Re: [Firebird-test] Firebird Testing - A plan for discussion

[Maverick <ye...@ho...>]

I agree with Martijn.  I am offering some time to test FB.  However, I did
not think that I would need to install a development environment for FB
itself.

I was thinking more along the lines of preparing a set of scripts to run
against my database and comparing the consistency of results and
performance.  My db is about 50MB in size, soon to grow to about 300MB, used
in a data-warehouse like environment (I know, it's not ideal, but the
performance is ok with aggregations done upfront).

My environment is Win2000/WinME/Peanut Linux 8.4 (small but enough).

--
Regards
Ray Mond

""Martijn Tonies"" <m.t...@up...> wrote in message
news:005e01c0dcc1$1bae9900$0100a8c0@seal...
> Hi all,
>
> I offered testing some time ago and already had a few email with Mark and
> Tord.
>
> One of the points I very much dislike (and actually restrain me from
> testing) is the need to install a (visual) C++ suite. This is a big no-no
> for me... Besides the fact that I have zero C++ (alright, just a little)
> knowledge, I do not want to install a complete development suite for
testing
> IB/Firebird.
>
> I would like to see a set of scripts/application that runs a few
> tests/durability tests/volume tests etc...
>
> In general, I think that (mind you, I don't have knowledge about TCS):
>
> - after every change, the programmer should think of what has changed
> opposed to the code before the change
> - if it was changed because of a bug:
>   1) is the normal behaviour still available
>   2) does the original 'buggy behaviour' still exists (probably not :))
>   3) do the tests that test extraordinary situations still run
> - new functionality:
>   1) the programmer should make a detailed description of how this should
> work
>   2) make a test plan, with extraordinary situations, the normal behaviour
> and possible buggy situations when changed (in the form: when I do this,
the
> output should be: <something>)
>   3) test everything before declaring it 'working' code
>
> Furthermore I think that:
> 1) nothing should be installed (except the IB/FB client)
> 2) there should be detailed instructions for volume testing as well as a
> filling procedure
> 3) there should be detailed instructions considering the output of the
tests
> 4) there should be an easy-to-use application/set of scripts that test all
> or parts of the engine
> 5) there should be ways that the whole testing process can be devided into
> parts that can be performed by several people (or groups of people)
instead
> of a single testbed for all tests
> 6) there should be applications for durability and load testing
>
> Well, that's about it.
>
> For starters, I can contribute with testing on Win2000, Linux (Suse 7.1)
> (both AMD CPUs).
>
>
> Comments anyone?
>
> Martijn Tonies
> InterBase Workbench - the developer tool for InterBase and Firebird
> http://www.interbaseworkbench.com
>
> Upscene Productions
> http://www.upscene.com
>
> "This is an object-oriented system.
> If we change anything, the users object."
>
>
>
>
>
> _______________________________________________
> Firebird-test mailing list
> Fir...@li...
> http://lists.sourceforge.net/lists/listinfo/firebird-test
>