hypercontent-users

[Hypercontent-users] Problem installing

[Carl B. <C.P...@hu...>]

Alex

I have installed HyperContent on our uPortal 2.1.4 install here at 
Hull.  It seems to have installed okay after some tweaking.  However,  I 
get the message  'You do not have access to any configured projects'  in 
the channel.   I am logged in as admin, I have checked that the 
repositories path in Project.xml  is set correct 
'hypercontent-repositories/project-definitions/'  I have also tried 
putting the full path in there as well. 
The log output is :
ERROR [tcpConnection-6502-2] root.[] Oct/18 09:29:07 - EXCEPTION: 
edu.columbia.filesystem.FileSystemException: The filesystem could not 
mounted due to faulty arguments - please check your configuration
edu.columbia.filesystem.FileSystemException: The filesystem could not 
mounted due to faulty arguments - please check your configuration

Do you have any idea what might be happening?
Thanks
Carl

-- 
************************************

Carl Barrow
Web Developer
e-Services
The University of Hull
Cottingham Road
Hull
HU6 7RX
Ext. 6838
 
************************************

[Hypercontent-users] New Project Button

[Kirk S. <sto...@mu...>]

There is no longer a "New" button for creating projects in the 
HyperContent Project Definitions project. Does anyone have any idea what 
happend to it, and how I can get it back?

I was logged in on a portal admin account, which I had used previously 
to created new projects.


Thanks,
Kirk

Re: [Hypercontent-users] Text Only Template

[Adam C. <ac...@co...>]

Carl,

Did you set up the following variables:

> <xsl:variable name="sourceDirectory"  
> select="/cms:wrapper/cms:source/@directory" />
> <xsl:variable name="sourceBaseName"  
> select="/cms:wrapper/cms:source/@basename" />


Adam

On Oct 12, 2004, at 6:32 AM, Carl Barrow wrote:

> Adam
>
> I have been somewhat distracted from setting hypercontent up due to  
> other work here.  However it is now as pressing as ever.   I just have  
> a couple of  things to clear up....
>
> You sent the code that you used for a link to a printable page which I  
> have set up and it works for my text only site,  my code below:
> <a class="topmenu"  
> href="{concat($baseURL,'/campus/ 
> text_only',$sourceDirectory,$sourceBaseName, '.html')}">Text Only</a>
>
> This works on the live site (http://port.hull.ac.uk/campus) but won't  
> work on my test/preview site, (  
> http://berk.lib.hull.ac.uk/uPortal/hulluni/build ). baseURL is passed  
> fine but it's the path after that I can't get fully working, I have to  
> hardcode the path after $baseURL in the link, so it's ether working on  
> one site or the other.  Have you come across this at all and do you  
> have any idea how I might pass that path instead of coding it?
>
> All help appreciated
> Thanks
> Carl
>
> Adam Carl wrote:
>
>> Carl,
>>
>> Many times I produce a 'printable' version of a page as well as the  
>> web version. This page is created for all pages
>> but I give the user the ability to hide the link to the printable  
>> page.  I create a variable in my page dtd and check for yes in my  
>> xsl.
>>
>> Here is some sample code:
>>
>> <xsl:variable name="sourceDirectory"  
>> select="/cms:wrapper/cms:source/@directory" />
>> <xsl:variable name="sourceBaseName"  
>> select="/cms:wrapper/cms:source/@basename" />
>>
>>     <xsl:if  
>> test="/cms:wrapper/cms:source/page/@include-printable-page-link  
>> ='yes'">
>>                        <div id="divPrintablePageLink">
>>             <a target="_blank" class="PrintablePageLink"  
>> href="{concat($baseURL,'/printable',$sourceDirectory,$sourceBaseName,  
>> '.html')}">View Printable Version</a>
>>             </div>
>>                                            </xsl:if>
>>
>> This assumes you pass in the base URL into the stylesheet as a  
>> parameter.
>>
>> <xsl:param name="baseURL" />
>>
>> Hope this helps.
>>
>> Adam
>>
>>
>
> -- 
> ************************************
>
> Carl Barrow
> Web Developer
> e-Services
> The University of Hull
> Cottingham Road
> Hull
> HU6 7RX
> Ext. 6838
> ************************************

Re: [Hypercontent-users] Text Only Template

[Carl B. <C.P...@hu...>]

Adam

I have been somewhat distracted from setting hypercontent up due to 
other work here.  However it is now as pressing as ever.   I just have a 
couple of  things to clear up....

You sent the code that you used for a link to a printable page which I 
have set up and it works for my text only site,  my code below:
<a class="topmenu" 
href="{concat($baseURL,'/campus/text_only',$sourceDirectory,$sourceBaseName, 
'.html')}">Text Only</a>

This works on the live site (http://port.hull.ac.uk/campus) but won't 
work on my test/preview site, ( 
http://berk.lib.hull.ac.uk/uPortal/hulluni/build ). baseURL is passed 
fine but it's the path after that I can't get fully working, I have to 
hardcode the path after $baseURL in the link, so it's ether working on 
one site or the other.  Have you come across this at all and do you have 
any idea how I might pass that path instead of coding it?

All help appreciated
Thanks
Carl

Adam Carl wrote:

> Carl,
>
> Many times I produce a 'printable' version of a page as well as the 
> web version. This page is created for all pages
> but I give the user the ability to hide the link to the printable 
> page.  I create a variable in my page dtd and check for yes in my xsl.
>
> Here is some sample code:
>
> <xsl:variable name="sourceDirectory" 
> select="/cms:wrapper/cms:source/@directory" />
> <xsl:variable name="sourceBaseName" 
> select="/cms:wrapper/cms:source/@basename" />
>
>     <xsl:if 
> test="/cms:wrapper/cms:source/page/@include-printable-page-link ='yes'">
>            
>             <div id="divPrintablePageLink">
>             <a target="_blank" class="PrintablePageLink" 
> href="{concat($baseURL,'/printable',$sourceDirectory,$sourceBaseName, 
> '.html')}">View Printable Version</a>
>             </div>
>                                    
>         </xsl:if>
>
> This assumes you pass in the base URL into the stylesheet as a parameter.
>
> <xsl:param name="baseURL" />
>
> Hope this helps.
>
> Adam
>
>

-- 
************************************

Carl Barrow
Web Developer
e-Services
The University of Hull
Cottingham Road
Hull
HU6 7RX
Ext. 6838
 
************************************

Re: [Hypercontent-users] About_hypercontent14_errors

[Alex V. <av...@co...>]

Responses are embedded.

On Oct 8, 2004, at 3:38 PM, Jevgenijs Sokolovs wrote:

> Hello Alex,
>
>   We are from University of Latvia and we have some problems with  
> Hypercontent 1.4:
>
> 1.      We have an error in log file :
> pool worker #8] root.[] 2004-10-08 13:51:39,815  -
> EXCEPTION: java.lang.NullPointerException
> java.lang.NullPointerException
>         at
> edu.columbia.hypercontent.contentmanager.CMModel.expandRepositoryNodeRe 
> cursive(CMModel.java:197)
>         at
> edu.columbia.hypercontent.contentmanager.commands.HighlightNode.execute 
> (HighlightNode.java:53)
>         at
> edu.columbia.hypercontent.contentmanager.CContentManager.setRuntimeData 
> (CContentManager.java:192)
>         at
> org.jasig.portal.MultithreadedChannelAdapter.setRuntimeData(Multithread 
> edChannelAdapter.java:61)
>         at
> org.jasig.portal.ChannelRenderer$Worker.run(ChannelRenderer.java:399)
>         at org.jasig.portal.utils.threading.Worker.run(Worker.java:88)
>
> a.
>
> What expandRepositoryNodeRecursive class does?  What does this error
> shows?

This method is used to open up the directory structure in the Content  
Manager channel to highlight a particular file or directory.  That  
exception indicates either an attempt to highlight a node that has been  
recently deleted by another user, or an attempt to perform the  
operation on project that has not been properly initialized. This is  
not a serious error - the best fix is to close and reopen the project.

>
> 2.      We are using xml with cvs(vCard) files to store people which
> work in university, we have more than 3000 people. Is it good to store  
> so many files in one directory?

I am not aware of any particular issues.  This is likely system  
dependent, so you should consult the appropriate resources for your  
server operating system.

> 3.      How large can be repository and is it possible to know how
> many files we can keep there (MB)?

There is no limit to the size of a repository, and no performance  
degradation as it grows.  HyperContent 2 adds a getContentLength method  
which will make it possible to ascertain the exact size of your  
repository.  If you use the "rebuild search index" command, you will  
see an accurate project file count when the command is completed.

>
>
> -- 
> Best regards,
>  Jevgenij Sokolov
>
> University of Latvia
> IT department
> Developer
> tel: 7034736
> mob.tel: +(371) 9720754
> email: Jev...@lu...
>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on  
> ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give  
> us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out  
> more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>

[Hypercontent-users] About_hypercontent14_errors

[Jevgenijs S. <Jev...@lu...>]

Hello Alex,

  We are from University of Latvia and we have some problems with Hypercontent 1.4:

1.      We have an error in log file :
pool worker #8] root.[] 2004-10-08 13:51:39,815  -
EXCEPTION: java.lang.NullPointerException
java.lang.NullPointerException
        at
edu.columbia.hypercontent.contentmanager.CMModel.expandRepositoryNodeRecursive(CMModel.java:197)
        at
edu.columbia.hypercontent.contentmanager.commands.HighlightNode.execute(HighlightNode.java:53)
        at
edu.columbia.hypercontent.contentmanager.CContentManager.setRuntimeData(CContentManager.java:192)
        at
org.jasig.portal.MultithreadedChannelAdapter.setRuntimeData(MultithreadedChannelAdapter.java:61)
        at
org.jasig.portal.ChannelRenderer$Worker.run(ChannelRenderer.java:399)
        at org.jasig.portal.utils.threading.Worker.run(Worker.java:88)

a.      

What expandRepositoryNodeRecursive class does?  What does this error
shows?

2.      We are using xml with cvs(vCard) files to store people which
work in university, we have more than 3000 people. Is it good to store so many files in one directory?
3.      How large can be repository and is it possible to know how
many files we can keep there (MB)?


-- 
Best regards,
 Jevgenij Sokolov

University of Latvia
IT department
Developer
tel: 7034736
mob.tel: +(371) 9720754
email: Jev...@lu...

Re: [Hypercontent-users] About_hyporconent2

[Alex V. <av...@co...>]

HyperContent 2.0 is not ready yet for use in production - while much of 
the code that is there is stable, I'd like to keep open the possibility 
of some API changes between now and the release.  As I have mentioned 
before, HyperContent 2 is backwards compatible with web sites built in 
the current version, so I continue to encourage use of HyperContent 1.4 
until 2.0 is released.  If you need to build a web application today, I 
would encourage you to consider something like Spring.  For building a 
static web site without form handling, HyperContent 1.4 is sufficient.
	As a framework, HyperContent 2.0 will offer a great deal of benefit in 
terms of establishing pipelines to generate a presentation layer bound 
to some form processing; however, it provides no database specific 
utilities, unlike Spring.   Berkeley DB is only used currently as a 
content caching mechanism, so that the output of a complex pipeline can 
be reused.  I have considered an IFileSystemManager implementation 
based on BerkeleyDB, but this is not currently a priority, as the 
existing repository implementation is stable and performs well.
	To use  MySQL or any other data source in HyperContent 2.0, there are 
3 possible levels of integration: an IFileSystemManager implementation 
would allow for a complete, generic repository to be implemented in a 
DB.  A custom pipeline stage could be used to load data from the DB 
into objects that could be rendered via Velocity to an HTML form, and 
bound to a FormProcessor which could persist changes back to the DB. A 
subclass of FileAsset could be used to load some data from the database 
based on the contents of a file in a normal HyperContent repository.

I have not analyzed what would be involved in a data conversion between 
HyperContent and Cocoon.  I have no plans to perform such an analysis, 
but if you would like to investigate, that information may be of some 
value to the community.

HyperContent 1.4 uses uPortal groups and permissions to evaluate 
whether a user can view or edit files.  When a user is a Portal 
Administrator, this logic is bypassed to always respond "yes".  Thus 
there is more processing involved in displaying file-browsing screens 
for non-administrators.  However, uPortal does cache permissions 
information, so the performance hit should not be too noticeable.  It 
may help to make sure you have indexes on the UP_GROUP_MEMBERSHIP and 
UP_PERMISSION tables.

Alex

	

On Oct 1, 2004, at 1:52 PM, Ingmars Rubenis wrote:

> Hello!
>
>
> Thanks for help to solve the problem about sftp and scp and thanks
> about tomcat memory recommendations. Now we are testing
> scp and sftp.
> About hyporconent2
> We are going to build new site. What do You think about using
> Hypercontent 2.0 for using their classes in that project.
> Now we are thinking about using MVC architecture (Spring) or pipeline
> model like Hypercontent 2.0 or Cocoon. Is possible to use hypercontent
> 2.0 as low level CMS (methods and classes will be used for building
> for example connection with database.For the presentaion logic will be
> used velocity) or may be we should use cocoon or anothe framework like
> Spring. What is your opinion about using hypercontent 2.0 as framework
> for web pages?
> As I understand You are going to use Berkley DB for Hypercontent 2.0.
> What will be the functions of Berkley DB?
> Is true that it will be possible to store data also in relational
> database like Mysql?
> Will it be possible to convert data or repository to Cocoon repository 
> from
> version Hypercontent 2.0 or  Hypercontent 1.4?
>
> We have another problem with  Hypercontent 1.4. I have to add all
> users whish ar adding information in Hypercontent 1.4 to portal
> administrators group, because it takes a lot of time to go through
> files in the project. I think it is not normal. What is your
> suggestion? When users are portal administrators then everything works
> fine.
>
>
>
>
>
>
>
>
>
>
>
>
>
> Best Wishes
> Ingmars Rubenis
> LU
> University of Latvia
> Developer
> IT department
> (371) 7034736
>
>
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on 
> ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give 
> us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out 
> more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users

[Hypercontent-users] About_hyporconent2

[Ingmars R. <ing...@lu...>]

Hello!


Thanks for help to solve the problem about sftp and scp and thanks
about tomcat memory recommendations. Now we are testing
scp and sftp.
About hyporconent2
We are going to build new site. What do You think about using
Hypercontent 2.0 for using their classes in that project.
Now we are thinking about using MVC architecture (Spring) or pipeline
model like Hypercontent 2.0 or Cocoon. Is possible to use hypercontent
2.0 as low level CMS (methods and classes will be used for building
for example connection with database.For the presentaion logic will be
used velocity) or may be we should use cocoon or anothe framework like
Spring. What is your opinion about using hypercontent 2.0 as framework
for web pages?
As I understand You are going to use Berkley DB for Hypercontent 2.0.
What will be the functions of Berkley DB?
Is true that it will be possible to store data also in relational
database like Mysql?
Will it be possible to convert data or repository to Cocoon repository from
version Hypercontent 2.0 or  Hypercontent 1.4?

We have another problem with  Hypercontent 1.4. I have to add all
users whish ar adding information in Hypercontent 1.4 to portal
administrators group, because it takes a lot of time to go through
files in the project. I think it is not normal. What is your
suggestion? When users are portal administrators then everything works
fine.













Best Wishes
Ingmars Rubenis
LU
University of Latvia
Developer
IT department
(371) 7034736

[Hypercontent-users] Can't publish single file in HyrepContent1.4

[<sd...@la...>]

Hello, Alex!

Last time I sent you an email with attachment, perhaps you couldn't see i=
t. The
problem was that HyperContent doesn't publish single file. I still can do
"Publish All". In shows the message which I stored online:
http://portaldv.lanet.lv/rss/index/index.html
What could cause such behaviour?
Some words about SCP using JSch and updated JCshFileSystemImpl. After we =
started
to use JSch instead of j2ssh buid process became more stable. It isn't
necessary to restart tomcat every day any nore. Looks like SCP realy work=
s
faster than SFTP, at least buid process is faster!=20

Thanks,=20
Juris Treibahs

Re: [Hypercontent-users] xmlns question

[Alex V. <av...@co...>]

You know, I saw this once before . . . weird Xalan bug . . . I ended up 
with

exclude-result-prefixes="xsl cms fs rdf dc vcard"

Excluding the xsl namespace may have been the key, but I could be 
wrong.  Let me know if that works.

Alex

On Sep 29, 2004, at 12:27 PM, Kirk Storer wrote:

> I am converting common.xsl to generate pages that are xhtml 1.0 
> transitional complient. I have it done except for a problem with 
> xmlns:xmlns:cms attributes being put into some of the tags. I have 
> used the following to remove all of the other namespaces
> exclude-result-prefixes="cms cfs rdf dc"
>
> However, I can't figure out what to add to keep the xmlns:xmlns:cms 
> attributes from being put in the output. I have tried putting 
> xmlns:cms and many other similar things, but none of them have helped.
>
>
> Thanks,
> Kirk
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: IT Product Guide on 
> ITManagersJournal
> Use IT products in your business? Tell us what you think of them. Give 
> us
> Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out 
> more
> http://productguide.itmanagersjournal.com/guidepromo.tmpl
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>

[Hypercontent-users] xmlns question

[Kirk S. <sto...@mu...>]

I am converting common.xsl to generate pages that are xhtml 1.0 
transitional complient. I have it done except for a problem with 
xmlns:xmlns:cms attributes being put into some of the tags. I have used 
the following to remove all of the other namespaces
exclude-result-prefixes="cms cfs rdf dc"

However, I can't figure out what to add to keep the xmlns:xmlns:cms 
attributes from being put in the output. I have tried putting xmlns:cms 
and many other similar things, but none of them have helped.


Thanks,
Kirk

Re: [Hypercontent-users] Can't publish in HyperoContent 1.4

[Alex V. <av...@co...>]

That error indicates that a filesystem was defined in your project XML  
without a "type" attribute, or with a "type" attribute that did not  
match any of the filesystem types configured in FileSystems.xml.  I  
didn't see an attachment on your message - was that stack trace it?

Cheers,
	Alex

On Sep 27, 2004, at 10:52 AM, sd...@la... wrote:

> Hello!
>
> Right now we are using HyperContent1.4 - we've compiled last version  
> of code
> from reposotory. It is under intensive for a couple of days.  The last  
> version
> looks more stable. For SFTP JSchSFTPFileSystemImpl is used. Today's  
> morning it
> refused to "Publish" (see attachment). The only way to publish is to  
> use
> "Publish All".
> Also, could you look at the following error from log files:
>
> ERROR [Thread-6] root.[] 2004-09-27 17:35:02,345  - EXCEPTION:
> edu.columbia.hypercontent.CMSException: There was an error while  
> processing
> this directive; A configuration error made it impossible to fulfil this
> request
> There was an error while processing this directive
> Recorded Exception:
> edu.columbia.filesystem.FileSystemException: A configuration error  
> made it
> impossible to fulfil this request
> 	at
> edu.columbia.filesystem.FileSystemFactory.getFileSystem(FileSystemFacto 
> ry.java:85)
> 	at
> edu.columbia.hypercontent.Project.getFileSystemFromElement(Project.java 
> :225)
> 	at edu.columbia.hypercontent.Project.initFileSystems(Project.java:215)
>
>
>
>
> Best wishes,
> Juris Treibahs.
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
> Project Admins to receive an Apple iPod Mini FREE for your judgement on
> who ports your project to Linux PPC the best. Sponsored by IBM.
> Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>

[Hypercontent-users] Can't publish in HyperoContent 1.4

[<sd...@la...>]

Hello!

Right now we are using HyperContent1.4 - we've compiled last version of c=
ode
from reposotory. It is under intensive for a couple of days.  The last ve=
rsion
looks more stable. For SFTP JSchSFTPFileSystemImpl is used. Today's morni=
ng it
refused to "Publish" (see attachment). The only way to publish is to use
"Publish All".
Also, could you look at the following error from log files:

ERROR [Thread-6] root.[] 2004-09-27 17:35:02,345  - EXCEPTION:
edu.columbia.hypercontent.CMSException: There was an error while processi=
ng
this directive; A configuration error made it impossible to fulfil this
request
There was an error while processing this directive
Recorded Exception:=20
edu.columbia.filesystem.FileSystemException: A configuration error made i=
t
impossible to fulfil this request
	at
edu.columbia.filesystem.FileSystemFactory.getFileSystem(FileSystemFactory=
.java:85)
	at
edu.columbia.hypercontent.Project.getFileSystemFromElement(Project.java:2=
25)
	at edu.columbia.hypercontent.Project.initFileSystems(Project.java:215)




Best wishes,
Juris Treibahs.
=20
=20

Re: [Hypercontent-users] using SCP

[Alex V. <av...@co...>]

Juris,
	You can try SCP in JSch, but after spending another day with it I have 
found that the SFTP implementation was very poor; much slower than 
j2ssh, especially the SFTP get command.  I have made improvements, 
which I will submit back to the JCsh project, so that it is now perhaps 
20% faster than j2ssh for both SFTP puts and gets.  You will see in the 
updated JCshFileSystemImpl that I have a method "storeSCP" which was an 
attempt to use SCP for puts.  It is much slower than the SFTP 
implementation, but perhaps you can look at that as a starting point 
for implementing with SCP. It was copied and pasted from the SCP code 
in Ant. I suspect we may not be able to achieve better performance with 
SCP and JCsh, because the principal performance benefit of SCP is 
supposedly that it runs on the SSH1 protocol, whereas JCsh only 
implements the slower SSH2.

The jcsh jar and JCshFileSystemImpl are updated in CVS with the 
changes.  If you were denied project access when you tried to change 
the classname, you should check the logs for an error; chances are the 
classname was not correctly spelled, or there was an instantiation 
error.

Cheers,
	Alex

On Sep 24, 2004, at 9:07 AM, sd...@la... wrote:

> Hello, Alex!
>
> We've got an idea of using SCP protocol using Java Secure Channel(jsch)
> distribution. That would work faster than SFTP in hyperContent1.4. 
> What do you
> think  of rewriting SFTPFileSystemImpl possibility to use SCP protocol?
>
>
> Thanks,
> Juris Treibahs
>
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
> Project Admins to receive an Apple iPod Mini FREE for your judgement on
> who ports your project to Linux PPC the best. Sponsored by IBM.
> Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>

[Hypercontent-users] using SCP

[<sd...@la...>]

Hello, Alex!

We've got an idea of using SCP protocol using Java Secure Channel(jsch)
distribution. That would work faster than SFTP in hyperContent1.4. What d=
o you
think  of rewriting SFTPFileSystemImpl possibility to use SCP protocol?


Thanks,=20
Juris Treibahs

[Hypercontent-users] RE:slow_hyperconent14

[<sd...@la...>]

Hello, Alex!
Thanks wery mych for recomendations. We've set all of them except one and=
 till
now there were no crashes. One thing I could not do is to use new class
JSchSFTPFileSystemImpl. After I change class name for sftp in
FileSystemConfig.xml and restart Tomcat, I don't have rights to access an=
y
projects I any more. It's strange, I didn't expect such relation. Where c=
ould
be the reason?
=20
Best wishes,=20
Juris Treibahs

Re: [Hypercontent-users] slow_hyperconent14

[Alex V. <av...@co...>]

Just an update on SFTP speed - some research reveals that there are =20
some fundamental issues with the SFTP protocol which limit its speed: =20=

SFTP implements its own packet technology, and every sent packet must =20=

be acknowledged before the the next packet is sent.  By one estimate, =20=

this ends up yielding a maximum throughput of only 17% of the available =20=

bandwidth, and is essentially unnecessary since the underlying TCP/IP =20=

protocol basically fulfills the same purpose.  JSch uses a 1k packet =20
size - by boosting this to 8k I saw a 15-20% speed increase.  Setting =20=

it much higher seems to yield errors.  SCP is free of this limitation, =20=

so I may look into a hybrid SFTP/SCP implementation, where SFTP is used =20=

for things like directory listings, and SCP for actual file transfer.

Alex

On Sep 22, 2004, at 11:44 AM, Ingmars Rubenis wrote:

> Hello!
>
> About version 1.4 hypercontent. We are daily building more than
> 1300 files about 15 times per day. We are going alive on the 28th of
> September. We are planing about 3000 http files. The site of =20
> university of Latvia is
> quit large site please see http://portal.lu.lv .
> The hyperconent 1.4 is very slow. My memory setting are:
> JAVA_OPTS=3D"-server -Xincgc -Xms512m -Xmx1024m -Xss128k =20
> -XX:MaxHeapFreeRatio=3D70 -XX:MinHeapFreeRatio=3D40 -XX:NewSize=3D256m =
=20
> -XX:MaxNewSize=3D256m -XX:SurvivorRatio=3D8
> -verbose:gc -XX:+UseLWPSynchronization"
> It is even difficult to go  through files in hypercontent. Build =20
> process
> hangs a couple times a day.  Also it is time consuming to upload gif
> files (it is ok with jpeg).  May be you
> could help us with that. I have updated ftp.jar from =20
> www.enterprisedt.com library with newer one.
> About j2ssh-core.jar. I think this is the slowest point  in your
> application (the build proces hangs because of this in my mind).  The =20=

> library is quite slow and old and there are a lot of
> bugs. We are going to replace the class SFTPFileSystemImpl with
> library from project http://sourceforge.net/projects/jsch/ . What do
> You thing? Here is rather old comparing between these libraries
>   =20
> http://nagoya.apache.org/eyebrowse/ReadMsg?listName=3Dcommons-=20
> de...@ja...&msgId=3D697314
>   Is it worth to try? Where is the main problem? Why is this
>   application so slow and build through sftp is not working as it
>   should?
> I think You should not use old libraries in Your project hypercontent
> 2.0 like j2ssh-core.jar. May be try
> http://sourceforge.net/projects/jsch/ instead.
> Why You are not using apache slide =20
> http://jakarta.apache.org/slide/index.html in the hypercontent 2 =20
> project?
> What do You think about apache cocoon and =20
> http://cocoon.apache.org/lenya/ ?
> How spring framework will act in the next release and velocity also?
> We would like that hypercontnet 2.0 would be as seperate applacition
> (war)(not bundled with tomcat). What is Your opinion? How we could =20
> help?
>
>
> About tomcat and uportal.
> As I understand from Your previos mail that tomcat is restarted daily?
> Is there any application startup script for uportal (after restart it
> is slow)? Is there any
> tomcat health checking system or may be .sh file?
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Best Wishes
> Ingmars Rubenis
> LU
> University of Latvia
> Developer
> IT department
> (371) 7034736
>
> =
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=D3=86+=12=17=F3^=B5=E9=9A=
=8AX=AC=B2=9A'=B2=8A=DEu=BC=FF`=E5=01=111=C4%@=20
> =
=C6=13=F0^=A2w=A8=7F^=F4>=BA#y=CB@vh=A7=B2=DA+y=C7=A2=BD=E6=A7=02=9Aez#=E8=
t=C8=A7=88TD=11=FA+=CA=8B=AB=8E=E7`zg=A7=B6=89=F0=86=8Ah=AE=DB2=A2=EA=E9=AE=
=88=DEr=DBh.=20
> )=EE=C4=F3=C2=B6=17=9Bz=CB=FFJ=9A'=B2=8A=DEu=BC=88=04=CF=C3y=A7e=8Aw=FFI=
=EAm=FFn=FF=1A=88^=AD=EF=E1=B6=DA=FF=FF=FB=1F=FEw=AD=FE=9A\=FD=CA'=B5=EB-=20=

> =FE=98i=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=C7=CA=97=ABr=89=EDz{=FF=BA=C7=AB=B2=
f=A2=96)=E0=96+-=1F*^=AD=20
> =CA'=B5=E9=ED=FE=EB=1E=AE=CF=E5=8A=CBl=FE=CA.=AD=C7=9F=A2=B8=1E=FEw=AD=86=
=DBi=B3=FF=FF=96+-=B3=FB(=BA=B7=1E~=8A=E0{=F9=DE=B7=F9b=B2=DB?=96+-=20
> =8Aw=E8=FE=1C=A9z=B7(=9E=D7=A7=B7=FB=AC

Re: [Hypercontent-users] slow_hyperconent14

[Alex V. <av...@co...>]

Here are our production settings:

  -Xms3072m -Xmx3072m -XX:NewSize=3D512m -XX:MaxNewSize=3D512m

What hardware configuration are you running on?

I recommend you turn off verbose garbage collection in production.  I =20=

have not tried LWPSynchronization, so I am not certain as to its =20
impact.

I don't have an exact count, but we are managing at least 10,000 files =20=

via HyperContent right now.  As often as possible, we mount the build =20=

and publish filesystems via NFS and treat it as a local filesystem, =20
which is much faster than SFTP.   I recommend you examine this option =20=

if it is possible in your environment.

I also recommend you check out and use the latest 1.x repository code.  =20=

Since the 1.4 release, I have made the build process more forgiving in =20=

low-memory situations through more extensive use of soft references.  =20=

Just today, I wrote a JSch based SFTP filesystem impl based on your =20
recommendation.  My initial tests do not show a performance boost over =20=

j2ssh, but perhaps you will see improved stability.  You will need the =20=

jsch-20040922.jar from cvs (I have changed one method to reduce the =20
number of network calls as the library is used in HyperContent, so you =20=

cannot currently use the distributed binary), and the new class =20
JSchSFTPFileSystemImpl.  To enable this instead of J2ssh, change the =20
classname for the "sftp" filesystem as specified in Filesystems.xml.

The speed of building has a lot to do with how you use XML Includes, =20
and whether you have optimized your XSL stylesheets.  If you could post =20=

a ZIP file of your content repository along with the project definition =20=

file to a web address, I will download them and have a look and see if =20=

I can recommend any optimizations.  It is important with XML Includes =20=

only to include what you need, and with XSL it is very beneficial to =20
performance to use keys.

Apache Slide may be integrated for WebDAV support in a 2.x release - =20
that decision will be discussed on the developer list when it is time =20=

to add WebDAV support, which is not currently planned until sometime =20
next year.

I think Cocoon is great, and in fact the pipeline model in HyperContent =20=

2.0 is loosely inspired by Cocoon's model.  Unfortunately, Cocoon's =20
pipeline is XML specific, and I wanted a mechanism which would support =20=

arbitrary chaining of operations on other types of files, such as =20
resizing and watermarking images. The use of FileAsset subclasses like =20=

XMLAsset makes it possible to maintain XML characteristics across =20
pipeline stages in HyperContent.

I am not sure if Spring will be used or not in 2.0 - I have not yet =20
found occasion to use it.  The MVC framework is attractive, but is tied =20=

to servlets.  HyperContent abstracts the notion of server request and =20=

response and is not tied to being implemented as a servlet.  I have =20
recently checked in a new form binding pipeline stage for use in MVC =20
style development, that allows you to bind form processors to inputs =20
via a special name prefix.  These bindings can be expired upon use to =20=

prevent resubmission of forms, and allow multiple form processors to be =20=

mapped to different inputs in a single HTML form.   This is critical to =20=

the component-based model of constructing authoring forms in 2.0, as =20
multiple forms can be combined without losing their distinct =20
processing.

Velocity is already integrated as a template language in 2.0, though =20
there are no examples in the repository yet.  This will be an =20
alternative to XSL for building forms or pages.

HyperContent 2.0 does follow the WAR packaging model - the "webapp" =20
directory in CVS is equivalent to the root of the WAR file.  Tomcat is =20=

only embedded via a single class in the CVS as a convenience.  There =20
will be an ant target in the future for building the war file, but you =20=

can do it manually by just zipping the webapp directory.

I recommend you check the portal list for discussion of tomcat =20
restarting  - I believe the issue has been discussed there.

Cheers,
	Alex

On Sep 22, 2004, at 11:44 AM, Ingmars Rubenis wrote:

> Hello!
>
> About version 1.4 hypercontent. We are daily building more than
> 1300 files about 15 times per day. We are going alive on the 28th of
> September. We are planing about 3000 http files. The site of =20
> university of Latvia is
> quit large site please see http://portal.lu.lv .
> The hyperconent 1.4 is very slow. My memory setting are:
> JAVA_OPTS=3D"-server -Xincgc -Xms512m -Xmx1024m -Xss128k =20
> -XX:MaxHeapFreeRatio=3D70 -XX:MinHeapFreeRatio=3D40 -XX:NewSize=3D256m =
=20
> -XX:MaxNewSize=3D256m -XX:SurvivorRatio=3D8
> -verbose:gc -XX:+UseLWPSynchronization"
> It is even difficult to go  through files in hypercontent. Build =20
> process
> hangs a couple times a day.  Also it is time consuming to upload gif
> files (it is ok with jpeg).  May be you
> could help us with that. I have updated ftp.jar from =20
> www.enterprisedt.com library with newer one.
> About j2ssh-core.jar. I think this is the slowest point  in your
> application (the build proces hangs because of this in my mind).  The =20=

> library is quite slow and old and there are a lot of
> bugs. We are going to replace the class SFTPFileSystemImpl with
> library from project http://sourceforge.net/projects/jsch/ . What do
> You thing? Here is rather old comparing between these libraries
>   =20
> http://nagoya.apache.org/eyebrowse/ReadMsg?listName=3Dcommons-=20
> de...@ja...&msgId=3D697314
>   Is it worth to try? Where is the main problem? Why is this
>   application so slow and build through sftp is not working as it
>   should?
> I think You should not use old libraries in Your project hypercontent
> 2.0 like j2ssh-core.jar. May be try
> http://sourceforge.net/projects/jsch/ instead.
> Why You are not using apache slide =20
> http://jakarta.apache.org/slide/index.html in the hypercontent 2 =20
> project?
> What do You think about apache cocoon and =20
> http://cocoon.apache.org/lenya/ ?
> How spring framework will act in the next release and velocity also?
> We would like that hypercontnet 2.0 would be as seperate applacition
> (war)(not bundled with tomcat). What is Your opinion? How we could =20
> help?
>
>
> About tomcat and uportal.
> As I understand from Your previos mail that tomcat is restarted daily?
> Is there any application startup script for uportal (after restart it
> is slow)? Is there any
> tomcat health checking system or may be .sh file?
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Best Wishes
> Ingmars Rubenis
> LU
> University of Latvia
> Developer
> IT department
> (371) 7034736
>
> =
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=D3=86+=12=17=F3^=B5=E9=9A=
=8AX=AC=B2=9A'=B2=8A=DEu=BC=FF`=E5=01=111=C4%@=20
> =
=C6=13=F0^=A2w=A8=7F^=F4>=BA#y=CB@vh=A7=B2=DA+y=C7=A2=BD=E6=A7=02=9Aez#=E8=
t=C8=A7=88TD=11=FA+=CA=8B=AB=8E=E7`zg=A7=B6=89=F0=86=8Ah=AE=DB2=A2=EA=E9=AE=
=88=DEr=DBh.=20
> )=EE=C4=F3=C2=B6=17=9Bz=CB=FFJ=9A'=B2=8A=DEu=BC=88=04=CF=C3y=A7e=8Aw=FFI=
=EAm=FFn=FF=1A=88^=AD=EF=E1=B6=DA=FF=FF=FB=1F=FEw=AD=FE=9A\=FD=CA'=B5=EB-=20=

> =FE=98i=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=
=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=FF=C7=CA=97=ABr=89=EDz{=FF=BA=C7=AB=B2=
f=A2=96)=E0=96+-=1F*^=AD=20
> =CA'=B5=E9=ED=FE=EB=1E=AE=CF=E5=8A=CBl=FE=CA.=AD=C7=9F=A2=B8=1E=FEw=AD=86=
=DBi=B3=FF=FF=96+-=B3=FB(=BA=B7=1E~=8A=E0{=F9=DE=B7=F9b=B2=DB?=96+-=20
> =8Aw=E8=FE=1C=A9z=B7(=9E=D7=A7=B7=FB=AC
 =20=

[Hypercontent-users] slow_hyperconent14

[Ingmars R. <ing...@lu...>]
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[Hypercontent-users] text only link site links

[Carl B. <C.P...@hu...>]

Hi Alex

RE: "I have set up a second text only high contrast site template and 
that site is built/published at the same time as the standard site in a 
directory called ‘text_only’ which is off the root of the standard site. 
I assumed that the navigation would be relative and work fine for that 
version of the site, however all links take you to the standard site 
rather the pages within the text only site. Have I missed something in 
setup, do I need to do something else? "

I sorted this out and it's working fine now.


Thanks

Carl

-- 
************************************

Carl Barrow
Web Developer
e-Services
The University of Hull
Cottingham Road
Hull
HU6 7RX
Ext. 6838
 
************************************

Re: [Hypercontent-users] Text Only Template

[Adam C. <ac...@co...>]

Carl,

Many times I produce a 'printable' version of a page as well as the web=20=

version. This page is created for all pages
but I give the user the ability to hide the link to the printable page.=20=

  I create a variable in my page dtd and check for yes in my xsl.

Here is some sample code:

<xsl:variable name=3D"sourceDirectory"=20
select=3D"/cms:wrapper/cms:source/@directory" />
<xsl:variable name=3D"sourceBaseName"=20
select=3D"/cms:wrapper/cms:source/@basename" />

	<xsl:if=20
test=3D"/cms:wrapper/cms:source/page/@include-printable-page-link=20
=3D'yes'">
		=09
			<div id=3D"divPrintablePageLink">
			<a target=3D"_blank" class=3D"PrintablePageLink"=20=

href=3D"{concat($baseURL,'/printable',$sourceDirectory,$sourceBaseName,=20=

'.html')}">View Printable Version</a>
			</div>
								=09
		</xsl:if>

This assumes you pass in the base URL into the stylesheet as a=20
parameter.

<xsl:param name=3D"baseURL" />

Hope this helps.

Adam


On Sep 20, 2004, at 11:30 AM, Carl Barrow wrote:

> Hi Alex
>
> I have set up a second text only high contrast site template and that=20=

> site is built/published at the same time as the standard site in a=20
> directory called =91text_only=92 which is off the root of the standard=20=

> site. I assumed that the navigation would be relative and work fine=20
> for that version of the site, however all links take you to the=20
> standard site rather the pages within the text only site. Have I=20
> missed something in setup, do I need to do something else?
>
> I need a link on each page of the standard site to the corresponding=20=

> page in the text only site. Do you have any idea how I might=20
> accomplish this? I assume I would be able to do it in xsl?
>
> The last question is for now=85.
>
> When the user is editing a site using hypercontent the files are=20
> visible on the left menu. Each file has an xml extension. Would it be=20=

> possible to remove the xml extension? While to me it makes sense, some=20=

> of our users here won't be able to cope with it.
>
> Many Thanks
> Carl
>
> --=20
> ************************************
>
> Carl Barrow
> Web Developer
> e-Services
> The University of Hull
> Cottingham Road
> Hull
> HU6 7RX
> Ext. 6838
> ************************************
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
> Project Admins to receive an Apple iPod Mini FREE for your judgement =
on
> who ports your project to Linux PPC the best. Sponsored by IBM.
> Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users

[Hypercontent-users] Text Only Template

[Carl B. <C.P...@hu...>]

Hi Alex

I have set up a second text only high contrast site template and that 
site is built/published at the same time as the standard site in a 
directory called ‘text_only’ which is off the root of the standard site. 
I assumed that the navigation would be relative and work fine for that 
version of the site, however all links take you to the standard site 
rather the pages within the text only site. Have I missed something in 
setup, do I need to do something else?

I need a link on each page of the standard site to the corresponding 
page in the text only site. Do you have any idea how I might accomplish 
this? I assume I would be able to do it in xsl?

The last question is for now….

When the user is editing a site using hypercontent the files are visible 
on the left menu. Each file has an xml extension. Would it be possible 
to remove the xml extension? While to me it makes sense, some of our 
users here won't be able to cope with it.

Many Thanks
Carl

-- 
************************************

Carl Barrow
Web Developer
e-Services
The University of Hull
Cottingham Road
Hull
HU6 7RX
Ext. 6838
 
************************************

[Hypercontent-users] About hypercontent1.4 errors

[<sd...@la...>]

Hello, Alex!

   So, we changed settings memory and other settings settings you replied=
 to
Ingmars. But build process still is unstable. The place where it crushes =
and
stops building seems to be in module responsible for SFTP. Do you have an=
y idea
where is the problem?


Best wishes,=20
Juris Treibahs

Re: [Hypercontent-users] Running hypercontent in uP2-1-4

[Alex V. <av...@co...>]

Well, you would probably need something like

<xsl:variable name="navigation" 
select="/cms:wrapper/child::*/navigation"/>
<xsl:variable name="sourceLoc" select="/cms:wrapper/cms:source/@path"/>

<xsl:template name="breadcrumb">
	<!-- by default, grab the topmost nav-item which is an ancestor of the 
nav-item corresponding to the current page -->
	<xsl:param name="item" 
select="$navigation/nav-item[//nav-item[@location=$sourceLoc]]"/>
	<!-- presumes a template which formats a link from a nav-item -->
	<xsl:call-template name="nav-item-link:>
		<xsl:with-param name="item" select="$item"/>
	</xsl:call-template>
	<xsl:if test="not($item/@location=$sourceLoc)">
		<!-- if we haven't reached the source nav-item yet, recurse -->
		<xsl:text> &gt; </xsl:text>
		<xsl:call-template name="breadcrumb">
			<!--grab the next ancestor nav-item -->
			<xsl:with-param name="item" 
select="$item/nav-item[//nav-item[@location=$sourceLoc]]"/>
		</xsl:call-template>
	</xsl:if>
</xsl:template>

On Sep 16, 2004, at 4:32 AM, Carl Barrow wrote:

> I am using a navigation document to build navigation....
>
> Thank
> Carl
>
> Alex Vigdor wrote:
>
>> HyperContent runs fine in uPortal 2.1.4 - that's what we use here at  
>> Columbia as well.  The same build script works fine to deploy it, as  
>> long as you configure the directories in build.properties  
>> appropriately.
>>
>> Setting up a bread crumb trail depends on how your site is configured 
>> -  are you using a navigation document (as described at  
>> http://hypercontent.sourceforge.net/docs/manual/develop/ 
>> navigation.html), or just using includes to build navigation?  If you 
>>  use the navigation editor, you can just walk up the navigation 
>> document  from the node corresponding to the current page.  If you're 
>> using  includes, you would have to define includes for the ancestor 
>> pages  (e.g. ../index.xml, ../../index.xml), and then look them up by 
>> pattern  in your XSL.  Let me know which way your site is configured, 
>> and  perhaps we can get Adam Carl to provide some sample code from 
>> one of  our sites.
>>
>> Alex
>>
>>
>> On Sep 15, 2004, at 7:34 AM, Carl Barrow wrote:
>>
>>> Hi Alex
>>>
>>> More questions......
>>>
>>> Now I have HyperContent up and running here at Hull on a test server 
>>> I  need to get some people using it.  To do this I'm probably going 
>>> to  have to run it though uPortal_2-1-4.  Is this easy to do?  Do 
>>> you have  any specific instructions to help out?
>>>
>>> You said you might be able to provide some code for setting up a 
>>> bread  crumb trail.  That would be very helpful.
>>>
>>> Thanks for all your help with this.
>>>
>>> Carl
>>>
>>> -- 
>>> ************************************
>>>
>>> Carl Barrow
>>> Web Developer
>>> e-Services
>>> The University of Hull
>>> Cottingham Road
>>> Hull
>>> HU6 7RX
>>> Ext. 6838
>>> ************************************
>>>
>>>
>>>
>>>
>>> -------------------------------------------------------
>>> This SF.Net email is sponsored by: thawte's Crypto Challenge Vl
>>> Crack the code and win a Sony DCRHC40 MiniDV Digital Handycam
>>> Camcorder. More prizes in the weekly Lunch Hour Challenge.
>>> Sign up NOW http://ad.doubleclick.net/clk;10740251;10262165;m
>>> _______________________________________________
>>> Hypercontent-users mailing list
>>> Hyp...@li...
>>> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>>>
>>
>>
>>
>> -------------------------------------------------------
>> This SF.Net email is sponsored by: thawte's Crypto Challenge Vl
>> Crack the code and win a Sony DCRHC40 MiniDV Digital Handycam
>> Camcorder. More prizes in the weekly Lunch Hour Challenge.
>> Sign up NOW http://ad.doubleclick.net/clk;10740251;10262165;m
>> _______________________________________________
>> Hypercontent-users mailing list
>> Hyp...@li...
>> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>
>
> -- 
> ************************************
>
> Carl Barrow
> Web Developer
> e-Services
> The University of Hull
> Cottingham Road
> Hull
> HU6 7RX
> Ext. 6838
> ************************************
>

Re: [Hypercontent-users] Running hypercontent in uP2-1-4

[Alex V. <av...@co...>]

HyperContent runs fine in uPortal 2.1.4 - that's what we use here at  
Columbia as well.  The same build script works fine to deploy it, as  
long as you configure the directories in build.properties  
appropriately.

Setting up a bread crumb trail depends on how your site is configured -  
are you using a navigation document (as described at  
http://hypercontent.sourceforge.net/docs/manual/develop/ 
navigation.html), or just using includes to build navigation?  If you  
use the navigation editor, you can just walk up the navigation document  
from the node corresponding to the current page.  If you're using  
includes, you would have to define includes for the ancestor pages  
(e.g. ../index.xml, ../../index.xml), and then look them up by pattern  
in your XSL.  Let me know which way your site is configured, and  
perhaps we can get Adam Carl to provide some sample code from one of  
our sites.

Alex


On Sep 15, 2004, at 7:34 AM, Carl Barrow wrote:

> Hi Alex
>
> More questions......
>
> Now I have HyperContent up and running here at Hull on a test server I  
> need to get some people using it.  To do this I'm probably going to  
> have to run it though uPortal_2-1-4.  Is this easy to do?  Do you have  
> any specific instructions to help out?
>
> You said you might be able to provide some code for setting up a bread  
> crumb trail.  That would be very helpful.
>
> Thanks for all your help with this.
>
> Carl
>
> -- 
> ************************************
>
> Carl Barrow
> Web Developer
> e-Services
> The University of Hull
> Cottingham Road
> Hull
> HU6 7RX
> Ext. 6838
> ************************************
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: thawte's Crypto Challenge Vl
> Crack the code and win a Sony DCRHC40 MiniDV Digital Handycam
> Camcorder. More prizes in the weekly Lunch Hour Challenge.
> Sign up NOW http://ad.doubleclick.net/clk;10740251;10262165;m
> _______________________________________________
> Hypercontent-users mailing list
> Hyp...@li...
> https://lists.sourceforge.net/lists/listinfo/hypercontent-users
>