refdb-users — List for RefDB users

[Refdb-users] Further citation styles available ?

[Marc B. <ba...@sm...>]

Hi,

now that I start getting close to production level
with refdb, I wonder, whether there are more citation
styles available than 

Eur.J.Pharmacol.xml
J.Biol.Chem.xml
bibtex-abbrev.xml
bibtex-full.xml

and whether there is some kind of central site to
collect those styles (and deposit new ones, should
I happen to write one) ? 

Cheers,
  Marc

-- 
 Dr. Marc Baaden  - Institut de Biologie Physico-Chimique, Paris
 mailto:ba...@sm...      -      http://www.marc-baaden.de
 FAX: +49 697912 39550  -  Tel: +33 15841 5176 ou +33 609 843217

Re: [Refdb-users] Refdb/SGML part II

[Marc B. <ba...@sm...>]

Hi Markus,

as I still had refdb 0.9.1 installed, I switched to 0.9.2a after
your email, to see whether it would improve things. Unfortunately
I still have more or less the same problems. Let me give you some
details:


CITING VIA CITATION KEY

>   please check again the syntax for using citation keys. Use something
>   like:
>   <citation role="REFDB">miller1999;jones2001</citation>
>   <citation role="REFDB">A:jones2001</citation>
>   i.e. leave out the "ID" from citation keys. Please check whether this
>   cures the multiple citation issue as well. If not, please provide an
>   example document and the style you use.

The citation via citation keys still doesn't work. Find attached the
very simple document and corresponding Makefile using the standard refdb
J.Biol.Chem. citation style, that generates the following error:

/usr/bin/openjade:t.sgml:8:68:X: reference to non-existent ID "IDWANG2002X"

The ID actually comes from the refdb refdbxp program, which processed my 
entry to "<citation role="REFDB"><xref linkend="IDwang2002A"/></citation>."
But even if I remove the leading ID string, it still doesn't work. Any ideas
what's wrong ? (anything to do with the fact that I use docbook 31 and sgml ?)


SGML CATALOG INCLUSION

 >>> I had to add a line "CATALOG /etc/sgml/catalog" to
 >>> /usr/local/share/refdb/dsssl/catalog to get the Public identifiers
 >>> mapped correctly.

>   This is a bit surprising as the Debian Jade/Openjade binaries
>   automatically use the standard catalog /etc/sgml/catalog. [..]

Same problem after 0.9.2a install, the error messages are:
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:1:72:W: cannot generate system identifier for public text "-//James Clark//DTD DSSSL Style Sheet//EN"
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:1:72:E: reference to entity "STYLE-SHEET" for which no system identifier could be generated
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:1:0: entity was defined here
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:1:72:E: DTD did not contain element declaration for document type name
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:3:12:E: element "STYLE-SHEET" undefined
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:4:20:E: element "STYLE-SPECIFICATION" undefined
/usr/bin/openjade:/usr/local/share/refdb/dsssl/citations.dsl:5:25:E: element "STYLE-SPECIFICATION-BODY" undefined
/usr/bin/openjade:E: specification document does not have the DSSSL architecture as a base architecture


OPENJADE WEIRDNESS

 >>> I also had to replace the 'sgmldecl="docbook.dcl" ' line by
 >>> 'sgmldecl=' in /usr/local/bin/runbib and in /usr/local/bin/refdbjade

>   This makes me suspicious. What version do you run?

Indeed, I don't seem to have this with 0.9.2a any more.
But now I have another problem (which has a simple workaround):
refdbjade uses opnejade by default, now, and it hangs indefinitely
when processing a document. Using jade instead works fine. But
maybe I haven't properly setup openjade ?? I'll have to check ..


CUSTOMIZING STYLES

 >>> 3) Other languages than english ?
 >>> As I am currently writing a text in French, I'd like to have the
 >>> "Reference List" text replaced with something in French (for instance
 >>> "Bibliographie" or so ..). How do I go about this ?

>   You'd have to edit the XML files that describe the bibliography
>   styles. Customize them as you see fit, change the style names, and load
>   them with the refdba:addstyle command.

Hmm. Editing the XML file was fine. But then when I try to use it, I get
runbib -d mainbib -S J.Biol.Chem.MBfr. -t db31 t.sgml
getbib error: could not access format information:

but when I rename it to something simpler, eg test or so, it works.
Is there anmy restriction on the filename of the xml file ??
(Actually with the "problematic" files, addstyle yields
0  dataset(s) added, 0 failed instead of 1 dataset added, but the
style nevertheless appears with liststyles ..)


CUSTOMIZING DOCBOOK BEHAVIOUR

>   [..] provide a private driver file with just your
>   personal customizations. [..] In addition you'll have to accomplish that the RefDB
>   stylesheets (print/docbook-refdb.dsl and html/docbook-refdb.dsl) pull
>   in your private stylesheet instead of the official docbook.dsl
>   version.

Well, actually I went the other way round. I took the dsl file generated by
refdb as template, renamed it, referred to that, and added my customization
to it. This seems to work fine :))


Regards,
Marc

NB: and sorry for being such a pain and having so many questions ...

Re: [Refdb-users] php, cgi, jsp and refdb?

[Bruce D'A. <bd...@fa...>]

On Sunday, March 30, 2003, at 01:23  AM, Alan Anderson wrote:

>> What it means from my Mac OS X user perspective is that applications
>> that support Rendezvous have automatic access to services on the
>> network.  If I open my web browser bookmark window, I have a list of
>> Rendezvous servers.  If I open my ftp or chat clients, the same thing.
>> So I naturally just thought it might be really use to have this in bib
>> database clients.
>
> This certainly sounds like an attainable goal. It might be important 
> to note, that it seems like Rendezvous is NOT a protocol for 
> communication with a service, but rather a protocol for service 
> discovery. Once discovered, the client will still (I assume) 
> communicate with the server using the typical protocol (FTP, HTTP, or 
> whatever).

Yes; I think you've got it.

> Definitely sounds interesting--wish I still had my PowerMac. ;)

Ah, but the source for Rendezvous is available, and is apparently plain 
C.  Of course, I doubt it has all of the client code for other 
platforms that OS X has (but who knows, maybe it does), but it still 
could be useful...

http://developer.apple.com/darwin/projects/rendezvous/

 From the site:

> If you have registered with this site, you can download the source 
> code components of Rendezvous:
>
> Multicast DNS Responder (mDNSResponder), the component that listens 
> for and responds to DNS-format query packets, sent via IP Multicast to 
> UDP port 5353.
>
> IP configuration (the bootp project, which also handles DHCP and 
> link-local addressing)
>
> The lower level C APIs to allow servers and clients to communicate 
> with the mdns subproject of Libinfo

Some guy put together an Apache module for this:

http://sourceforge.net/projects/modrendezvous/

Bruce

RE: [Refdb-users] php, cgi, jsp and refdb?

[Alan A. <al...@ru...>]

Bruce,

Thanks for the interesting discussion...

> > Again, I have not looked specifically at Apple's contribution, but
> > this sounds
> > exactly the same as J2EE and .Net, i.e. a web services framework.  So
> > other than
> > providing yet-another way to develop web applications, I don't see
> > that Apple's
> > contribution is new or unique.
>
> I'm getting out of my realm of understanding here, which is part of the
> reason I asked the question, but IF I understand any of this correctly,
> Rendezvous is not equivalent to a web services API, but is simply a
> networking protocol.  A couple links:
>
> http://www.macdevcenter.com/pub/a/mac/2002/11/08/cocoa.html
> http://www.macdevcenter.com/pub/a/mac/2002/11/15/cocoa.html
> http://www.zeroconf.org/

You are correct that it is not equivalent to a web services API, although it proposes to do similar
things.  I didn't get to read too much about it, but it is an enhancement to current protocols (DNS,
ARP) to simplify network administration.  Definitely sounds interesting--wish I still had my
PowerMac. ;)


> What it means from my Mac OS X user perspective is that applications
> that support Rendezvous have automatic access to services on the
> network.  If I open my web browser bookmark window, I have a list of
> Rendezvous servers.  If I open my ftp or chat clients, the same thing.
> So I naturally just thought it might be really use to have this in bib
> database clients.

This certainly sounds like an attainable goal.  It might be important to note, that it seems like
Rendezvous is NOT a protocol for communication with a service, but rather a protocol for service
discovery.  Once discovered, the client will still (I assume) communicate with the server using the
typical protocol (FTP, HTTP, or whatever).  I would guess that this means the bib client might
support Rendezvous, but it will still need to be implemented using whatever protocol is chosen--in
the case of RefDB, this is the internal refdb protocol or whatever Markus calls it.


> In any case, as I say, I don't fully understand this stuff, so am happy
> to let others more knowledgeable correct me!

I certainly don't profess to be more knowledgeable than anyone, but I certainly enjoy learning new
stuff.  Thanks.

Al

RE: [Refdb-users] php, cgi, jsp and refdb?

[Alan A. <al...@ru...>]

Markus,

Thanks again for the good work; I'll have to look at the Perl module when I find some time.  I'm
attaching my Perl code, so others can critique it.  It is mostly working. It adds, deletes, updates,
and queries references.  The extra field choices and personal list options are not implemented--I
think these should be trivial.  Also, the "search for keywords, authors, and journals" feature has
not been implemented.

I'm also attaching an RIS DTD that I put together last week.  It is my FIRST attempt at a DTD, but I
needed to learn it as part of another project, so it seemed like a easy starting point.  It
validates a couple of small test files (also attached).  I was hoping it would provide for easy
input using jEdit's XML module (reference a previous thread about using jEdit (I thought using a DTD
might be an easy way make reference creation a little easier using jEdit), but for some reason
nothing beyond the "ris" tag is available--probably something dumb on my part.  I was hoping to test
it more, but since you mentioned that you were working on it, I thought you might find it useful.

Hope this helps.

Al

RIS DTD Files:
	ris.dtd -- the DTD
	ristest.xml -- test file for a single record
	multitest.xml -- test file for multiple records

Perl Files:
	index.cgi -- application instance script-this is the one accessed by the client
	refdb.tmpl -- template file for the pages
	RefDB_CGI.pm -- the module containing my Perl application

A few notes on the use of this module...

First, you need to have CGI::Application module installed (I think this one requires CGI module),
and you'll need the HTML::Template module installed (I think this is embedded into the newer
CGI::Application modules, but currently I don't take advantage of this).

Second, this app uses the same cookie as the standard RefDB CGI application, so there is no need to
worry about running the two side by side (I did this a lot during initial development).

Third, you need to have your web server configured correctly, but I won't go into that now.

Fourth, I have only addressed a few of reference types in the interactive forms (BOOK, CHAP, CONF,
INPR, JFULL, JOUR, PCOMM, RPRT, UNPB) and of those only BOOK, CONF, and JOUR have their forms
defined--it is relatively easy to add new forms, so if anyone wants to add a few, I'd be happy to
explain how to do it.

Fifth, the "Enter raw format" input option only works with RIS right now, and the BibXML option
should probably be removed.

Sixth, and hopefully last, the cell containing the username and database isn't updated immediately
after logging in for the first time, so please ignore it as it will show up as soon as the screen is
refreshed.


> -----Original Message-----
> From: ref...@li...
> [mailto:ref...@li...]On Behalf Of Markus
> Hoenicka
> Sent: Friday, March 28, 2003 4:45 PM
> To: ref...@li...
> Subject: Re: [Refdb-users] php, cgi, jsp and refdb?
>
>
> Hi,
>
> I'd be happy to see something like a web service grow around
> RefDB. Currently this stuff is above my head, but I'm willing to get
> into it if I can contribute something useful.
>
> On a much smaller level I'd like to point out that I've finished a
> first draft of a Perl client module (currently available in CVS
> only). This allows to develop Perl scripts (which in turn can be used
> as CGI applications if you wish) that directly access the refdbd
> server. I'm currently adding a RIS-based XML data format that will be
> well suited for data mangling in Perl.
>
> Similar client libraries could be written in next to any language. I'm
> pondering an Emacs Lisp implementation, just for the heck of it.
>
> regards,
> Markus
>
> --
> Markus Hoenicka
> mar...@ca...
> (Spam-protected email: replace the quadrupeds with "mhoenicka")
> http://www.mhoenicka.de
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by:
> The Definitive IT and Networking Event. Be There!
> NetWorld+Interop Las Vegas 2003 -- Register today!
> http://ads.sourceforge.net/cgi-bin/redirect.pl?keyn0001en
> _______________________________________________
> Refdb-users mailing list
> Ref...@li...
> https://lists.sourceforge.net/lists/listinfo/refdb-users
>
>
> ---
> Incoming mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.455 / Virus Database: 255 - Release Date: 2/13/2003
>

[Refdb-users] General Stylesheet customizations with refdb/docbook ?

[Markus H. <mar...@mh...>]

Hi Marc,

this goes pretty deep into the innards of SGML. The easiest, though
not recommended way to fix this is to edit the dbparam.dsl file of the
DocBook stylesheets. It is not recommended because you'd have to apply
these changes everytime you update the stylesheets.

A better way is to provide a private driver file with just your
personal customizations. Instructions to write driver files are
available in the online version of "DocBook - The definitive guide" by
Norm Walsh. In addition you'll have to accomplish that the RefDB
stylesheets (print/docbook-refdb.dsl and html/docbook-refdb.dsl) pull
in your private stylesheet instead of the official docbook.dsl
version. If you want to go this way, I'd require some time to fiddle
on a spare box to work out more specific instructions.

regards,
Markus

Marc Baaden writes:
 > 
 > Hi,
 > 
 > another one .. if I want to customize something for the stylesheet,
 > let's say eg I want a justified output, which you usually get by
 > adding (define %default-quadding% 'justify) to a customization
 > stylesheet, where/what do I need to modify ?
 > 
 > Should I modify the bibliography style in the current directory
 > (eg J.Biol.Chem.dsl -- which I tried to no avail) or do I need
 > to give any specific options to eg refdbjade, or indicate a second
 > dsl file ??
 > 
 > Thanks again,
 >   Marc 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] Refdb/SGML part II

[Markus H. <mar...@mh...>]

Hi again,

please check again the syntax for using citation keys. Use something
like:

<citation role="REFDB">miller1999;jones2001</citation>
<citation role="REFDB">A:jones2001</citation>

i.e. leave out the "ID" from citation keys. Please check whether this
cures the multiple citation issue as well. If not, please provide an
example document and the style you use.

regards,
Markus

Marc Baaden writes:
 > 
 > Hi,
 > 
 > I was too optimistic, I ran into further problems:
 > so far I used ID numbers to cite, which is fine.
 > But Key's don't work. The error I get is
 > 
 > [..]
 > /usr/bin/jade:rapport.sgml:16:117:X: reference to non-existent ID "ID1X"
 > /usr/bin/jade:rapport.sgml:15:63:X: reference to non-existent ID "IDWANG2002X"
 > [..]
 > server error: database error while reading xml file:
 > 0 reference(s) formatted, 0 failed
 > 
 > 
 > As you might have guessed, the keyword I was trying to cite is Wang2002,
 > and it exists:
 > ID*:1 (2002)
 > Key: Wang2002
 > [..]
 > 
 > Also, the Y: and Q: modifiers seem not to work. For Q: the error is
 > [..]
 > /usr/bin/jade:rapport.sgml:15:63:X: reference to non-existent ID "IDQ:2X"
 > [..]
 > server error: database error while reading xml file:
 > 0 reference(s) formatted, 0 failed
 > 
 > and with Y: it passes, but doesn't show the year.
 > 
 > There is also a slight problem with multiple citations. If I cite 1,2,3
 > and then 1,4,5 it shows as 1-3 (ok) and 1;45 (missing "-" between 4 and 5).
 > 
 > 
 > But for now the most pressing thing for me would be to get the cite by
 > key rather than ID number going. 
 > 
 > Thanks in advance,
 >   Marc 
 > 
 > 
 > -------------------------------------------------------
 > This SF.net email is sponsored by:
 > The Definitive IT and Networking Event. Be There!
 > NetWorld+Interop Las Vegas 2003 -- Register today!
 > http://ads.sourceforge.net/cgi-bin/redirect.pl?keyn0001en
 > _______________________________________________
 > Refdb-users mailing list
 > Ref...@li...
 > https://lists.sourceforge.net/lists/listinfo/refdb-users
 > 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] refdb and SGML/docbook issues

[Markus H. <mar...@mh...>]

Hi Marc,

Marc Baaden writes:
 > I had to add a line "CATALOG /etc/sgml/catalog" to
 > /usr/local/share/refdb/dsssl/catalog to get the Public identifiers
 > mapped correctly.
 > 

This is a bit surprising as the Debian Jade/Openjade binaries
automatically use the standard catalog /etc/sgml/catalog. In any case,
if this should turn out to be necessary (I'll be able to test this in
a week or two - got a used bargain box for testing), I'd do it the
other way round - include the RefDB catalog in the master catalog.

 > I also had to replace the 'sgmldecl="docbook.dcl" ' line by
 > 'sgmldecl=' in /usr/local/bin/runbib and in /usr/local/bin/refdbjade
 > 

This makes me suspicious. What version do you run? 0.9.2 uses its own
copies of the declaration files unless you specify different paths on
the ./configure command line.

 > Now it seems to work fine.
 > 
 > 2) First time error
 > 
 > I guess it is normal/OK that in the first pass, when the *.bib.sgml
 > file does not yet exist, there is an error message about this ?
 > 
 > /usr/bin/jade:rapport.sgml:168:0:E: cannot find "rapport.bib.sgml"; [..]
 > 
 > as well as the "errors" for each citation reference
 > 
 > /usr/bin/jade:rapport.sgml:16:66:X: reference to non-existent ID[..]
 > 
 > Or am I missing something ?
 > 

No, this is expected (and described in the manual). You're
bootstrapping, so neither the bib file nor the IDs it contains exist
during the first run.

 > 3) Other languages than english ?
 > 
 > As I am currently writing a text in French, I'd like to have the
 > "Reference List" text replaced with something in French (for instance
 > "Bibliographie" or so ..). How do I go about this ?
 > 

You'd have to edit the XML files that describe the bibliography
styles. Customize them as you see fit, change the style names, and load
them with the refdba:addstyle command.

 > 4) Multiple entities
 > 
 > I have tried to follow the hint in the documentation about using
 > sgmlnorm for documents with multiple entities. It works, but my
 > solution seems particularly ugly .. maybe I did overlook something ?
 > I actually use a custom header with Docbook ref and inclusion of
 > bibliography, then paste the output of sgmlnorm to that, then
 > add a &bibliography reference to the entity before the final
 > </book>,</article>  or whatever.
 > 
 > Is there a better way ?
 > 

There is a better way, but I can't provide it at the moment. The
solution would be to rewrite refdbxp based on the OpenSP library and
make it fully SGML-aware. Unfortunately I'm not much of a C++
programmer, so this project is pretty close to the bottom of my todo
list. I'm afraid we'll have to live with half-assed solutions for the
time being.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] General Stylesheet customizations with refdb/docbook ?

[Marc B. <ba...@sm...>]

Hi,

another one .. if I want to customize something for the stylesheet,
let's say eg I want a justified output, which you usually get by
adding (define %default-quadding% 'justify) to a customization
stylesheet, where/what do I need to modify ?

Should I modify the bibliography style in the current directory
(eg J.Biol.Chem.dsl -- which I tried to no avail) or do I need
to give any specific options to eg refdbjade, or indicate a second
dsl file ??

Thanks again,
  Marc 

[Refdb-users] Refdb/SGML part II

[Marc B. <ba...@sm...>]

Hi,

I was too optimistic, I ran into further problems:
so far I used ID numbers to cite, which is fine.
But Key's don't work. The error I get is

[..]
/usr/bin/jade:rapport.sgml:16:117:X: reference to non-existent ID "ID1X"
/usr/bin/jade:rapport.sgml:15:63:X: reference to non-existent ID "IDWANG2002X"
[..]
server error: database error while reading xml file:
0 reference(s) formatted, 0 failed


As you might have guessed, the keyword I was trying to cite is Wang2002,
and it exists:
ID*:1 (2002)
Key: Wang2002
[..]

Also, the Y: and Q: modifiers seem not to work. For Q: the error is
[..]
/usr/bin/jade:rapport.sgml:15:63:X: reference to non-existent ID "IDQ:2X"
[..]
server error: database error while reading xml file:
0 reference(s) formatted, 0 failed

and with Y: it passes, but doesn't show the year.

There is also a slight problem with multiple citations. If I cite 1,2,3
and then 1,4,5 it shows as 1-3 (ok) and 1;45 (missing "-" between 4 and 5).


But for now the most pressing thing for me would be to get the cite by
key rather than ID number going. 

Thanks in advance,
  Marc 

[Refdb-users] refdb and SGML/docbook issues

[Marc B. <ba...@sm...>]

Hi,

first of all, let me say that I am a real docbook/SGML beginner,
so I might say studip things or overlook something obvious.

1) Getting it to work

The last days I fiddled with getting refdb+docbooc+jade going on
my Debian woody system. For that I had to modify some things in
the refdb distribution (only minor changes, but when you don't
know where you need to interfere ..)

I had to add a line "CATALOG /etc/sgml/catalog" to
/usr/local/share/refdb/dsssl/catalog to get the Public identifiers
mapped correctly.

I also had to replace the 'sgmldecl="docbook.dcl" ' line by
'sgmldecl=' in /usr/local/bin/runbib and in /usr/local/bin/refdbjade

Now it seems to work fine.

2) First time error

I guess it is normal/OK that in the first pass, when the *.bib.sgml
file does not yet exist, there is an error message about this ?

/usr/bin/jade:rapport.sgml:168:0:E: cannot find "rapport.bib.sgml"; [..]

as well as the "errors" for each citation reference

/usr/bin/jade:rapport.sgml:16:66:X: reference to non-existent ID[..]

Or am I missing something ?

3) Other languages than english ?

As I am currently writing a text in French, I'd like to have the
"Reference List" text replaced with something in French (for instance
"Bibliographie" or so ..). How do I go about this ?

4) Multiple entities

I have tried to follow the hint in the documentation about using
sgmlnorm for documents with multiple entities. It works, but my
solution seems particularly ugly .. maybe I did overlook something ?
I actually use a custom header with Docbook ref and inclusion of
bibliography, then paste the output of sgmlnorm to that, then
add a &bibliography reference to the entity before the final
</book>,</article>  or whatever.

Is there a better way ?

I attach my slightly modified Makefile and header and footer as
example.

Thanks in advance for any hints,
  Marc Baaden

[Refdb-users] Re: Web Access Problems

[Daniel H. <da...@se...>]

Thanks Markus,

That seemed to help with the login issue.  Now am getting an error 
"Could not loag cgi templates.".  I think this is probably due to
my apache setup, which i will check when i have a little more time.


From: Markus Hoenicka <markus@mh...> 
Re: Web Access Problems   
2003-03-26 12:57  
 Daniel,
 
 thanks for the additional information. I think I've figured out what's
 going wrong, and it appears to be a stupid bug that happens to be
 silent on my systems.
 
 refdbc checks for a valid username before it even reads the CGI data
 from the web server. This is not a problem as long as your system
 provides a username via getlogin(), as the real username used in the
 queries is read afterwards from the CGI data. But your system
 apparently can't provide a username via getlogin() for processes
 running as CGI apps, so you're clearly screwed.
 
 The short-term fix is to provide a username entry in the
 /usr/local/etc/refdb/refdbcgirc configuration file. The name is not
 really important, there just needs to be one to bypass this bug.
 
 Let me know whether this helps on your system.
 
 regards,
 Markus
 
 Daniel Hyduke writes:
  > 
  > Thanks Markus.  I tried that and discovered that nothing happend when I
  > tried the web interface.  So I wrote a quick perl script to see if
  > something was happening with the username/password/database and cgi;
  > everything appeared to check out here.  Next, I modified the refdbc.c
  > code (line 3517) to see if I could print the username when the refdbc is
  > called.  Now when refdbc is called via cgi the error message prints:
  > 	'' is an incorrect username
  > 	incorrect username.
  > 
  > 3516  else if (!strcmp(varname, "username")) {
  > 3517    fprintf(stderr,"\ '%s' is an incorrect username\n",username); //daniel's Code
  > 3518    if (!username[0] && getlogin()) {
  > 
  > As I know very very little about CGI and what is happening in refdbc.c, I would
appreciate any input.
 
 -- 
 Markus Hoenicka
 markus.hoenicka@ca...
 (Spam-protected email: replace the quadrupeds with "mhoenicka")
 http://www.mhoenicka.de
 

Re: [Refdb-users] php, cgi, jsp and refdb?

[Markus H. <mar...@mh...>]

Hi,

I'd be happy to see something like a web service grow around
RefDB. Currently this stuff is above my head, but I'm willing to get
into it if I can contribute something useful.

On a much smaller level I'd like to point out that I've finished a
first draft of a Perl client module (currently available in CVS
only). This allows to develop Perl scripts (which in turn can be used
as CGI applications if you wish) that directly access the refdbd
server. I'm currently adding a RIS-based XML data format that will be
well suited for data mangling in Perl.

Similar client libraries could be written in next to any language. I'm
pondering an Emacs Lisp implementation, just for the heck of it.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] RIS titles, etc.

[Markus H. <mar...@mh...>]

Hi Marc,

I quickly glanced over this document, and the appendix looks pretty
helpful to me. I'll wade through these tables early next week and see
whether they mandate changes in the current RefDB code.

regards,
Markus

Mar...@en... writes:
 > Google found this RIS guide for me: it documents how this university has
 > standardized the use of RIS fields, including T1,T2,T3
 > 
 > <http://www.msd.ucl.ac.uk/demist/pdfdocs/rpsguide.pdf>
 > 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] php, cgi, jsp and refdb?

[Bruce D'A. <bd...@fa...>]

On Friday, March 28, 2003, at 01:40  PM, al...@ru... wrote:

> Again, I have not looked specifically at Apple's contribution, but 
> this sounds
> exactly the same as J2EE and .Net, i.e. a web services framework.  So 
> other than
> providing yet-another way to develop web applications, I don't see 
> that Apple's
> contribution is new or unique.

I'm getting out of my realm of understanding here, which is part of the 
reason I asked the question, but IF I understand any of this correctly, 
Rendezvous is not equivalent to a web services API, but is simply a 
networking protocol.  A couple links:

http://www.macdevcenter.com/pub/a/mac/2002/11/08/cocoa.html
http://www.macdevcenter.com/pub/a/mac/2002/11/15/cocoa.html
http://www.zeroconf.org/

An excerpt from the latter (on which Rendezvous is based):

> The charter of the Zeroconf Working Group is to enable Zero 
> Configuration IP Networking. That means making it possible to take two 
> laptop computers, and connect them with a crossover Ethernet cable, 
> and have them communicate usefully using IP, without needing a man in 
> a white lab coat to set it all up for you. We're not limiting the 
> network to just two hosts, but we want to take that as the starting 
> point.

[...]

> It is important to understand that the purpose of Zeroconf is not 
> solely to make current personal computer networking easier to use, 
> though this is certainly a useful benefit. The long-term goal of 
> Zeroconf is to enable the creation of entirely new kinds of networked 
> products, products that today would simply not be commercially viable 
> because of the inconvenience and support costs involved in setting up, 
> configuring, and maintaining a network to allow them to operate.

What it means from my Mac OS X user perspective is that applications 
that support Rendezvous have automatic access to services on the 
network.  If I open my web browser bookmark window, I have a list of 
Rendezvous servers.  If I open my ftp or chat clients, the same thing.  
So I naturally just thought it might be really use to have this in bib 
database clients.

In any case, as I say, I don't fully understand this stuff, so am happy 
to let others more knowledgeable correct me!

Bruce

Re: [Refdb-users] php, cgi, jsp and refdb?

[<al...@ru...>]

Let me start by saying that I am NOT an expert in web services, but I have
looked into them several times, and I have drawn some of my own conclusions
based on what I have read and understood.


> The current RefDB web client uses CGI, which Markus and others have 
> suggested is limited, PHP being preferred.  So my question: why?  What 
> does PHP allow that CGI doesn't?
> 
> And what about JSP?

CGI is just the interface between a web browser and the application running on
the server, and it is communicated through the HTTP protocol.  Whether the
server application (in this case RefDB) is written in C, C++, Perl, PHP, JSP,
FORTRAN, Basic, Java, or anything else is pretty much a matter of choice by the
developer.  In the end they all have to send and receive their information in
the same way--using CGI.

There is, as I understand it, nothing special about JSP in terms of what it can
and cannot do WRT server applications--the difference is that the web pages are
generated using Java code instead of Perl, PHP, or something else.

I've been watching this list for a while and contributing when I can, but I do
not recall a concensus on PHP being preferred.  I know someone was working on a
PHP client for a specialized application (Howard, maybe?), but there has also
been work on a Perl client.  What PHP provides is a relatively easy to learn
language that provides some OOP and caters to web development and database
access (along with some cross-platform compatibility).

I have more comments regarding the client below...


> The question is prompted by the release of this free GUI front-end to 
> Tomcat:
> 
> http://www.micromata.de/en/M_tec/schroedinger/index.html

Here we add a little bit of a twist when you start to look at application
servers.  Tomcat is an application server that supports J2EE.  J2EE is a web
services framework that defines a way to advertise services to allow the
objectification (in an OOP sense) and simplification of development through the
reuse of code and services.  There are lots of application servers out there,
Tomcat is an open source one developed by the Apache group (IBM, Oracle, Sun,
and many others have their own versions).  I have not had a chance to see what
Schroedinger is doing, but it sounds like they are trying to simplify
adminstration and development when using Tomcat.  Sun already offers the
Community Edition of Forte for Java which integrates with "their"
development-level application server (it is actually Tomcat).  Forte provides an
IDE for developing web applications in the J2EE environment.

I should note a small contradiction to a statement above.  When JSP is used in
this sense, it has access to the J2EE functionality, so there are additional
services that JSP can provide.  In the end, J2EE is just a mechanism for
distributing services in a multi-tier architecture.  PHP/Perl communicating with
a database is effectively a multi-tier architecture also.

 
> While I'm at it, I also have a vague sense that the new Rendezvous 
> networking protocol that Apple has adopted, and which will be (or maybe 
> already has been) open sourced, might have some relevance to refdb:
> 
> http://developer.apple.com/macosx/rendezvous/
> 
> One of the pieces of sample code here is described as follows:
> 
> > This sample demonstrates how to use NSNetServices to advertise a 
> > service using Rendezvous. When used in conjunction with the 
> > PictureSharingBrowser sample code, PictureSharingBrowser clients will 
> > be able to discover and view pictures being shared by a PictureSharing 
> > server.
> 
> If I understand right, then, it ought to be possible for an institution 
> to set up a refdb server that advertised itself via Rendezvous, and 
> which any compatible device (desktop computer, wireless device, etc.) 
> would b able to auto-discover, with zero configuration.

Again, I have not looked specifically at Apple's contribution, but this sounds
exactly the same as J2EE and .Net, i.e. a web services framework.  So other than
providing yet-another way to develop web applications, I don't see that Apple's
contribution is new or unique.  I'd be willing to bet that if you looked under
the hood, it is using a form of J2EE, Corba, COM, RMI, SOAP, or someother
distributed architecture--in the end all web services are roughly the same, it's
just the details that are different.

To comment further on the development of clients in the vein of web services, I
believe there is a potential need for what you suggest in your last paragraph--a
web services implementation of the RefDB client.  I seem to recall another short
discussion about multi-tier architecture for RefDB, but the discussion stalled
around which tier would retain which part of the logic (couldn't find it in the
archive, so maybe it was a different discussion).  One of the difficulties will
be a need to support different web services implementations, SOAP (not yet
mentioned, but still frequently discussed), J2EE, .Net, XML-RPC (another simple
web-service framework).  Obviously, one could just choose a framework to start
and others could develop for different frameworks as required.

Another complication is the additional modules to be maintained.  The web
service, in effect, defines its own API, so you're maintaining the API, the web
service code, the client code, AND RefDB core code.  However, you do add the
ability to provide information to clients OTHER than web clients.  Once the web
service is defined, it would be "trivial" to create a Java, KDE, GTK, or other
client GUI to interact with the web service--no more web browsers required (note
that this may depend on which web service is chosen--J2EE is implicitly Java,
but SOAP is at a lower level, more of an RPC than web service, and can be used
from many different languages).  Also, the web service could be integrated into
a larger application, e.g. a departmental website or document managment
system--they only need to know the API presented by the web service.  The design
needs to be carefully considered though--if the logic is put in the wrong place,
it can be a nightmare to maintain.

I'm not sure I've cleared up anything for you.  I would certainly be interested
in helping (as much as possible given my limited free time) design and/or
develop a web services implementation of a RefDB client.

Hope this helps.

Al


-------------------------------------------------
This mail sent by http://webmail.rushmore.com

[Refdb-users] php, cgi, jsp and refdb?

[Bruce D'A. <da...@mu...>]

This is just a curious question for someone (me) not too up on web 
technologies:

The current RefDB web client uses CGI, which Markus and others have 
suggested is limited, PHP being preferred.  So my question: why?  What 
does PHP allow that CGI doesn't?

And what about JSP?

The question is prompted by the release of this free GUI front-end to 
Tomcat:

http://www.micromata.de/en/M_tec/schroedinger/index.html

While I'm at it, I also have a vague sense that the new Rendezvous 
networking protocol that Apple has adopted, and which will be (or maybe 
already has been) open sourced, might have some relevance to refdb:

http://developer.apple.com/macosx/rendezvous/

One of the pieces of sample code here is described as follows:

> This sample demonstrates how to use NSNetServices to advertise a 
> service using Rendezvous. When used in conjunction with the 
> PictureSharingBrowser sample code, PictureSharingBrowser clients will 
> be able to discover and view pictures being shared by a PictureSharing 
> server.

If I understand right, then, it ought to be possible for an institution 
to set up a refdb server that advertised itself via Rendezvous, and 
which any compatible device (desktop computer, wireless device, etc.) 
would b able to auto-discover, with zero configuration.

Thoughts?

Bruce

Re: [Refdb-users] RIS titles, etc.

[Bruce D'A. <bd...@fa...>]

On Friday, March 28, 2003, at 05:38  AM, Mar...@en... wrote:

> Google found this RIS guide for me: it documents how this university 
> has
> standardized the use of RIS fields, including T1,T2,T3
>
> <http://www.msd.ucl.ac.uk/demist/pdfdocs/rpsguide.pdf>

Ah, the tables at the very end.  This is useful; thanks.  I'm unclear 
on how the "code" in the first column relates to the actual tags.  For 
example, code 11 (which is used for newspaper and magazine and journal 
titles) is what tag?

I just sent Markus a similar thing to show Endnote fields.  It might be 
worthwhile to figure out such a similar document for RefDB, to 
encourage best practices in data import/export.  I'll take a look and 
compare RefMan and Endnote. I suspect they're very similar, which would 
facilitate movement of data.

It's interesting to note, Markus, that recipient for PCOMM is indeed 
mapped to A2.

Bruce

[Refdb-users] RIS titles, etc.

[<Mar...@en...>]

On 27 Mar, ref...@li... wrote:
>  > 
>  > 1)  What RIS fields should magazine and newspaper titles (Newsweek, New 
>  > York Times, etc.) be mapped to?  I run into the same issue with things 
>  > like academic departments for theses.  In all of these cases, Endnote 
>  > maps this to the BT tag (aka T2).
>  > 
>  > 2)  BT should be T2?  A bit odd RIS has redundant tags like this.  But 
>  > RefDB does not support BT or other similar fields, right?
>  > 
> 
> RefDB recognizes all tags and synonyms as of Reference Manager
> 8.01. This is the most recent "specification" of RIS that I have
> access to. RIS is weird and ugly, no doubt.
> 
> The RIS spec is not entirely comprehensible to me with regard to the
> levels of bibliographic information. My guess is that TI/T1 is
> equivalent to the title at the analytic level (work in bibx speak),
> BT/T2 is equivalent to the title at the monographic level
> (publication), and T3 should then map to the series level (set). This
> is how RefDB currently deals with these tags.


Google found this RIS guide for me: it documents how this university has
standardized the use of RIS fields, including T1,T2,T3

<http://www.msd.ucl.ac.uk/demist/pdfdocs/rpsguide.pdf>

[Refdb-users] Re: Web Access Problems

[Markus H. <mar...@mh...>]

A quick follow-up. I guess you'll need to provide a dummy password as
well.

regards,
Markus
-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] How to Get 17,169 Visitors a Day to Any WebSite!

[<mar...@mi...>]

<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=windows-1252">
<META content="MSHTML 6.00.2800.1126" name=GENERATOR></HEAD>
<BODY>
<P align=center><B><FONT size=4>Massive Hits to your 
Website...</FONT></B><BR><IMG src="http://people.freenet.de/f1network/picture03.jpg"


<P> 
<P>
<CENTER><B><FONT face="MS Sans Serif" color=navy size=4>How to Get 17,169 
Visitors a Day to Any WebSite!</FONT></B> </CENTER><P>
<CENTER><B><FONT face=Arial color=#cc0000 size=3>Follow these Step-by-Step 
instructions! </FONT></B></CENTER>
<CENTER><B><FONT face=Arial color=#cc0000 size=4>&nbsp;</FONT><FONT face=Arial 
color=#000000 size=3>Starting Now, You can have a </FONT><FONT face=Arial 
color=#cc0000 size=3>phenomenal advertising machine </FONT><FONT face=Arial 
color=#000000 size=3>that will literally build </FONT><FONT face=Arial 
color=#cc0000 size=3>any </FONT><FONT face=Arial color=#000000 size=3>online 
business.</FONT></B><BR>
<P>
<P>
<CENTER>
<TABLE width=450>
  <TBODY>
  <TR>
    <CENTER>
    <P align=justify><B>Many people believe<FONT color=#0033cc> (There Is No 
    Such Thing As A Free Lunch)</FONT><BR>Guess what?<FONT color=#cc0000> THERE 
    IS!</FONT><BR></B>
    <P align=justify>And you do not need to outlay a single, solitary<B> dollar, 
    pound,<BR>punt, rand, mark or euro</B> for this plan to work for YOU!<BR>
    <P><B>And even better</B> -<B> you can earn while you learn!</FONT></B><BR>
    <P align=justify><B>MPAM (Massive Passive Advertising Machine)</B> is a<B> 
    Viral Marketing Strategy</B><BR>that out-performs anything you have 
seen.<BR>
    <P>It's all here for free<B>(which most people charge for)</B> and you will 
    learn how to combine the <B>SYNERGISTIC POTENTIAL</B> of over 50 FREE 
    programs</B> . 
    <P align=justify><B>MPAM</B> has had over <B>33 MILLION</B> page views since 
    it began in August 2001<BR>this means <B>EVERY SECOND</B> of <B>EVERY 
    DAY</B> for over <B>14 MONTHS</B><BR>someone somewhere has viewed an 
    <B>MPAM</B> page.<BR>
    <P><B>THIRTY THREE MILLION</B> page views!! And over <B>14 MILLION</B> 
    visitors! <B><BR>24 visitors EVERY MINUTE</B>, <B>EVERY DAY</B> for 
    over&nbsp; <B>14 MONTHS</B>!!</B>.<BR>Does the<B> MPAM</B> system work? No 
    hype needed - no fancy advertising <BR>just <B>FACTS</B> and <B>PROVEN 
    STATISTICS!</B> </P>
    <P>
    <P align=center><B><FONT size=4>Without MPAM traffic is 
    flat...</FONT></B><BR><IMG height=180 src="http://mryamamoto.topcities.com/pictmission/9w-jersey-freeway.gif" 
    width=300>
    <P>
    <P align=center><B><FONT size=4>With MPAM traffic is 
    Massive!</FONT></B><BR><IMG src="http://genesis.sannet.gov/infospc/templates/cd1/images/traffic.jpg" <BR>
    <P>
    <P align=justify>You won't be charged<B> ANYTHING</B> for the<B> <FONT 
    color=#cc0000>FREE TRAFFIC</B></FONT> lessons on this site.<BR><B>NO STRINGS 
    ATTACHED, EVER!</B><BR>
    <P>Your Success <B>Will</B> cost you <B>Time</B> and <B>Effort</B> - and you 
    <STRONG>MUST</STRONG> be willing to put in that <B>Effort!</B> </P>
    <P align=justify><B>And as an extra benefit</B> - not only is the 
    system<B><FONT color=#cc0000> FREE</B></FONT> but&nbsp;I will show <BR>You 
    how to link all systems together!</STRONG></FONT></P>
    <P align=justify>These lessons will teach you how to trigger a 
    <B>CHAIN-REACTION</B><BR>by linking these<B><FONT color=#cc0000> 
    FREE</B></FONT> programs together.<BR>
    <P>You will learn how to plug one program into another to create 
    a<BR><B>Self-Feeding,</B> <B>Never-Ending</B></FONT> stream of prospects 
    into an<BR><B><FONT color=#cc0000>Exponential Explosion of Hits</B></FONT> 
    to your website.<BR>Recruit <B>Thousands of affiliates</B>, make 
    <B>Thousands of sales!</B></P>
    <P align=justify>I'm talking about a <B>PERPETUAL-MOTION HIT 
    MACHINE</B>.<BR>You cannot comprehend it until you study the entire SERIES 
    of LESSONS.<BR>And in the end, you will be convinced that nobody anywhere 
    can promise You more hits in the years ahead.</P>
   
    <P>
    <DIV align=right></DIV></P></TR></TBODY></TABLE>
<TABLE height=20 width="98%" border=0>
  <TBODY>
  <TR>
    <TD width="60%">
      <P align=center><FONT color=#0033cc><B>You can have as much as<B><FONT 
      color=#cc0000> 60%</B></FONT> increase every month.<BR>Check out the chart 
      at right and watch this...&nbsp;</FONT></B></P>
      <P align=justify>If you get only a <B>TWO PERCENT INCREASE</B> in hits 
      daily and funnel them correctly,<BR>you can produce approximately <B>SIXTY 
      PERCENT INCREASE</B> in hits to your website<BR><B>Every Month!</B> </P>
      <P>This system<B> CAN ABSOLUTELY</B> give you that<B> TWO PERCENT INCREASE 
      DAILY!</B> </P>
      <P><B><FONT size=3>Do the calculations on<FONT color=#cc0000> 60%</FONT> 
      monthly growth...</FONT></B> 
      <P>This <B>Realistic Approximation</B> shows how fast your website can 
      begin receiving<BR><FONT color=#cc0000><STRONG>THOUSANDS OF 
      HITS</STRONG></FONT>! </P></TD>
    <TD width="40%">
      <TABLE border=1>
        <TBODY>
        <TR>
          <TD align=middle width="50%" bgColor=#ffff00 height=19><B><FONT 
            face=Arial size=4>Month</FONT></B></TD>
          <TD align=middle width="50%" bgColor=#ffff00 height=19><B><FONT 
            face=Arial size=4>DAILY HITS</FONT></B></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>1</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>100</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>2</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>160</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" bgColor=#ffff66 height=17><FONT 
            face=Arial size=2>3</FONT></TD>
          <TD align=middle width="50%" bgColor=#ffff66 height=17><FONT 
            face=Arial size=2>250</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>4</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>400</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>5</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>640</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" bgColor=#ffff66 height=19><FONT 
            face=Arial size=2>6</FONT></TD>
          <TD align=middle width="50%" bgColor=#ffff66 height=19><FONT 
            face=Arial size=2>1025</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>7</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>1,640</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>8</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>2,620</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" bgColor=#ffff66 height=19><FONT 
            face=Arial size=2>9</FONT></TD>
          <TD align=middle width="50%" bgColor=#ffff66 height=19><FONT 
            face=Arial size=2>4,192</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>10</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>6,707</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>11</FONT></TD>
          <TD align=middle width="50%" height=19><FONT face=Arial 
            size=2>10,731</FONT></TD></TR>
        <TR>
          <TD align=middle width="50%" bgColor=#ffff00 height=22><FONT 
            face=Arial size=4>12</FONT></TD>
          <TD align=middle width="50%" bgColor=#ffff00 height=22><B><FONT 
            face=Arial 
size=4>17,169</FONT></B></TD></TR></TBODY></TABLE></TD></TR></TBODY></TABLE>
<DIV></DIV>
<P>
<CENTER>
<TABLE width=450>
  <TBODY>
  <TR>
    <CENTER>
    <P align=left><B><I><FONT color=#cc0000 size=4><FONT color=#0033cc>What will 
    these massive hits do for your INCOME?</B></FONT><BR></FONT></I>How much 
    will this program be worth in <B>2 to 5 years?</B> Will it compensate you 
    for your efforts over the next few months?</B>&nbsp;&nbsp; <B><FONT 
    color=#cc0000>YES!</FONT></B> 
    <P></P>
    <P align=left><B><I><FONT color=#0033cc>WHY DO YOU NEED THOUSANDS AND 
    THOUSANDS OF HITS?</B></FONT></I> <ALIGN=JUSTIFY>Because <B>ADVERTISING</B> 
    is the engine that builds business, recruits affiliates into money-making 
    programs, and makes sales.<BR>You will learn the absolutely easiest way to 
    become a<B> MASTER </B>at mass advertising. </P>
    <P align=justify>One year from now,<B> YOUR MASSIVE PASSIVE ADVERTISING 
    MACHINE</B> will be producing enough hits to insure the sale or promotion 
    of<BR><B>ANY LEGITIMATE Program</B> with <B>PHENOMENAL SUCCESS.</B> </P>
    <P align=justify>You will be able to <B>build a downline in a new program 
    within DAYS instead of YEARS</B> by merely plugging in its URL for a few 
    days!<BR>You will be able to build a new income every few days, then move on 
    to another program immediately. </P>
    <P align=justify>Once you have your<B><FONT color=#0033cc> MASSIVE PASSIVE 
    ADVERTISING MACHINE </B></FONT>running at full throttle, I will then show 
    you how to turn those <B>VISITORS</B> into<B> CASH</B> in your 
    pocket!</STRONG></FONT></P>
    <P align=justify><B>But first of all, you need to get your own MPAM web 
    site.<BR>placing your own referral codes in all locations.</B></P>
    <CENTER></P></TR></TBODY></TABLE>
<TABLE>
  <TBODY>
  <TR>
    <TD vAlign=top align=left bgColor=white cellspacing="4">
      <TABLE cellPadding=2 bgColor=white border=0>
        <TBODY>
        <TR>
          <TD 
          style="BORDER-RIGHT: black 1px outset; BORDER-TOP: black 1px outset; BORDER-LEFT: black 1px outset; BORDER-
BOTTOM: black 1px outset" 
          noWrap align=middle bgColor=navy>
<A title="YOUR referral codes will be inserted instead of mine so you can build your downline!" style="TEXT-DECORATION: 
none" 
    <A href="mailto: mpa...@mi...?subject=Send MPAM please">
<FONT face="Arial, Geneva, Helvetica" color=yellow size=3><U><B>CLICK 
            HERE NOW for Your own  FREE MPAM!</B></U></A>
            
<P><center>
<a href="mailto:rem...@mi...?subject=REMOVE"><font color="#C40000" size=2><b>Sorry Uninterested Click 
Here</b></font></A> 
</center></body>

<P></P></FONT></FONT></FONT></TR></TBODY></TABLE></TR></TBODY></TABLE></CENTER></CENTER></
CENTER></CENTER></CENTER></CENTER></BODY></HTML>

[Refdb-users] Re: Web Access Problems

[Markus H. <mar...@mh...>]

Daniel,

thanks for the additional information. I think I've figured out what's
going wrong, and it appears to be a stupid bug that happens to be
silent on my systems.

refdbc checks for a valid username before it even reads the CGI data
from the web server. This is not a problem as long as your system
provides a username via getlogin(), as the real username used in the
queries is read afterwards from the CGI data. But your system
apparently can't provide a username via getlogin() for processes
running as CGI apps, so you're clearly screwed.

The short-term fix is to provide a username entry in the
/usr/local/etc/refdb/refdbcgirc configuration file. The name is not
really important, there just needs to be one to bypass this bug.

Let me know whether this helps on your system.

regards,
Markus

Daniel Hyduke writes:
 > 
 > Thanks Markus.  I tried that and discovered that nothing happend when I
 > tried the web interface.  So I wrote a quick perl script to see if
 > something was happening with the username/password/database and cgi;
 > everything appeared to check out here.  Next, I modified the refdbc.c
 > code (line 3517) to see if I could print the username when the refdbc is
 > called.  Now when refdbc is called via cgi the error message prints:
 > 	'' is an incorrect username
 > 	incorrect username.
 > 
 > 3516  else if (!strcmp(varname, "username")) {
 > 3517    fprintf(stderr,"\ '%s' is an incorrect username\n",username); //daniel's Code
 > 3518    if (!username[0] && getlogin()) {
 > 
 > As I know very very little about CGI and what is happening in refdbc.c, I would appreciate any input.

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] RIS titles, etc.

[Markus H. <mar...@mh...>]

Hi Bruce,

Bruce D'Arcus writes:
 > More questions on RIS fields:
 > 
 > 1)  What RIS fields should magazine and newspaper titles (Newsweek, New 
 > York Times, etc.) be mapped to?  I run into the same issue with things 
 > like academic departments for theses.  In all of these cases, Endnote 
 > maps this to the BT tag (aka T2).
 > 
 > 2)  BT should be T2?  A bit odd RIS has redundant tags like this.  But 
 > RefDB does not support BT or other similar fields, right?
 > 

RefDB recognizes all tags and synonyms as of Reference Manager
8.01. This is the most recent "specification" of RIS that I have
access to. RIS is weird and ugly, no doubt.

The RIS spec is not entirely comprehensible to me with regard to the
levels of bibliographic information. My guess is that TI/T1 is
equivalent to the title at the analytic level (work in bibx speak),
BT/T2 is equivalent to the title at the monographic level
(publication), and T3 should then map to the series level (set). This
is how RefDB currently deals with these tags.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[Refdb-users] RIS titles, etc.

[Bruce D'A. <bd...@fa...>]

More questions on RIS fields:

1)  What RIS fields should magazine and newspaper titles (Newsweek, New 
York Times, etc.) be mapped to?  I run into the same issue with things 
like academic departments for theses.  In all of these cases, Endnote 
maps this to the BT tag (aka T2).

2)  BT should be T2?  A bit odd RIS has redundant tags like this.  But 
RefDB does not support BT or other similar fields, right?

Bruce

[Refdb-users] Re: Web Access Problems

[Daniel H. <da...@se...>]

Thanks Markus.  I tried that and discovered that nothing happend when I
tried the web interface.  So I wrote a quick perl script to see if
something was happening with the username/password/database and cgi;
everything appeared to check out here.  Next, I modified the refdbc.c
code (line 3517) to see if I could print the username when the refdbc is
called.  Now when refdbc is called via cgi the error message prints:
	'' is an incorrect username
	incorrect username.

3516  else if (!strcmp(varname, "username")) {
3517    fprintf(stderr,"\ '%s' is an incorrect username\n",username); //daniel's Code
3518    if (!username[0] && getlogin()) {

As I know very very little about CGI and what is happening in refdbc.c, I would appreciate any input.
-----------
refdb0.9.2a
Mozilla 1.1
GNU/Linux Mandrake 9.0
apache-1.3.26
mysql-3.23.52

> 
> 
> 
> 
> From: "Markus Hoenicka" <mar...@mh...>
> Date: Tue, 25 Mar 2003 00:08:40 +0100
> To: ref...@li...
> Subject: [Refdb-users] web access problems
> 
> Hi,
> 
> you can run refdbd with log level set to 7 (e.g. refdbd -s -e 0 -l
> 7). This will show the username/password that refdbc sent to the
> server.
> 
> regards,
> Markus
> 
> Daniel Hyduke writes:
>  > I have just started using refdb (0.9.2a), on GNU/Linux Mandrake 9.0 with
>  > apache-1.3.26 and mysql-3.23.52, and am having problems with the web
>  > interface.  The client works perfectly from the commandline but when I
>  > try to access the database through the web interface I get an internal
>  > server error.  I have inserted the tail of my webserver and refdbd logs
>  > below.  The cookie and the username, password, and database are
>  > correct.  I have the correct grant for MySQL.  I have checked the
>  > mailing list, but none of the suggestions worked for me.  Does anybody
>  > have any ideas?  Also, is there anyway to see the command that is issued
>  > when refdbc is called so that I can see if the correct
>  > usr/paswword/database is being passed?
>  > 
>  > *********httpd log**********
>  > incorrect username
>  > [Mon Mar 24 14:02:17 2003] [error] [client 128.97.83.249] Premature end
>  > of script headers: /usr/local/www/internal/cgi-bin/refdbc
>  > incorrect username
>  > [Mon Mar 24 14:05:36 2003] [error] [client 128.97.83.249] Premature end
>  > of script headers: /usr/local/www/internal/cgi-bin/refdbc
>  > incorrect username
>  > [Mon Mar 24 14:07:00 2003] [error] [client 128.97.83.249] Premature end
>  > of script headers: /usr/local/www/internal/cgi-bin/refdbc
>  > incorrect username
>  > [Mon Mar 24 14:07:09 2003] [error] [client 128.97.83.249] Premature end
>  > of script headers: /usr/local/www/internal/cgi-bin/refdbc
>  > incorrect username
>  > [Mon Mar 24 14:19:27 2003] [error] [client 128.97.83.249] Premature end
>  > of script headers: /usr/local/www/internal/cgi-bin/refdbc
>  > 
>  > *********refdbd.log**********
>  > 6:pid=10299:Mon Mar 24 22:09:01 2003:serving client on fd 6 with
>  > protocol version 1
>  > 6:pid=10299:Mon Mar 24 22:09:01 2003:child finished client on fd 6
>  > 6:pid=3205:Mon Mar 24 22:09:01 2003:parent removing client on fd 6
>  > 6:pid=3205:Mon Mar 24 22:09:01 2003:server waiting n_max_fd=5
>  > 6:pid=3205:Mon Mar 24 22:09:01 2003:child exited with code 0
>  > 6:pid=3205:Mon Mar 24 22:09:01 2003:server waiting n_max_fd=5
>  > 6:pid=3205:Mon Mar 24 22:09:29 2003:adding client on fd 6
>  > 6:pid=3205:Mon Mar 24 22:09:29 2003:server waiting n_max_fd=6
>  > 6:pid=10305:Mon Mar 24 22:09:29 2003:serving client on fd 6 with
>  > protocol version 1
>  > 6:pid=10305:Mon Mar 24 22:09:29 2003:child finished client on fd 6
>  > 
>  > -- 
>  > Daniel Hyduke <da...@se...>
>  > 
>  > 
>  > 
>  > -------------------------------------------------------
>  > This sf.net email is sponsored by:ThinkGeek
>  > Welcome to geek heaven.
>  > http://thinkgeek.com/sf
>  > _______________________________________________
>  > Refdb-users mailing list
>  > Ref...@li...
>  > https://lists.sourceforge.net/lists/listinfo/refdb-users
>  > 
-- 
Daniel Hyduke <da...@se...>