You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lenya.apache.org by Andreas Hartmann <an...@apache.org> on 2007/05/22 08:17:12 UTC
Status of proxying issue for 1.4
Hi Lenya devs,
what's the current status of the proxying issue?
Which changes are necessary so we can retry the RC?
My latest info was that we postpone all code changes
and add some documentation what to achieve with the
current codebase.
Could anyone imagine to take the lead in this mission?
TIA!
-- Andreas
--
Andreas Hartmann, CTO
BeCompany GmbH
http://www.becompany.ch
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Thorsten Scherler <th...@apache.org>.
On Tue, 2007-05-22 at 23:05 +0200, Joern Nettingsmeier wrote:
...
> my points are these:
> * everything should be done just once, in one place.
agree
> * publication boilerplate code is bad.
agree
>
> fixing proxying with the current sitemap structure means patching things
> up in a number of places,
actually the transformer is a catch all links and proxy them approach.
The only thing is that it should be used in the main match **.html just
before the serializer. Nothing more to patch.
> and it means that users will have to remember
> to use a lot of boilerplate code correctly in their publications. i
> attempted to fix proxying that way some weeks ago, and concluded that it
> was not worth it.
Maybe you want to look into the transformer I guess I may have taken a
different route.
Do not get me wrong I am the first one to claim that a pub sitemap
should not be needed at all. Like you said implementing it in the last
step of the global sitemap is the best solution. BUT if one is
overriding this last step (like we do in default) one needs to add the
transformer to the end.
>
> still, if you can get proxying to work, that's great for our users and i
> really welcome it. but i still feel we need fundamental changes, which
> might make all the work you are doing now obsolete, and that would be
> wasteful.
Do not thing so. Like said first have a look in the patch and then you
may review this statement. I as well do not like the concept of reserved
url spaces and area magic at all but even if we get rid of it we can
still use the proxyTransformer. Further I am trying to help to fix this
issue to get 1.4 out of the door which is first priority and after such
a long time waiting for 1.4 we cannot allow us to release with blocker
issues.
> but you are right: a working solution for 1.4 is worth a lot. it's just
> that i was not motivated to tackle it, because an intermediate fix would
> go against the ideas i have in mind for future releases of lenya.
>
The thing is not all code that have been in lenya before is bad because
we do not understand it. I do not say this to piss somebody off, just in
the light of recent threads. Some concepts we have in lenya needs
rethinking I totally agree (see my answers in the 1.6 thread) but some
are only need to polish a bit that they shine.
That reminds me on the locationmap in forrest that have been in the code
for ages and we finally understood the concepts behind it recently.
The problem of proxy code in 1.4 is that it did not mature from 1.2
where it is originally coming from. In 1.2 there where no reserved url
spaces such as /lenya and /module that is the reason why proxying an
area may made sense. Now we need to enhance the existing code to meet
the new demands.
salu2
--
Thorsten Scherler thorsten.at.apache.org
Open Source Java consulting, training and solutions
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Joern Nettingsmeier <ne...@folkwang-hochschule.de>.
Thorsten Scherler wrote:
> On Tue, 2007-05-22 at 13:19 +0200, Jörn Nettingsmeier wrote:
>> Andreas Hartmann wrote:
>>> Hi Lenya devs,
>>>
>>> what's the current status of the proxying issue?
>>> Which changes are necessary so we can retry the RC?
>>> My latest info was that we postpone all code changes
>>> and add some documentation what to achieve with the
>>> current codebase.
>>>
>>> Could anyone imagine to take the lead in this mission?
>> will do that tomorrow - i'm travelling atm.
>> i'll try and document what can be done, but i still think we should be
>> honest and label lenya's proxy support as broken ("works sometimes under
>> not-precisely-defined optimal conditions" is hardly helpful).
>> what little proxying can be accomplished atm mostly relies on web server
>> feature, not on lenya.
>
> -1
>
> Andreas and I are trying to fix it and IMO it is not that bad at all.
> Sure took some time to have a fix but proxying after the latest patches
> is nearly back.
>
>> i'm opposed to adding quick hacks to tackle some corner cases, because
>> the fact remains that proxying is fundamentally broken and cannot be
>> fixed in a sane way.
>
> I disagree, I do not see what you state.
>
>> the one correct solution i can imagine is to have
>> one global pipeline that gets to handle every page that comes out of
>> lenya, and that's where link rewriting according to lenya-document://
>> and proxying semantics must happen. which is certainly 1.5 stuff.
>
> It is not the problem since the publication sitemap is doing it. See the
> proxyTransformer that I just added, it is missing 2 small things and
> proxying will work.
thanks for your efforts!
my comments were not intended to de-value your work (i didn't even have
a chance to test and review it yet).
my points are these:
* everything should be done just once, in one place.
* publication boilerplate code is bad.
fixing proxying with the current sitemap structure means patching things
up in a number of places, and it means that users will have to remember
to use a lot of boilerplate code correctly in their publications. i
attempted to fix proxying that way some weeks ago, and concluded that it
was not worth it.
still, if you can get proxying to work, that's great for our users and i
really welcome it. but i still feel we need fundamental changes, which
might make all the work you are doing now obsolete, and that would be
wasteful.
but you are right: a working solution for 1.4 is worth a lot. it's just
that i was not motivated to tackle it, because an intermediate fix would
go against the ideas i have in mind for future releases of lenya.
regards,
jörn
--
jörn nettingsmeier
home://germany/45128 essen/lortzingstr. 11/
http://spunk.dnsalias.org
phone://+49/201/491621
Kurt is up in Heaven now.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Thorsten Scherler <th...@juntadeandalucia.es>.
On Tue, 2007-05-22 at 13:19 +0200, Jörn Nettingsmeier wrote:
> Andreas Hartmann wrote:
> > Hi Lenya devs,
> >
> > what's the current status of the proxying issue?
> > Which changes are necessary so we can retry the RC?
> > My latest info was that we postpone all code changes
> > and add some documentation what to achieve with the
> > current codebase.
> >
> > Could anyone imagine to take the lead in this mission?
>
> will do that tomorrow - i'm travelling atm.
> i'll try and document what can be done, but i still think we should be
> honest and label lenya's proxy support as broken ("works sometimes under
> not-precisely-defined optimal conditions" is hardly helpful).
> what little proxying can be accomplished atm mostly relies on web server
> feature, not on lenya.
-1
Andreas and I are trying to fix it and IMO it is not that bad at all.
Sure took some time to have a fix but proxying after the latest patches
is nearly back.
>
> i'm opposed to adding quick hacks to tackle some corner cases, because
> the fact remains that proxying is fundamentally broken and cannot be
> fixed in a sane way.
I disagree, I do not see what you state.
> the one correct solution i can imagine is to have
> one global pipeline that gets to handle every page that comes out of
> lenya, and that's where link rewriting according to lenya-document://
> and proxying semantics must happen. which is certainly 1.5 stuff.
It is not the problem since the publication sitemap is doing it. See the
proxyTransformer that I just added, it is missing 2 small things and
proxying will work.
salu2
>
> regards,
>
> jörn
>
>
--
Thorsten Scherler thorsten.at.apache.org
Open Source Java consulting, training and solutions
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Jörn Nettingsmeier <ne...@apache.org>.
Andreas Hartmann wrote:
> Hi Lenya devs,
>
> what's the current status of the proxying issue?
> Which changes are necessary so we can retry the RC?
> My latest info was that we postpone all code changes
> and add some documentation what to achieve with the
> current codebase.
>
> Could anyone imagine to take the lead in this mission?
will do that tomorrow - i'm travelling atm.
i'll try and document what can be done, but i still think we should be
honest and label lenya's proxy support as broken ("works sometimes under
not-precisely-defined optimal conditions" is hardly helpful).
what little proxying can be accomplished atm mostly relies on web server
feature, not on lenya.
i'm opposed to adding quick hacks to tackle some corner cases, because
the fact remains that proxying is fundamentally broken and cannot be
fixed in a sane way. the one correct solution i can imagine is to have
one global pipeline that gets to handle every page that comes out of
lenya, and that's where link rewriting according to lenya-document://
and proxying semantics must happen. which is certainly 1.5 stuff.
regards,
jörn
--
Jörn Nettingsmeier
Kurt is up in heaven now.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Michael Wechner <mi...@wyona.com>.
Andreas Hartmann wrote:
>Andreas Hartmann schrieb:
>
>
>>Hi Lenya devs,
>>
>>what's the current status of the proxying issue?
>>Which changes are necessary so we can retry the RC?
>>
>>
>
>[...]
>
>After diving deeper into this issue, I'll try to summarize my
>understanding of the core problem:
>
>
>The root of a publication is usually
>
> http://host:port/contextPrefix/pubId/area
>
>where contextPrefix might be an empty string.
>This would work quite well with proxying. You'd just need
>to put the proxy URL in front of all your internal links
>and could end up with
>
> http://mysite.com/...
>
>serving the URL space of the publication.
>
>
>But now we have URLs of the form
>
> http://host:port/contextPrefix/modules/...
>
>which are located outside the URL space of the publication.
>
>
>One approach would be to move all these URLs into the
>publication URL space:
>
> http://host:port/contextPrefix/pubId/area/modules/...
>
>IIUC the major argument against this is that it would require
>a reserved URL space within the URL space of the publication
>(e.g., /modules/... or /lenya/modules/...). OTOH, if we add
>some proxy rules to map this space into the proxy URL space
>
> http://mysite.com/modules/...
>
>we have masked the URL space anyway. IMO it would be more
>user-friendly to move the /modules URL space into the publication
>URL space in Lenya, instead of requiring to configure the proxy
>accordingly.
>
>WDYT?
>
>
+1 for introducing a reserved prefix below the publication id.
One could make this prefix configurable such that one could avoid conflicts
or choose a name, which is hardly ever to generate a conflict, e.g.
lenya-modules-324234432
whereas I think making it configurable makes more sense.
Cheers
Michi
>-- Andreas
>
>
>
>
--
Michael Wechner
Wyona - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
+41 44 272 91 61
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Andreas Hartmann <an...@apache.org>.
Andreas Hartmann schrieb:
> Thorsten Scherler schrieb:
>>> IMO it would be more
>>> user-friendly to move the /modules URL space into the publication
>>> URL space in Lenya, instead of requiring to configure the proxy
>>> accordingly.
>>>
>>> WDYT?
>> Not sure, when we now chance the urls how much custom code will bounce
>> and blow up. IMO to close this bug for now and get the release out, I
>> prefer adding proxy rules then changing the linking.
>>
>> As soon as we get rid of areas the whole proxy stuff needs rethinking
>> anyway.
>
> OK, I see your point. The major advantage would be that no further
> testing is required for non-proxy or simple proxy environments.
> But I think that the changes for the reserved URL space would be small
> enough to be tested quite quickly.
BTW, this issue is especially problematic when you export your
site statically. In this case you can't set any proxy rules,
you have to rely on a self-contained URL space.
-- Andreas
--
Andreas Hartmann, CTO
BeCompany GmbH
http://www.becompany.ch
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Andreas Hartmann <an...@apache.org>.
Thorsten Scherler schrieb:
>> IMO it would be more
>> user-friendly to move the /modules URL space into the publication
>> URL space in Lenya, instead of requiring to configure the proxy
>> accordingly.
>>
>> WDYT?
>
> Not sure, when we now chance the urls how much custom code will bounce
> and blow up. IMO to close this bug for now and get the release out, I
> prefer adding proxy rules then changing the linking.
>
> As soon as we get rid of areas the whole proxy stuff needs rethinking
> anyway.
OK, I see your point. The major advantage would be that no further
testing is required for non-proxy or simple proxy environments.
But I think that the changes for the reserved URL space would be small
enough to be tested quite quickly.
-- Andreas
--
Andreas Hartmann, CTO
BeCompany GmbH
http://www.becompany.ch
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Thorsten Scherler <th...@juntadeandalucia.es>.
On Wed, 2007-05-23 at 12:02 +0200, Andreas Hartmann wrote:
> Andreas Hartmann schrieb:
> > Hi Lenya devs,
> >
> > what's the current status of the proxying issue?
> > Which changes are necessary so we can retry the RC?
>
> [...]
>
> After diving deeper into this issue, I'll try to summarize my
> understanding of the core problem:
>
>
> The root of a publication is usually
>
> http://host:port/contextPrefix/pubId/area
>
> where contextPrefix might be an empty string.
> This would work quite well with proxying. You'd just need
> to put the proxy URL in front of all your internal links
> and could end up with
>
> http://mysite.com/...
>
> serving the URL space of the publication.
>
>
> But now we have URLs of the form
>
> http://host:port/contextPrefix/modules/...
>
> which are located outside the URL space of the publication.
>
>
> One approach would be to move all these URLs into the
> publication URL space:
>
> http://host:port/contextPrefix/pubId/area/modules/...
>
> IIUC the major argument against this is that it would require
> a reserved URL space within the URL space of the publication
> (e.g., /modules/... or /lenya/modules/...). OTOH, if we add
> some proxy rules to map this space into the proxy URL space
>
> http://mysite.com/modules/...
>
> we have masked the URL space anyway. IMO it would be more
> user-friendly to move the /modules URL space into the publication
> URL space in Lenya, instead of requiring to configure the proxy
> accordingly.
>
> WDYT?
Not sure, when we now chance the urls how much custom code will bounce
and blow up. IMO to close this bug for now and get the release out, I
prefer adding proxy rules then changing the linking.
As soon as we get rid of areas the whole proxy stuff needs rethinking
anyway.
salu2
--
Thorsten Scherler thorsten.at.apache.org
Open Source Java consulting, training and solutions
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Andreas Hartmann <an...@apache.org>.
Andreas Hartmann schrieb:
> Hi Lenya devs,
>
> what's the current status of the proxying issue?
> Which changes are necessary so we can retry the RC?
[...]
After diving deeper into this issue, I'll try to summarize my
understanding of the core problem:
The root of a publication is usually
http://host:port/contextPrefix/pubId/area
where contextPrefix might be an empty string.
This would work quite well with proxying. You'd just need
to put the proxy URL in front of all your internal links
and could end up with
http://mysite.com/...
serving the URL space of the publication.
But now we have URLs of the form
http://host:port/contextPrefix/modules/...
which are located outside the URL space of the publication.
One approach would be to move all these URLs into the
publication URL space:
http://host:port/contextPrefix/pubId/area/modules/...
IIUC the major argument against this is that it would require
a reserved URL space within the URL space of the publication
(e.g., /modules/... or /lenya/modules/...). OTOH, if we add
some proxy rules to map this space into the proxy URL space
http://mysite.com/modules/...
we have masked the URL space anyway. IMO it would be more
user-friendly to move the /modules URL space into the publication
URL space in Lenya, instead of requiring to configure the proxy
accordingly.
WDYT?
-- Andreas
--
Andreas Hartmann, CTO
BeCompany GmbH
http://www.becompany.ch
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Re: Status of proxying issue for 1.4
Posted by Thorsten Scherler <th...@juntadeandalucia.es>.
On Tue, 2007-05-22 at 08:17 +0200, Andreas Hartmann wrote:
> Hi Lenya devs,
>
> what's the current status of the proxying issue?
> Which changes are necessary so we can retry the RC?
> My latest info was that we postpone all code changes
> and add some documentation what to achieve with the
> current codebase.
I had a quick look on this issue and the only thing that occurred to me
is to add a new proxy tag:
<proxy prefix="/lenya" ssl="true" url="https://www.host.com/ssl/lenya"/>
This would then be picked up by the UuidToUrlTransformer.
wdyt?
salu2
--
Thorsten Scherler thorsten.at.apache.org
Open Source Java consulting, training and solutions
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org