[whatwg] framesets

[Eduard Pascual]

On Sat, Oct 10, 2009 at 6:12 AM, Peter Brawley <pb at artfulsoftware.com> wrote:
> [lots...]

Now, your last mail does describe the use-cases and their presumed
requirements. Your wording is maybe a bit messy, but you have at least
provided something worth discussing.
Just to make sure I (and others) are understanding your proposal as
intended, I'll try to paraphrase it in a more structured way (please,
if I got something wrong, just let me know):

Use case: displaying tree-based content (editable or not). (Note: I'm
omitting the database aspect because it only matters on the server
side: once on the client, the page doesn't care whether the data comes
from/goes to a database, a collection of files, or anywhere else).
Requirements:
1) The display will use multiple subwindows, such as "header",
"tree-view", and "content" (the names are arbitrary, hope they are
descriptive and concise).
2) The subwindows (or some of them) need to be independently scrollable.
3) The subwindows must be resizable.
4) The tree structure being displayed may be protected through some
permission mechanism, so each user only gets to see or interact with
the nodes such user has permissions for.
5) The "back button" and "bookmark" features shouldn't work.

So far, so good. Just if you had gone a bit further...
That's a use-case with a series of requirements. This is a good
beginning, but let's go to the next steps.
Requirement justification: you haven't provided any (it's the
requirements, not the use-case itself, what needs to be justified; a
use-case is inherently justified by the real-world needs it
addresses).
For requirements 1 to 3, I can assume as an implicit justification
something like "this is the behavior every user would expect" or
anything like that, so let's move forward.
For requirement 4, things are a bit weird: should authentication be
handled on the server or on the client side? It seems that it has to
be handled on the server side, so the nodes a user is not allowed to
see should never be sent to the client (otherwise, the user could look
at the source, hack through grease-monkey scripts, or override the
permission system in many ways). If it's handled by the server, then
it isn't relevant to HTML: HTML is a client-side language. If there is
some need to handle (part of) the authentication issue from the
client, you should provide more details and justify such need;
otherwise the requirement can just be omitted as it wouldn't be
relevant.
Requirement 5 does need some justification. In my opinion, there is no
need on your use case for these features (back button and bookmarks)
to work, but is there any real need for them to break? If there is,
please justify it; if there isn't, then that's not a requirement (it
would just be the absence of a requirement for these features to
work). Not needing them to work is *not* the same as needing them to
break.

Now, leaving aside the issues with the last two requirements, we can
move forward. The next step would be to see which requirements are met
by currently existing solutions, and which aren't. I will be ignoring
for now Requirement 4 because it's unclear what the actual requirement
would be.
First, let's look at what the currently existing solutions are: I may
be missing some, but I hope the range is descriptive enough:
A) <table>+<iframe>: This meets requirements 1, 2, and 5 out of the
box. Requirement 3 could be achieved with some javascript.
B) CSS position:fixed + overflow:auto: Again, this meets requirements
1, 2, and 5. Requirement 3 would also be achievable with a bit of
scripting.
C) Insane <div>s + CSS + Scripting: This essentially meets all
requirements (maybe excluding 4, depending on what the actual
requirement is); although at a high development cost. (This would be
the "MSDN style" approach.)
D) HTML4 Frameset + HTML5 documents for frame contents: this meets
requirements 1, 2, 3, and 5 out of the box, it's an almost trivial
upgrade from any HTML4 web-app that takes a similar approach, and is
relatively easy to implement.

Finally, once it is shown that currently existing solutions can't
handle the use-case (which hasn't been shown: C and D both can, and A
and B can as well if a bit of scripting is added to handle the
resizing requirement), it would only be left to verify that your
proposal actually meets the requirements. There is a problem here,
however: What is your proposal? I'm not sure why you haven't described
your proposal. If you want the editor to change the spec, it's quite a
good idea to describe the changes you want to be made on the spec.

In summary, what you need to do is:
1) Correct me if I made any mistakes when trying to synthesize your use-case.
2) Clarify and justify Requirement 4.
3) Justify Requirement 5.
4) Describe your proposal.
5) Show how does your proposal meet all of the requirements for the use-case.

As far as this has gone, my impression is that you want a HTML5
Frameset document type that is exactly identical to the HTML 4
version. It is pointless to "update" a version of a standard to a new
version that includes no changes at all.
Keep in mind that Frameset and content are two different document
types, with different content models (for example, a frameset page has
no <body>). HTML5 currently replaces/updates the Transitional/Strict
document types; it doesn't deal with Frameset because nothing is being
changed on it, so HTML4 Frameset stays valid as the "newest" (despite
its age) standard for frameset "master" pages.

Regards,
Eduard Pascual