commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Rall <>
Subject Re: [Submission] commons.util.http enhancements
Date Wed, 21 Nov 2001 21:31:29 GMT
Lavandowska <> writes:

> --- Daniel Rall <> wrote:
> > I finally had some time to look at your patch, Lance.  The build.xml
>> file didn't have the JSDK JAR in the classpath at compile time,
>> ExtendedBrowserDetector includes all of the content of
>> BrowserDetector (which it sub-classes) and didn't compile.  
> I'm still largely mystified by the build.xml stuff, and probably have
> the JSDK in my system classpath because some other project wouldn't
> compile without it there (masking this issue).  I'm curious as to any
> other reason it wouldn't compile.  It sounded like there was a time
> issue before, so I hadn't taken the time to truly subclass
> BrowserDetector.  When I get a chance I'll rectify that.

I've done so in the EBD code (along with fixing the compile errors and
correcting the build.xml file) -- I can my modified version over, if
you like.

>> It might make more sense
>> to have a single "flags" int member in ExtendedBrowserDetector and do
>> bitwise ORs instead of having so many boolean flags.  
> I was just following the "design philosophy" of the original
> BrowserDetector.  I think the idea was that the end user of the
> BrowserDetector bean could just fetch the relevant boolean.  Having
> fewer members would be good, however.

Since there are accessor methods, using a "flags" member might be the
way to go, then.

>> I didn't look at
>> RequestUtils because I tired of fixing problems.
> This is the one I'm really interested in contributing.  If there are
> problems with it (I'm guessing the JSDK issue), please just let me know
> directly (off-line) rather than wearing yourself out fixing them - I
> need to get more familiar with the issues of Apache contribution (yes,
> I've read all the docs & examined code from various projects).

I'll look over the versino you sent me (hopefully over this weekend),
and get back to you.

To unsubscribe, e-mail:   <>
For additional commands, e-mail: <>

View raw message