[webkit-qt] Implementing my cache- control flow

[Tarandeep Singh]

On Fri, May 21, 2010 at 2:04 PM, Benjamin Meyer <ben at meyerhome.net> wrote:

> Do you plan on putting your db cache implementation up online?  I would be
> interested in checking it out (I wrote the simple disk one in Qt).  Have you
> written benchmarks yet?
>
>
In my case, speed is not that important. But having a centralized place for
cache is more imp. We are running webkit on multiple computers and having a
central place to cache the cookies, jss, css is useful as different webkit
instances running on different computers can share this cache.

Further, instead of using db, I am putting these things in memcache ( we
already have couple of memcache servers running so it made more sense to put
them there).

Regarding putting the implementation online, right now I don't have the
bandwidth to contribute the code back to open source and maintain it, but I
will be happy to help anyone with code snippets etc.

-Tarandeep


> -Benjamin Meyer
>
> On May 21, 2010, at 1:15 PM, siddharth.mathur at nokia.com wrote:
>
> > Tarandeep, Benjamin,
> >
> > You can reduce your client's chattiness with the server if you also set
> the QNetworkRequest::PreferCache flag on outbound HTTP requests:
> >
> > request2.setAttribute(QNetworkRequest::CacheLoadControlAttribute,
> QNetworkRequest::PreferCache);
> >
> > The default is PreferNetwork, which means that even though the content
> was originally sent with an "expires" head, the cache will still check with
> the server if it has a newer copy. This is bad for performance, but good for
> correctness.
> >
> > http://doc.trolltech.com/4.6/qnetworkrequest.html#CacheLoadControl-enum
> >
> > http://doc.trolltech.com/4.6/qnetworkdiskcache.html#details
> >
> >
> > Siddharth
> >
> >> -----Original Message-----
> >> From: webkit-qt-bounces at lists.webkit.org [mailto:webkit-qt-
> >> bounces at lists.webkit.org] On Behalf Of ext Benjamin Meyer
> >> Sent: Thursday, May 20, 2010 12:03 AM
> >> To: Tarandeep Singh
> >> Cc: webkit-qt at lists.webkit.org
> >> Subject: Re: [webkit-qt] Implementing my cache- control flow
> >>
> >> All of the html caching is done outside of webkit and in the QNetwork
> >> classes.  Referring to the docs for the QNetworkCache interface is best.
> >> What you have it about correct, often times even if the cache is ok it
> >> will update headers if I remember.
> >>
> >> -Benjamin Meyer
> >>
> >>
> >> On Apr 13, 2010, at 2:25 PM, Tarandeep Singh wrote:
> >>
> >>> Hi,
> >>>
> >>> I am trying to implement a Database cache for webkit for my project.
> >> Can
> >>> someone please confirm if this control flow is correct-
> >>>
> >>> 1) For a given url, webkit will get its metadata by calling the
> >> function
> >>>    QNetworkCacheMetaData metaData( Url)
> >>>
> >>> 2) webkit will check the last modified date, expiration date stored in
> >> the
> >>> metadata and if these dates are fine (compared to the actual dates
> >> found on
> >>> the site) then it will get the data by calling the function-
> >>>    QIODevice * data( Url)
> >>>
> >>> 3) else it will get the file from network and then store in cache by
> >> first
> >>> calling prepare( ) and then insert( )
> >>>
> >>> -Tarandeep
> >>> _______________________________________________
> >>> webkit-qt mailing list
> >>> webkit-qt at lists.webkit.org
> >>> http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt
> >>
> >> _______________________________________________
> >> webkit-qt mailing list
> >> webkit-qt at lists.webkit.org
> >> http://lists.webkit.org/mailman/listinfo.cgi/webkit-qt
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-qt/attachments/20100521/1c2c3533/attachment.html>