all of this). I deleted
> this package and I got the same message... I reinstalled using the
> ./configure --enable-float command and the result was the same. Some
> light please...

You probably installed fftw in /usr/local. The gromacs configure script
doesn't know to look in /usr/local for your fftw stuff. It even tells you
this :-)

Should fftw be a proper darwinport, and should it be a dependency of
gromacs?

A quick way forward for you is to remove your fftw installation, and rebuild
it using:

    ./configure --prefix=/opt/local [whatever else]

Then the gromacs configure script should find it.

Cheers,
(Continue reading)

Permalink | Reply | 4 comments |
headers
James Berry | 1 Dec 17:32

Re: Building Tomcat5 with darwinports

I haven't looked at the dependency checking code, but I wonder if in  
these cases there is a discrepancy between the PATH as understood by  
port, and the PATH ultimately used by the shell invoked to do the  
build.

-jdb

On Dec 1, 2004, at 8:22 AM, Markus W.Weissmann wrote:

> Hi folks,
>
> I had this problem some days ago with someone building jython; there  
> was a build-dep on ant
> and he also had ant installed in /usr/bin, though his build failed  
> with an "ant not found" error;
> after fiddling around, he somehow managed to build if w/o _any_  
> modifications on the Portfile.
> Either this is some weird race condition or I dont know what.
> If someone can reproduce this, please tell me!
> $ port build some_port
> -> "ant not found"
> $ which ant
> /usr/bin/ant
>
> I'm cc'ing this to the darwinports mailing-list as it seems to me that  
> this could be a bug in port(1);
>
>
> cheers,
>
(Continue reading)

Permalink | Reply | 5 comments |
headers
Markus W. Weissmann | 1 Dec 17:22

Re: Building Tomcat5 with darwinports

Hi folks,

I had this problem some days ago with someone building jython; there  
was a build-dep on ant
and he also had ant installed in /usr/bin, though his build failed with  
an "ant not found" error;
after fiddling around, he somehow managed to build if w/o _any_  
modifications on the Portfile.
Either this is some weird race condition or I dont know what.
If someone can reproduce this, please tell me!
$ port build some_port
-> "ant not found"
$ which ant
/usr/bin/ant

I'm cc'ing this to the darwinports mailing-list as it seems to me that  
this could be a bug in port(1);

cheers,

Markus

On Dec 01, 2004, at 15:45, James Berry wrote:

> Hi Marco,
>
> Thanks. As you suspected, I have already seen and fixed the issue with  
> the struts build.
>
> I'm somewhat puzzled by the build failure for junit. The junit
(Continue reading)

Permalink | Reply | 6 comments |
headers
Gregory Wright | 1 Dec 17:50
Picon

Re: Help


Hi,

On Dec 1, 2004, at 11:16 AM, Chris Ridd wrote:

> On 1/12/04 3:18 pm, Ismael <ismael <at> laguna.fmedic.unam.mx> wrote:
>
> [...]
>> That is also necessary to do if your compiler doesn't search
>> /usr/local/include and /usr/local/lib by default.
>> You can find information at www.gromacs.org, or in the INSTALL file.
>>
>>
>>
>>
>> I installed the fftw 3.0.1 by myself (before all of this). I deleted
>> this package and I got the same message... I reinstalled using the
>> ./configure --enable-float command and the result was the same. Some
>> light please...
>
> You probably installed fftw in /usr/local. The gromacs configure script
> doesn't know to look in /usr/local for your fftw stuff. It even tells 
> you
> this :-)
>
> Should fftw be a proper darwinport, and should it be a dependency of
> gromacs?

fftw is a proper darwinport (both versions 2 and 3). See math/fftw and 
math/fftw-3.
(Continue reading)

Permalink | Reply | 3 comments |
headers
Kevin Walzer | 1 Dec 18:06

Re: Questions re: DarwinPorts API for TkDarwinPorts


Ernie,

I'm still not entirely clear on what you're suggesting, so let me
outline my current understanding and you can clarify what I'm missing.

Here's TkDawrinPort's current structure, with each layer wrapping the
lower layer:

TkDarinports (TkAqua GUI)--> port (tcl script providing CLI -->
DarwinPorts API (manages client-level access) --> Ports API (DP primitives)

This seems to me to be a pretty clean separation of interface from
back-end logic. Here's my understanding of the direction I'm being asked
to take:

TkDarwinPorts --> DarwinPorts API --> Ports API

Re-doing TkDarwinPorts in this manner effectively re-implements the
entire interface logic, rather than simply wrapping the interface
provided by port.

Speaking candidly, I have two concerns about this:

1. I'm not sure I'm yet proficient enough in Tcl to do this: I'm pretty
sure that port.tcl is better-written than anything I could come up with
right now. In any event, this will cause a fairly substantial delay in
getting TkDP  into a usable form because I will have to spend a great
deal of time on the learning curve.
(Continue reading)

Permalink | Reply | 3 comments |

Gmane