[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#215549: Additional commentary on IRC


        The people participating are:

 ron is Ron Lee
 pusling is Sune Vuorela 
 Manoj is Manoj Srivastava 

Manoj> config script is called first, before stuff happens, and passed
Manoj> configure. Then the postinst is run, and the config script runs
Manoj> again, but should not ask the questions a second time
Manoj> if then the user does dpkg-reconfigure, the config script is run
Manoj> again -- but this time it should ask questions
Manoj> (the postinst is only run twice, so is not relevant here)
Manoj> well, the only relevance is, how the config script is called

ron> debconf already automatically doesn't ask the questions again if
ron> they were preconfigured, and does ask them again if you
ron> dpkg-reconfigure ... 

Manoj> ron: but before I ask questions, I might do some lengthy
Manoj> calculations

ron> as I see this it's really only if postinst should do something
ron> _different_ ...  and so far we can't think of a case where it
ron> should that isn't already handled, can we? 

Manoj> right. why should the postinst care if it is being configured or
Manoj> reconfigured? 

ron> I can only imagine that joeyh imagined there might be some really
ron> expensive operation that you do only every want to do once ... but
ron> I can't think of such a thing that you _can't_ later change with
ron> dpkg-reconfigure ... that would be too much like the persistant-foo
ron> udev evil for me ... even probing for hardware is something that
ron> may change (you can always drop an installed disk image into some
ron> totally different machine, and should be able to reconfigure it for
ron> that easily) 


pusling> Manoj: if reconfigured, it might have a exit level at each step
pusling> to just reuse the already present information from the previous
pusling> configure run.

ron> pusling: but then you'd need a dpkg-reconfigure --harder or
ron> something if you really do want to update those things ... 

Manoj> pusling: the whole idea of reconfiguring is not to use the old
Manoj> information  

pusling> Manoj: if the installation script requires 20 user entered
pusling> steps, and you only want to change step2 ? 

ron> then you are abusing debconf?

Manoj> pusling: and you have the postinst ask the user at every step
Manoj> "Are we there yet? Are we there yet?" 

pusling> Manoj: there could just be "next", "back" and "done" buttons ?

ron> how can you know if you are done if you haven't seen the rest of
ron> the questions yet? if debconf can solve the halting problem that
ron> would be much cooler than dpkg sudoku 

Where humor is concerned there are no standards -- no one can say what
is good or bad, although you can be sure that everyone will. -- John
Kenneth Galbraith
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Reply to: