Language integration via xsltproc

[Joerg Kurt Wegner]

I just got the git package and use
xsltproc.exe --xinclude fvx-html.xsl foaf.xml > foaf.html

Still, I get only some of the FOAF entries, and especially the
language dependency of the field names is not resolved properly.

I do not see why how the language packages in the subdirectory i18n
are integrated. Has anyone every figured that out?

Finally, is anyone interested helping me making some modifications in
a way that we all could benefit from them?

[Wojciech Polak]

On Aug 2, 10:53 pm, Joerg Kurt Wegner <joergkurtweg...@gmail.com>
wrote:

> I just got the git package and use
> xsltproc.exe --xinclude fvx-html.xsl foaf.xml > foaf.html
> Still, I get only some of the FOAF entries, and especially the
> language dependency of the field names is not resolved properly.
> I do not see why how the language packages in the subdirectory i18n
> are integrated. Has anyone every figured that out?

They are resolved by a script which I haven't committed yet
into the repo.

> Finally, is anyone interested helping me making some modifications in
> a way that we all could benefit from them?

I am :) What would you like to improve particularly?

Regards,
Wojciech

[Joerg Kurt Wegner]

Wojciech, great !
CC: Mark

There are three development options I am personally highly interested
in

1.) More information gathering and transparency on fvx: Getting more
clarity on the XSL details, since I think that fvx is potentially the
best designed FOAF visualization out there. I contributed a while to
FOAFER, and its a long way to realize what I would like to get with
it, besides the Python/GoogleAppEngine solution is less generic than
fvx XSL.

2.a) There are various minor things I would like to get (mainly some
stuff I find useful in FOAFER). One thing I would really find useful
is getting an activity stream for some of my accounts, potentially
using XSL as well (fastest?). RSS for accounts is part of FOAFER and
the sorting is described here
http://kpumuk.info/development/sorting-rss-feed-by-date-using-xslt/

2.b) I really would love to merge the information of my personal FOAF
http://github.com/joergkurtwegner/snippets/blob/master/xml/foaf.xml
with e.g. the information automatically gathered by http://semantictweet.com
and Steve will look into this option, too
http://twitter.com/semantictweet/status/19268737533
BTW, they do not have a visualization option, so maybe this alone
would be a fantastic collaboration option.

3.) I just joined http://www.science3point0.com as community manager,
Mark if hosting the platform which is based on PHP and we would like
to create FOAF profiles for each user. The key is not only creating
them based on the platform information, but especially finding an
elegant way integrating the FOAF visualization into StackOverFlow,
BTW, the science3point0 platform was just open sourced.
http://github.com/mhahnel/science3point0
On that note, I have >10 years Java experience, and speak various
other dialects, e.g. Python, XML, and XSL, but I must admit not in the
detail you do *bow*. Anyway, I do not know to much about PHP by now,
we have a technical role for the platform filled with a developer,
though he also needs to get started with PHP. If you can recommend
anyone, which would be willingly and supportive along the things
discussed, we would be glad seeing new team and board members to join
the platform. Since semantic and FOAF are *big* in science communities
we are looking forward in getting this done.
Finally, we are also looking into collaboration with OSF (the Open
Science Foundation), which is using more or less the same platform, so
I want to highlight the impact options here, which are *ehm nice*;-)

Thanks for listening, and looking forward to your answer.

Cheers, Joerg
http://www.joergkurtwegner.eu/

[Wojciech Polak]

On Aug 3, 11:45 am, Joerg Kurt Wegner <joergkurtweg...@gmail.com>
wrote:

> 1.) More information gathering and transparency on fvx: Getting more
> clarity on the XSL details, since I think that fvx is potentially the
> best designed FOAF visualization out there.

Well, I won't write any docs about the XSLT details due to my time
limits now, but I'll try to answer any questions about it,
asked here.

> 2.a) There are various minor things I would like to get (mainly some
> stuff I find useful in FOAFER). One thing I would really find useful
> is getting an activity stream for some of my accounts, potentially
> using XSL as well (fastest?).

I don't plan to write any activity stream visualizator within FOAF
format. There is already one existing, provided by a friendly project,
at http://foaf.me/ You might be interested in it.

Anyway, XSLT is fast, but it's not that important. Gathering your
accounts streams from different sources slows down everything
and mixing them into one stream is crucial. Visualizing different
formats with XSLT can be painful. I once wrote a quite big and complex
XSLT for all kinds of webfeeds. It's the core part of the Cheetah News
project (GPL licensed) and works very well, but today I wouldn't
repeat that, I guess. I would choose Python and Feedparser.

> 3.) I just joined http://www.science3point0.comas community manager,

> Mark if hosting the platform which is based on PHP and we would like
> to create FOAF profiles for each user. The key is not only creating
> them based on the platform information, but especially finding an
> elegant way integrating the FOAF visualization into StackOverFlow,

Regarding FOAF.Vix: It's possible to use it in an embedded mode,
on any site.

Regards,
Wojciech

[Joerg Kurt Wegner]

The foaf.me activity stream is extremely slow and not correct pulling
the wrong streams, so this is not an option. I also was not interested
in pulling all the content, just the first x titles of the streams
allowing users to click-through. I will have to think about this.

> Well, I won't write any docs about the XSLT details due to my time
> limits now, but I'll try to answer any questions about it,
> asked here.

Do you have any solution I can use locally for playing around a bit.
The web-service is great for integrating it, but it is of little use
for poking around a bit with the XSL.

I would really appreciate getting a local sandbox I can use, since I
like as open source'ler the "running code" paradigm ;-)
Asking questions takes longer and it less collaborative, besides,
digging into your XSL without getting it running myself is not too
much fun.

Cheers, Joerg