nnslashdot

nnslashdot

[Lars Magne Ingebrigtsen]

I wanted to do some real programming, so I wrote nnslashdot.

It's pretty simple now -- if you put `(nnslashdot "")' in your
gnus-secondary-select-methdods, you'll get a new group for every new
article that appears on Slashdot.  When you enter these groups, you
can read the responses.

And that's pretty much it for now.  The version in CVS doesn't even
download the actual article, but that should be simple to fix, but I
need to sleep now, I think.  Or at least not type for a few minutes.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Or at least not type for a few minutes.

The group names look awful now, though.  But the title of the topics
could be put in the group description, but then it wouldn't be shown
unless you have the %D spec in gnus-group-line-format.  Hm...

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> I wanted to do some real programming, so I wrote nnslashdot.
> 
> It's pretty simple now -- if you put `(nnslashdot "")' in your
> gnus-secondary-select-methdods, you'll get a new group for every new
> article that appears on Slashdot.

I'm not getting any nnslashdot groups.  `F' doesn't help.  The server
is open, but when I browse the server I get an empty buffer.  Any
idea what could be going wrong?  

Dan

-- 
Dan Christensen
jdc@math.jhu.edu

Re: nnslashdot

[Shenghuo ZHU]

>>>>> "Dan" == Dan Christensen <jdc@jhu.edu> writes:

Dan> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>> I wanted to do some real programming, so I wrote nnslashdot.
>> 
>> It's pretty simple now -- if you put `(nnslashdot "")' in your
>> gnus-secondary-select-methdods, you'll get a new group for every new
>> article that appears on Slashdot.

Dan> I'm not getting any nnslashdot groups.  `F' doesn't help.  The
Dan> server is open, but when I browse the server I get an empty
Dan> buffer.  Any idea what could be going wrong?

Updating to the latest CVS version would help.

-- 
Shenghuo ZHU

Re: nnslashdot

[Dan Christensen]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:
> >>>>> "Dan" == Dan Christensen <jdc@jhu.edu> writes:
> 
> Dan> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> >> I wanted to do some real programming, so I wrote nnslashdot.
> >> 
> >> It's pretty simple now -- if you put `(nnslashdot "")' in your
> >> gnus-secondary-select-methdods, you'll get a new group for every new
> >> article that appears on Slashdot.
> 
> Dan> I'm not getting any nnslashdot groups.  `F' doesn't help.  The
> Dan> server is open, but when I browse the server I get an empty
> Dan> buffer.  Any idea what could be going wrong?
> 
> Updating to the latest CVS version would help.

No difference.  Any other ideas?

Also, gnus freezes if I start it when not connected to the net.
Is there any way to avoid this, or at least to catch `C-g' and
continue with the other servers?

Dan

-- 
Dan Christensen
jdc@math.jhu.edu

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc@jhu.edu> writes:

> I'm not getting any nnslashdot groups.  `F' doesn't help.  The server
> is open, but when I browse the server I get an empty buffer.  Any
> idea what could be going wrong?  

Hm...

It could be a case of `case-fold-search' sloppiness.  If you have
case-fold-search set to nil, I've now fixed this in CVS.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Dan Christensen]

Matt Pharr <mmp@graphics.stanford.edu> writes:
> I've got the latest and it doesn't do anything for me, either.  I have
> added it to my secondary select methods.
> 
> It does indicate that it's contacting slashdot.org, though my 
> " *nnslashdot *" buffer is empty.  I've got the following in " *Message-Log*"
> 
> Reading active file via nnslashdot...
> Contacting slashdot.org:80
> Retrieval complete.
> Reading active file via nnslashdot...
> Reading active file via nnslashdot...done

Essentially the same as me.  Here are relevant lines from my *Messages*:

  Reading active file via nnslashdot...
  Opening nnslashdot server...
  Contacting slashdot.org:80
  Request sent, waiting for response...
  Retrieval complete.
  
  Reading active file via nnslashdot...
  Reading active file via nnslashdot...done
  
  Connecting to ...
  Reading active file...
  Contacting slashdot.org:80
  Request sent, waiting for response...
  Retrieval complete.

Is that "Connecting to ..." (with no hostname) a problem?

> To followup to myself, I do have a "slashdot/groups" file in my mail
> directory, which is full of relevant looking stuff.

Hmm, I have that file too, but all it contains is "nil".

Dan

-- 
Dan Christensen
jdc@math.jhu.edu

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc@jhu.edu> writes:

> Also, gnus freezes if I start it when not connected to the net.

And slashdot is sometimes down, as well.

> Is there any way to avoid this, or at least to catch `C-g' and
> continue with the other servers?

The version now in CVS now catches C-g's in `gnus-read-active-file'. 

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> > It could be a case of `case-fold-search' sloppiness.  If you have
> > case-fold-search set to nil, I've now fixed this in CVS.
> 
> I just updated and re-made my elc files and still no luck.  Anything else
> to look at?

Is your case-fold-search nil?

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc@jhu.edu> writes:

> Essentially the same as me.  Here are relevant lines from my *Messages*:

After updating to the latest CVS, what's in your " *nntpd*" buffer
after saying `B nnslashdot RET RET'?

> Is that "Connecting to ..." (with no hostname) a problem?

Nope.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> To followup to myself, I do have a "slashdot/groups" file in my mail
> directory, which is full of relevant looking stuff.

If you go to the server buffer and hit `RET' on the nnslashdot server, 
but you get anything?  What's in your " *nntpd*" buffer after doing
that? 

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> Yep, that works great, like the browsing stuff.  Same bug out after I exit
> though, when I hit G in the group buffer.
> 
> Signaling: (error "Selecting deleted or non-existent buffer")
>   gnus-read-active-file()

Hm.  I've found a bug there.  Check out the CVS, and try again.  :-)

If you get the same bugout again, could you `(load "gnus-start.el")'
to load an uncompiled Gnus?  That will give me a much more detailed
backtrace.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Matt Pharr]

Dan Christensen <jdc@jhu.edu> writes:
> Shenghuo ZHU <zsh@cs.rochester.edu> writes:
> > >>>>> "Dan" == Dan Christensen <jdc@jhu.edu> writes:
> > 
> > Dan> I'm not getting any nnslashdot groups.  `F' doesn't help.  The
> > Dan> server is open, but when I browse the server I get an empty
> > Dan> buffer.  Any idea what could be going wrong?
> > 
> > Updating to the latest CVS version would help.
> 
> No difference.  Any other ideas?

I've got the latest and it doesn't do anything for me, either.  I have
added it to my secondary select methods.

I thought I was on to something when I changed my gnus-check-new-newsgroups
from nil to (setq gnus-check-new-newsgroups '((nnslashdot ""))), but that
didn't help any either.  I've hit "g", I've hit "F", and nothing nothing
nothing. :-(.

It does indicate that it's contacting slashdot.org, though my 
" *nnslashdot *" buffer is empty.  I've got the following in " *Message-Log*"

Reading active file via nnslashdot...
Contacting slashdot.org:80
Retrieval complete.
Reading active file via nnslashdot...
Reading active file via nnslashdot...done

Anyone?

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Matt Pharr]

[-- Attachment #1: Type: text/plain, Size: 123 bytes --]


To followup to myself, I do have a "slashdot/groups" file in my mail
directory, which is full of relevant looking stuff.


[-- Attachment #2: slashdot/groups file --]
[-- Type: application/octet-stream, Size: 2948 bytes --]

(("Interview: Grill John Vranesevich of AntiOnline (99/11/08/1021204)" 239 "99/11/08/1021204") ("Mainstream Media on Slashdot and Microsoft (99/11/08/1226255)" 214 "99/11/08/1226255") ("Candidates for 1999 GNU Free Software Award (99/11/08/116242)" 116 "99/11/08/116242") ("The Post-Microsoft Era (99/11/07/1750211)" 399 "99/11/07/1750211") ("IBM,Sun and Others to Help Apache Support XML (99/11/08/1133211)" 84 "99/11/08/1133211") ("User Friendly:  The Book (99/11/02/1115243)" 22 "99/11/02/1115243") ("Amazon.com switches to Apache (99/11/05/0947259)" 82 "99/11/05/0947259") ("Legal Actions Against Linux-DVD authors (99/11/05/1158259)" 344 "99/11/05/1158259") ("The Battle That Could Lose Us The War (99/11/05/144251)" 549 "99/11/05/144251") ("Linux on a Magazine Cover? (99/11/03/2329249)" 234 "99/11/03/2329249") ("SGI announces Linux Kernel Crash Dumps (LKCD) (99/11/05/1521233)" 208 "99/11/05/1521233") ("Microsoft == Monopoly says Judge (99/11/05/1841214)" 927 "99/11/05/1841214") (
 "!
!
!
Slashdot's \"Instant\" Legal Analysis of the MS Ruling (99/11/05/2016212)" 457 "99/11/05/2016212") ("Geeks In Space: Easy Listening (99/11/05/2129207)" 48 "99/11/05/2129207") ("The Do-It-All Remote? (99/11/03/2245248)" 236 "99/11/03/2245248") ("Vote in a CNN Poll on the DOJ MS Ruling (99/11/05/2312254)" 214 "99/11/05/2312254") ("Rick Moen on LinuxOne's IPO (99/11/06/1016209)" 120 "99/11/06/1016209") ("Checkpoint Porting Firewall-1 to Linux (99/11/06/1018228)" 132 "99/11/06/1018228") ("Microsoft Adresses World (99/11/06/140237)" 604 "99/11/06/140237") ("RealPlayer Uploads Your ID Too (99/11/06/1347228)" 164 "99/11/06/1347228") ("Communicator Is Losing The War..... (99/11/06/1747232)" 555 "99/11/06/1747232") ("First Journaling FS for Linux (99/11/06/1758233)" 287 "99/11/06/1758233") ("Online Romance - For Good or Evil? (99/11/03/2321243)" 505 "99/11/03/2321243") ("New Genetic Information Web Portal (99/11/07/0833206)" 69 "99/11/07/0833206") ("Lightning On Demand (99/11/07/08402
 0!
!
!
0)" 161 "99/11/07/0840200") ("The JFC Swing Tutorial (99/11/02/1052223)" 93 "99/11/02/1052223") ("Digital Television Transmission Standards (99/11/07/0845212)" 228 "99/11/07/0845212") ("Lotus Domino to ship RSN (99/11/07/127219)" 81 "99/11/07/127219") ("Debian Freeze Rescheduled (99/11/07/1611205)" 231 "99/11/07/1611205") ("David Bowie talks about Technology and Music (99/11/07/1913240)" 149 "99/11/07/1913240") ("Finding an Intellectual Property Patent Lawyer? (99/11/03/2141246)" 73 "99/11/03/2141246") ("Applications Service Providers May Change Your Life (99/11/08/080201)" 78 "99/11/08/080201") ("Does ATi Have a GeForce 256 Killer? (99/11/08/082223)" 94 "99/11/08/082223") ("U.S. Military Grapples With Cyber Warfare Rules (99/11/08/0812254)" 140 "99/11/08/0812254") ("TRUSTe Decides Its Own Fate Today (99/11/05/1021214)" 108 "99/11/05/1021214") ("Coming to a Desktop near you: Tempest Capabilities (99/11/08/093250)" 98 "99/11/08/093250"))

[-- Attachment #3: Type: text/plain, Size: 122 bytes --]


-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Matt Pharr]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Dan Christensen <jdc@jhu.edu> writes:
> 
> > I'm not getting any nnslashdot groups.  `F' doesn't help.  The server
> > is open, but when I browse the server I get an empty buffer.  Any
> > idea what could be going wrong?  
> 
> Hm...
> 
> It could be a case of `case-fold-search' sloppiness.  If you have
> case-fold-search set to nil, I've now fixed this in CVS.

I just updated and re-made my elc files and still no luck.  Anything else
to look at?

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Matt Pharr]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Matt Pharr <mmp@graphics.stanford.edu> writes:
> > > It could be a case of `case-fold-search' sloppiness.  If you have
> > > case-fold-search set to nil, I've now fixed this in CVS.
> > 
> > I just updated and re-made my elc files and still no luck.  Anything else
> > to look at?
> 
> Is your case-fold-search nil?

Nope; it's t.

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Matt Pharr]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Dan Christensen <jdc@jhu.edu> writes:
> 
> > Essentially the same as me.  Here are relevant lines from my *Messages*:
> 
> After updating to the latest CVS, what's in your " *nntpd*" buffer
> after saying `B nnslashdot RET RET'?

Now this is getting weird.  For me, browsing it worked great.  Whoo-hoo!
Then I quit from the foreign server browser buffer, hit G in my normal
group buffer, and got:

Signaling: (error "Selecting deleted or non-existent buffer")
  gnus-read-active-file()
  gnus-group-get-new-news(nil)
  call-interactively(gnus-group-get-new-news)

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Matt Pharr]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Matt Pharr <mmp@graphics.stanford.edu> writes:
> 
> > To followup to myself, I do have a "slashdot/groups" file in my mail
> > directory, which is full of relevant looking stuff.
> 
> If you go to the server buffer and hit `RET' on the nnslashdot server, 
> but you get anything?  What's in your " *nntpd*" buffer after doing
> that? 

Yep, that works great, like the browsing stuff.  Same bug out after I exit
though, when I hit G in the group buffer.

Signaling: (error "Selecting deleted or non-existent buffer")
  gnus-read-active-file()
  gnus-group-get-new-news(nil)
  call-interactively(gnus-group-get-new-news)

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Dan Christensen <jdc@jhu.edu> writes:
> 
> > I'm not getting any nnslashdot groups.  `F' doesn't help.  The server
> > is open, but when I browse the server I get an empty buffer.  Any
> > idea what could be going wrong?  
> 
> It could be a case of `case-fold-search' sloppiness.  If you have
> case-fold-search set to nil, I've now fixed this in CVS.

This fixed part of the problem.  Now I get nnslashdot groups created
automatically, and I can enter them and see summary lines.  However,
if I select an article I get things like:

Search failed: "^<p>.*A href=http://slashdot.org/article.pl"
Search failed: "<a name=\"4\">"
Search failed: "<a name=\"31\">"
Search failed: "<a name=\"6\">"

The " *nnslashdot*" buffer contains capital A's in places of the small
a's.

case-fold-search's value is 
nil
Local in buffer  *nnslashdot *; global value is 
nil

I just updated to the latest cvs version a couple of minutes ago.

Dan

Re: nnslashdot

[Shenghuo ZHU]

>>>>> "Matt" == Matt Pharr <mmp@graphics.stanford.edu> writes:

Matt> Some more information.  It turns out that all of the slashdot
Matt> articles were zombified.  Preivously I had
Matt> gnus-subscribe-newsgroup-method as the default
Matt> gnus-subscribe-zombies; the only other subscribing variable I
Matt> had set was

Matt> (setq gnus-subscribe-hierararchical-interactive t)

Matt> I was able to do A z in the group buffer and subscribe to those
Matt> guys and successfully read slashdot stuff--wheee!

Matt> I've now changed gnus-subscribe-newsgroup-method to
Matt> gnus-subscribe-randomly; I suspect that that will fix things and
Matt> will report back once slashdot has posted something new and I
Matt> see if that's the case.

Lars did a trick `gnus-subscribe-topics' here. 

I use it as following,
1. (setq gnus-subscribe-newsgroup-method 'gnus-subscribe-topics)
2. (add-hook 'gnus-group-mode-hook 'gnus-topic-mode) ;; be sure using topics
3. `T n' to create a topic named "slashdot"
4. `G p' on the topic
5. Put ((subscribe . "nnslashdot")) as topic parameters

For details, check the info node.

Matt> Now my next question: how to reliably make the original slashdot
Matt> article be the first thing in the summary buffer?

The original article should be article number 1 in the group. If
you've read the article, it may not be the first one in the summary
buffer.

-- 
Shenghuo

Re: nnslashdot

[Matt Pharr]

Some more information.  It turns out that all of the slashdot articles were 
zombified.  Preivously I had gnus-subscribe-newsgroup-method as the default 
gnus-subscribe-zombies; the only other subscribing variable I had set was

(setq gnus-subscribe-hierararchical-interactive t)

I was able to do A z in the group buffer and subscribe to those guys and
successfully read slashdot stuff--wheee!

I've now changed gnus-subscribe-newsgroup-method to
gnus-subscribe-randomly; I suspect that that will fix things and will
report back once slashdot has posted something new and I see if that's the
case.

Now my next question: how to reliably make the original slashdot article be
the first thing in the summary buffer?

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc@jhu.edu> writes:

> The " *nnslashdot*" buffer contains capital A's in places of the small
> a's.

Ok; this should be fixed in CVS now.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Kevin Ryde]

Dan Christensen <jdc@jhu.edu> writes:
> 
> Also, gnus freezes if I start it when not connected to the net.
> Is there any way to avoid this, or at least to catch `C-g' and
> continue with the other servers?

If that's with FSF emacs it could be open-network-stream retrying
gethostbyname forever, see src/process.c.  I posted bug-gnu-emacs
about this not so long ago, although probably it would've been known
already.  xemacs limits its retries, and has done for a while.

Re: nnslashdot

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
> Dan Christensen <jdc@jhu.edu> writes:
> 
> > The " *nnslashdot*" buffer contains capital A's in places of the small
> > a's.
> 
> Ok; this should be fixed in CVS now.

Yup, thanks.

Re: nnslashdot

[Matt Pharr]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:
> 
> Lars did a trick `gnus-subscribe-topics' here. 
> 
> I use it as following,
> 1. (setq gnus-subscribe-newsgroup-method 'gnus-subscribe-topics)
> 2. (add-hook 'gnus-group-mode-hook 'gnus-topic-mode) ;; be sure using topics
> 3. `T n' to create a topic named "slashdot"
> 4. `G p' on the topic
> 5. Put ((subscribe . "nnslashdot")) as topic parameters
> 
> For details, check the info node.

Yes, that did it--thanks!  Two questions and another bug now.

Question: I'd like to combine the behavior of gnus-subscribe-randomly and
gnus-subscribe-topics.  That is, now that all the slashdot groups are
appearing in my slashdot topic, I'd like to have the new ones appear at the 
top rather than the bottom as they do now.  Anyone have any ideas?

Question: Now with the addition of the slashdot backend, I have some
backends that I want to check every once in a while and some that I want to 
check every single time I hit 'g'.  Is group levels the way to do this?
Are there other options--what if I wanted to check slashdot no more than
once per 6 hours, say?

Bug: the first time I enter a slashdot group, all is well--I can read
articles, read comments, etc.  But then later on, when gnus reports new
messages, if I try to enter the group it immediately says "No unread news"
and reports no new messages.  It does this quite quickly, so I suspect that 
it isn't even contacting slashdot.org.

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Kai Großjohann]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> Question: Now with the addition of the slashdot backend, I have some
> backends that I want to check every once in a while and some that I want to 
> check every single time I hit 'g'.  Is group levels the way to do this?
> Are there other options--what if I wanted to check slashdot no more than
> once per 6 hours, say?

Yes, group levels are the way to go.  Set gnus-activate-level to 4,
say, then type `g' to update the important groups and `5 g' to update
all groups.  (If your important groups are at level less equal 4 and
the others at level 5.)

Dunno about the `no more often than once every six hours' part,
though.  You could add a demon which did it automatically every six
hours, but that's not what you're asking for.  Or you could train
yourself not to hit `5 g' more often than once per six hours...

kai
-- 
This gubblick contains many nonsklarkish English flutzpahs,
but the overall pluggandisp can be glorked from context. -- David Moser

nnslashdot is great, shame about the content

[David Maslen]

nnslashdot is great. I stopped reading slashdot comments because of
the web interface. I read all about how netscape suxs and ie5 is
terrific. Someone there thought perhaps Netscape should release their
code Open Source :)

I think I'm going to need some extreme scoring methods. Which is going
to be a bit of a challenge because the group only exist for each
topic.

Has anyone else considered scoring for nnslashdot groups.

Re: nnslashdot is great, shame about the content

[Laura Conrad]

>>>>> "David" == David Maslen <david.maslen@iname.com> writes:

    David> nnslashdot is great. I stopped reading slashdot comments
    David> because of the web interface. 

I used to think that was why I didn't read them, but then I got
nnslashdot and discovered I also didn't want to read them because
they're boring.

Skimming Freshmeat in gnus is really good, though.


-- 
Laura Conrad	mailto:lconrad@gamesville.com
(617) 673-1104	fax: (617) 673-1199

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> Question: I'd like to combine the behavior of gnus-subscribe-randomly and
> gnus-subscribe-topics.  That is, now that all the slashdot groups are
> appearing in my slashdot topic, I'd like to have the new ones appear at the 
> top rather than the bottom as they do now.  Anyone have any ideas?

Well, you can just alter `gnus-subscribe-topics'...

> Question: Now with the addition of the slashdot backend, I have some
> backends that I want to check every once in a while and some that I want to 
> check every single time I hit 'g'.  Is group levels the way to do this?

Yup.

> Bug: the first time I enter a slashdot group, all is well--I can read
> articles, read comments, etc.  But then later on, when gnus reports new
> messages, if I try to enter the group it immediately says "No unread news"
> and reports no new messages.  It does this quite quickly, so I suspect that 
> it isn't even contacting slashdot.org.

Have you upgraded to the latest CVS?

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot is great, shame about the content

[Raja R Harinath]

David Maslen <david.maslen@iname.com> writes:
> nnslashdot is great. I stopped reading slashdot comments because of
> the web interface. I read all about how netscape suxs and ie5 is
> terrific. Someone there thought perhaps Netscape should release their
> code Open Source :)
> 
> I think I'm going to need some extreme scoring methods. Which is going
> to be a bit of a challenge because the group only exist for each
> topic.
> 
> Has anyone else considered scoring for nnslashdot groups.

Slashdot has a pretty decent scoring system built-in.  It would be
nice if we could incorporate those scores into the nnslashdot article
scores. 
 
- Hari
-- 
Raja R Harinath ------------------------------ harinath@cs.umn.edu
"When all else fails, read the instructions."      -- Cahn's Axiom
"Our policy is, when in doubt, do the right thing."   -- Roy L Ash

Re: nnslashdot

[Matt Pharr]

> > Bug: the first time I enter a slashdot group, all is well--I can read
> > articles, read comments, etc.  But then later on, when gnus reports new
> > messages, if I try to enter the group it immediately says "No unread news"
> > and reports no new messages.  It does this quite quickly, so I suspect that 
> > it isn't even contacting slashdot.org.
> 
> Have you upgraded to the latest CVS?

Yep, just a few minues ago.  Now I'm getting the following on all slashdot
articles.

Signaling: (args-out-of-range "" 0 3)
  nnslashdot-date-to-date(" Thursday November 11, @11:39AM")
  nnslashdot-retrieve-headers((1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 ...) "Samba 2.06 Released (99/11/11/1220207)" "" some)
  gnus-retrieve-headers((1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 ...) "nnslashdot:Samba 2.06 Released (99/11/11/1220207)" some)
  gnus-cache-retrieve-headers((1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 ...) "nnslashdot:Samba 2.06 Released (99/11/11/1220207)" some)
  gnus-retrieve-headers((1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 ...) "nnslashdot:Samba 2.06 Released (99/11/11/1220207)" some)
  gnus-select-newsgroup("nnslashdot:Samba 2.06 Released (99/11/11/1220207)" nil nil)
  gnus-summary-read-group-1("nnslashdot:Samba 2.06 Released (99/11/11/1220207)" nil nil nil nil nil)
  gnus-summary-read-group("nnslashdot:Samba 2.06 Released (99/11/11/1220207)" nil nil nil nil nil nil)
  gnus-group-read-group(nil nil nil)
  gnus-topic-read-group(nil)
  call-interactively(gnus-topic-read-group)

-matt
-- 
Matt Pharr                                   mmp@graphics.stanford.edu
<URL:http://graphics.stanford.edu/~mmp>

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Matt Pharr <mmp@graphics.stanford.edu> writes:

> Yep, just a few minues ago.  Now I'm getting the following on all slashdot
> articles.
> 
> Signaling: (args-out-of-range "" 0 3)
>   nnslashdot-date-to-date(" Thursday November 11, @11:39AM")

This is now fixed in CVS.

On the "missing theading due to lacks of cid's" problem -- Malda
didn't want to fix that.  Because that would hurt the banner ads.  So
I offered to include the banner ads in the articles, but didn't hear
back from him after that.  (Does XEmacs do animated gifs, by the way?
:-)

So, there's a new variable in nnslashdot -- nnslashdot-threaded.  If
it's t, nnslashdot will retrieve absolutely all comments each time you 
enter the group, so it'll be slow, but pretty.  

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Norbert Koch]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> (Does XEmacs do animated gifs, by the way? :-)

Andy Piper has brought something up in 21.2 ... I've never tested it
though. 

norbert.

Re: nnslashdot is great, shame about the content

[Lars Magne Ingebrigtsen]

Raja R Harinath <harinath@cs.umn.edu> writes:

> Slashdot has a pretty decent scoring system built-in.  It would be
> nice if we could incorporate those scores into the nnslashdot article
> scores. 

I use the following nnslashdot:all.SCORE file:

(("subject"
  (" (5" nil nil s)
  (" (4" nil nil s)
  (" (3" nil nil s)))

So I score up articles that have been moderated up.

The moderators don't seem to have much taste, though.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Hrvoje Niksic]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> On the "missing theading due to lacks of cid's" problem -- Malda
> didn't want to fix that.  Because that would hurt the banner ads.

He didn't want to fix a bug for fear of hurting the banner ads?

Tsk tsk tsk.

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Hrvoje Niksic <hniksic@iskon.hr> writes:

> He didn't want to fix a bug for fear of hurting the banner ads?

Well, to be more precise -- he kinda didn't like the non-banner bit
about the, er, project, and apart from that, he didn't think it was a
bug since when you're displaying articles in a non-threaded manner,
they have no parents.

\/\/hatever.

> Tsk tsk tsk.

Well.  I'm not surprised.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Steinar Bang]

>>>>> Lars Magne Ingebrigtsen <larsi@gnus.org>:

> Hrvoje Niksic <hniksic@iskon.hr> writes:
>> He didn't want to fix a bug for fear of hurting the banner ads?

> Well, to be more precise -- he kinda didn't like the non-banner bit
> about the, er, project, and apart from that,

Hm... maybe I should post a feature about pgnus nnslashdot to slashdot 
and see if it is posted...? ;-)

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Steinar Bang <sb@metis.no> writes:

> Hm... maybe I should post a feature about pgnus nnslashdot to slashdot 
> and see if it is posted...? ;-)

:-)

We should perhaps wait until pgnus 0.99 has been released, though.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Jack Twilley]

>>>>> "ZSH" == Shenghuo ZHU <zsh@cs.rochester.edu> writes:

ZSH> Lars did a trick `gnus-subscribe-topics' here.

[... cookbook instructions on setting up gnus-subscribe-topics ...]

ZSH> For details, check the info node.

Here's the relevant part of the info node:

`gnus-subscribe-topics'
     Put the groups into the topic that has a matching `subscribe' topic
     parameter (*note Topic Parameters::.).  For instance, a `subscribe'
     topic parameter that looks like

          "nnslashdot"

     will mean that all groups that match that regex will be subscribed
     under that topic.

I tried this, and when a new group came up in nnslashdot as seen on
the server menu, it didn't come up subscribed in the Slashdot topic.
Why? 

Jack.
-- 
Jack Twilley
jmt at nycap dot rr dot com
http colon slash slash jmt dot dhs dot org slash tilde jmt slash

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Jack Twilley <jmt+usenet@nycap.rr.com> writes:

> I tried this, and when a new group came up in nnslashdot as seen on
> the server menu, it didn't come up subscribed in the Slashdot topic.

Where did it end up, then?

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: nnslashdot

[Jack Twilley]

>>>>> "Lars" == Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

Jack> I tried this, and when a new group came up in nnslashdot as seen
Jack> on the server menu, it didn't come up subscribed in the Slashdot
Jack> topic.

Lars> Where did it end up, then?

It, umm, didn't end up anywhere.  I was in the Slashdot topic when I
did 'A z', and it came up.  Is this the default behavior?  Is it
possible for me to fix it on my end so mine just appear in the right
place?  

Jack.
-- 
Jack Twilley
jmt at nycap dot rr dot com
http colon slash slash jmt dot dhs dot org slash tilde jmt slash

Re: nnslashdot

[Lars Magne Ingebrigtsen]

Jack Twilley <jmt+usenet@nycap.rr.com> writes:

> It, umm, didn't end up anywhere.  I was in the Slashdot topic when I
> did 'A z', and it came up. 

If you have `gnus-subscribe-topics' as your subscription method, then
it should work.  It doesn't?

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen