W3C home > Mailing lists > Public > whatwg@whatwg.org > April 2011

[whatwg] Application Cache Improvement

From: Edward Gerhold <edward.gerhold@googlemail.com>
Date: Tue, 12 Apr 2011 20:23:42 +0200
Message-ID: <BANLkTin-AGgDgyH5LEd=QGTx4=ezUWm0JQ@mail.gmail.com>
Dear Subscribers,

from the beginning on, i provided almost all functions you would need for a
good appcache which is able to cache dynamic websites, e.g. the joomla
content manager and the articles behind.
The function complex i personally forgot is the "rename urls" complex.
Sometimes we would have to rename the url for the use of the page in the
cache. Randy?s reply enlightened me, he told
me, that sometimes cms systems use a index.php for the whole pages they put
out. I remember viewtopic.php from the forums, which always appeared in my
old browser, if i tried to save
pages, they hadn?t been saved with the name of the head title. I answered
something, that i wouldn?t see the values if they are posted, said that a
get request with index?
postid=1024&catid=512 should always lead to the same place. I could be
wrong, i don?t test that, i don?t need for, but i remember no urls in the
cache with any paremeters saved, i added none
to the manifest. Maybe the string is threatened complete and parameters
could also be used, maybe it?s not. Anyways, ...

For the extension of the Application Cache.
Not caching the Master File is sometimes neccessary, too. For quick changes
to the Cache before HTML5 it would be an improvement to let me enter the
manifest in Joomla!?s index.php,
to blacklist the index.php, but to cache the scripts and styles, and to add
pages, i can access from the frontpage on, or try out some fallbacks.
CACHE MANIFEST NOMASTER, where NOMASTER is my new token for you.
Now a little control is needed for all of the files. I still can?t cache the
rest of the pages.

How can i cache the files behind the frontpage?
The cache with add, del, mod, get for the pages. And
begin,seek,pos,next,prev,end for the Cache List. And an event fired after
each (including the iterator) of the operations.The event is allowing the
programmer to make final changes to the cache. For the iterator a writeback
is useful and saves time. If you don?t want to make changes to the cache,
just copy the event data over and use the other variable, or just make no
changes. If you want to, the writeback of the event data on changes of the
data is very useful.

Now the new.

What is the point, Randy brought you on?
The URLs have sometimes to be changed.
The cache should have additionally own internal identifiers.

There have changes to be made to the URLs for caching the files.
Old Url and new Url have to be kept. Both.
User Agents SHOULD be able to change and use the old and new urls.
User Agents MUST be capable of to manipulate the links the the cached file
and the files pointing to.

For the extension of the appcache, the has to be
1. The NOMASTER token
2. Additional Functions and an enumerator.

And:
3. Rename the Url for Local Use and have Local and Network Url stored.
The user agents have to be capable to store cached objects with a new name
automatically.
The programmer MUST be allowed to manipulate URLs for caching files.
He HAS TO be allowed to replace the links pointing to the correct cached
file by his Web Application.
3. Let the cache objects have a local name.

The GET parameters have to be saved with, the POST data fields could be
saved with, (for the List of all urls saved).
Leads me to another idea.

4. The Cache Object SHOULD save the whole HTTP Request made to!
Exception? On secure pages. You should turn that off on the bank, or? But
you won?t cache your bank pages, or?
I hope 4. is not making your to do list to long now. But saving all behind
index.php (?&=#) and POST is to be considered for comparation
of urls in the cache or needed by a good offline reader. Simply writing down
the whole request and splitting it up into fields is easier.
4. Write down the request and split it into fields for each cache object.

"I have to rewrite all this again." I think i have stated the three (four!)
things clear. Enjoy the new issue of the application cache improvement.
I?ll keep you up to date with a better formulation, if i can. If you would
like to talk about. My channel is open, this channel is, too.
With the above i could extend my Joomla! so far, that i could cache the
whole CMS. With the actual application cache i can save a boring page,
or a good offline application basing on a boring page and much much design,
grafix and fancy scripts. But i couldn?t cache a cms and that is
shit for the Offline Cache, coz i can?t take my favourite site how far i
want to the cache. That?s better than bookmarking, or saving html pages
with no connection. The Offline Cache with my saved pages are linked with
each other, i can browse it normally, how far i?ve saved it. That?s
the aim of my application cache improvement again.

3. Renameability of Urls for the Cache is a MUST. Local and Network Name
have to be kept both.
4.Another thing.  Write down the whole HTTP Request and split it into
fields. Extend the Cache Item with the HTTP fields.

Ok and now.
Don?t forget the add and enumerate functions for me and my Joomla!,
Wordpress, phpBB, Wiki. I could read linked wikipedia offline with my
offline reader application.
Which is installable over the web with an index.html and a manifest, later.
My offline reader would be using the application cache. Which it couldn?t
today. Like i can not
cache my Joomla!, which is "weak" coz it?s THE cache (THE cache of HTML5).
Comments are welcome.

Bye

With friendly regards
Edward Gerhold
Received on Tuesday, 12 April 2011 11:23:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 30 January 2013 18:48:03 GMT