Re: Automatic testing of Debian packages

2005-04-25 Thread Goswin von Brederlow
Mads Lindstrøm <[EMAIL PROTECTED]> writes: >... > Do there exist, any other kind of automatic tests that I have not > mentioned here? > > Anything else I have missed? > > Which errors are typical for Debian packages? This is important as > certain kinds of error will properly newer be caught autom

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-14 Thread Stephen Birch
Josselin Mouette([EMAIL PROTECTED])@2005-04-12 09:20: > Why? When you don't know Perl, and you feel like improving a software in > Perl is like eating oysters with skiing gloves, LOL > rewriting the software in Python so that you can work on it seems > like the best solution. An even better solu

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Nico Golde
Hello Josselin, * Josselin Mouette <[EMAIL PROTECTED]> [2005-04-12 11:53]: > Le mardi 12 avril 2005 à 08:31 +0200, Andreas Tille a écrit : > > The reason: "I just rewrite an application because the language it is > > written in." sounds a very stupid reason to me. > > Why? When you don't know Per

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Kevin Mark
On Tue, Apr 12, 2005 at 09:20:28AM +0200, Josselin Mouette wrote: > Le mardi 12 avril 2005 à 08:31 +0200, Andreas Tille a écrit : > > The reason: "I just rewrite an application because the language it is > > written in." sounds a very stupid reason to me. > > Why? When you don't know Perl, and you

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Nico Golde
Hello Emanuele, * Emanuele Rocca <[EMAIL PROTECTED]> [2005-04-12 07:53]: > * [ 11-04-05 - 22:03 ] Nico Golde <[EMAIL PROTECTED]> wrote: > > * Martin Schulze <[EMAIL PROTECTED]> [2005-04-11 22:00]: > > > Why are there Vi and Emacs? > > > Why are there Perl and Python? > > [...] > > But thats

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Thijs Kinkhorst
On Mon, April 11, 2005 22:26, Emanuele Rocca said: > Well some differences came out: > - linda has proper l10n strings for most errors (in German) > - different output formats > - different test sets > - linda is faster So, linda is better than lintian? Faster and localized, that sound like good

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Andreas Tille
On Tue, 12 Apr 2005, Alexander Schmehl wrote: It doesn't for me after I heard, that foo was orhpaned. None volunteered to adopt it, but someone volunteered to write something, that does work in his prefered language. Sometimes it is easier to restart from scratch than to try to adopt an foreign t

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Alexander Schmehl
* Andreas Tille <[EMAIL PROTECTED]> [050412 08:31]: > >Guess some people have preferences for either languages or other. > The reason: "I just rewrite an application because the language it is > written in." sounds a very stupid reason to me. It doesn't for me after I heard, that foo was orhpaned.

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-12 Thread Josselin Mouette
Le mardi 12 avril 2005 à 08:31 +0200, Andreas Tille a écrit : > The reason: "I just rewrite an application because the language it is > written in." sounds a very stupid reason to me. Why? When you don't know Perl, and you feel like improving a software in Perl is like eating oysters with skiing g

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-11 Thread Andreas Tille
On Mon, 11 Apr 2005, Martin Schulze wrote: Why are there Vi and Emacs? Different taste of the users - valid reason. Why are there Perl and Python? Different taste of programmers - valid reason. Why are there viewcvs and cvsweb? Why are there cvs-syncmail and cvs-mailcommit? Why are there tkirc and

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-11 Thread Emanuele Rocca
* [ 11-04-05 - 22:03 ] Nico Golde <[EMAIL PROTECTED]> wrote: > * Martin Schulze <[EMAIL PROTECTED]> [2005-04-11 22:00]: > > Why are there Vi and Emacs? > > Why are there Perl and Python? > [...] > But thats not my problem. The programs etc. you showed are very > different in using, look etc

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-11 Thread Nico Golde
* Martin Schulze <[EMAIL PROTECTED]> [2005-04-11 22:00]: > Nico Golde wrote: > > There is another thing that I don't understand often. Why > > are there linda and lintian? > > In my opinion this makes things difficulter. Both have to > > coordinate themselves and keep their policy rules up to > > d

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-11 Thread Martin Schulze
Nico Golde wrote: > There is another thing that I don't understand often. Why > are there linda and lintian? > In my opinion this makes things difficulter. Both have to > coordinate themselves and keep their policy rules up to > date. Why are there Vi and Emacs? Why are there Perl and Python? Why

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-10 Thread Nico Golde
Hallo Emanuele, * Emanuele Rocca <[EMAIL PROTECTED]> [2005-04-10 19:21]: > * [ 08-04-05 - 15:38 ] Nico Golde <[EMAIL PROTECTED]> wrote: > > There is another thing that I don't understand often. Why > > are there linda and lintian? > > In my opinion this makes things difficulter. Both have to >

Re: lintian & linda (was: Automatic testing of Debian packages)

2005-04-10 Thread Emanuele Rocca
* [ 08-04-05 - 15:38 ] Nico Golde <[EMAIL PROTECTED]> wrote: > There is another thing that I don't understand often. Why > are there linda and lintian? > In my opinion this makes things difficulter. Both have to > coordinate themselves and keep their policy rules up to > date. [...] > The on

Re: Automatic testing of Debian packages

2005-04-10 Thread Ola Lundqvist
Hello When it comes to testing you have to ask yourself what you want to test. We have a couple of different types of tests: * Component test: Tests if a program component (object, function etc) is working as specified. Normally this kind of tests are implemented in the same programming la

lintian & linda (was: Automatic testing of Debian packages)

2005-04-08 Thread Nico Golde
Hallo Mads, * Mads Lindstrøm <[EMAIL PROTECTED]> [2005-04-08 14:08]: > I am writing this email as I think there could be done more to > automatically test Debian packages. This would improve quality, as > discovering errors is the first step in correcting them. I am not a > Debian maintainer, but

Re: Automatic testing of Debian packages

2005-04-08 Thread Pierre THIERRY
Scribit Michael Schiansky dies 08/04/2005 hora 11:44: > None of my maintained packages do provide such things. For all of them > any automatic test is impossible as they require user interaction. apt-get install expectk > IMHO it's not worth the efford to implement such an autotest-system > just

Re: Automatic testing of Debian packages

2005-04-08 Thread Michael Schiansky
Hi Mads! On Fri, Apr 08, 2005 at 10:37:27AM +0200, Mads Lindstrøm wrote: > It would be nice if packages were automatically tested under different > setups, like language settings or other things that may affect a package. > Also each package maintainer could supply a script which should test if t

Re: Automatic testing of Debian packages

2005-04-08 Thread Petter Reinholdtsen
[Mads Lindstrøm] > Do there exist, any other kind of automatic tests that I have not > mentioned here? More automatic testing would be good. Some packages run self-tests as part of the build process. More packages should do that. A long time ago, I wrote a prototype for testing package upgrades

Automatic testing of Debian packages

2005-04-08 Thread Mads Lindstrøm
Hi I am writing this email as I think there could be done more to automatically test Debian packages. This would improve quality, as discovering errors is the first step in correcting them. I am not a Debian maintainer, but have been using Debian for 4+ years. As I am not an insider, I am likely t