C++ / Qt / Mac / Win / Delphi / JavaScript

Google analytics event tracking: outbound links

2017-02-19T02:20:00.000-08:00

Ok, if you want to track outbound links (when user clicks external link) -- and you are using jQuery -- add this somewhere (after importing jquery):

$(function() {
    //add event tracking just if we have google analytics loaded
    ga(function(){
        $("a.track").click(function(e) {
            var url = $(this).attr("href");
            ga("send", "event", "outbound", "click", url, {"hitCallback":
                function () {
                    document.location = url;
                }
            });

            e.preventDefault();
        });
    });
});

Then just add track class to your links:

<a href="http://something.com" class="track"> outbound link </a>

Find certificate expiration date and other info of android keystore file

2017-02-06T16:30:00.002-08:00

Easy as single command:

keytool -list -v -keystore release-key.keystore -alias mydomain

- replace release-key.keystore with your keystore file name

- mydomain with the certificate alias

Upgrade / Downgrade PHP on Mac OS X

2016-11-24T13:01:00.003-08:00

Quick and easy:

https://php-osx.liip.ch/

And for more complicated setup ;-):

https://getgrav.org/blog/macos-sierra-apache-multiple-php-versions

Firebird auto increment

2016-10-21T08:37:00.000-07:00

Ouch, it happens that Firebird doesn't support auto increment fields out of the box.
Quick look on the web shows the same solution everywhere: triggers & counters.
But I lost over a hour trying to figure why it didn't work in my case (Lazarus+embedded Firebird). So, finally, I found the right syntax:

Create auto-increment field in Firebird:

CREATE TABLE ctmaster (
 id integer NOT NULL,
 name varchar(128) not NULL PRIMARY KEY,
   shtiftoredove INTEGER not NULL
)
CREATE SEQUENCE ctmaster_id_sequence;

CREATE TRIGGER trgTTEST_BI_V3 for ctmaster
active before insert position 0
as
begin
 if ((new.id is null) or (new.id = 0)) then
 begin
 new.id = gen_id( ctmaster_id_sequence, 1 );
 end
end

(the field that will get the auto-increment value is ID)

git - discard your master branch and replace with other

2016-10-17T16:04:00.003-07:00

If you get into situation (like me) to find your master branch totally messed, but with some other branch that fixed the mess, and there is no way on earth to merge them without losing your mind, here is the solution: make the other branch to be the master without merging and just discard the old master...

Funny, but -- thanks lords -- stack overflow will help:
http://stackoverflow.com/questions/2763006/change-the-current-branch-to-master-in-git

quote from the page above:

I have a repository in git. I made a branch, then did some changes both to the master and to the branch.

Then, tens of commits later, I realized the branch is in much better state than the master, so I want the branch to "become" the master and disregard the changes on master.

I cannot merge it, because I don't want to keep the changes on master. What should I do?

and the answer:

The problem with the other two answers is that the new master doesn't have the old master as an ancestor, so when you push it, everyone else will get messed up. This is what you want to do:

git checkout better_branch
git merge --strategy=ours master    # keep the content of this branch, but record a merge
git checkout master
git merge better_branch             # fast-forward master up to the merge

If you want your history to be a little clearer, I'd recommend adding some information to the merge commit message to make it clear what you've done. Change the second line to:

git merge --strategy=ours --no-commit master
git commit          # add information to the template merge message

Lazarus / Delphi ToolButton drop-down menu on button click

2016-10-15T01:13:00.002-07:00

Yeah... I am now (again, after few years) working on Delphi, but this time -- using Lazarus; Still, on the day one, I found one crazy problem I spent over 15 minutes to solve.

The problem: Delphi (and Lazarus) TToolButton has a style property you can set to tbsDropDown. In such case you get a button with small nice little triangle in the right part and when you click the triangle you get the attached PopUpMenu. But I want the button itself (no matter where you clicked) to open the popup menu while by default it is set to invoke some default action on click.

My first idea was to create onClick handler where to force the popup menu to show with specified x,y position. But that's clumsy and little tricky. And it turned out there is way easier method:

Just add onClick event handler to the ToolButton and use CheckMenuDropdown:

-------

procedure TForm1.tbTableClick(Sender: TObject);
begin
tbTable.CheckMenuDropdown;
end;

-------

(tbTableClick is the event handler for tbTable button onClick event; the toolbutton itself is tbTable)

BTW -- I do like how much Lazarus looks and feels like the real old Delphi (without the crazy price tag that made me to quit using Delphi on first place)...

Natural Sort in JavaScript

2014-10-12T13:06:00.004-07:00

Here is something I am saving here for future reference: a funny and working way to do natural sorting in JavaScript. See the original thread here: http://stackoverflow.com/questions/15478954/sort-array-elements-string-with-numbers-natural-sort

[v.1]: will sort items starting with numbers first, letters second
[ 1,1a,1b ... 5a,5b, a1,a2...]
//----------------------------------------------


function strcmp(a, b) {

    return a > b ? 1 : a < b ? -1 : 0;

}



function natcmp(a, b) {

    var x = [], y = [];



    a.replace(/(\d+)|(\D+)/g, function($0, $1, $2) { x.push([typeof($1) == 'undefined' ? Infinity : $1, $2]) })

    b.replace(/(\d+)|(\D+)/g, function($0, $1, $2) { y.push([typeof($1) == 'undefined' ? Infinity : $1, $2]) })



    while(x.length && y.length) {

        var xx = x.shift();

        var yy = y.shift();

        var nn = (xx[0] - yy[0]) || strcmp(xx[1], yy[1]);

        if(nn) return nn;

    }



    if(x.length) return -1;

    if(y.length) return +1;



    return 0;

}



var ary = ["5б", "555а", "12б", "12в", "8г"];



ary.sort(natcmp)



console.log(ary);

//----------------------------------------------
[v.2]: will sort items starting with letters first, numbers second
[ a1,a2 ... 1,1a,12 ]
//----------------------------------------------

function strcmp(a, b) {

return a > b ? 1 : a < b ? -1 : 0;

}

function natcmp(a, b) {

var x = [], y = [];

// а, а1, а2, ... 1, 1а, 1б:

a.replace(/(\d+)|(\D+)/g, function($0, $1, $2) { x.push([$1 || 0, $2]) })

b.replace(/(\d+)|(\D+)/g, function($0, $1, $2) { y.push([$1 || 0, $2]) })

while(x.length && y.length) {

var xx = x.shift();

var yy = y.shift();

var nn = (xx[0] - yy[0]) || strcmp(xx[1], yy[1]);

if(nn) return nn;

}

if(x.length) return -1;

if(y.length) return +1;

return 0;

}

var ary = ["5б", "555а", "12б", "12в", "8г"];

ary.sort(natcmp)

console.log(ary);

//----------------------------------------------

Opencart 1.5.5.1 -- fixing menu bug in IE10

2013-02-28T02:02:00.001-08:00

Yep, it did it again: Internet Explorer 10 gets things wrong with the default OpenCart menu.

It should be:

But in my brand new Internet Explorer 10 (Windows 8 / 64) it looks like this:

The folks from opencart team are saying they'll fix it in the next release, but for those who need the hack right now -- here it is:

1) Open the common.js file in catalog\view\javascript folder

2) Locate this piece of code:

// IE6 & IE7 Fixes
if ($.browser.msie) {
   if ($.browser.version <= 6) {
      $('#column-left + #column-right + #content, #column-left + #content').css('margin-left', '195px');
        
      $('#column-right + #content').css('margin-right', '195px');
  
      $('.box-category ul li a.active + ul').css('display', 'block');  
   }
     
   if ($.browser.version <= 7) {
      $('#menu > ul > li').bind('mouseover', function() {
         $(this).addClass('active');
      });
        
      $('#menu > ul > li').bind('mouseout', function() {
         $(this).removeClass('active');
      });  
   }
}

3) Replace the above snippet with this:

// IE Fixes
if ($.browser.msie) {
if ($.browser.version <= 6) {
$('#column-left + #column-right + #content, #column-left + #content').css('margin-left', '195px');

$('#column-right + #content').css('margin-right', '195px');

$('.box-category ul li a.active + ul').css('display', 'block');
}

if ( ($.browser.version <= 7) || ($.browser.version == 10) ) {
$('#menu > ul > li').bind('mouseover', function() {
$(this).addClass('active');
});

$('#menu > ul > li').bind('mouseout', function() {
$(this).removeClass('active');
});
}

}

4) Save and upload to your server.

5) Reload

http://wonderwebware.com/blog/2013/02/28/how-to-fix-opencart-1-5-menu-bug-in-internet-explorer-10/

chrome web store snapshot: 2012

2013-02-08T05:05:00.002-08:00

This is just a collection of links with quotes (as they are at the time of writing).

What I saw on the chrome web store battleground the last year? Pretty much what follows:

1) Wikipedia still says: "TechCrunch reported that sales of paid web apps had by January 2011 sunk, with the best-selling app earning just $165 a week." ("Sales Are At A Trickle On Google’s Chrome Web Store") nobody changed or improved this statement in 2012 ;-)

2) Medianama reported the stats (and that's what matters most for me right now):
- Google Chrome now has 310 million active users
- users have installed 750 million apps from the Chrome web store till now.
Compare this with the statement that "Gmail currently has 425 million active users" and you can get to believe in the numbers.
Source: http://www.medianama.com/2012/06/223-the-lowdown-google-io-2012-day-2-310m-chrome-users-425m-gmail-more/

3) chromeosapps.org is still the nice all-in-one resource it's supposed to be

--- bonus, no important stuff below ---

And just for those who want more -- quick history and hints about the chrome web store from the past (how loooong ago it was, 2-3 years ago...):

Chrome Web Store Passes 30 Million Users, But How Bright Is The Future?

Sales Are At A Trickle On Google's Chrome Web Store

The Chrome Web Store: get paid to make things

http://www.quora.com/Is-anyone-making-money-on-the-Chrome-Web-Store

http://www.lucidchart.com/blog/2012/10/31/500000-chrome-web-store-installs-what-weve-learned/

Qt 4.8.3 / QtCreator 2.6 -- missing rcc.exe fix

2012-11-13T14:58:00.003-08:00

I found strange problem today (installed the latest QtCreator and Qt 4.8.3 libs) -- for some weird reason rcc compiler wasn't starting in the build process. It turned out to be messed path or something like that. The fix:

1) open the qmake.conf of your platform (for me its C:\Qt\4.8.3\mkspecs\win32-g++)

2) search for QMAKE_IDC

3) add the following line after the QMAKE_IDC line:

QMAKE_RCC = $$[QT_INSTALL_BINS]$${DIR_SEPARATOR}rcc.exe

4) save file and clsoe

5) restart all your dev tools and remove all auto generated Makefiles

6) recompile

This and other workarounds available here: https://bugreports.qt-project.org/browse/QTBUG-27237

How to fix Qt app GUI in Windows 8

2012-11-13T11:14:00.002-08:00

I found that my app looks ugly on the new windows 8; Since digia is not yet ready with the Win'8 support, here is what you can do to fix the ugly controls:

for example, in the main.cpp file, add this:

QApplication app(argc, argv);

...

#ifdef Q_WS_WIN //if Windows -- force using XP style GUI

    QApplication::setStyle(new QWindowsXPStyle);

#endif

...

Hope that helps.

All the app stores in one place

2012-09-28T00:43:00.003-07:00

This is nice list with app stores (and being a developer you may wish to check it):
http://blog.inner-active.com/2011/04/a-complete-list-of-app-stores-across-all-mobile-platforms/

Fixing Awesome Ads wordpress plugin...

2012-07-11T00:07:00.000-07:00

I use Awesome Ads plugin on some of my wordpress sites (and some of my client's sites) and I found the plugin author tends to change DNS settings from time to time (moving sites?) or just there are some server downtimes on his end. And that results in failed /wp-admin access. When trying to login in the WordPress backend, you just get a long wait and crash messages (php can't connect to plugin's site and with this failed call-back-home funtctionality it crashes the whole wordpress admin panel).

Luckily, for any php developer it is an easy fix. Still, if you don't know what to do, here it is:

1) Find and open this file:

/wp-content/plugins/awesome-ads/libs/functions.php

2) Go to line 146 and add the two lines in bold below right after that line, it should look like this:

function verify_adult_content() {

ini_set("allow_url_fopen", 1);

return false;

3) Upload the changed file to the server

CloudFlare vs. Google Analytics vs. AWStats/Webalizer/Analog Stats

2012-05-08T14:26:00.001-07:00

I found some questions flying around today about how CloudFlare counts pageviews (some people seems to think the cloudflare stats aren't real compared to GA for example... So I decided to check with real site what all these different stats will say. Pretty much as I expected, everything looks right for me, but here are the numbers (for the last 7 days):

CloudFlare:

35,546 pageviews
26,090 from regular traffic
2,560 from crawlers/bots
6,896 from threats

AWStats:

Viewed traffic -- 23,785

Not viewed traffic* -- 20,973

Webalizer:

Total Pages -- 20505

Analog:

pages: 18841

Google Analytics:

Pageviews: 6,637

So, go get it! AWStats reports over 44,000 pages, CloudFlare -- over 35,000, Webalizer -- 20,000, Analog -- 18,000 and Google Analytics just below 7,000. Still, when it comes to regular (viewed, non-bot traffic) -- things are comming closer. CloudFlare and AWStats are saying pretty much the same (23/26K sounds OK, given the fact they are saving/processing logs in different times and ways), and if we assume Webalizer and Analog are showing only the "viewed" pages -- it will looks like all these are showing the "actual" server stats. And then it is all fine with Google Analytics, cause it's quite real to get 3 times difference in this site when measuring with JavaScript instead of server logs.

I can confirm that CloudFlare stats are real when it comes to page views.

Chrome Web Store App Developers: don't do that!

2012-04-17T01:05:00.002-07:00

I've made one very interesting observation and I wish to make it public for anyone who may accidentally find this blog post. It is something hidden in the way Chrome Web Store works, but you may get a nasty surprise if this happens. Here is the story:

A week ago I made an update in my experimental title in Chrome Store (it is a Calculator app, mentioned several times before in this blog). The experiment was initially meant to test jQueryMobile in the real life, but somehow it turned to be a useful chrome app and made thousands of users in the chrome web store. So, I wanted to give my support to this tool and from time to time I am updating the app to add user-requested features. This time it was relatively serious update -- with copy/paste support, among other things -- and I was happy that I had the time to make it. What I didn't expect was how this will reflect to the app popularity at all.

Before the update there were above 31,000 users of this app, as reported by Chrome Web Store. Suddenly, a week later, the store says the same app have 20,000 users. Crazy, isn't it? Is it realistic for an app to lose 1/3 of it's users in just a week? Well, there are 2 main hypothesis here: that something is/was wrong with the chrome stats (unlikely) and that something in that update forced current users to remove the app. I tend to believe the second one is the right one. Read above -- I added copy/paste functionality to the app. It was nice addition and some users wanted it. But in Chrome world that means permissions. So, on the very next "silent" update (you know -- chrome updates all apps "silently" behind the scene) a nasty pop-up message showed, informing the users that this new version needs more permissions. A single message, that -- I believe -- destroyed 30% of the current user base.

What's the moral?

1) Never change the permissions required by a chrome app or extension. That will force the browser to ask for user acceptance once again, but not during the initial app installation. Months later the end user may not even remember what is that app and will just forbid the change, thus meaning will actually deinstall the app. Set the app/extension permissions from the beginning and don't touch them in any future update, even if users are requesting new feature that rely on these changed permissions.

2) If you need to do that, be prepared to lose a lot of current users. In my case that single change cost me 30% drop.

Update 1 [18 april]: I wrote this post this morning. Since then things have changed again. Now, exactly 21:21 my time the chrome web store dashboard says the app have 16,329 users. This is weird. Two days ago they were almost 32,000. Then, last night they dropped to just above 20,000. Now, less than 8 hours later, they are 16,000. I believe it is quite possible in the next days the app to completely disappear from the world ;-)

The strange thing is, however, that my Analytics stats (you know -- these google analytics we all use) are not showing any such drop in the usage and installs. I have analytics tracking added to this app, and it shows almost the same daily uses as always... Well, will give that phenomena some time and will report with more later.

Update 2 [19 april]: Another day -- another drop. This morning chrome web store shows 14,656 users which means another (almost) 2 thousand users were lost in the upgrade.

Update 3 [20 april]: It became a habit to update this post (and I will until the drop turns into growth again). Today chrome web store reports weekly users = 14,041, which means only 600 users have been lost for the last 24 hours. Seems like the big wave passed already? We'll see...

Update 4 [22 april]: 13,417 left, and that makes me think I will end with less than 1/3 of the initial number ;-) Oh, yes, I also wrote about this in my "official" blog:

Chrome Web Store for Developers: How to Destroy your App’s User Base with Just One Update

JavaScript doesn't react well on double-pluses

2012-04-11T15:03:00.000-07:00

What's wrong with this code?:

copy1 = ""+( ('0' + d.getFullYear()).slice(-2) + '-'
+ ('0' + d.getMonth()).slice(-2) + ''
+ ('0' + d.getDate()).slice(-2) + '*' + ''
+ ('0' + hours).slice(-2) + ':' +
+ ('0' + mn).slice(-2) + '' );

Well, after almost 20 minutes of thinking (as an excuse -- remember that this was just a piece of code inside several hundred lines) I found what was wrong and how to fix it:

+ ('0' + hours).slice(-2) + ':' + ''
+ ('0' + mn).slice(-2) + '' );

A simple garbage not collected (the second plus) and puff -- the slicing wasn't working and time got wasted. See what happens when you put the wrong code in one line:

[...] + ('0' + hours).slice(-2) + ':' + + ('0' + mn).slice(-2) + '' );

Did you see it coming? I didn't even know JavaScript can handle such case, but in any case, it doesn't handle it as one would expect. Firebug didn't warn with these nice red errors in the console. And the programmer was lazy (by cutting some piece of code w/o checking what follows)...

So, please, delete the unneeded + symbols when you cut something from a list ;-)

It's JavaScript, not c plus plus, so doubled pluses aren't welcome ;-) /not that they are in cpp/

JavaScript num padding (external link)

2012-04-05T16:00:00.002-07:00

Here it is:
http://www.electrictoolbox.com/pad-number-zeroes-javascript-improved/

A nice example of why there are different ways to do the same thing ;)

JavaScript Namespaces?

2012-04-04T15:39:00.002-07:00

From this thread:
http://stackoverflow.com/questions/881515/javascript-namespace-declaration#answer-3588712

var yourNamespace = {
foo: function() {
},
bar: function() {
}
};
...
yourNamespace.foo();

or:

var ns = new function() {
var internalFunction = function() {
};
this.publicFunction = function() {
};
};

and naturally, the "pro" solution: to use separate library:
https://github.com/smith/namespacedotjs/blob/master/example/sandbox.js

Well, I am often mixing the 1st two and never actually tried the 3rd option. But for some really big project, definitely opt for the last one...

Whanna be a micro ISV?

2012-03-28T14:45:00.001-07:00

A must-read blog post on independent software vendor startup:

http://codebetter.com/patricksmacchia/2008/11/03/advices-to-developers-on-starting-an-independent-software-vendor-isv-business/

And don't forget to check these two links (posted as a comment under the above post):

http://www.userscape.com/blog/...
http://www.startupspark.com/ia...

and http://successfulsoftware.net/2009/01/05/running-a-microisv/ (for some business hints)

And if you ever start into that game: may the force be with you!

how long it takes to create new version of your software?

2012-03-28T03:21:00.000-07:00

I have a site where I am publishing some web design tools (wonderwebware.com). And right this morning I received an email from a user who requested simple new feature for one of my programs (duplicates Finder). It was definitely a quick and easy change and it took me exactly 21 minutes to add the requested new features, test and re-pack (including all these signing, editting setup files, rebuilding setup etc.).

Now, can you guess how long it took with the additional extra work (that isn't in any case related to programming)?

Here is what my time-tracker says:
1) preparing for work -- 45 min. -- this includes searching for the latest version in my never-ending messed archives, copying/extracting files etc.
2) programming, packaging -- 21 minutes -- including writing code itself, code signing exe's, re-building setup, signing the setup and initial test (on the dev.machine)
3) testing, backup -- 5 min. -- these are tests on different boxes (XP and 7) and archiving the new version
4) uploading to wonderwebware.com, adding to the news, fixing version info on site... -- 11 min. -- these are just the usual fast changes (to change version numbers, to add a newsflash line on the front page and similar)
5) final test (download and install on separate computers) -- 4 minutes -- only to go to the site, download the program (as a user), install, test...

So, for anyone curious, it took me 86 minutes to get the new version online, and half of that time went in useless searching in messy archives. What's the moral? -- Always keep your archives organized and always backup the dev. projects in the same folder structure, keeping the main master backup location the same for all your projects!

Anyway, if I wasn't so lazy in my internal organization, I wouldn't waste 45 minutes for nothing. But still, it would take about 5 minutes (presuming that backup itself takes 5 minutes to be created and stored). So, we still get about half of the time spent for "additional" tasks (not development).

And now, imagine that this is just the beginning. We should now start promoting the new version -- read: submitting to download sites, writing some news articles, promoting the news articles, fb, twitter... You get the idea. I see another 2-3 hours spent in happy submitting only. Ughrr. Next time someone asks me "how long it takes to create new version of the X program", I'll not be wrong if I say: a day...

JavaScript should be short!

2012-03-23T18:08:00.000-07:00

I was playing with some JavaScript timers today and used to find this code sample. (It was a function to return the current time):

function getTime() {

var datetime = new Date();

var hours = datetime.getHours();

var minute = datetime.getMinutes();

var seconds = datetime.getSeconds();

var period = "AM";

if (hours > 12) {

period = "PM"

}

else {

period = "AM";

}

hours = ((hours > 12) ? hours - 12 : hours);

return hours + ":" + minute + ":" + seconds + " " + period

}

Although that will work and is simple to understand (and I was trying to find nice tutorials on first place) -- I can't stop thinking of the so underestimated common sense when it comes to performance. It's meant to be downloaded (that's why we have minifying tools) and interpreted (that's why we want faster processors). It's so simple. Why are we giving such a bad example to the kids here? What the else block does above? Why we are using variables for minutes and seconds when we could just invoke the getMinute / getSeconds methods when needed (only once)? What about something like this:

function getTime() {

var dt = new Date(),
hours = dt.getHours(),
period = "AM";

if (hours > 12) { period = "PM" }

hours = ((hours > 12) ? hours - 12 : hours);
return hours + ":" + dt.getMinutes() + ":" + dt.getSeconds() + " " + period
}

Isn't it just as readable as the original?

But then, there is always bad practice in the other end of the specter:

function gT() {
var d = new Date(), h = dt.getHours(), p = "AM";
if (h > 12) { period = "PM" }
return ((h > 12) ? h - 12 : h) + ":" + d.getMinutes() + ":" + d.getSeconds() + " " + p
}

So, may be it's the right balance after all?

CloudFlare for your site...

2012-03-06T13:25:00.000-08:00

Yesterday I finally decided to give www.cloudflare.com a look*. And this post is just a quick note to remind myself to write more about these guys in future. Keep in mind that I am not affiliated in any way with them (and the link above is not an affiliate link or something like that). I was just trying to find a good CDN for my needs without paying the big fat prices one will imagine. But CloudFlare is free (and even the paid version costs $20 / month). And they offer a lot of security options too...

Here is a screenshot of the statistics panel in my account after one single day of usage (the first one):

And as you see, there is a lot of sense in using their service.

They are not only caching and distributing content around their servers, but also stopping hackers, spammers, zombie machines etc.

The setup is piece of cake (the only thing you will need to do is to replace the dns servers in your registrar's account with theirs (for GoDaddy it took me about 4 hours for the dns change to be propagated)

Good work, people! If the service works as so far, you have one webmaster with a bunch of sites to pay for the "pro" version!

*Funny quote from a forum post:
"Basically, CloudFlare is a distributed nginx caching reverse proxy network with real-time firewall protection"
(http://www.webhostingtalk.com/archive/index.php/t-1115500.html)

WordPress Plugins Collision: W3 Total Cache crashes WP-Filebase

2012-03-06T00:14:00.002-08:00

I found a strange thing happening when installed and enabled W3 Total Cache plugin in my WordPress blog. Right after the installation, another plugin -- WP-Filebase -- just stopped working. The plugin wasn't affected, but the download links was not functional anymore (producing 404 Error page instead of starting the download). If you have the same problem (tested for these two plugins but may work for other download plugins as well) -- here is what I did:

1) Go to W3 Total Cache general settings (click "Performance" menu item from the WordPress admin panel) and make sure "Browser Caching" is enabled. I suppose it is (because here the problem lies, but still it doesn't hurt to check). It should look like that:

2) Now, click the "Browser Cache" menu to go to the browser cache settings page:

3) Uncheck the "Prevent caching of objects after settings change " checkbox and click the "Save" button:

Now, go to your blog, reload the page and try again. In my case that was the only option changed in W3 Total Cache that fixed the broken downloads in WP-Filebase.

PS: Also, better keep "Database Cache" option disabled in W3 Total Cache performance settings.

A must follow link for HTML5 developers

2012-03-02T06:35:00.001-08:00

Everyone and his mother must keep an eye on this man's blog:

http://coenraets.org/

A lot of code, examples and wonderful style of writing.

Multicore is Mainstream!

2012-03-01T10:01:00.001-08:00

Single-core CPUs are doomed. It is time for anyone seeking performance to check the options to adopt multi-threading in it's software. A link to start:

http://software.intel.com/en-us/articles/planning-for-parallel-optimization/

(And look for the chart on the page above)