You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by Jon Svede <js...@yahoo.com> on 2013/10/11 19:14:10 UTC
using the webDAV component
Hi,
Our Jackrabbit app is using a repo that runs within our application. We load our repository.xml in our code to create access. Lately, I've been looking for some tools to help explore our content and all of them talk about being able to use webDAV or RMI. The docs talk about RMI not being very performant, so I guess I'd like to set up the webDAV.
How do I setup a webDAV repo with the one I already have? Can I deploy the webDAV war file and point it to our repo? There don't seem to be any docs I can use (but if there are please forward the links). The docs that exist on the Jackrabbit site are somewhat vague (at least to me).
Thanks,
Jon
Re: using the webDAV component
Posted by Julian Reschke <ju...@gmx.de>.
On 2013-10-16 12:15, Christoph Läubrich wrote:
> Hi Julian,
>
> I just can tell my experience. I'm using Apache SVN Webdav module and
> that is working with linux, windows and "standard" Webbrowsers. So I can
> assume tha there ARE ways of handling this compfortable and reliable for
> users that don't care if they are using the "right" client. I evaluated
> JR exactly for that purpose: Giving access to a Datastore via Webbrowser
> and/or system integrated methods without need for special software.
Understood.
But there are problems Jackrabbit can't fix as it doesn't control them.
For instance, when the client makes the incorrect assumption that the
server's root must be WebDAV-enabled.
> > And when you say "Webbrowser" -- what does this have to do with WebDAV?
>
> WebDav is an extension to HTTP. Webbrowsers can talsk http. And at least
> the Apache SVN Webdav module can be used with simple browsers to
> a) Browse a directory tree
> b) Download (the latest revision) of a file
That's not WebDAV; that's simply an HTML based UI.
> And with JR Webdav this is also possible to browse the repro (of course
> as with the SVN webdav it is not possible to use ALL features through
> simple http GET), even if I don't know that this is REQUIRED and/or
> COVERED by the WebDav spec. But IMO thats something most ppl assume.
It has nothing to do with WebDAV.
> As the original author of the discussion I'd love to see an working
> WebDav example, but the common answer is: Try Sling, and even with that
> it (as described) din't work that smooth that I would recommend this
> over RMI (what worked for browsing the repro out-of-the-box).
I successfully used Jackrabbit's WebDAV last week with MacOS (builtin
client) and Cyberduck (standalone client).
> Again don't take this as an offence or statement that it is not working
> at all! Just some impressions I got when I tried this latetely.
>
> BTW: In the end I'm using this project
> (http://webdav-servlet.sourceforge.net/) for presenting relevant content
> as WebDav and it worked at least with the mentioned defaultclients of
> Windows,Linux,Webbrowser.
>
> regards Christoph
> ...
If that's "just" a servlet, and it works where Jackrabbit does not, then
we should obtain traces and find out what's different.
Best regards, Julian
Re: using the webDAV component
Posted by Christoph Läubrich <la...@googlemail.com>.
Hi Julian,
I just can tell my experience. I'm using Apache SVN Webdav module and
that is working with linux, windows and "standard" Webbrowsers. So I can
assume tha there ARE ways of handling this compfortable and reliable for
users that don't care if they are using the "right" client. I evaluated
JR exactly for that purpose: Giving access to a Datastore via Webbrowser
and/or system integrated methods without need for special software.
> And when you say "Webbrowser" -- what does this have to do with WebDAV?
WebDav is an extension to HTTP. Webbrowsers can talsk http. And at least
the Apache SVN Webdav module can be used with simple browsers to
a) Browse a directory tree
b) Download (the latest revision) of a file
And with JR Webdav this is also possible to browse the repro (of course
as with the SVN webdav it is not possible to use ALL features through
simple http GET), even if I don't know that this is REQUIRED and/or
COVERED by the WebDav spec. But IMO thats something most ppl assume.
As the original author of the discussion I'd love to see an working
WebDav example, but the common answer is: Try Sling, and even with that
it (as described) din't work that smooth that I would recommend this
over RMI (what worked for browsing the repro out-of-the-box).
Again don't take this as an offence or statement that it is not working
at all! Just some impressions I got when I tried this latetely.
BTW: In the end I'm using this project
(http://webdav-servlet.sourceforge.net/) for presenting relevant content
as WebDav and it worked at least with the mentioned defaultclients of
Windows,Linux,Webbrowser.
regards Christoph
Am 16.10.2013 09:19, schrieb Julian Reschke:
> On 2013-10-15 15:50, Christoph Läubrich wrote:
>> > "stable" and "working with every client" are two very different
>> things.
>> You are right, sorry for using inadequate wording :-)
>>
>> > Out of curiosity: what was the problem that you encountered?
>>
>> I wasn't able to mount the WebDav neither under Linux nor Windows
>> releiable (while "other" WebDavs like Apache worked) sometimes it worked
>> for one of them (I think I finally got it working under Windows using
>> Sling, but not JackRabbit Standalone) , in a Webbrowser it worked but
>> all file nodes where not accessible as files (there was a posting in the
>> newsgroup as well).
>
> Well, it can be a problem of the client; I recall that the Windows
> WebDAV driver makes incorrect assumptions about servers, and there's
> little Jackrabbit can do about that
> (<http://www.greenbytes.de/tech/webdav/webdav-redirector-list.html#issue-server-discovery>).
>
>
> And when you say "Webbrowser" -- what does this have to do with WebDAV?
>
> Best regards, Julian
Re: using the webDAV component
Posted by Lukas Kahwe Smith <sm...@pooteeweet.org>.
> On 16.10.2013, at 09:19, Julian Reschke <ju...@gmx.de> wrote:
>
>> On 2013-10-15 15:50, Christoph Läubrich wrote:
>> > "stable" and "working with every client" are two very different things.
>> You are right, sorry for using inadequate wording :-)
>>
>> > Out of curiosity: what was the problem that you encountered?
>>
>> I wasn't able to mount the WebDav neither under Linux nor Windows
>> releiable (while "other" WebDavs like Apache worked) sometimes it worked
>> for one of them (I think I finally got it working under Windows using
>> Sling, but not JackRabbit Standalone) , in a Webbrowser it worked but
>> all file nodes where not accessible as files (there was a posting in the
>> newsgroup as well).
>
> Well, it can be a problem of the client; I recall that the Windows WebDAV driver makes incorrect assumptions about servers, and there's little Jackrabbit can do about that (<http://www.greenbytes.de/tech/webdav/webdav-redirector-list.html#issue-server-discovery>).
>
> And when you say "Webbrowser" -- what does this have to do with WebDAV?
>
> Best regards, Julian
indeed the reality of webdav is that in order to be truly you need to spend a lot of time adding special code to deal with various broken client implementations.
regards,
Lukas
Re: using the webDAV component
Posted by Julian Reschke <ju...@gmx.de>.
On 2013-10-15 15:50, Christoph Läubrich wrote:
> > "stable" and "working with every client" are two very different things.
> You are right, sorry for using inadequate wording :-)
>
> > Out of curiosity: what was the problem that you encountered?
>
> I wasn't able to mount the WebDav neither under Linux nor Windows
> releiable (while "other" WebDavs like Apache worked) sometimes it worked
> for one of them (I think I finally got it working under Windows using
> Sling, but not JackRabbit Standalone) , in a Webbrowser it worked but
> all file nodes where not accessible as files (there was a posting in the
> newsgroup as well).
Well, it can be a problem of the client; I recall that the Windows
WebDAV driver makes incorrect assumptions about servers, and there's
little Jackrabbit can do about that
(<http://www.greenbytes.de/tech/webdav/webdav-redirector-list.html#issue-server-discovery>).
And when you say "Webbrowser" -- what does this have to do with WebDAV?
Best regards, Julian
Re: using the webDAV component
Posted by Santiago Gala <sa...@gmail.com>.
On Tue, Oct 15, 2013 at 3:50 PM, Christoph Läubrich
<la...@googlemail.com>wrote:
> > "stable" and "working with every client" are two very different things.
> You are right, sorry for using inadequate wording :-)
>
>
> > Out of curiosity: what was the problem that you encountered?
>
> I wasn't able to mount the WebDav neither under Linux nor Windows
> releiable (while "other" WebDavs like Apache worked) sometimes it worked
> for one of them (I think I finally got it working under Windows using
> Sling, but not JackRabbit Standalone) , in a Webbrowser it worked but all
> file nodes where not accessible as files (there was a posting in the
> newsgroup as well).
>
Under linux I have experienced serious problems with the gvfs gnome
component. I remember mounting jackrabbit some years ago, so either a
ubuntu or jackrabbit regression is at work. I was able to mount jackrabbit
using davfs2, but it downloads the whole file on any access (GET /file
without any range request is what I see on the wire), placing a big strain
on the network and the client and tremendous latency when one manipulates
a number of 1GB files, which is my use case.
Regards
Santiago
>
> Am 14.10.2013 16:36, schrieb Julian Reschke:
>
> On 2013-10-12 13:03, Christoph Läubrich wrote:
>>
>>> Last time I tried using WebDav it was not very stable (e.g. not working
>>> with every client) and it was not that easy to set it up "from scratch"
>>>
>>
>> "stable" and "working with every client" are two very different things.
>>
>> Out of curiosity: what was the problem that you encountered?
>>
>> with an existing repro/installation.
>>> So I would recommend you first try the RMI aproach and the decide if the
>>> performance is enough for you (I think the objectexplorer uses RMI and
>>> that worked out-of-the-box).
>>>
>>> Am 11.10.2013 19:14, schrieb Jon Svede:
>>>
>>>> Hi,
>>>>
>>>> Our Jackrabbit app is using a repo that runs within our application.
>>>> We load our repository.xml in our code to create access. Lately, I've
>>>> been looking for some tools to help explore our content and all of
>>>> them talk about being able to use webDAV or RMI. The docs talk about
>>>> RMI not being very performant, so I guess I'd like to set up the webDAV.
>>>>
>>>> How do I setup a webDAV repo with the one I already have? Can I deploy
>>>> the webDAV war file and point it to our repo? There don't seem to be
>>>> any docs I can use (but if there are please forward the links). The
>>>> docs that exist on the Jackrabbit site are somewhat vague (at least to
>>>> me).
>>>>
>>>> Thanks,
>>>>
>>>> Jon
>>>>
>>>
>>>
>>>
>>
>
Re: using the webDAV component
Posted by Christoph Läubrich <la...@googlemail.com>.
> "stable" and "working with every client" are two very different things.
You are right, sorry for using inadequate wording :-)
> Out of curiosity: what was the problem that you encountered?
I wasn't able to mount the WebDav neither under Linux nor Windows
releiable (while "other" WebDavs like Apache worked) sometimes it worked
for one of them (I think I finally got it working under Windows using
Sling, but not JackRabbit Standalone) , in a Webbrowser it worked but
all file nodes where not accessible as files (there was a posting in the
newsgroup as well).
Am 14.10.2013 16:36, schrieb Julian Reschke:
> On 2013-10-12 13:03, Christoph Läubrich wrote:
>> Last time I tried using WebDav it was not very stable (e.g. not working
>> with every client) and it was not that easy to set it up "from scratch"
>
> "stable" and "working with every client" are two very different things.
>
> Out of curiosity: what was the problem that you encountered?
>
>> with an existing repro/installation.
>> So I would recommend you first try the RMI aproach and the decide if the
>> performance is enough for you (I think the objectexplorer uses RMI and
>> that worked out-of-the-box).
>>
>> Am 11.10.2013 19:14, schrieb Jon Svede:
>>> Hi,
>>>
>>> Our Jackrabbit app is using a repo that runs within our application.
>>> We load our repository.xml in our code to create access. Lately, I've
>>> been looking for some tools to help explore our content and all of
>>> them talk about being able to use webDAV or RMI. The docs talk about
>>> RMI not being very performant, so I guess I'd like to set up the
>>> webDAV.
>>>
>>> How do I setup a webDAV repo with the one I already have? Can I deploy
>>> the webDAV war file and point it to our repo? There don't seem to be
>>> any docs I can use (but if there are please forward the links). The
>>> docs that exist on the Jackrabbit site are somewhat vague (at least to
>>> me).
>>>
>>> Thanks,
>>>
>>> Jon
>>
>>
>
Re: using the webDAV component
Posted by Julian Reschke <ju...@gmx.de>.
On 2013-10-12 13:03, Christoph Läubrich wrote:
> Last time I tried using WebDav it was not very stable (e.g. not working
> with every client) and it was not that easy to set it up "from scratch"
"stable" and "working with every client" are two very different things.
Out of curiosity: what was the problem that you encountered?
> with an existing repro/installation.
> So I would recommend you first try the RMI aproach and the decide if the
> performance is enough for you (I think the objectexplorer uses RMI and
> that worked out-of-the-box).
>
> Am 11.10.2013 19:14, schrieb Jon Svede:
>> Hi,
>>
>> Our Jackrabbit app is using a repo that runs within our application.
>> We load our repository.xml in our code to create access. Lately, I've
>> been looking for some tools to help explore our content and all of
>> them talk about being able to use webDAV or RMI. The docs talk about
>> RMI not being very performant, so I guess I'd like to set up the webDAV.
>>
>> How do I setup a webDAV repo with the one I already have? Can I deploy
>> the webDAV war file and point it to our repo? There don't seem to be
>> any docs I can use (but if there are please forward the links). The
>> docs that exist on the Jackrabbit site are somewhat vague (at least to
>> me).
>>
>> Thanks,
>>
>> Jon
>
>
Re: using the webDAV component
Posted by Christoph Läubrich <la...@googlemail.com>.
Last time I tried using WebDav it was not very stable (e.g. not working
with every client) and it was not that easy to set it up "from scratch"
with an existing repro/installation.
So I would recommend you first try the RMI aproach and the decide if the
performance is enough for you (I think the objectexplorer uses RMI and
that worked out-of-the-box).
Am 11.10.2013 19:14, schrieb Jon Svede:
> Hi,
>
> Our Jackrabbit app is using a repo that runs within our application. We load our repository.xml in our code to create access. Lately, I've been looking for some tools to help explore our content and all of them talk about being able to use webDAV or RMI. The docs talk about RMI not being very performant, so I guess I'd like to set up the webDAV.
>
> How do I setup a webDAV repo with the one I already have? Can I deploy the webDAV war file and point it to our repo? There don't seem to be any docs I can use (but if there are please forward the links). The docs that exist on the Jackrabbit site are somewhat vague (at least to me).
>
> Thanks,
>
> Jon
>