archimag / cliki2 Goto Github PK

Go to http://74.207.228.11/site/history?title=zlib
or
http://74.207.228.11/site/recent-changes

A message with the content "1" is printed when some entries are loaded. Those appear to be (a subset of)

Tue, 13 Dec 2011 08:51:19 GMT 178.137.21.73 ()
(prev) Tue, 13 Dec 2011 08:49:59 GMT 178.137.21.73 ()
(prev) Tue, 13 Dec 2011 08:49:18 GMT 178.137.21.73 ()
(prev) Tue, 13 Dec 2011 08:36:03 GMT 178.137.21.73 ()
(prev) Tue, 13 Dec 2011 08:27:51 GMT 178.137.21.73 ()

Hopefully a good place to report this?

The minion IRC bot tries to retrieve the source for cliki pages to extract a description. On the old cliki it could do so by appending ?source to a page's URL. There seems to be a route for this in archimag/cliki2, but i couldn't find anything similar in vsedach/cliki2 or on the actual cliki.net site.

(in-package :cliki2)
(ensure-directories-exist "/var/tmp/cliki/lisp.not.org")
(start-cliki-server
8081
"/var/tmp/cliki"

`("wiki.alu.org"
,(make-wiki "lisp.not.org" "Not Association of Lisp Users"
"/var/tmp/cliki/lisp.not.org" "[email protected]")))

@vsedach

I know that we can already see the edit summary from the edit history page before going to the diff page, but it would be nice to also have this valuable piece of contextual information directly on the diff page itself.

One scenario where this would be specially useful is if someone links to a diff directly.

It might be nice to change the site title from "CLiki: index" to something like "CLiki: a Common Lisp Wiki".

It would move it up in Google searches and be easier to notice - currently if you search "lisp wiki" in Google it's the 5th result, and I initially skipped over it, not realizing what CLiki was.

CLiki: index
www.cliki.net/‎
Welcome to CLiki. CLiki is a Common Lisp wiki. It contains resources for learning about and using the programming language Common Lisp, and information ...

Apparently, it's only possible to change the email address (via http://cliki.net/site/preferences accessed by clicking on one's name after logging in) or "reset password", which generates a random password and sends it to the email address associated with the account.

It would be nice to be able to set a new password of one's own choosing ("enter old password, enter new password, confirm new password"). If there's already a way to do that, I didn't find it.

If user want to add some page he press the link - http://cliki2.lisper.ru/edit/dfs1g. 'edit' is action. But we can change the link - http://cliki2.lisper.ru/add/dfs1g - and got error 404 by Hunchentoot. I think it would be better if Cliki will check the action property is incorrect.

Greetings

Today I visited CLiki.net with my mobile phone, and found that i could not read the contents of the site, due to the text being very small. To fix this, one might add the viewport meta-tag in the header of the document, as described on the MDN site

I'll try to take a look at creating a PR doing this within a couple of days.

noticed this last night: sbcl-internals.cliki.net is down

I'm using asdf-install to (attempt to) install a package which depends on trivial-gray-streams. It looks like it's trying to get some information from http://www.cliki.net/TRIVIAL-GRAY-STREAMS?download, which is returning a 500 error.

Here's the transcript from SBCL (I replaced the install directory with a fake one below):

debugger invoked on a ASDF-INSTALL::DOWNLOAD-ERROR in thread
#<THREAD "main thread" RUNNING {10029C0EB3}>:
  Server responded 500 for GET http://www.cliki.net/TRIVIAL-GRAY-STREAMS?download

Type HELP for debugger help, or (SB-EXT:EXIT) to exit from SBCL.

restarts (invokable by number or by possibly-abbreviated name):
  0: [RETRY] Retry installation
  1: [ABORT] Exit debugger, returning to top level.

(ASDF-INSTALL::DOWNLOAD-FILES-FOR-PACKAGE
 "TRIVIAL-GRAY-STREAMS"
 #P"/home/user/fake/TRIVIAL-GRAY-STREAMS.asdf-install-tmp")
0] 0
Downloading package TRIVIAL-GRAY-STREAMS, required by yason

debugger invoked on a ASDF-INSTALL::DOWNLOAD-ERROR in thread
#<THREAD "main thread" RUNNING {10029C0EB3}>:
  Server responded 500 for GET http://www.cliki.net/TRIVIAL-GRAY-STREAMS?download

Type HELP for debugger help, or (SB-EXT:EXIT) to exit from SBCL.

restarts (invokable by number or by possibly-abbreviated name):
  0: [RETRY] Retry installation
  1: [ABORT] Exit debugger, returning to top level.

To help update dead urls, it would be nice if it was possible to search strings in urls too (eg. search a domain name or the name of a library would give all the pages where an url containing this domain name or the library name in the path is present).

So, I know this one is a bit pie-in-the-sky and complex to do, but here goes:

When making one kind of change repeatedly across many pages, it would be nice to be able to "open a transaction" (a persistent set of related unapplied changes), then add many changesets to this transaction (possibly across days), and then at the end review all the changes and apply them all in one go. The user would be notified if conflicting changes were applied to some of the targetted pages in the meantime.

The changelog would report the apply of the "transaction" only, not each individual change, but of course it would be possible to click through to see all the changes applied by the transaction.

This would avoid "spamming" the changelog for big sets of related changes, help better plan such changes, help consistency by avoiding having some of the targetted pages updated while others not updated yet, and encourage consistency-improving "large-scale" changes by helping users in the aforementioned ways as well as preventing inhibition stemming from knowing the big set of changes would "spam" the changelog.

It would also be possible to "publish" in-progress transactions to advertise one's intentions in advance and sollicit public comments for improvements and such.

I hope this is all clear.

It'd be great if the inner converter (from <code lang="common lisp"> to some <div> structure) would work as if the code tag had in a style attribute "white-space: pre;", simply makes code tags work like pre tags

Steps to reproduce:

1. Register new user and fill username with:
   < script >alert(1);< /script >
2. Finish registration.
3. You will be redirected to home page.

You can see popup window with text "1".
Browser: Firefox 8.0

Steps to reproduce:

1. Edit some article and fill Edit summary with:
   < script >alert(1);< /script >
2. Save article.
3. Go to Recent changes page.

You can see popup window with text "1".
Browser: Firefox 8.0

The README you should contain the URL of the current website.

I mistakenly created a Wiki page without a title. The page now has no link afterwards, yet it is still in the wiki.

See first line here:
https://www.cliki.net/Java

Packages that help with CL/Java integration:
- A simple example/wannabe-lib on how to use Swing UI components from Common Lisp (using ABCL)

I request, please, this wiki entry to be deleted, and a guard put in place to prevent further pages with no title being created.

It's not really a bug, but it's annoying that the whole area of menu item isn't clickable.
// The header & footer stripes

Given an internal page named “Lorem ipsum (dolor sit amet)”, how does one link to it with _() syntax? “Text formatting” page doesn't seem to provide an answer (I tried placing slashes in various ways to no avail), and “CLiki Style” page has no warnings suggesting to avoid such names.

The bare minimum would be to add a maxlength="128" attribute to the input element in the edit page.

The link in the menu bar Text Formatting is wrong, a 't' is missing in the URL.

i.e. [http://74.207.228.11/Text Formatting](http://74.207.228.11/Text Formatting) vs [http://74.207.228.11/Text Formating](http://74.207.228.11/Text Formating)

I try to sign in using my account name (marcoxa) and and says that the account does not exist. Same if I try to reset the pw.

If I try to create an account with that name (marcoxa) cliki tells me that the account name is already taken.

Can you help?

Thanks

MA

Major security bug. User can save js-code in Cliki page and code be executed on every page view. For example: save some page with text <script>alert(1);</script> and open it. You can see alert with 1. See - http://cliki2.lisper.ru/history/test/843685ae3f3b09ea6195d07869634033dac6d5b7

Steps to reproduce:

1. Change id for view-revision URL. Example: http://74.207.228.11/site/view-revision?id=4+4546
2. Got Internal Server Error.

Details:
junk in string "4 4546"

P.S. Also this happens while processing /site/compare-revisions request. See - http://74.207.228.11/site/compare-revisions?old=4+4537&diff=44538

When user try to add new Cliki page he can see preview of new page. If user filled Summary of changes, check Minor edit and press Preview button, Summary of changes field is cleared.