> the only boards that use them are image boards where you have to load images anyway.

Correction: http://www.akatsukimanga.com/kareha/

>>196
Actually, a solution to >> links with static pages is to simply make them reference a certain point on a certain page number for that thread (ie, http://wakaba.c3.cx/sup/1129153864/index2.html#197).

>>91 Ohshi-, time paradox!

>>323-324
The 2channel moderation request forms that are mods of 0ch use 強制リモホ, which is Forced_Remote_Host, more or less. And it makes sense, there.

Forced_IP is only enabled on the "Siberia super-news flash" board: http://etc3.2ch.net/siberia/

And I have no idea what that board is about...

> metadata

Not sure, that would require a database redesign and I don't want to force people with a current install to do that. Also, it seems something like that would work better for a whole new script, properly designed around the idea.

> config.pl parameter for a generic image that takes the place of a deleted image (ie, Hello Kitty)

Ah, good, been meaning to do, forgot about.

> Fine-grained banning options that let you choose whether or not the user is blocked from reading a board, posting to a board, or both. Another parameter defines the duration of his ban ('0' for permaban), and another defines a reason/message displayed when the user tries to access a board.

None of those seem useful to me, because I'm of the opinion that bans are to prevent abuse, not to punish users.

> Replace HTML error pages with dialog box equivalents using JavaScript.

Would require a bunch of hidden-iframing and such. I'd like to do a complete re-design full of javascript trickery, and this idea would fit better in such a context... That is to say, I'm lazy and the current version is robust, and I'm loathe to go around changing it, since it would introduce new problems.

> Kill user deletion. I can't see any case for when it'd have constructive uses.

On image boards, it has a very definite use - people do fuck up and post in the wrong thread, or create new threads. It's better if they can clean up after themselves. In Kareha, you can already disable deletion.

> Conversion to mod_perl?

As far as I know, it should work in mod_perl already, modulo some prototype bugs. I'll try to get those fixed.

> The standalone thumbnailer project is a great idea too. As a suggestion, how about adding functionality to also read and thumbnail document files like TXT, PDF, and DOC?

That would require a LOT of code, especially when you don't want external dependencies, so it's a bit iffy.

Also, wouldn't making capcodes even more prevalent be considered A Bad Thing®? If anything, the role of capcodes should be minimized or altogether eradicated, in favor of ninja moderation.

Another question: would FUDGE_BLOCKQUOTES be considered deprecated by now, or are there still CSS styles out there that require it?

>>32

Your browser momentarily regressed to an old bug and then got better? Who can tell?

>>22

Running in pure perl would be ideal, portability-wise, but in practice implementing a JPEG loader and saver from scratch in Perl is both a lot more work than anyone wants to do, and the result will also be too slow.

As was already stated, making a statically linked executable lets you distribute pre-compiled binaries that people can just upload along with the script.

>>157
So the functions need to be hardcoded to a post element one way or another? If I wanted to, let's say, create a template for gazo-box or Shiichan (both use checkboxes for sage), I'd need to slightly modify kareha.pl to check that new checkbox input instead of the Link string input?

An interesting limitation. Thanks for explaining.

About permasaging/deleting after a certain thread filesize is reached: would this be the same as a limit on the total number of characters in a thread? Or would we also include WakabaMark formatting, hyperlinks (including navigation), name/date/title headers, reply boxes, and non-Unicode characters in the formula?

>>158
Chances are that most if not all major/fundamental changes made to Kareha's core scripts will clash with the philosophy of most people here (including WAHa), and they won't care for them anyway. There really aren't all that many big-bang end-user features left to be implemented in Kareha before it loses its minimalist charm.

P.S. Reminder for >>85

> Wait, why should l50 links be indexed/cached?

Because the only way for the search engine to find the old threads is to go through the l50 links. I'm taking the advice of >>322 though.

> The CSS in the All threads page is unsightly. Is there a way to properly wrap the outer color border(s) around the table of threads?

You need to explain what you're talking about before I can do anything about that. What style, and what does "wrap the outer color border(s) around" mean?

> I still say that the "Navigation: " text is extraneous when people can clearly see what the links do. Also still partitioning for 0ch-style error pages (with displayed user host and all).

It's there because freefloating unlabelled links look weird.

> Now that we do have filesize indicators in the backlog page of mode_message, do you still find it useless to have the red bold filesizes near the bottom of thread subpages?

They're there, but only if you enable pruning by size.

> Does mode_message now work in PAGE_GENERATION => 'paged'?

No. I'm too lazy to figure what that's supposed to do, and I don't think anybody actually wants to use that in the first place.

> Idea: forced anonymous/sage/ID/fusianasan by IP/thread/board/whole site (some of these combinations already exist, I know)?

There's no database to keep IP data in, and I'd prefer to keep the script completely agnostic to IP addresses.

> Finally, I imagine that the permasage/close/delete functions in kareha.pl will be easily interchangeable among the conditions in post_stuff(). Can you confirm this?

No, because I don't know what you mean.

>>16

Yes, it's throwing Javascript errors for me if I use that character. Gonna look into that some more.

About rel=nofollow: What links should have it? Obviously not the "entire thread" link, but the l50 links in the thread list sort of need it, otherwise the search engine will never find them in the first place. But that means the l50 links will end up in the index.

If you want to have a look at what the code actually does to dig out flaws, here is the current version:

sub sanitize_html($%)
{
	my ($html,%tags)=@_;
	my (@stack,$clean);
	my $entity_re=qr/&(?!\#[0-9]+;|\#x[0-9a-fA-F]+;|amp;)/;

	while($html=~/(?:([^<]+)|<([^<>]*)>?)/g)
	{
		my ($text,$tag)=($1,$2);

		if($text)
		{
			$text=~s/$entity_re/&amp;/g;
			$text=~s/>/&gt;/g;
			$clean.=$text;
		}
		else
		{
			if($tag=~m!^\s*(/?)\s*([a-z0-9_:\-\.]+)(?:\s+(.*?)|)\s*(/?)\s*$!si)
			{
				my ($closing,$name,$args,$implicit)=($1,lc($2),$3,$4);

				if($tags{$name})
				{
					if($closing)
					{
						if(grep { $_ eq $name } @stack)
						{
							my $entry;

							do {
								$entry=pop @stack;
								$clean.="</$entry>";
							} until $entry eq $name;
						}
					}
					else
					{
						my %args;

						$args=~s/\s/ /sg;

						while($args=~/([a-z0-9_:\-\.]+)(?:\s*=\s*(?:'([^']*?)'|"([^"]*?)"|['"]?([^'" ]*))|)/gi)
						{
							my ($arg,$value)=(lc($1),defined($2)?$2:defined($3)?$3:$4);
							$value=$arg unless defined($value);

							my $type=$tags{$name}{args}{$arg};

							if($type)
							{
								my $passes=1;

								if($type=~/url/i) { $passes=0 unless $value=~/(?:^$protocol_re:|^[^:]+$)/ }
								if($type=~/number/i) { $passes=0 unless $value=~/^[0-9]+$/  }

								if($passes)
								{
									$value=~s/$entity_re/&amp;/g;

									if($value=~/"/) { $value="'$value'" }
									else { $value="\"$value\"" }

									$args{$arg}=$value;
								}
							}
						}

						my $cleanargs=join " ",map { "$_=$args{$_}" } keys %args;

						$implicit="/" if($tags{$name}{empty});

						push @stack,$name unless $implicit;

						$clean.="<$name";
						$clean.=" $cleanargs" if $cleanargs;
						$clean.=" $implicit" if $implicit;
						$clean.=">";
					}
				}
			}
		}
	}

	my $entry;
	while($entry=pop @stack) { $clean.="</$entry>" }

	return $clean;
}

> Making them configurable from site to site is really dumb, because it would create an unthinkable usability mess.

Why? Let people figure out things themselves, if they are so keen on changing their keywords. They can get together in their own webmaster threads and figure this out. I don't see why this should be solved here.
Of course I think this is a dumb idea in the first place. Nobody needs to know what fusianasan and sage are. Write a FAQ with two sentences about it and/or let your oldtime users tell newbies. Two frickin' words, and you people talk about it as if it were something like making up a new system of romanization!

> Then why not simply boil it all down to the comment field, with trigger strings for inputting the name, e-mail, sage, ID:Heaven, and fusianasan? You can get a lot more minimal with the current interface.

That's a bit too much. You want to talk about sage and fusianasan in the comment field, not trigger it with it.
I suspect you are joking here, though. Design is about what you can take away and still remain optimal conveniency/efficiency on the user part, not about taking as much away as you are technically capable of.

I had a number of good old threads from this board bookmarked so I could read them later and wrap my head around their ideas, but then I lost them all in a hard drive crash (strike two, Western Digital!). I also had a number of Japanese BBS's linked from this board bookmarked so I could take a look at their software's functionality and get some other ideas.

Anyway, these are all non-template suggestions:

• Topical category metadata in new thread creation so only one instance is necessary per site (further fine-grained behavior with PATH_INFO would let you nest one category inside another and view multiple categories of threads at once).

• Thorough search functionality a-la notchan, using PATH_INFO. This might not be possible without implementing a per-post metadata system though.

• Elaborating on the above, per-post metadata for "sage" status, for future-proofing in case someone wants to remove the functionality from the e-mail field (onto a checkbox, for example).

• config.pl parameter for a generic image that takes the place of a deleted image (ie, Hello Kitty)

• Fine-grained banning options that let you choose whether or not the user is blocked from reading a board, posting to a board, or both. Another parameter defines the duration of his ban ('0' for permaban), and another defines a reason/message displayed when the user tries to access a board.

• Replace HTML error pages with dialog box equivalents using JavaScript.

• Kill user deletion. I can't see any case for when it'd have constructive uses.

• Conversion to mod_perl?

I also have an early idea swimming around in my head about only bumping threads a few positions up, instead of to the top. Another idea is actively bumping threads down, either by a few positions or to the bottom. I'm not exactly sure yet what good it'd be for.

Also...

>* I'll add thread closing to Kareha, but I was thinking of setting the default behaviour to never permasage or close threads.

I think this is ideal for the time, until we have enough statistical data to derive thorough auto-permasage and auto-delete/archive algorithms. Just add the functionality for mods to manually set these statuses, but remove the "permasage at 1000" behavior.

The standalone thumbnailer project is a great idea too. As a suggestion, how about adding functionality to also read and thumbnail document files like TXT, PDF, and DOC?

About rel=nofollow: What links should have it? Obviously not the "entire thread" link, but the l50 links in the thread list sort of need it, otherwise the search engine will never find them in the first place. But that means the l50 links will end up in the index.

> but I don't quite see why there should be a "First 100" at the top but not at the bottom.

Probably to avoid clutter and because of the assumption that if you arrive at the bottom of a page, you can do without the "First 100" link. "First 100" seems to be a navigational aid for beginners who are new to the thread, so it makes sense to only have it at the top.

More information on the all threads page, date of the last post? file size?

A quote button that puts >>n and puts the post prefixed by > in the reply box

Different secret strings for different functions (e.g. one for ID generation and one for secure tripcodes)

First thought: It would eliminate the concept of sageing as a protest entirely.

A 1001th post would be a bother.

n is implemented, but not for >> yet.

Also, >>1 is, as it is, only added to URLs of the form xx-yy and lxx. 2ch doesn't add >>1 for single-reply URLs, and if you're using commas, I figure you can add >>1 yourself if you want it. I'm not sure if this is the best behaviour, but that's how it works at the moment.

Anyway, test with |

>>180

The one with encoded Javascript that makes you post with fusianasan is cute.

Well, there are some issues to consider here:

• Hardly anybody needs to ever use fusianasan. It's a gimmick. Design decisions should not be made around it.
• The strings may be strange, and combining fields isn't the best design possible, but this is a 0ch clone, after all. If I were designing something from scratch, I'd do things differently, but as it is, people are expecting 0ch behaviour, and it would confuse them if the script worked like 0ch in some ways but not others.
• I might consider adding additional strings that trigger sage and fusianasan, but I'm not sure what they should be.

In the end, people actually enjoy the 0ch quirkiness. I know I do. I know about designing good interfaces, but there's something fun about an interface that is a little bit quirky, as long as it doesn't get in your way, and these things don't.

/-100 shows the first post two times.

Funky, works... I am pretty sure the error has something to do with the characters preceeding the "¦" in the unprocessed tripcode. It begins with a "`"

>>216
I remember at least one or two boards on 2ch that used it, though I can't remember which (moon language and such, you see).

I apologize for the dumb question I made at the end of >>218. I forgot that Kareha permasages (not closes) a thread after the limit is exceeded, so there's no need for a hypothetical 1001th post anyway! orz

> forced fusianasan

that seems kind of evil.

>>247

Well, you have your chance to try and abuse it over on the test page. Although the list of allowed tags there doesn't exactly match what would be allowed here.

Also, the point is to make the type of markup selectable, so you can pick WakabaMark or HTML or none at all.

How about adding a link to 2ch in footer.html called "2ch mode"?