6

We have a Joomla site (http://guru99.com) running over dedicated server. Currently, Rockettheme's gantry based template is installed on the site. So, to optimize my site's performance I have tested various templates of different frameworks with same content on my homepage and then I checked its Loadtime, Pagesize and No of requests using http://webpagetest.org I found following difference in them with respect to my current template's parameters

Template Framework : Loadtime(Sec) / Pagesize(Kb) / No of Requests

Warp : 0.792 / 130 / 14

T3 : 0.708 / 61 / 19

Helix : 0.225 / 106 / 24

I have tested all these frameworks with their basic free themes, So Now my Question is "am I getting a big difference in them with respect to my current template", and do I need to look for different template for increased performance considering these differences?

  • Are you sure webpagetest.org is the proper tool to do performance tests? – Valentin Despa Sep 30 '14 at 12:16
  • What were the results for Gantry? – Neil Robertson Sep 30 '14 at 13:00
  • 1
    Thanks @NeilRobertson, These are the differences between the results of Gantry and other template frameworks as i mantioned. For Gantry loadtime : 4.116, Pagesize : 949 kb, Requests : 46 – Ram Guru99 Oct 1 '14 at 6:46
3

On a side note, if you're worried about page speed, I noticed a couple of improvements you can do on your images.

You have several .png images (with and without transparent background) on your page. One example is this page. As your background is white, you could very well convert most of your png images to .jpg format and reduce page load time a lot. In case you need to use .png images, I recommend tinypng.com to reduce the size of any png image between 50-70%, even with transparency.

Examples:

Original image: http://cdn.guru99.com/images/ETL_Testing/ETLTesting_5.png - 57.6 KB

  • Compressed using tinypng.com - 25.1 KB
  • Converted to jpg - ~22 KB

Original image: http://cdn.guru99.com/images/logo_funlearn.png - 102 KB

  • Compressed using tinypng.com - 34.5 KB
  • Converted to jpg - N/A (transparent background)

In addition, some of your images larger than they need to be, this also increases your page load time. Try reducing your images to the actual size they'll have on your website (using an image editor).

Example:

Display size vs natural size

Finally, another site to test website speeds, with several tools and reports, is http://tools.pingdom.com.

  • Thanks @johanpw, A little favor, what do you mean by 'with and without transparent concept'? You suggested us to reduce the images size to actual size, I want to know, is there any programmatic way to do it? – Ram Guru99 Oct 1 '14 at 6:55
  • gtmetrix.com is also good for performance checking and recommendations. – Neil Robertson Oct 1 '14 at 7:04
  • @RamGuru99 - with and without transparent background. Basically he's saying that if the image being used does not have a transparent background, then it should be a jpg. For even better compression, you might also want to consider using some gif formats. – Lodder Oct 1 '14 at 8:39
  • @Lodder explained the transparent background well. In addition, you have transparent images .png files allow transparency, but generally creates bigger files than .jpg when the image has many details. This article has more info. I'm not aware of any way to automatically reduce the image dimensions, you'll have to do it manually. Always take a backup before replacing your images – johanpw Oct 1 '14 at 10:51
2

Overall your site performance is very good.

All scripts and CSS files are being delivered via CDN. What I would consider now is GZipping all of these assets and this will increase performance even more. For this, I would recommend JCH Optimize. I've used this before so can confirm it does the job very well.

However (there's always a "however"), I've noticed some scripts sometimes forget closing brackets at the end, therefore when compressing them, it results in a console log error. If you do get this, you will have to exclude each JS file one by one to find the culprit. The plugin does allow you to single out specific files.

The one thing you have to bare in mind when testing page speeds for different frameworks is that the templates that come with them may require different assets. For example, the Helix framework as you have seen is nearly 4x faster to load than Warp, however the Master theme that comes with Warp may contain more images and CSS files, thus more HTTP requests, thus slower loading speeds.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.