[EMAIL PROTECTED] wrote:
I think it's probably unrealistic to create and maintain a big archive
with all of this. However, I'm sure few of you will be surprised when I'm
now going to suggest how we partially use the wiki for helping with this.
So here's one possible approach:
>
> ...
If this makes sense, I'd be happy to set it up. Something like this would
also have made it much easier for me to test LyX when it was time for 1.4.
I like your idea, if you could set this up easily please do so.
We could have automated tests but many things must be tested manually.
For example my math manual is build upon sequences to create a formula,
e.g. the sequence:
A_u uparrow ^2
should create an "A" with the subscript "u" and the superscript "2". To
have a look how it should look there is a PDF-file for the testers to
compare. If a sequence doesn't produce the expected results you found a
bug. (Btw. this case is bug 2581).
That's my way of testing features. The same can be done using my table
manual in the wiki and I'm sure we could produce more test LyX-files
with sequences how to do something to achieve a certain result and a
PDF-file to compare. This procedure cost a bit of time but you only need
to do this one time before a major release.
regards Uwe
