Hi Thomas

> very nice. Could you open one or three enhancement bugs for this?!

Yep, http://bugzilla.qooxdoo.org/show_bug.cgi?id=3837 and
http://bugzilla.qooxdoo.org/show_bug.cgi?id=3838

>> myTest.methodThatDoesNotExist();    // No such method
>No class attribute checking currently. But this would be questionable, as
>JS classes are "open", and attributes can be added at any time. A warning
>would be in place.

Definitely a tricky thing to catch correctly without generating too many
spurious warnings; this is kind of linked to the suggestion about using
comment annotations to indicate types etc and would require that code was
written "properly", i.e. declared methods and properties.

>NB: We had interesting feedback on warnings so far. Some people are
>disturbed by them, mistaking them as errors and find it hard to judge
>their relevance correctly. Others are annoyed and are not interested in
>them at all. <sigh>...

Interesting; the extra warnings could be reserved for the "lint" job so they
are there for those that want and appreciate it?

>> Coming from a world of JIT compilers and IDEs that do constant builds in
>> the
>> background to check your code and do code completion, it's another area
>> where aspects of JavaScript development are like stepping back into the
>> dark ages
>
>That's the price you pay for working with novel technologies. And once
>these mature, and environments and tools reach the level of comfort, new
>things and platforms pop up, and the cycle restarts (just think of
>mobile). I'm having flash-backs every other day...

:) Sure, I remember vi and VT220 terminals and I'm really glad for what we
have now.  What started me thinking was if the generator was kept in memory
and constantly compiled (eg under Jython in an Eclipse plugin) warnings
could be continuously fed into the IDE etc and we would be a big step
closer.

>> How about using the comment annotations to specify the arguments and
>> return
>> types and check that values are returned or add type validations during
>> "source" builds?
>Nice one -> bug.

Did you mean a bug?  I did it as a feature request,
http://bugzilla.qooxdoo.org/show_bug.cgi?id=3839

John



------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
qooxdoo-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

Reply via email to