10x website acceleration with WEBO Site SpeedUp

Store Builder Questions
61

Lee, we use WEBO Site SpeedUp to perform all optimization actions for our website. Due to some reasons (static CDN over dynamic engine) it’s not possible (right now of course) to enable data:URI / mhtml techniques, so we use CSS Sprites only.


[quote name=‘Lee Li Pop’]Hello Webo,



I have two questions.



1 - Why don’t you combine images to save HTTP requests, your website could load faster yet?

2 - Why don’t you use CSS sprites?

[/QUOTE]



Pingdom results can’t be accessible as precise. It parses CSS files, finds all links to CSS images, and counts them. To keep backward comatibility there are IE6 rules (due to PNG32 usage) w/o CSS Sprites. The main CSS Sprite contains about 100 images, and actually there is a few out of this file

[URL=“http://www.webogroup.com/cache/webo.be302eb665.wo1284902835.png”]http://www.webogroup.com/cache/webo.be302eb665.wo1284902835.png[/URL]



Also WEBO Site SpeedUp can use CSS Sprites for HTML images (we call this techniques HTML Sprites), but due to static HTML pages for all users ths approach don’t work for our corporate website (I think it’s a temporary situation, and we add HTML Sprites support even for such cases).



Maybe our corporate website, www.webogroup.com, can be loaded a bit faster, but we are inventing some generic approaches to improve our products and make all websites faster, not just only ours :slight_smile:

62

Hello Webo,



I tried WEBO free Edition on one of my website Vs. another without WEBO, just optimized with Smartoptimizer.



In fact, I have many CS-Cart websites, with same release, same skin (CSS) same webhosting, and so on… So, it’s easy for me to run a contest.



I add this in the beginning of my index.php file:


require('/usr/www/users/xxxx/web-optimizer/web.optimizer.php');



I add this in the end of my index.php file:


$web_optimizer->finish();



So, here are results:



Without WEBO => 1 => Smartoptimizer Optimized

With WEBO => 2 => Extreme WEBO Optimized



*****

Step 1:



Google Speed Page:



1: 80

2: 82



Yslow:



1: B; 87

2: B; 85



*****



Step 1 Conclusions:



Nearly ex aequo



*****



Real Time to Loading in Browser (IE, Chrome, Firefox, Opera)



Average Product Page Cold (not in Broswer cache. Never seen) Loading



1: 3 seconds

2: 12 seconds



Average Product Page Hot (in browser cache. Already seen) Loading



1: 3 seconds

2: Less than 1 second: With display mistakes (French characters mistakes PLUS CSS mistakes under IE)



*****



Average Category Page Cold (not in Broswer cache. Never seen) Loading



1: 9 seconds

2: 18 seconds



Average Category Page Hot (in Browser cache. Already seen) Loading



1: 3 seconds

2: Less than 1 second: With display mistakes (French characters mistakes PLUS CSS mistakes under IE)



*****



Second Step Conclusion:



On cold request (ie. from Search Engine to Website) WEBO is TOOOO long, up to 3 times longer than Smartoptimizer.



On hot request (ie. seen before) WEBO is 3 times faster than Smartoptimizer, but it has mistakes PLUS CSS mistakes under IE.



Smartoptimizer Wins.



*****



Final Conclusion :



Smartoptimizer wins thanks to its fastness and stability.



I think WEBO could be an excellent product if correctly tweak, provided that ALL pages are present in the cache of the site, ie. in the /web-optimizer/cache directory, but I do not know if WEBO is capable of put every cached pages automatically by itself.







Lee Li Pop

63

I’m not saying WEBO is bad. I simply prefer SO because it doesn’t brake anything when upgrading CS. I was able to move from 1.3.5 to 2.0.15 with no issues. Then for testing purposes only, from 2.0.15 to 2.1.1 in a temp directory. SO works perfect for me and it takes 3min to install it and tweak it. Not hours of scratching my head. Time is money.



I believe, WEBO may be good but with so many options and possibilities, it doesn’t look like it’s good for beginners or just PHP-uneducated e-store owners. I prefer a simple solution: good quality VPS + SO + some add-ons disabled and that’s it.



The final result you get with WEBO, for the trouble, time and money you spend on it…hmm, you decide.

64

Hello Noman,



Yes, WEBO should be a good product, once correctly tweak.



Moreover, it is an essential tool, vital, for a site of e-commerce, here’s why:



1 - Google is using site speed in web search ranking:



[url]http://googlewebmastercentral.blogspot.com/2010/04/using-site-speed-in-web-search-ranking.html[/url]



2 - Revenue drop on slower site:



3 - Performance Related Changes and their User Impact:



Phil Dixon of Shopzilla for his part noted in dropping the loading time of his site from 7 seconds to 2 seconds:



25% of page views and more

[SIZE=“3”]7 - 12% of sales more[/SIZE]



[url]http://velocityconference.blip.tv/file/2279751/[/url]



4 - Time is Money



From one to five tenth second: This is the limit to which we have unconsciously believe immediate action on what is handled. Word, MacOS or Windows or a telephone interface must respond to the area to be assessed. The brain interprets, orders, receives, as if the tool used was an extension of the body it controls directly.



From five to tenths of one second: The perception is changing, we no longer feel in control “directly”. The brain assimilates the information and no longer has the illusion of control the interface as it would control a member of the body.



From 1 to 5 seconds is the usual time of loading sites. The user agrees in general fairly well. If a load must be longer than 5 seconds, it is interesting to occupy the visual space of the Internet to keep his attention (a loading bar, a percentage). Keep the attention more than 10 seconds without taking a meaning is illusory.



From 10 seconds: The brain awaits the response and wanders. It wonders if it will not simplest to go to another competitor’s site, if he has paid its taxes, if the webmaster sends the pages by mail or on the backs of hamsters.



In short, the sale will probably not concluded.



A grandmother can wait 20 seconds the web page of the site to book their holidays. Bob, fifteen year old, beyond the two seconds, he wondered if the pages are sent to him by carrier pigeon or hamster back.



So, if correctly tweak, WEBO is really a service that can really fatten your wallet.





Lee Li Pop

65

Hello Noman,



Your CSS files are affected by the Smartoptimizer disease, ie. it changes the URLs of images by their equivalents in Base64, overweighting up the CSS files, turning light files in very big files…



For example into your CSS files:



/image/back.gif



Becomes



background: # fff1a8 url (‘data: image / gif base64 R0lGODlhEAAQAPQAAP/xqD8AAPnrpGk1JaJ9V0MFA1skGd/Ljb2fb08UDpdvTotgQ rYlrGQZNO7gnVEL39ROAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH + + + GkNyZWF0ZWQgd2l0aCBhamF4bG9hZC5pbmZvACH5BAAKAAAAIf8LTkVUU0NBUEUyLjADAQAAACwAAAAAEAAQAAAFUCAgjmRpnqUwFGwhKoRgqq2YFMaRGjWA8AbZiIBbjQQ8AmmFUJEQhQGJhaKOrCksgEla KIkYvC6SJKQOISoNSYdeIk1ayA8ExTyeR3F749CACH5BAAKAAEALAAAAAAQABAAAAVoICCKR9KMaCoaxeCoqEAkRX3AwMHWxQIIjJSAZWgUEgzBwCBAEQpMwIDwY1FHgwJCtOW2UDWYIDyqNVVkUbYr6CK o2eUMKgWrqKhj0FrEM8jQQALPFA3MAc8CQSAMA5ZBjgqDQmHIyEAIfkEAAoAAgAsAAAAABAAEAAABWAgII4j85Ao2hRIKgrEUBQJLaSHMe8zgQo6Q8sxS7RIhILhBkgumCTZsXkACBC + 0 cwF2GoLLoFXREDcDlkAojBICRaFLDCOQtQKjmsQSubtDFU/NXcDBHwkaw1cKQ8MiyEAIfkEAAoAAwAsAAAAABAAEAAABVIgII5kaZ6AIJQCMRTFQKiDQx4GrBfGa4uCnAEhQuRgPwCBtwK kCNFgjh6QlFYgGO7baJ2CxIioSDpwqNggWCGDVVGphly3BkOpXDrKfNm/4AhACH5BAAKAAQALAAAAAAQABAAAAVgICCOZGmeqEAMRTEQwskYbV0Yx7kYSIzQhtgoBxCKBDQCIOcoLBimRiFhSABYU5gIgW01pLUBYkRItAYAqrlhYiwKjiWAcDMWY8QjsCf4DewiBzQ2N1AmKlgvgCiMjSQhACH5BAAKAAUALAAAAAAQABAAAAVfICCOZGmeqEgUxUAIpkA0AMKyxkEiSZEIsJqhYAg + + + boUFSTAkiBiNHks3sg1ILAfBiS10gyqCg0UaFBCkwy3RYKiIYMAC RAxiQgYsJdAjw5DN2gILzEEZgVcKYuMJiEAOwAAAAAAAAAAAA ==’)



You can cure these mistakes by changing by hand every infected files and Tar Gz them after.





Lee Li Pop

66

[quote name=‘Lee Li Pop’]Hello Noman,



Your CSS files are affected by the Smartoptimizer disease, ie. it changes the URLs of images by their equivalents in Base64, overweighting up the CSS files, turning light files in very big files…



Lee Li Pop[/QUOTE]



Thanks Lee



I don’t really pay much attention to it as it doesn’t really do any harm. I will look into this maybe tonight. I remember touching SO from the back so this could be the reason. Other than that, Google tool says my website is faster than 96% of websites so I’m cool :wink: Thanks again.

67

[quote name=‘Noman’]my website is faster than 96% of websites [/QUOTE]







You too, Noman, you too :smiley:





Lee Li Pop

68

[quote name=‘Noman’]I’m not saying WEBO is bad. I simply prefer SO because it doesn’t brake anything when upgrading CS. I was able to move from 1.3.5 to 2.0.15 with no issues. Then for testing purposes only, from 2.0.15 to 2.1.1 in a temp directory. SO works perfect for me and it takes 3min to install it and tweak it. Not hours of scratching my head. Time is money.



I believe, WEBO may be good but with so many options and possibilities, it doesn’t look like it’s good for beginners or just PHP-uneducated e-store owners. I prefer a simple solution: good quality VPS + SO + some add-ons disabled and that’s it.



The final result you get with WEBO, for the trouble, time and money you spend on it…hmm, you decide.[/quote]

Absolutelly right…

69

[quote name=‘indy0077’]Absolutelly right…[/QUOTE]



Please Indy, read again my test result:


[QUOTE]Less than 1 second: With display mistakes (French characters mistakes PLUS CSS mistakes under IE)[/QUOTE]



1 - Mistakes could easily be resolved with a perfectly tweaked WEBO by a pro.

2 - I tried a Free Basic WEBO, that can load pages in less than 1 second.



SmartOptimizer (SO) cannot be faster, because it does not play in the same League as WEBO:



A - SO can GZIP only CSS files and JS scripts, WEBO can do it too.

B - SO do not GZIP HTML files, WEBO can do it.

C - SO do not manage CDN WEBO can do it.

D - SO cannot combine CSS files, WEBO can do it

E - SO cannot combine JavaScript files, WEBO can do it

F - SO cannot Gzip fonts, WEBO can do it

G - SO cannot check for gzip possibility via cookies, WEBO can do it

H - SO cannot use deflate instead of gzip for IE6/7, WEBO can do it

I - SO cannot cache images, WEBO can do it

J - SO cannot cache fonts, WEBO can do it

K - SO cannot cache video files, WEBO can do it

L - SO cannot cache static assets, WEBO can do it

M - SO cannot cache HTML (GZIP, see B), WEBO can do it

N - SO cannot setup a default timeout to cache HTML, WEBO can do it

O - SO cannot change itself .htaccess file, WEBO can do it automaticaly

P - SO cannot manage data:URI, WEBO can do it

Q - SO cannot manage mhtml, WEBO can do it

R - SO cannot manage CSS Sprites, WEBO can do it

S - SO cannot exclude IE6 from CSS Sprites, WEBO can do it

T - SO cannot manage DB cache, WEBO can do it

U - SO cannot include combined JavaScript file before , WEBO can do it

V - SO cannot move all JavaScript code to , WEBO can do it

W - SO cannot distribute images, WEBO can do it

X - SO cannot check hosts’ availability automatically, WEBO can do it

Y - SO cannot distribute CSS files, WEBO can do it

Z - SO cannot distribute JavaScript files, WEBO can do it



AA - And more…



So, remember, I tried a Free Basic WEBO.



This Free Basic WEBO is ex equao with Full SO.



I use SO for a long time. It’s a really basic tool.



With WEBO, when setup by a pro, you can load EVERY pages for EVERY customers THREE time faster than SO.



WEBO may seem expensive at first glance, but you can refund your WEBO purchase by dozens of times in the first year:


[QUOTE]Phil Dixon of Shopzilla for his part noted [SIZE=“3”]in dropping the loading time of his site from 7 seconds to 2 seconds[/SIZE]:



25% of page views and more

[SIZE=“3”]7 - 12% of sales more[/SIZE]

Source: [url]http://velocityconference.blip.tv/file/2279751/[/url][/QUOTE]



Free Basic WEBO for CS-Cart, can load page under 1 second. SO cannot be so fast :wink:





Lee Li Pop

70

[quote name=‘Lee Li Pop’]So, it’s easy for me to run a contest.[/QUOTE]



I’m not sure that WEBO Site SpeedUp was properly tuned + results are collected when all files are in cache. The very first-time page load can take sufficient time to perform all optimization actions. After this page loads really faster.



I’m sure that the test results can be fantastic if we perform WEBO Site SpeedUp tuning for the test website (and there won’t be any glitches).

71

[quote name=‘webo’]I’m not sure that WEBO Site SpeedUp was properly tuned + results are collected when all files are in cache. The very first-time page load can take sufficient time to perform all optimization actions. After this page loads really faster.



I’m sure that the test results can be fantastic if we perform WEBO Site SpeedUp tuning for the test website (and there won’t be any glitches).[/QUOTE]



Hello WEBO,



Yes, I agree with you, however, it would be simpler to give us some demo, for example an optimized demo for CS-Cart. We could compare speed between yours optimized with WEBO and the official demo:



[url]Instant Demo - CS-Cart Multi-Vendor Demo Try Free for 15 days



Like you do with Joomla:



[url]http://demo.sitespeedup.com/[/url]



In the same way, you could setup a Magento demo. Why Magento? Because Magento is well known to be one of the slowest website engine on the market!



http://demo.magentocommerce.com





Lee Li Pop

72

[quote name=‘Lee Li Pop’]an optimized demo for CS-Cart.[/QUOTE]



Lee, of course we will prepare such demo websites. But we don’t nave examples to install WEBO Site SpeedUp on them. There was one for CS-Cart

webo.cscart-usa.com (in comparison to demo.cscart-usa.com/), but for Magento we have nothing to deploy a test website with example data, and it’s optimized copy.

73

Hello All,



I just found two new services to analyze performance of your website.



The first one is GTMetrix, quite useful and helpful:



[url]http://gtmetrix.com/[/url]



For example:



Basic CS-Cart [url]Ecommerce Solutions and Software for all types of business - CS-Cart is rated E - C:



YSlow Grade (55%)

Page Speed Grade (71%)



[url]http://gtmetrix.com/reports/demo.cscart-usa.com/ITtJGSI6[/url]



Optimized CS-Cart [url]Ecommerce Solutions and Software for all types of business - CS-Cart is rated B - B:



YSlow Grade (84%)

Page Speed Grade (85%)



[url]http://gtmetrix.com/reports/webo.cscart-usa.com/pnzrFuuN[/url]



You can find a Top1000:



[url]http://gtmetrix.com/top1000.html[/url]



You can find some recommendations too:



[url]http://gtmetrix.com/recommendations.html[/url]



Nota Bene:



For your knowledge, a static white empty HTML page is rated A - A:



YSlow Grade (100%)

Page Speed Grade (95%) in my case, I don’t minify HTML



Here is the code of this static white empty HTML page:






Here is the only recommendation (grade B - 88%) about my static white empty HTML page:


[QUOTE]Minifying the following HTML resources could reduce their size by 22B (50% reduction).[/QUOTE]



Here is the optimized version of my static white empty HTML page optimized by GTMetrix.com:






I upload this optimized static white empty HTML



Optimized static white empty HTML page is rated A - A:



YSlow Grade (100%)

Page Speed Grade (96%)



Here is the optimized code of this static white empty HTML page:






I get 100% everywhere. Page load time: 0.25s - Total page size: 22B - Total number of requests: 1 - but the “Page Speed Grade” can’t reach 100%, remaining blocked at 96%.



*****



The second one is zOompf.com, more useful and helpful with more detailed review, with a very helpful “Summary of Issues”:



[url]http://zoompf.com/[/url]



Free test is there:



[url]http://zoompf.com/free[/url]



For example:



Basic CS-Cart [url]Ecommerce Solutions and Software for all types of business - CS-Cart is rated 72.32:



[url]http://zoompf.com/scans/97aacf144ec069820f8d17ac84671c29/report.html[/url]



Optimized CS-Cart [url]Ecommerce Solutions and Software for all types of business - CS-Cart is rated 92.84:



[url]http://zoompf.com/scans/fbc7cdb051907b3fa33d318670a64aea/report.html[/url]



Nota Bene:



For your knowledge, a static white empty HTML page is rated 100.00



Here is the code of this static white empty HTML page:






Zero issue.



Same result with optimized HTML page that is rated 100.00:






Zero issue.



*****



These two analyzes with two different services, shows us the incontestable WEBO speed-up, boosting from E - C to B - B with the first one, and from 72.32 to 92.84 with the second one.





Lee Li Pop

74

Hello All,



Here is the Comparison Report between CS-Cart and WEBO Optimized CS-Cart:



[url]http://gtmetrix.com/compare/ZMJ98EpQ/2fEDX6iE[/url]



The Numbers Speak for Themselves.



WEBO wins hands down.



Please, now, click on the “Add Another URL” button on the upper left and fill the field with your URL. So, what is your result?





Lee Li Pop

75

Hello All,



Here are three incredible webistes for testing your website:



WebPageTest:



[url]http://www.webpagetest.org/[/url]



The “Full Optimization Checklist” table checks nicely your page in detail.



For cons, it’s a little bit long to get results.



WebWait:



[url]http://webwait.com/[/url]



WebWait is a website timer. Use it to benchmark your website or test the speed of your web connection.



The Performance Grader:



[url]NameBright - Coming Soon



This script will analize and report on the health of your website, and web server, as well as how the server compares to others. It will also grade the performance of your site.





Lee Li Pop

76

Hello All,



Many people here believe in Pingdom:



[url]http://tools.pingdom.com/[/url]



However, in my experience, Pingdom is useless, because too basic. I prefer WebPageTest which provides more detailed and powerful results:



[url]http://www.webpagetest.org/[/url]



For those that want to go further in website boosting, use the freeware dynaTrace Ajax:



[url]Digital experience monitoring (DEM) | Dynatrace



Here’s a video that explains how works dynaTrace Ajax:



[url]Digital experience monitoring (DEM) | Dynatrace



This blog is also a great source: [url]http://www.phpied.com/[/url]



For example, you can see, with WebPageTest, the positive effect of a “Flush ()” well placed in your Smarty template. In my case, just with a “Flush ()” well positioned, the loading of my pages is reduced by 0.5 seconds.



You can easily get a clear boost of your site, about 2 to 3 times faster. This is achieved, first and foremost, by reducing the “so beautiful” enlightment elements of your template, while reducing the CSS file. Then, by configuring your “Htaccess” file perfectly. By implementing a CSS Sprites. By combining and minifying Javascript and CSS files. By turning ON the “mod_deflate”. Do you know you can “Turn ON” the Gzip compression for your website WITHOUT Apache “mod_deflate” ? It’ easy, ask me!



By following the advices given by GTMetrix site:



[url]http://gtmetrix.com/[/url]



I need to understand how to set up a CDN (Content Delivery Network) with CS-Cart, and how to minify HTML with Smarty. Be aware! Minify is NOT Gzip compression. They are 2 additional and different things!



Currently, here are the GTMetrix results of my site:



YSlow: 92%

Google Page Speed: 94%

Page load time: 1.73 seconds



According to the website WebPageTest, my homepage is came (before optimization) of a load time of 5.874 seconds to a load time of 1.955 seconds.



Now, my website is 3 times faster! :smiley:



Everything done by hand. Two weeks of hard job.



Aternative solution: Install WEBO add-on. 20 minutes of pleasure!





Lee Li Pop

77

I agree with you about Pingdom.



Since you are determined to divulge all of this information, why don’t you just “spill the beans” and tell us exactly what you have done. :smiley:



Edit:



*

A

Keep-alive Enabled

*

A

Compress Text

*

F

Compress Images

*

D

Cache Static Content

*

F

Combine JS/CSS Files

*

F

Use of CDN



First View 3.508s

Repeat View 1.250s



Edit2: You may also want to start an new thread since this has nothing to do with WEBO

78

Hello The Tool,



Still, I want successful my minifying HTML before. After that, I will create a new topic. For cons, I want to emphasize several points:



1 - Successful optimization is a long and very difficult task.



2 - My solution does NOT have cached pages (as WEBO)



So:



1 - I can NOT optimize your website for yourself. I think especially to the “.htaccess” file which is the most difficult and most complicated to tune. I can NOT help you. No way. It’s tooo long. In my case, 20 hours of hard job to find the best configuration for the only “.htaccess”! Tweak your “.htaccess” is widely dependent of your web hosting company. Each case has its own particularities. So, DIY: Do it yourself.



2 - Although perfectly optimized, my site does NOT as fast as WEBO because WEBO has a cache, and I have NO cache.



I want to say, even if my results to YSlow and Google Page Speed are higher than those of WEBO, my site is slower than WEBO.



[SIZE=“3”]So, definitively, WEBO is the best solution in terms of ultimate optimization. Much more, it’s as easy as a mouse click.[/SIZE]



Not mine.





Lee Li Pop



PS: In addition, I want to say:



1 - My CSS file is NOT the original. It is much lighter! Only 13kb.



2 - My optimization started 2 years ago now.



Read these topics for example:



[url]http://forum.cs-cart.com/showthread.php?t=14032[/url]



[url]http://forum.cs-cart.com/showthread.php?t=7212[/url]



Besides, I never published “Speed Up CS-Cart - Level 2” and up…



Just to tell you that optimizing a site is an ongoing task.

79

[quote name=‘Lee Li Pop’]For example, you can see, with WebPageTest, the positive effect of a “Flush ()” well placed in your Smarty template. In my case, just with a “Flush ()” well positioned, the loading of my pages is reduced by 0.5 seconds.[/QUOTE]



Lee, flush() function prohibits gzip usage, so it can be used only with complete understanding of the situation.

80

Hello WEBO,



I acknowledge that my researches are based on an incredible series of tests. In fact, these are truly empirical researches.



However, WEBO on this blog, is noted that the use of flush() is a good thing:


[QUOTE][SIZE=“3”]The head flush()[/SIZE]



While your server is busy stitching the HTML from different sources - web services, database, etc - the browser (and hence the user) just sits and waits. Why don’t we make it work and start downloading components we know will be absolutely needed, such as the logo, the sprite, css, javascript… While the server is busy, you can send a part of the HTML, for example the whole of the document. In there you can put the references to external components such as the CSS, which then the browser can head-start downloading while waiting for the whole HTML response.[/QUOTE]



Source 1: [url]http://www.phpied.com/progressive-rendering-via-multiple-flushes/[/url]



Source 2: [url]http://www.phpied.com/15-minutes-could-save-you/[/url]



Smarty code needs to be added BEFORE the opening of:


```php


and after:


```

Smarty code is this one:

```php {php}ob_start("ob_gzhandler");{/php}
...

{php}ob_flush(); flush(); {/php} ```

The "chunkview" works perfectly with my site:

[url]http://tools.w3clubs.com/chunkview/chunkview.php[/url]

I get the same results (2 "chunk size") that the example set by PHPied.com:

[url]http://tools.w3clubs.com/chunkview/chunkview.php?q=http%3A%2F%2Fwww.phpied.com%2F&gzip=yep&go=Submit+Query[/url]

[B]Finally, with this technique alone, my pages load faster. The exact loading gained time of my pages is 0.5 second faster.[/B]

[B][SIZE="3"][U]Be Aware![/U][/SIZE][/B]

As I said before, this is widely dependent of your web hosting company... What works for me, couldn't works for other. You need test, test and test again and again, with IE, FF, Opera, Safari, Chrome, under Win2K, Vista, W7, Linux, etc...

Optimizing a site is an ongoing task. ;-)

But the game worth the candle!

Sincerely yours,


Lee Li Pop

PS: You can see the difference "before/after" of this technic with [url]http://www.webpagetest.org/[/url]