Hi Christian and Andrea,

On Oct 20, 2010, at 12:05 PM, Christian Lohmaier wrote:
> On Wed, Oct 20, 2010 at 1:26 PM, Andrea Pescetti
> <[email protected]> wrote:
>> Marc Paré wrote:
>> 
>> A proper discussion would not fit the deadlines and tools the Document
>> Foundation has now: if a decision has to be taken now, it will miss the
>> bigger scope. If the Document Foundation must choose a tool that will be
>> flexible enough to rebuild all the current OpenOffice.org infrastructure
>> on it
> 
> No. I strongly disagree. Not one tool that can do all.

I'm with Andrea here--I think reasonable consolidation of web assets will make 
for simpler administration (less effort over time), especially because upgrades 
will only have to be done to one core system. I also think it provides a better 
user experience via consistent design, domains and user accounts.

>> (e.g.: OOo site; Extensions; TCM; QATrack),
> 
> And even that: Duplicating the extensions site would be a waste of
> time and efforts. There are already two repositories, the OOo one, the
> FSF one. It would be bad if there would be another one, just for the
> sake of having it.
> LO should be compatible to OOo in that regard, Extensions should run
> on OOo, LO, other derivates, thus a dedicated site is a nogo.

Extensions that work on OOo and LibO would be great, but I can't imagine that 
will be the case over the longterm. I think it would be very risky to rely on 
the existing OOo Extension repository to work for LibO for very long... but on 
the other hand, one repository with a parameter for each extension to indicate 
whether it works on OOo, LibO, or both, might be pretty useful.

>> then Drupal is that
>> tool; I can't imagine how to rebuild the Extensions site and all
>> processing in Silverstripe, for one.
> 
> It would involve creating an appropriate module. But again, I'd not
> see that as a good idea.

If we're worried about finding enough admins for our CMS, this is just 10x more 
complex and we should be even more cautious to avoid the situation.

> The extensions site has a different target user group, and a different
> contributers group, so I don't see any reason to try to cover it with
> the same tool that is used for the website.
> 
>> 2) Moving to a database-based CMS can imply loss of traceability of
>> changes. The current CVS infrastructure, as bad as it can be, allows to
>> see a full log of changes very easily. Drupal has a killer feature here:
>> site settings can be exported to PHP code, shared among a distributed
>> development team through any revision control system (SVN, git,
>> whatever) and applied in a safe way to the running site. It is very
>> important that we are able to answer the question "Who enabled
>> this permission, when and why?"
> 
> Why is that a killer feature of drupal, why do you assume you can't
> put silverstripe's configuration under version control?
> What makes you think silverstripe wouldn't have a change history for the 
> pages?

Most CMSs store configuration information in their databases and do not keep a 
record of config changes made in the web admin (for things outside of the 
content themselves). This is a killer feature because Drupal offers a way to 
track those changes as well. It's not the pages we're talking about here at 
all. If SilverStripe has something like this, I have not seen it.

-Ben

Benjamin Horst
[email protected]
646-464-2314 (Eastern)
www.solidoffice.com


--
E-mail to [email protected] for instructions on how to unsubscribe
List archives are available at http://www.libreoffice.org/lists/website/
All messages you send to this list will be publicly archived and cannot be 
deleted

Reply via email to