1

(16 replies, posted in PunBB 1.2 show off)

Reines wrote:
titansolaris wrote:

omg, they tore off the punbb copyright, it doesnt say richard did it, how HAPPY. a big company pwned pun. this could be called pwndBB

You are allowed to remove the copyright, aslong as you don't replace it with your own.

It's illegal to remove Rickard's copyright notice from the source code but you can remove it from the HTML output (although he asks that you don't).

See FAQ: Copyright.

2

(8 replies, posted in Feature requests)

Rickard wrote:

Anyone? Shawn, how are you at regular expressions?

My regex is pretty good. I'd be happy to help with this.

3

(8 replies, posted in Feature requests)

I think that this type of internal rearrangement is a step in that direction. A master list of pattern?replacement pairs could allow for an easy Textile mod.

The BBCode patterns would be replaced with the Textile patterns and the replacement values would get a little tweak:
    BBCode --> $masterList['\\[url=([^\\]]+)\\](.*?)\\[\\/url\\]'] = '<a href="$1">$2</a>';
      Textile --> $masterList['"([^"]*)":http://([^ ]*)'] = '<a href="http://$2">$1</a>';

Note: The patterns above have not been tested and could be cleaned up quite a bit (I haven't been regexing in PHP for a while).

4

(8 replies, posted in Feature requests)

I think it could be useful to rewrite some of the bbcode functions in /include/parser.php. Now, I'm not clamoring to see these changes right away but I think they would be helpful in the future.

Here's what I'm thinking:

(1) Move much of what preparse_bbcode() is doing into do_bbcode() or into some other function. (Look at dropping preparse_bbcode() altogether?tricked out RegExps can take up the slack and process the text with less preg_replace() calls.)

(2) Rearrange things so that there's only one copy of the bbcode patterns (lets say, in a new bbcode_patterns() function).

(3) Now, do_bbcode() can get its pattern arrays by calling bbcode_patterns().


So what's the point?:

These changes will help PunBB better cooperate with preview and WYSIWYG editor mods. PunBB will act the same as before, but future mods will have easy access to a master bbcode pattern list. A JavaScript editor mod could start sharing code with PunBB's standard editor.

I think this will be useful in the future because, as web apps mature, WYSIWYG editors are going to be the norm. People are going to expect them and anything less will be seen as old and unfriendly.

Later when a mature JavaScript WYSIWYG mod emerges, it could be offered as a convenient plug-in or even wrapped into the PunBB core.

In the case of JavaScript plug-ins: There are differences between PHP's and JavaScript's RegExp patterns but they could be converted by the mod.

I was thinking about this and decided to write a demo: As-you-type Preview

As is, the script isn't ready to use in a live forum--it would need to be part of a mod package. If someone (bedroom?) wants to use this code in a future mod, go right ahead.

If using the innerHTML property is an option, JavaScript can parse bbcode just as easily as PHP can?and maybe in fewer lines of code than PunBB's parser.php file. Of course innerHTML isn't a W3C standard property and if bedroom is a standards purist, this may not be an option.

On the other hand, it is a de facto standard, its faster than DOM methods and enjoys wide browser support. Gmail makes good use of it.

Some people at PunRes have reported that PunJS New Posts doesn't work with the newer versions of PunBB. And I'm sorry to say that development of this plug-in/mod has been discontinued.

Please see Miles' Mark topics as read mod or wait for PunBB 1.3.

8

(132 replies, posted in General discussion)

Decentralize! - Co-create! - Remix! - Put the 2 in Web 2.0!

Q: Foundation Attributes + Experience Attributes = ?
A: A valuable new service with a fast-growth business model!!

Sadly, this Adaptive Path article doesn't appear to be joking: Buzzword Burnout

Credit: Saw this at http://37signals.com/svn/archives2/heck … e_path.php.

9

(40 replies, posted in PunBB 1.2 bug reports)

Rickard wrote:

1.3 will have proper "mark as read" functionality. Yes, I give up smile

Now we need a volunteer to make the PunBB 1.3 mod that restores classic new-message indicator behaviour.

10

(40 replies, posted in PunBB 1.2 bug reports)

Rickard wrote:

Don't forget about ShawnBrown's PunJS New Posts. It's a drop-in Javascript solution. I don't know if it works with 1.2.5, but I can't see any reason it wouldn't.

Ahh, I didn't see this post when it was new:

Some people at punres have reported that PunJS New Posts doesn't work with the newer versions of PunBB. I want to release an update but I haven't had the time. Miles' mod is better maintained at this point.

erissiva wrote:

I'm not sure I want a JS getting loaded every time that a page is loaded.

A side note:

The script is ran with every page load but it isn't downloaded every time. When you include a .js file with the SCRIPT tag, it's only downloaded once (like a .css file). After that, the server returns a "304 Not Modified" status and the browser uses a locally cached copy. This is partly how GMail gets away with it's 20,000 line/230k+ JavaScript (though they're using IFRAME tricks too).

11

(7 replies, posted in General discussion)

Some interesting PHP work: Toogle

http://www.c6.org/toogle/index.php?phra … b.org+logo
http://www.c6.org/toogle/index.php?phrase=maclogo
http://www.c6.org/toogle/index.php?phrase=MS+Windows

12

(132 replies, posted in General discussion)

Rickard wrote:

You've implemented a PHP script interpreter in Javascript? Amazing! But why? smile

I know your question is rhetorical since you, of all people in this forum, know why the computing industry desperately needs client-side PHPScript. But I'll spell it out for others who may not see the answer so clearly:

PHPScript:
...results in code that is far cleaner and easier to read than code written in other languages.
...lets us take advantage PHP's elegant function conventions.
...prevents code that relies on reference pointers (which are simply a crutch for weak programmers).
...leverages PHP's superior object model in the client-side world (nothing in client-side use today can even begin to match it).

Client-side PHPScript is guaranteed to transform the world of computing. It's a new technology that will revitalize the IT industry.

Put simply, PHPScript is a radically new concept that breaks away from the old paradigm of stagnant, flawed and less capable browser scripting technologies that we suffer with now. Today's scripting technologies (VBScript, JavaScript, HyperCardScript, JythonScript, AppleLinkScript, RubyScript, VMS/DCLScript, FireFox's buggy Tcl/TkScript and JavaScript's dreaded offspring JavaScriptScript) are too limited and obviously the wrong tools for any type of client-side job. Period.

13

(132 replies, posted in General discussion)

zaher in the "PHP 6.0 discussion" wrote:

I think, we need php in client side (WebBrowser) like as JavaScript or VBScript, let us name it as PHPScript roll

I'm almost finished with an alpha release of this.

The PHPScript interpreter is written in JavaScript and imports the PHP source files via XMLHttpRequest/ActiveX-XMLHTTP (synchronously though).

There are four main steps:
1. The Main file loads and calls phpScript.js.
2. phpScript.js instantiates an interpreter object and fetches the PHPScript file (.phps).
3. From here, the interpreter processes the script (rendering the markup in one pass).
4. Finally, the output is sent to the DOM via the BODY tag's innerHTML property (doing this with W3C DOM methods is just too much to deal with).

As of now, I only have about 15% of PHP 4's functions implemented?at least that's a start. I don't think PHPScript will have decent database support any time soon though (sorry about that).

It's important to note that "viewing source" will not show the rendered output or your source files since the elements are written to the DOM via JavaScript. But you can't rely on this to protect your source files; the unprocessed scripts are sent to the client before the PHPScript interpreter processes it.

Remember: PHPScript is client-side scripting. The unprocessed scripts are stored in the user's browser cache. So simply looking in the cache folder reveals your source code. So be careful what you put in there.

14

(12 replies, posted in General discussion)

hcgtv's VectorLinux article wrote:

"...VectorLinux was designed primarily to run well on older hardware..."

Looks nice, the article makes it sound even better. I might have to install this on my extra machine instead of Knoppix.

That reminded me of this Longhorn article on ExtremeTech mentioning the use of "virtual pixels".

ExtremeTech's article wrote:

This support for high-DPI displays is done by creating "virtual pixels" that have a resolution of 96dpi. These virtual pixels are mapped onto the actual display resolution, so that objects will look the same size and render correctly, even on displays with dense pixel resolutions.

15

(12 replies, posted in General discussion)

Some higher resolution screenshots: Windows Vista Virtual Pressroom. The status bar and the file-associated keywords are interesting.

I hope there's a simple way to turn off the eye candy.

16

(12 replies, posted in General discussion)

No more slashes in the file path ? [JPEG]

Vista Screenshots
Related Article

17

(1 replies, posted in PunBB 1.2 discussion)

caj wrote:

wannabies??

More like cybersquatters.

Previous mentions of punbb.com:
? "...someone had registered punbb.com"
? punbb.COM

18

(5 replies, posted in Programming)

Testing under Linux, I was able to fix the problem for both Konqueror and Safari. I received Mac/Safari confirmation from middleground this morning (thanks!).

The sticking point was Safari/Konqueror's handling of XMLHttpRequests that violate the same origin security policy.

If anyone is interested in the details, I've documented these cross-browser differences on SourceForge's wiki (see Same Origin Violation Behaviour).

19

(5 replies, posted in Programming)

Excellent! hcgtv, middleground: thanks for the browser test!

Bert's (hcgtv's) Debian test has inspired me to get a copy of Knoppix. I'm using Konqueror on Knoppix to post this message.

Hopefully I'll be able to test with this setup and fix the problem for both Linux/Konqueror and Mac/Safari users--since Safari is based on Konqueror, and it looks like you've both reported the same problem (middleground here in the forum and hcgtv via email).

Other windows users: If you're looking for a low-hassle way to test for Safari/Konqueror, you might want to have a look at Knoppix. You download an .iso image (for free of course) and burn it onto a CD-ROM. The CD is a bootable install of Linux with KDE and other goodies ready to go. It automatically mounts your existing drives and doesn't muck with your Windows setup. It seems to be working great--after I figured out how to make it use my antiquated network adapter.

20

(5 replies, posted in Programming)

Hello... it's been a while since I've posted.

I'm working on a JavaScript to embed a remotely hosted lookup service with a simple script tag.

I don't have access to a Mac but I don't want to neglect Mac users. Could someone with a Mac (and ~90 seconds to spare) check out my test pages using IE for Mac and/or Safari/Konqueror?

If you can, try to load resources "1" and "2" on each of the following pages and let me know how the script behaves:
Remotely Hosted [link removed]
Locally Hosted [link removed]

I would greatly appreciate any help with this!

(For the test, the "resources" are simply text strings.)

I'm travelling right now so I don't have regular access to a decent computer. It might be a week or more before I can address these issues properly.

Widell: I'm almost certain this is a problem with the script. The compression program I used didn't like double quotes nested within single quotes. To compensate for this I replaced the double quotes with their UNICODE-escaped equivalent (which works fine on my English-language test forum under IE). I think this is what's causing the problem... and it will probably be an issue for other non-English forums until I get a chance to correct it. The nested quotes only occur once in the script--when it alters the visibility rules of the first cells in the stylesheet.

Jérémie: I'm not sure what why it isn't working for your forum. I might have to send you a version of the script with debugging turned on so we can see what is working and what isn't working.

Gary13579: I assume you're experiencing the same problem that Jérémie is?

The packer that I used might also be causing some odd problems of its own (with the high ASCII and required ISO-8859-1 encoding). I'm going to put a lighter-weight but unpacked version up when I get the chance. But like I said before, it will be a while before I can take care of these issues.

Update: I've completed a new update of the script. This version has been compressed with the packer script suggested by Mart--it's now ~8k. The script can be downloaded at the same location (Working Update: new_posts.js 1.0.1dev).

The issue Smartys was experiencing was simply the result of a mod he had made to his forum. On the other hand, the problem that Gary13579 reported was an oversight on my part and the issue has been fixed. (I had forgotten to integrate last-post tracking with the index page functions.) The index page indicators should work fine now.

Gary13579: it sounds like the script is wrongly turning the new indicators off but I'm not sure why. I'd like to test this and see exactly what the script is doing. I'll IM you on ICQ and see if we can figure it out.

24

(200 replies, posted in General discussion)

The most noticeable difference I've seen with Gmail (thanks to Bwongar's invite) is that the experience of using it feels smoother. You can tell that Google put a lot of work into the mental ergonomics of the service. The archive, filtering, tab system, spellcheck and message view are all organized in a way that's more natural to use. Gmail also uses JavaScript/DOM DHTML in a satisfying way ... and the page loads are kept small by loading information dynamically only as it's needed (via JavaScript httpRequest). The other web-mail services I've used look like legacy designs when compared to gmail.

EDIT: I should note that I never received two messages that I knew were sent to my gmail account. But I don't know if the problem was with gmail or somewhere earlier in the chain of things.

Gary13579 wrote:

NOO sad
My one bug, and its already been fixed in a newer version.

Ah I forgot to mention the added Sticky support in the 1.0.1dev announcement... it's there now. The Sticky support went hand-in-hand with the middot.