[Borges-users] Patches, Proposals: Patch for current CVS

[Kaspar Schiess]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello Eric,

Thanks for considering my monster post.

| I'll do my best to comment on these, but it'd be easier if there was one
| email per issue.  It also makes it easier to commit the patches I
| accept right away.
Ok. Sometimes things accumulate ;)

|>a) Introduction of text_area_on analoguous to text_input_on.
patch is called 040401-borges-textarea.diff

|>b) Security patch to input type=* value=... and to text_area: Values
| Ok.
| Seaside's #text now encodes the text be default.  I'd like to do this as
| well.  Could this problem be solved in a more general way by simply
| encoding all attribute values?
I don't think it's a good idea to encode all attribute values (or
anything at all) by default. Let the user choose what he wants to do,
and do the obvious where there is just one thing he can want to do.

I think it is very important to have a way of outputting direct html; so
the difference between #text and #encode_text should stay there.

I have memories from attribute values that did not support having &..;
encoded things in them, like alt tags of images. Maybe this has changed
and we can pull the stunt with the attribute values: Most often the
problem surfaces where user input is stuffed in attributes, and for that
the <input and <textarea tags are known.

What is the html version we are aiming at ?

Patch is named 040401-borges-value-security.diff

|>c) Introduction of more verbous error output if Borges cannot find a
|>continuation to call.
| Ok, tentatively.  When do you run into this problem?
Problem surfaces when user forgets (silly me) the use of radio_group.
While this should throw an error, it should be less cryptic than what
the code now outputs. What I submit here is what helped me find the bug
in my code.
040401-borges-verbosity.diff

| This will generate invalid HTML, since <form> is a block element,
| <p><form>...</form></p> will throw a validation error on the unopened
| <p>.
Thank you for enlightening me on this point: Tidy says the same. Did not
know that.

|>e) Borges::Path should now work. Can anyone explain to me how to use
|>this properly ?
|
|
| Ugh, it suffers from being to Smalltalkish.  I can clean up your patch,
| too.
|
| I think you can use Borges::Path for breadcrumbs.
I am using it for that, however the application architecture that one
needs for this is not clear to me. I submit my path.rb in whole, for you
to clean it up. (040401-borges-path.diff)

|>g) Introduction of 'HtmlRenderer#action': like 'default_action', but
| Ok.
040401-borges-action.diff

Some other small issues to make your CVS copy identical to my working
copy are in 040401-borges-various.diff. You will give me your opinion on
those; especially the introduction of xhtml doctype is something to talk
about.

All patches relative to todays cvs.

kaspar - code philosopher

- -- stolen off the net --
When you are a Bear of Very Little Brain, and you Think of Things, you
find sometimes that a Thing which seemed very Thingish inside you is
quite different when it gets out into the open and has other people
looking at it.
		-- A. A. Milne, "The House at Pooh Corner"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAa948h6tlx0BWOuARAtHNAKChQqqs6TO88dbLH84LtHrXtK+PGwCfcZgr
3S1BaaR8NhEF5h9aAQoc0D4=
=B9MV
-----END PGP SIGNATURE-----
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 040401-borges-patches.tar.gz
Type: application/gzip
Size: 2711 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/borges-users/attachments/20040401/257c261e/040401-borges-patches.tar.bin