cxf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Tomasz Oponowicz <tomasz.oponow...@gmail.com>
Subject Re: [GSOC] Simple and lightweight HTML-based browser for surveying Apache CXF logs
Date Wed, 07 Jul 2010 22:22:33 GMT
Hi Sergey,

On Wed, Jul 7, 2010 at 11:51 PM, Sergey Beryozkin <sberyozkin@gmail.com> wrote:
> Hi Tomasz
>
> you're doing very well and there's absolutely no need to apologize :-).
> Please take your time, this GSOC project is an important project but I
> personally do not expect the result be of the very top quality, no one can
> do it from the first try :-)

Yeah, I know - I just don't like make a "crap" ;-)

> See, I'm not proposing to write a new servlet implementation. First, I'd
> like to understand what you were trying to achieve by introducing
> StaticContentServlet - I can see it's about serving the static content - but
> what type of content, when exactly it is needed ? Next, I'd like to see if
> either AbstractHttpServlet [1] or BootstrapStorage endpoint can be enhanced
> to do what  StaticContentServlet does ?

At the moment we've got java classes and *.ui.xml files, which are
changed respectively into *.js and *.css files by GWT compiler (it
change java code to javascript).
To be more specified it's done by 'gwt-maven-plugin'. The GWT
dependency has compilation scope - we don't use GWT at runtime. All
these static files are packed into jar file (static flies takes about
0.5MB).  Because classes from package
"org.apache.cxf.management.web.logging.browser.client.*" are no longer
needed, they aren't packed into jar file.

Before user can use our LogBrowser, he has to retrieve all HTML, JS
and CSS files - it can be simply done by go to URL (for example):

<localhost>/logbrowser/LogBrowser.html

This content is served by StaticContentServlet - the static files are
retrieve from jar file and send as HTTP response. These happens only
once (per browser) because all of these files are highly cached. After
that user sign in and retrieve settings using BootstrapStorage
endpoint.

However StaticContentServlet and BootstrapStorage endpoints do similar
job - they "init" application. I think extending BootstrapStorage (to
add StaticContentServlet functionality) will be the best solution.

What do you think about that?

> thanks, Sergey
>
> [1]
> http://svn.apache.org/repos/asf/cxf/trunk/rt/transports/http/src/main/java/org/apache/cxf/transport/servlet/AbstractHTTPServlet.java
>
>
> On Wed, Jul 7, 2010 at 10:27 PM, Tomasz Oponowicz <
> tomasz.oponowicz@gmail.com> wrote:
>
>> Hi Sergey,
>>
>> On Wed, Jul 7, 2010 at 10:40 PM, Sergey Beryozkin <sberyozkin@gmail.com>
>> wrote:
>> > thanks - I'll be evaluating the demo these weekends, sorry, won't be able
>> to
>> > get to it earlier, but I'll do it in time so that we can prepare a mid
>> term
>> > evaluation report
>>
>> I hope you will enjoy.
>>
>> > I'd like to clarify the role of StaticContentServlet.java. I'm a bit
>> > concerned about users having to add this servlet entry in addition to two
>> > other ones (one for handling AtomPullServer and the other one for
>> > BootstrapStorage).
>> >
>> > What exactly does this servlet do that can not be done by enhancing say
>> > BootstrapStorage or (CXF) AbstractHttpServlet ?
>>
>> Yes, you've got absolute right. Of course, for end user, easier
>> configuration is better.
>>
>> I've done some mistake - I wrote StaticContentServlet by myself
>> instead of extending AbstractHttpServlet or DefaultServlet. I will
>> fixed it. However I'd like to join StaticContentServlet and
>> BootstrapStorage class into one single class (as you mentioned).
>>
>> This all happened because I've been working in a hurry. Now I'd like
>> to make some major clean up (for example join StaticContentServlet and
>> BootstrapStorage into one class and extending AbstractHttpServlet
>> rather than write code by myself or remove useless dependency - I mean
>> XStream) before mid-term mid-term evaluations deadline.
>>
>> Once again sorry for that mess - at the moment I'm working hard to
>> improve code quality.
>>
>> >
>> > cheers, Sergey
>> >
>> > On Tue, Jul 6, 2010 at 3:00 PM, Tomasz Oponowicz <
>> tomasz.oponowicz@gmail.com
>> >> wrote:
>> >
>> >> Hi,t
>> >> I've just committed basic version of logbrowser [1] (described by my
>> >> proposal).
>> >>
>> >> I also created sample application [2] for easy testing all parts of
>> >> the project.
>> >> However it's more or less what user have to do to integrate logbrowser
>> >> into his application.
>> >> Detailed description, how to run sample application, is here [3].
>> >>
>> >> I'm staring work on [4] and [5] (i.e. refactoring, clean up and
>> >> writing documentation)
>> >> and I'm going to finished those tasks before mid-term evaluations
>> >> deadline (i.e. 16.07.2010).
>> >>
>> >> Next I will add optional features [6].
>> >>
>> >> [1] http://svn.apache.org/viewvc?rev=960895&view=rev
>> >> [2] ./distribution/src/main/release/samples/logbrowser
>> >> [3] ./distribution/src/main/release/samples/logbrowser/README.txt
>> >> [4] https://issues.apache.org/jira/browse/CXF-2881
>> >> [5] https://issues.apache.org/jira/browse/CXF-2882
>> >> [6] https://issues.apache.org/jira/browse/CXF-2883
>> >>
>> >> On Mon, Jun 21, 2010 at 6:55 PM, Tomasz Oponowicz
>> >> <tomasz.oponowicz@gmail.com> wrote:
>> >> > Hi,
>> >> >
>> >> > I've just commit changes at repository [1]. Changes from last version:
>> >> >
>> >> > - Created Maven project structure;
>> >> > - Configured 'maven-gwt-plugin' for automatic execute GWT compiler
(to
>> >> > convert Java to JavaScript);
>> >> > - Added draft implementation of 'BootstrapStorage';
>> >> > - Create sample configuration of 'AtomPullServer';
>> >> > - Added support for HTTP basic authentication (temporary solution,
at
>> >> > the end I will replace it with UsernameToken);
>> >> > - Added support for JSON web service (client-side - in GWT);
>> >> > - Added support for XML web service (client-side - in GWT);
>> >> > - Removed GWT-RPC mock web services;
>> >> >
>> >> > Additionally you can easily run application by downloading source code
>> >> and type:
>> >> >
>> >> > $ mvn gwt:run
>> >> >
>> >> > Authentication details:
>> >> >
>> >> > username: "admin"
>> >> > password:"admin";
>> >> >
>> >> > It'll start in debug mode - it's much more slower than production
>> mode.
>> >> >
>> >> > At the moment I'm working to move all source code to Apache CXF
>> sandbox.
>> >> >
>> >> > [1] svn checkout http://cxf-log-browser.googlecode.com/svn/trunk/
>> >> > cxf-log-browser-read-only
>> >> >
>> >> > On Fri, Jun 18, 2010 at 3:29 PM, Daniel Kulp <dkulp@apache.org>
>> wrote:
>> >> >> On Thursday 17 June 2010 5:15:24 pm Tomasz Oponowicz wrote:
>> >> >>> Hi,
>> >> >>>
>> >> >>> On Thu, Jun 17, 2010 at 7:43 PM, Daniel Kulp <dkulp@apache.org>
>> wrote:
>> >> >>> > On Thursday 17 June 2010 1:31:30 pm Sergey Beryozkin wrote:
>> >> >>> >> Just for the record, Tomasz's project is also going
well....
>> >> >>> >
>> >> >>> >Well, for the record, I haven't seen anything on the dev
list about
>> it
>> >> >>> >yet. Thus, from CXF/Apache's point of view, it's not going
well.
>> >> >>> >Definitely encourage him to get involved here and not send
you
>> >> anything
>> >> >>> >off list.
>> >> >>>
>> >> >>> I have created project site and blog, few weeks ago.
>> >> >>> Sorry that I didn't announce about it at mailing list.
>> >> >>>
>> >> >>> I will store documentation and source code on Google Code Project
>> [1].
>> >> >>> Additionally I will announce important events and changes through
>> blog
>> >> >>> [2], which I created for this occasion.
>> >> >>
>> >> >> One of the aspects of GSoC, and in particular Apache's involvement
in
>> >> it, that
>> >> >> is often missed is teaching the students how the community works.
>> It's
>> >> not
>> >> >> supposed to be "just do this coding project".    For Apache,
it's
>> also
>> >> getting
>> >> >> the student involved in the community and keeping the entire
>> community
>> >> up to
>> >> >> date with what is happening.
>> >> >>
>> >> >> I would strongly suggest that any "import events" and any questions
>> and
>> >> stuff
>> >> >> get posted here on the dev list.   You can add it to the blog
as
>> well,
>> >> but
>> >> >> make sure they appear on the dev list.
>> >> >>
>> >> >> There is a saying in Apache:  "if it didn't happen on the dev
list,
>> it
>> >> didn't
>> >> >> happen".
>> >> >>
>> >> >> Dan
>> >> >>
>> >> >>
>> >> >>
>> >> >>
>> >> >>> Currently repository [1] contain draft of all features (on
my
>> roadmap
>> >> >>> it's milestone 1).
>> >> >>> There is also "Software Requirements Specification" document
>> >> >>> and  "Software Design Description" document. At the moment
I'm
>> working
>> >> to
>> >> >>> fix and refactor source code.
>> >> >>>
>> >> >>> When I finish I will move all source code to CXF sandbox.
>> >> >>>
>> >> >>> Once again sorry for my misunderstanding.
>> >> >>>
>> >> >>> [1] http://code.google.com/p/cxf-log-browser/
>> >> >>> [2] http://cxf-log-browser.blogspot.com/
>> >> >>>
>> >> >>> Best regards,
>> >> >>> Tomasz Oponowicz
>> >> >>
>> >> >> --
>> >> >> Daniel Kulp
>> >> >> dkulp@apache.org
>> >> >> http://dankulp.com/blog
>> >> >>
>> >> >
>> >> >
>> >> >
>> >> > --
>> >> > Pozdrawiam,
>> >> > Tomasz Oponowicz
>> >> >
>> >>
>> >>
>> >>
>> >> --
>> >> Best regards,
>> >> Tomasz Oponowicz
>> >>
>> >
>>
>>
>>
>> --
>> Best regards,
>> Tomasz Oponowicz
>>
>



-- 
Best regards,
Tomasz Oponowicz

Mime
View raw message