The interesting question is, does it have any effect? Pål Nedregotten, Channel Manager at API, which has 50 news papers with IE6 warning under their wings, posted some stats which he keeps updating with numbers for hour by hour and per hostname. The graphs show IE6 usage Friday 13th compared to Friday 20th of February.

From the updates we can see that there actually is a decrease in IE6 usage. (Red numbers are good) At the same time we know that there are winter holidays in Norway this week, and thus we will not know for sure before 1-2 weeks have passed.

It’s also important to remember that the impact of the campaign is not not likely to be visible from day one. As has been mentioned numerous places, the biggest job isn’t to make average Joe update his machine, but getting the big companies’ IT departments to update their machines. We can clearly see this with numbers from Budstikka.no, where I work. On a typical work day we see 23-25% with IE6, while on Saturdays and Sundays we see a drop down to 10% IE6 usage.

Big companies are of course reluctant to upgrade their machines. They have everything working right now, both intranet and internet. They have also applied all possible patches to IE6 to make it more secure, so security isn’t that much of an issue. Why would they want to upgrade?

The hope of this campaign is to make direct and indirect pressure on the IT departments to update their machine parks. By putting this banner out on all possible sites, users and IT departments will be constantly reminded that using an eight-year old browser is unacceptable, especially when you think about how much www has evolved.

Hopefully this will be the push most companies need, and that they will start planning to upgrade. This however requires testing and patching before it can be rolled out. I only hope that before the summer, the IE6 numbers will be so low that we can ignore IE6 completely.

I can also reveal that on budstikka I’ve been given green light for ignoring IE6 in our new design. The reason we can do this is that we will have the old design, which supports IE6, running too, and we can just forward IE6 users there automatically by doing some user agent sniffing.

Being on a Mac, I have no easy way to actually see these warnings. I thought it would be cool to see, so I started asking around for screen shots. The result can be seen here.

Apparently this campaign also spread to Sweden and Australia, so now I’ve created this post to try to keep track of international pages working for the same campaign: To push IE6 percentages so low that you don’t have to take it into consideration when you create web sites anymore.

axbom.se is a Swedish site about web strategies and web trends:

travelfish.org is an Indonesian travel intelligence site

mappic.org is a site where users can add pictures from places around the globe

travelfusion.com is a search engine related to travel information

aftonbloadet.se is one of the biggest news sites in Sweden

travellerspoint.com is a site to share and plan travels

flashmagazine.com is a site striving to give the best information about flash and related productions. Offices in Norway

whoa.nu is the biggest hip hop site in Sweden

easyimagesoftware.com is a site for a product site for easy image software

Willhaben.at is a big online market place is Austria, eqiuvalent of Norwegian finn.no

tv2.dk is the danish TV2 channel

hogyan.org is an Hungarian how-to site

Anyone missing? Please send me a message on twitter with a screen shot. (Having a hard time doing them myself, as I don’t have access to IE6 on my mac)

Being on a Mac, I have no easy way to actually see these warnings. I thought it would be cool to see, so I started asking around for screen shots. In this blog post I’ve collected as many images as I can to show how the differnet sites are “edutcating” the Norwegian masses.

For international sites, you can look here.

finn.no is the biggest online buy and sell market in Norway. Their blogpost. They were the first to suggest doing the campaign, on Twitter

tu.no were first to publish a post about it

gamer.no is the biggest gaming site in Norway, and have a whole 47% of their users on Firefox. Still they want to help out

nettby.no is the biggest norwegian community online. Owned by vg.no

nextgentel.no is one of the bigger ISPs in Norway, they also run broadpark.no which has the same message

pwned.no is the biggest online game tournament site in Norway:

rb.no is a local news paper owned by the umbrella corporation A-pressen interaktiv. API has about 50 online papers, all with the same warning

sesam.no is a norwegian search engine and portal, who also have their own mapping service. They are a part of finn.no

start.no is an online entertainment arena for youth in Norway

studentersamfundet.no is the Norwegian student society

vg.no is the biggest news site in Norway. They don’t like IE6 either

db.no is the second biggest news site in Norway. Although late, they are now participating in the campaign. They also have an article

sol.no is a large news portal forwarding clients to articles on other sites

dn.no is a big site about business news

foreldreportalen.no is a parent portal

lavkarbo.no is a portal for taklking about different diets

nwo.no plans to take over the world and have one global unitarian government. They have a slightly different approach, where they actually thank people for using IE6 as the insecure browser it is, making it easier for nwo to accomplish their goal

gulesider.no Yellow pages

telefonkatelogen.no online phone book. Owned by gulesider.no

keyteq.no does online marketing

mac1.no is a mac portal

linux1.no is a linux portal

ipod1.no is an ipod portal

kreativ1.no is a portal for creative work and people. They also have an article about the IE6 campaign

aftenposten.no is the third biggest news site in Norway

morgenbladet.no is a news paper with emphasis on culture, art, literature and politics

sysedata.no is one of Norway’s biggest hosting providers and ISP

Stack Overflow is a collaboratively edited question and answer site for programmers — regardless of platform or language. Jump in and share your software engineering expertise! No registration or account required.

So head over and start collaborating and taking advantage of collective knowledge: Stackoverflow.com

Update: TextMate 2 should be in private alpha in a month’s time, and public beta in maybe six months. Read more here.

I’ve earlier covered what we can expect to come in TextMate 2, and this time I’m listing what’s come from Allan himself.
All quotes are of the author of TextMate, Allan Odgaard.

Search & replace, regex:

• Named captures will be availble as $variable.

  The named captures will be available as $variables. Source.

• Better syntax for using captures in replacement string. Is 123$456 $4, or $45, or $456? With the new syntax, there’ll be no doubt: 123${4}56.

  There will however be a significant update to this syntax in… the future This will allow using ${1} etc. like in snippets. Source.

• Introduction of “Find all”, which will also allow you to do further operations on found items.

  Also, I have another feature coming in 2.0 which will allow you to effectively do the further replacements on the results of a ‘find all’ (so even w/o SmartyPants.pl, there will be a simple way to do the same). Source.

Drawer:

Note: I think the drawer will be replaced by a panel in 2.x, but I have no sources for this at the moment.

• Menu shortcuts in context menu in drawer will work (better).

  The only solution is to have the items in the top menu and let the Carbon Menu Manager do all the magic keyboard handling that Cocoa applications can’t do — I will do this in the 2.0 relayout of things. Source.

• It will likely be possible to sort files in the drawer.

  Not possible. There will (most likely) be some sorting options in the future. Source.

• No more virtual TextMate projects. A folder will be the project.

  Because the file system is there to group files into folders. Having TextMate do the exact same thing is IMO conceptually wrong. It also complicates things like for example having commands run on a project and generally just makes your project “groupings” non-portable. Source. I would strongly recommend against getting too dependent on the virtual grouping that tmproj files allow, as I presently have no intention of keeping that in 2.0. Source.

Bundles and bundle development

The bundle editor as we know it will very likely be gone in 2.x. I’ll see if I can dig up a source for this later. I’m not sure what will replace it either, but seing how important bundles are to TextMate, I’m sure we can expect some kind of improvement. My guess would be that you work with the bundle as any other folder and take it from there instead of the bundle editor

• Dependency checking for bundles.

  That said, I am looking into a more formalized dependency system so e.g. the entire Git bundle (sans Help) can be marked to require git, or we can have a bundle require Leopard etc. Source.

• Command line interface to the TextMate parser.

  I will provide a CLI interface to the TextMate parser in the future, that way we can fully automate such testing. And yes, it really should be sort of a requisite then, to include tests with the grammars Source.

• CocoaDialog will likely be gone.

  We are deprecating the use of CocoaDialog (and I am not 100% sure TM 2 will ship with it), so it would be good to have a TextMate::UI wrapper for these, so we can switch the implementation to use DIALOG when we add file requester support here. Source.

• Self-aware projects.

  Yes, there are thoughts toward that Source.

Working with text

• Better indentation.

  I have these notes from Jacob: http://pastie.textmate.org/189818 Source.

• Improvements to overwrite mode

  There are a few things in the overwrite mode that doesn’t as they should. I hope to address all of that in 2.x. Source.

• Possibility to highlight discontinious text.

  I do however feel TM needs a way to highlight discontinuous text segments, and technically this can be made using dynamic scopes, which in practice can be used by a command to “parse out” the selection… Source.

• Ability to disable soft-wrap for specific parts of a document.

  No, but it’s certainly something I would like to have myself as well. Source.

• Select Current Scope will be improved.

  Some cases I am aware of and will address in the future, some cases probably cannot be solved at all, and some cases might automatically be solved with new grammar features. Source.

• Highlight of paragraph. not just current line.

  Not currently possible. It is not entirely unlikely that it might be possible in a future release. Source.

Other

• Find in project will work in a separate thread, to avoid hanging TextMate.

  In a future update, yes… Source.

• Logging of tool tips and hence the possibility to show them again by command if necessary.

  I think a log is an excellent idea. So basically all (command) tool tip output goes to ~/Library/Logs/ with a date stamp or similar, preferably with some newline escaping so a “repeat last tool tip” command would just tail -n1 «log», and similarly for a copy. Source.

• Submenu titles in inline menu (When you press a key equivalent that matches several items). Example: Source.

  I’ve added a TODO to the relevant source Source.

• TextMate will remember your session, open files, modified files etc. between sessions.

  Complete session restore, including state of modified files, is OTOH pretty high on the list. Source.

• TextMate clipboard will have better heuristics for intermixing third party copy into the clipboard.

  The heuristic used will be better in next major update, that might solve this compatibility issue. Source.

• TextMate 2.0 is in a state where it can be used.

  What I can say is that a) TM 2 is in a state where I use it myself Source.

Unlikely to be included

• Selecting different parts of text in a document at the same time and use it for copy and paste. Note: The unlikeliness refers to the copy and paste part

I have no real plans of this. I think my last comments on the topic was something like this not being very useful because of how it would not be easy accessible via keyboard only and it would introduce a lot of complexity in basically every single thing that potentially works with the selection, in many cases having to define how a transformation should work in the case of discontinuous selections (which isn’t always intuitive) — so hardly worth it. Source.

Final words

Of course this is not a complete list, it’s actually just what I cared to read through in a bored moment. For instance I know split-windows are requested a lot, too much for Allan to completely ignore it, but I have no idea when they will/if it will be included. I also think/know he has some other great ideas he just can’t talk about to avoid the competition snatching them. Most of the competition is probably subscribed to the mailing list.

I’d just like to add this final quote from Allan, which I think is very important that everyone respect and understand.

I have taken note of the request, it might be added to a future version, I can’t really say for sure before I have actually implemented it and shipped the version where it has been added.

I put emphasis here on shipped as well because sometimes a feature is pulled again for various reasons (could be that it turns out not to work so well, might be because it complicates the implementation of other features, etc.). This is one of the reasons that I am not commenting on specifics wrt 2.0. Source.

What would you like to be included?

I installed the gem,

sudo gem install passenger

and everything went just according to the plan. The next step was to run the installation script,

sudo passenger-install-apache2-module

and I was even greeted with this promising message:

This installer will guide you through the entire installation process. It shouldn’t take more than 3 minutes in total.

Here’s what you can expect from the installation process:

1. The Apache 2 module will be installed for you.
2. You’ll learn how to configure Apache.
3. You’ll learn how to deploy a Ruby on Rails application.

Don’t worry if anything goes wrong. This installer will advise you on how to solve any problems.

Press Enter to continue, or Ctrl-C to abort.

“Shush you. Just take me to the solution, damn it!”

Problems

Ok, so even if something was to go wrong, the installer would help me.

Checking for required software…

• GNU C++ compiler… found at /usr/bin/g++
• Ruby development headers… found
• OpenSSL support for Ruby… found
• RubyGems… found
• Rake… found at /usr/bin/rake
• Apache 2… found at /usr/sbin/httpd
• Apache 2 development headers… found at /usr/sbin/apxs
• Apache Portable Runtime (APR) development headers… found at /Developer/SDKs/MacOSX10.5.sdk/usr/bin/apr-1-config
• fastthread… found
• rack… found

--------------------------------------------

--------------------------------------------

Compiling and installing Apache 2 module… cd /Users/peterhaza/Compile/passenger-2.0.3 rake clean apache2 (in /Users/peterhaza/Compile/passenger-2.0.3) rake aborted! Could not find ‘apachectl’ or ‘apache2ctl’. /Users/peterhaza/Compile/passenger-2.0.3/rakefile:35 (See full trace by running task with –trace)

--------------------------------------------

It looks like something went wrong

Please read our Users guide for troubleshooting tips:

/Users/peterhaza/Compile/passenger-2.0.3/doc/Users guide.html

If that doesn’t help, please use our support facilities at:

http://www.modrails.com/

We’ll do our best to help you.

LIES

Lies, lies and more lies. The installer didn’t help me, the user guide didn’t either, and the support facilities cost money. Damn it!

Well, so I set out to fix it myself. which behaved nicely and returned what I expected,

peterhaza@Farquaad passenger-2.0.3$ which apachectl
/usr/local/bin/apachectl

so it’s definitely there. I started to scan through the installer script, which was just a nice ruby script, and after trying to add the correct path at the top of the script (where it already adds some stuff to $PATH) and fail, I decided to dig a bit deeper.

Solution

In lib/passenger/platform_info.rb I found a method called self.determine_apache2_bindir. It does some fancy look-up stuff, but instead I just told it to return the path where I knew my apachectl was.

def self.determine_apache2_bindir
  "/usr/local/bin"
# if APXS2.nil?
#   return nil
# else
#   return `#{APXS2} -q BINDIR 2>/dev/null`.strip
# end
end

Voila! After this change, I ran the script again and would you know:

Lots of compiling messages…

The Apache 2 module was successfully installed.

Please edit your Apache configuration file, and add these lines:

LoadModule passenger_module /Users/peterhaza/Compile/passenger-2.0.3/ext/apache2/mod_passenger.so PassengerRoot /Users/peterhaza/Compile/passenger-2.0.3 PassengerRuby /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/bin/ruby

After you restart Apache, you are ready to deploy any number of Ruby on Rails applications on Apache, without any further Ruby on Rails-specific configuration!

Press ENTER to continue.

More messages…

So there you go!

ph^: allan: Now that Duke Nukem forever has released a new gameplay-video … :p

Anyway. Being a regular on the irc channel and mailing list for some time now, I’ve gained quite much information on what’s in development and even bits and pieces of what TM2 will offer.

Update: TextMate 2 should be in private alpha in a month’s time, and public beta in maybe six months. Read more here.

Update: More TM2 speculation here

DISCLAIMER: None of this is officially released, supported or confirmed by Allan/macromates.com. Some or all of this may never be part of the TextMate distribution at all. It will however be cool if it is!

Dialog2:

Dialog2 is the successor of Dialog, which intent was to allow bundles to use dialogs to request strings, show choices etc. With Dialog2, Ciaran Walsh is bringing the Dialog to the next level by amongst other code changes is introducing a set of new window helpers. It’s also written in a way that makes it a lot easier to extend and add more features later. There already is a ruby wrapper in place(Support/lib/ui.rb) which makes it very easy to take advantage of the new Dialog.

Prettier tool-tips

By using HTMLTips helper from Dialog2 you get easy access to creating custom tool tips in HTML and CSS. The result, here showed with “Documentation for Word” from the PHP Bundle:

Code completion

The new Dialog can also be used for code completion. Some footage shows how Dialog2 suggests functions as you type, and inserts the function as a snippet where you can tab to the next parameter. With the new aforementioned ruby wrapper for the new Dialog it should be a walk in the park for the bundle developers to add this. (I managed!)

Intelligent code completion / Intellisense

One of the biggest griefs in the community with TextMate has been the lack of intelligent code completion, often referred to as intellisense. Currently you can use ⎋ (esc) to toggle thru your previously written words and a predefined set of words defined by the bundle author, but it offers no way to see what parameters the functions and methods expects. There have been attempts by some bundle authors, but now there’s work in progress for making that easier too. Currently it requires the bundle author to write a lexer. Some example footage can be seen here(PHP), 2, 3 and here(Ruby).

Better handling of command line input (or lack thereof)

Another big problem in TM has been with shell commands that might need input. The first example that comes to mind is the subversion bundle. There is no way for the bundle to know whether you’re setup with ssh keys, ssh agent, or have to type your password every time you commit. Hence if subversion is asking for a password on STDIN while the command is executing, it will fail or even stall TextMate. Luke “alkemist” Daley set out to sort this, and with some help from allan he came up with tm_interactive_input. This basically injects into the read function in OS X and with some black magic kung fu figures out if the read process belongs to TM. If so, it will open up a dialog requesting the input.

Pretty bundles

Thomas “subtleGradient” Aylott is working on what he refers to as:

a rough framework for making gui applications built inside textmate command windows

and it looks like this:

It allows a bundle developer to create a GUI using HTML/CSS & Javascript. And the buttons in the GUI can run shell scripts and pass data to them interactively. With the possibly coming terminal stuff in TM2 it should also gain the ability to run any TextMate bundle commands, macros & snippets. .

TextMate 2

There isn’t a whole lot to say about TM2 yet, other than it’s coming. Allan has said several times that he does most of the TM2 development in TM2, so it’s there and it’s mostly working, it’s just not for the rest of us to see and use yet.

One feature that might make it into TM2 is the ability for bundles to get a copy of the current text buffer, which will allow a bundle to do some heavy scanning in the buffer, without affecting the responsiveness of TM. This is great news for auto-completion and other commands which needs to scan the whole document.

Another often requested feature is for bundles to have better access to TextMate’s scoping system. This could replace the need for a lexer for things like intellisense, and thus lower the bar for implementing such. If this makes it into TM2 only Allan knows, but I really do hope so.

I don’t really know what people expect from TM2, but your expectations should most likely be pointed towards bundles, not TextMate itself. TM2 is afaik a complete rewrite of TM, with a lot of new “invisble” features for the bundles to take advantage of, not the users directly. TextMate’s strength has never been the editor on its own, but the extensibility it supports, and thus bundles.

Other interesting things

• Yehuda Katz, with help from others on github.com has released a ruby gem that allows you to install, uninstall and list TextMate bundles like normal ruby gems. This makes it very easy for those out there who aren’t friends with subversion. You can read more about it here.
• There has been a lot discussion on how to best maintain bundles in a central repository. There has been many requests for a issue tracking system for the bundles, and I recall that both Jira and Lighthouse have been in the spotlight, but I think eventually Redmine pulled the longest straw due to its free and open nature, and that Allan can host it on macromates.com. But either way, nothing is written in stone, and there has been no final decision.
• There has also been talk about moving from subversion to git, but I’m not quite sure this is going to happen. Due to the nature of bundles, where each bundle has one or more main maintainers, but also super maintainers who maintain all over, Allan has mentioned that subversion is doing this pretty good already, while git might be difficult here. Nothing is written in stone here either, but it is on the agenda.

$form->getWidgetSchema()->setFormFormatterName('list');

The question is, “where do you put this to alter sfGuard’s login form?”

In the case of sfGuard you just need to create a new folder under apps/frontend/modules/ (replace frontend with your app name) called sfGuardAuth, in this you create a folder called actions, and within this again a file called actions.class.php.

apps/
-frontend/
--modules/
---sfGuardAuth/
----actions/
-----actions.class.php

Inside the actions.class.php file you put this:

<?php
    require_once sfConfig::get('sf_plugins_dir').
    '/sfGuardDoctrinePlugin/modules/sfGuardAuth/lib/BasesfGuardAuthActions.class.php';

/**
* Extend to use list schema formatter
*/
class sfGuardAuthActions extends BasesfGuardAuthActions
{
    public function executeSignin($request)
    {
        parent::executeSignin($request);
        $this->form->getWidgetSchema()->setFormFormatterName('List');
    }
}
?>

We manually include BasesfGuardAuthActions.class.php because symfony’s black magic autoloader doesn’t pick this up. Voila. Nice and easy!

If you wanna make your completely own sfWidgetFormSchemaFormatter you can read about it over at notjosh’s blog.

For anyone who has played with Ruby on Rails, symfony will be very familiar. You got a nice command line interface to do your scaffolding, migrations and application building, and of course the well-known and diffuse MVC coding paradigm to seperate business logic from presentation.

Symfony is very well documented in english, something I didn’t expect as the author company, Sensio labs is a french web development company. It totally blew me away.

You do your configuration in XML or YAML whereas the latter is recommended because it’s alot easier on the eye.

So far I’ve only scratched the surface of symfony, but as I’m doing a full-scale application in symfony these days, I expect to get more familiar and have more to present later.