Re: Multi-site provisioning with a shared SPIP kernel

Re: Multi-site provisioning with a shared SPIP kernel

Hi,

multi-site has been particularly really improved recently : it's now
even included in the installation process.
A plugin helps that : you can read it's documentation here :
http://www.spip-contrib.net/La-mutualisation-facile
But it's recent and still requires the last development version.
If possible, I suggest you to upgrade to this version. If not, you
should post your question to the mailing list spip@... : even if
it's the french list, there will be certainly more answers (in french
or english) than on this list (and personally I lack experience to
answer correctly to your request)

.Gilles
---
On 11/4/07, Barnie <spip@...> wrote:
>
> Hi
>
> After some months of inactivity I've to come back to my problem of
> multi-domain or multi-site provisioning with a shared SPIP  kernel.
>
> My problem is described in the "Original message" below. The answer of Mark
> Baber encouraged my understanding of  the idea behind all that rewrite fuss.
> However, due to my ignorance it did not solve my problem.
>
> After completing the procedure of
> http://www.spip.net/en_article3581.html I simply don"t know
> how to get access the sites or domains.
>
> Barnie
>
>
> On 2007 Mar 3, at 18:16, Mark Baber wrote:
>
> If the dependent sites are at the same URL, then the .htaccess or Apache
> configuration file is used to rewrite the incoming URLs so that they go look
> at spip for their source material. I think that's covered in the article.
>
> If the dependent sites are NOT at the same URL, then redirecting those
> sites to your new SPIP root site, and add new rewrite rules to that same
> .htaccess so it will recognise them and send them to the same SPIP kernel as
> above.
>
> Does that relate to your question ? I'm not sure I understood the question,
> actually.
>
> Rewrite rules aren't basic, for sure, but there's lots of material on the
> www about it, and it's mostly a matter of copy/paste with the example URL
> rewrites. There is an extensive document in SPIP (or maybe SPIP CONTRIB)
> about rewrites .. I haven't gotten around to translating that yet.
>
> Let me know if there's still confusion... I wouldn't be surprised, I found
> it quite confusing too to start with, and am still not sure.
>
>
> ----- Original message -----
>
> I am trying to establish "Multi-site provisioning with a shared SPIP
> kernel" accordingly to
> http://www.spip.net/en_article3581.html.
>
>  What still puzzles me is
> how to tell the dependent sites that they are from now on
> administered by spip. E.g., I have the following structure -
>
> www.master.com/spip/  with the spip kernel and all provisions made
> for existing -
>
> www.master.com
> www. dependent-1.com
> www. dependent-2.com
> www. dependent-3.com
>
> The article m.a. does not contain any clue what to put into the
> dependent domains. I am not talking about a redirect from, say,  www.
> dependent-1.com to www.master.com/spip/dependent-1.com.
> Redirecting
> any site to a paricular spip rubrique is no big deal.
>
> _______________________________________________
> spip-en@... -
> http://listes.rezo.net/mailman/listinfo/spip-en
>
>

The proper use of #TEXTE

Hi all,

After my last problem prompted so fast (and helpful) a response, I  
thought I'd ask here about the proper use of the #TEXTE.

I'm working on another SPIP based site and have seen some behaviour  
that surprised me (and we all know about the principle of least  
surprise): the type of output I get from #TEXTE seems to depend on  
the content. If the article contains more than one block level  
element (paragraph, heading, etc.) then #TEXTE gives me back some  
HTML that fits nicely into my template. If, however, the article  
contains a single paragraph of text, #TEXTE gives me some bare text  
which I need to wrap with <p>'s. I can use the `paragrapher` filter  
to make it work properly but I find the need to do so unusual (given  
my experiences with markdown and textile).

Now the question: can someone explain the reason for this behaviour  
(or if it is just a bug)?

Cheers,

Thomas Sutton

Re: The proper use of #TEXTE

Hi,

this is not a bug, but I recognize that it's quite disturbing :
The basic idea is that we don't know wich usage will be used with some
text fields like #PS, etc..
Sometime the use of <p> would break templates where #PS must contain
one line only.

The 'paragrapher' filter enables to force the use of <p> even if it's
a one paragraphe-only content.
Historically, this behaviour already existed before  this filter was
created to solve situations like yours. When the problem was
discussed, it was decided not to break the existing templates. That's
why it was decided not to change the default treatment.

.Gilles
---
On Nov 5, 2007 3:49 AM, Thomas Sutton <thomas@...> wrote:
> Hi all,
>
> After my last problem prompted so fast (and helpful) a response, I
> thought I'd ask here about the proper use of the #TEXTE.
>
> I'm working on another SPIP based site and have seen some behaviour
> that surprised me (and we all know about the principle of least
> surprise): the type of output I get from #TEXTE seems to depend on
> the content. If the article contains more than one block level
> element (paragraph, heading, etc.) then #TEXTE gives me back some
> HTML that fits nicely into my template. If, however, the article
> contains a single paragraph of text, #TEXTE gives me some bare text
> which I need to wrap with <p>'s. I can use the `paragrapher` filter
> to make it work properly but I find the need to do so unusual (given
> my experiences with markdown and textile).
>
> Now the question: can someone explain the reason for this behaviour
> (or if it is just a bug)?
>
> Cheers,
>
> Thomas Sutton
> _______________________________________________
> spip-en@... - http://listes.rezo.net/mailman/listinfo/spip-en
>

Re: The proper use of #TEXTE

Hi,

personally, I never understood that behaviour, but as Gilles says, it 
stayed in SPIP to keep a backward compatibility with older templates and 
sites.

If you know all the #TEXTE tags should went through the |paragrapher 
filter, then you can add the line:
$table_des_traitements['TEXTE'][] = "paragrapher(%s)";
in your mes_options.php file.

This will apply the filter automatically to all #TEXTE tags. Just beware 
that this will apply to all loops with a #TEXTE tag, not only the 
ARTICLES one.

Pierre

PS: Say Hi to Niek ;)

Thomas Sutton wrote:
> Hi all,
> 
> After my last problem prompted so fast (and helpful) a response, I  
> thought I'd ask here about the proper use of the #TEXTE.
> 
> I'm working on another SPIP based site and have seen some behaviour  
> that surprised me (and we all know about the principle of least  
> surprise): the type of output I get from #TEXTE seems to depend on  
> the content. If the article contains more than one block level  
> element (paragraph, heading, etc.) then #TEXTE gives me back some  
> HTML that fits nicely into my template. If, however, the article  
> contains a single paragraph of text, #TEXTE gives me some bare text  
> which I need to wrap with <p>'s. I can use the `paragrapher` filter  
> to make it work properly but I find the need to do so unusual (given  
> my experiences with markdown and textile).
> 
> Now the question: can someone explain the reason for this behaviour  
> (or if it is just a bug)?
> 
> Cheers,
> 
> Thomas Sutton

Re: The proper use of #TEXTE

What can you move manually from tmp/cache

Well just wondering what can you move manually from tmp/cache over ftp..
My cache is getting big ( +10MB) in a matter of minuttes..
and especially round midnite.. the 00.00 files sems to give me  
problems..
If I dont delete several times a day, my cache goes up to 124MB ( today)
And that is a bit to much ??
Tom

Re: What can you move manually from tmp/cache

Hi,

which version of Spip are you using ?
A typo in the code was responsible of such problem :

It was corrected with the commit [9430] :
http://trac.rezo.net/trac/spip/changeset/9430
(it's between 1.92b and 1.92c)

So check that you have the correct line :
$t = spip_fetch_array($s);
in ecrire/inc/invalideur.php (two lines after the comment "il semble y
avoir une desynchro ici")

Of course such correction isn't guaranteed to work : if possible you
should upgrade spip at least to the latest official release,
spip1.9.2c
(which corresponds to the release 10268)

But if you are already working with this version, the problem may be
somewhere else..

.Gilles
---

On Nov 7, 2007 2:05 AM, tom@... <tom@...> wrote:
> Well just wondering what can you move manually from tmp/cache over ftp..
> My cache is getting big ( +10MB) in a matter of minuttes..
> and especially round midnite.. the 00.00 files sems to give me
> problems..
> If I dont delete several times a day, my cache goes up to 124MB ( today)
> And that is a bit to much ??
> Tom
>
>
> _______________________________________________
> spip-en@... - http://listes.rezo.net/mailman/listinfo/spip-en
>
>

Re: What can you move manually from tmp/cache

Hello,

you can try to put a limit on the cache size, just add a line:

$quota_cache=10;

in your mes_options.php file. That should limit the cache to 10Mb.

Also check that you are not doing weird inclusions in your templates 
that would multiply the number of cached files.
There is one cache file per template per possible parameter value, so if 
  you have too many liberties in your inclusion parameters, it can 
exploder your cache.

Pierre

tom@... wrote:
> Well just wondering what can you move manually from tmp/cache over ftp..
> My cache is getting big ( +10MB) in a matter of minuttes..
> and especially round midnite.. the 00.00 files sems to give me problems..
> If I dont delete several times a day, my cache goes up to 124MB ( today)
> And that is a bit to much ??
> Tom
> 

Re: What can you move manually from tmp/cache

On 11/7/07, tom@... <tom@...> wrote:
> Tnx I will go for the 1.92c
> There is no number in my spip ( just says 192b ( no number)
> Will try the newest one
>

spip192b corresponds to rev 9379
look at changelog here :  http://trac.rezo.net/trac/spip

.Gilles

> thanx
> Tom
>
> Den 07/11/2007 kl. 2.49 skrev Gilles Vincent:
>
> > Hi,
> >
> > which version of Spip are you using ?
> > A typo in the code was responsible of such problem :
> >
> > It was corrected with the commit [9430] :
> > http://trac.rezo.net/trac/spip/changeset/9430
> > (it's between 1.92b and 1.92c)
> >
> > So check that you have the correct line :
> > $t = spip_fetch_array($s);
> > in ecrire/inc/invalideur.php (two lines after the comment "il semble y
> > avoir une desynchro ici")
> >
> > Of course such correction isn't guaranteed to work : if possible you
> > should upgrade spip at least to the latest official release,
> > spip1.9.2c
> > (which corresponds to the release 10268)
> >
> > But if you are already working with this version, the problem may be
> > somewhere else..
> >
> >
> > .Gilles
> > ---
> >
> > On Nov 7, 2007 2:05 AM, tom@... <tom@...> wrote:
> >> Well just wondering what can you move manually from tmp/cache over
> >> ftp..
> >> My cache is getting big ( +10MB) in a matter of minuttes..
> >> and especially round midnite.. the 00.00 files sems to give me
> >> problems..
> >> If I dont delete several times a day, my cache goes up to 124MB
> >> ( today)
> >> And that is a bit to much ??
> >> Tom
> >>
> >>
> >> _______________________________________________
> >> spip-en@... - http://listes.rezo.net/mailman/listinfo/spip-en
> >>
> >>
> > _______________________________________________
> > spip-en@... - http://listes.rezo.net/mailman/listinfo/spip-en
> >
>
>
>