Documents are now stored without extension

Andreas Hartmann wrote:

> Hi Lenya devs,
>
> after countless workarounds and lots of outbursts of
fury I have now
> changed the persistence layer so that documents are
stored without
> extension:
>
>   foo/index
>   foo/index.meta
>
> instead of
>
>   foo/index.xml


I think this should be decided by the resource itself

>   foo/index.xml.meta
>
>
> There's a simple migration target in
migration-build.xml:
>
>   ./build.sh migrate-extensions -Dpublication=default
>
> but it works only for .xml extensions and it doesn't
work for
> "outsourced" content.
>
> I think it doesn't make sense to generalize this
target, it is
> more efficient if you modify the target to fit your
needs:
>
>
>   <target name="migrate-extensions">
>     <property name="publication"
value=""/>
>     <move
todir="${build.webapp}/lenya/pubs">
>       <fileset
dir="${build.webapp}/lenya/pubs/$"&
gt;
>         <include name="content/**"/>
>         <exclude
name="content/*/site.xml"/>
>         <exclude
name="content/*/sitetree.xml"/>
>       </fileset>
>       <compositemapper>
>         <mapper type="glob"
from="*.xml" to="*"/>
>         <mapper type="glob"
from="*.xml.meta" to="*.meta"/>
>       </compositemapper>
>     </move>
>   </target>
>
>
> Please complain if problems occur (this will basically
happen if
> there still are pieces of code which bypass the
repository layer).
>
> Sorry for the inconvenience,


well, can't we discuss these things before we just go
ahead?!

It's not about inconvenience but rather about *ignoring*
other people's 
opinion completely ...

I think it's very wrong that Lenya forces people to adapt a
certain path 
structure and the reason I think this is completely wrong is
because 
Lenya is not the only application on this planet handling
data. Various 
applications should be able to share data and this means the

applications need to be customizable and not the data.
In the case of Lenya this doesn't seem to be the case
anymore or do I 
completely misunderstand something?

Michi

>
> -- Andreas
>


-- 
Michael Wechner
Wyona      -   Open Source Content Management   -    Apache
Lenya
http://www.wyona.com     
                http://lenya.apache.org
michael.wechnerwyona.com                        michiapache.org
+41 44 272 91 61


------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org

Michael Wechner wrote:
> Andreas Hartmann wrote:
> 
>> Hi Lenya devs,
>>
>> after countless workarounds and lots of outbursts
of fury I have now
>> changed the persistence layer so that documents are
stored without
>> extension:
>>
>>   foo/index
>>   foo/index.meta
>>
>> instead of
>>
>>   foo/index.xml
> 
> 
> I think this should be decided by the resource itself

That is impossible without fundamental changes to the
architecture
which I would object.

The document is identified by document ID and language.
These two
bits of information are used to find the meta data. From the
meta
data, the resource type can be determined.

With the current architecture, the meta data are attached to
the document source - i.e., the document source has to be
known
to get the meta data. That's why the extension can not be
part of
the document source. There were various complex and risky
workarounds in the code to make it work with source
extensions.
The only reasonable way for the moment was to remove the
extension
from the source.

We already agreed that we want to identify documents using
UUIDs,
so the next step would be to remove the hierarchy from the
storage
structure, e.g.:

content/<area>/<uuid>_<language>


[...]

>> Sorry for the inconvenience,
> 
> 
> well, can't we discuss these things before we just go
ahead?!

I agree, I should have discussed these changes before.


> It's not about inconvenience but rather about
*ignoring* other people's 
> opinion completely ...

Yes, in this case I ignored your opinion, and I apologize
for that.
My mistake was that, in this moment, I was driven by my
ambition
and valued code higher than community.


> I think it's very wrong that Lenya forces people to
adapt a certain path 
> structure and the reason I think this is completely
wrong is because 
> Lenya is not the only application on this planet
handling data.

The preferrable way to deal with this is offer import/export
interfaces.


> Various 
> applications should be able to share data and this
means the 
> applications need to be customizable and not the data.

This means that the applications need interfaces to exchange
data, or that they share their data using a central
repository
with a well-defined interface.


> In the case of Lenya this doesn't seem to be the case
anymore or do I 
> completely misunderstand something?

Sharing data has to be done through published interfaces,
not
on the base of an internal data store. How data are
persisted
is a concern of the framework. Of course, it could offer to
implement custom persistence managers, but this would rather
allow flexibility regarding performance than to allow
sharing
data.

-- Andreas


-- 
Andreas Hartmann
Wyona Inc.  -   Open Source Content Management   -   Apache
Lenya
http://www.wyona.com     
                http://lenya.apache.org
andreas.hartmannwyona.com                     andreasapache.org


------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org

On Mon, 2006-05-22 at 23:23 +0200, Michael Wechner wrote:
> Andreas Hartmann wrote:
[...]
> > Please complain if problems occur (this will
basically happen if
> > there still are pieces of code which bypass the
repository layer).
> >
> > Sorry for the inconvenience,
> 
> 
> well, can't we discuss these things before we just go
ahead?!
> 
> It's not about inconvenience but rather about
*ignoring* other people's 
> opinion completely ...

Just wanted to mention that this issue actually has been
discussed on
this list, see e.g. [1].
Andreas proposed to remove the source extension, and there
was not much
feedback from the community. In a case like that it seems ok
to me for a
developer to go ahead and implement a proposed change.

- Josias

[1]
http://www.nabble.com/Removing+source+extensions+%28wa
s%3A
+-blog-+creating+new+entries%29-p3878696.html




------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org

Josias Thöny wrote:
> On Mon, 2006-05-22 at 23:23 +0200, Michael Wechner
wrote:
>> Andreas Hartmann wrote:
> [...]
>>> Please complain if problems occur (this will
basically happen if
>>> there still are pieces of code which bypass the
repository layer).
>>>
>>> Sorry for the inconvenience,
>>
>> well, can't we discuss these things before we just
go ahead?!
>>
>> It's not about inconvenience but rather about
*ignoring* other people's 
>> opinion completely ...
> 
> Just wanted to mention that this issue actually has
been discussed on
> this list, see e.g. [1].

Josias, thank you for this pointer.

> Andreas proposed to remove the source extension, and
there was not much
> feedback from the community. In a case like that it
seems ok to me for a
> developer to go ahead and implement a proposed change.

I agree to that (in trunk) - what do the others think?

Do we want to require a specific announcement mail before
changes
which require adapting existing content or code, with a
certain
period of delay so that others can state their objections?

-1 from me, because that would include practically all
changes at
the moment because client code could be based on any of our
code.

-- Andreas

-- 
Andreas Hartmann
Wyona Inc.  -   Open Source Content Management   -   Apache
Lenya
http://www.wyona.com     
                http://lenya.apache.org
andreas.hartmannwyona.com                     andreasapache.org


------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org

Josias Thöny wrote:
> On Mon, 2006-05-22 at 23:23 +0200, Michael Wechner
wrote:
>   
>> Andreas Hartmann wrote:
>>     
> [...]
>   
>>> Please complain if problems occur (this will
basically happen if
>>> there still are pieces of code which bypass the
repository layer).
>>>
>>> Sorry for the inconvenience,
>>>       
>> well, can't we discuss these things before we just
go ahead?!
>>
>> It's not about inconvenience but rather about
*ignoring* other people's 
>> opinion completely ...
>>     
>
> Just wanted to mention that this issue actually has
been discussed on
> this list, see e.g. [1].
>   

well, I wouldn't call it a discussion, but rather stating
an opinion. Also
it was more than  a month ago and part of a reply to another
email topic.

I try very hard to follow most emails, but don't manage to
catch everything
and I think removing the suffixes is a major thing so I
think one could 
rectify a
more recent post.

Also if the UUID is used within the sitetree and done as
Andreas 
describes it within his
email then I don't think I have a problem with it, because
the suffix 
could be a part of the UUID
(doesn't matter if this makes sense or not). But this
doesn't seem to be 
the case and I really don't understand
why this shouldn't be possible. But I will check it again,
before I 
might say something which might be just a misunderstanding.
> Andreas proposed to remove the source extension, and
there was not much
> feedback from the community. In a case like that it
seems ok to me for a
> developer to go ahead and implement a proposed change.
>   
if it's recent and marked clearly then yes, but  I don't
think that was 
the case.
(see my arguments above for this case).

Michi
> - Josias
>
> [1]
> http://www.nabble.com/Removing+source+extensions+%28wa
s%3A
> +-blog-+creating+new+entries%29-p3878696.html
>
>
>
>
>
------------------------------------------------------------
---------
> To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
> For additional commands, e-mail: dev-helplenya.apache.org
>
>
>   


-- 
Michael Wechner
Wyona      -   Open Source Content Management   -    Apache
Lenya
http://www.wyona.com     
                http://lenya.apache.org
michael.wechnerwyona.com                        michiapache.org
+41 44 272 91 61


------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org

Michael Wechner wrote:

[...]

> Also if the UUID is used within the sitetree and done
as Andreas 
> describes it within his
> email then I don't think I have a problem with it,
because the suffix 
> could be a part of the UUID

That would mean either redundant information in the UUID

   <uuid> ::= <unique-part> - <extension>

or that entirely different documents can share a
"unique part"

   12345-xml
   12345-odt

I don't see any advantages if the extension is part of the
UUID.
It is sufficient to store this information in the meta data.


> (doesn't matter if this makes sense or not).

Why doesn't it matter?

> But this doesn't seem to be 
> the case and I really don't understand
> why this shouldn't be possible.

It is probably possible, but at the moment I see only
disadvantages (see above) ...

-- Andreas



-- 
Andreas Hartmann
Wyona Inc.  -   Open Source Content Management   -   Apache
Lenya
http://www.wyona.com     
                http://lenya.apache.org
andreas.hartmannwyona.com                     andreasapache.org


------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribelenya.apache.org
For additional commands, e-mail: dev-helplenya.apache.org