Hi Mike, Nadim,
At 04:04 11-11-2015, Jules Mike Giovanni wrote:
>On another note, I feel that all the blocks are too big. The overall
>design feels a bit like a metro design. Sometimes the blocks are so
>big that the content that is supposed to be in them does not fit
>properly. If we look at the "Featured" (why is it in English is
>another thing) block, we can notice that most of the pictures are of
>low quality and this creates this annoying "pixelated" effect.
The images appear as low quality. Some of images are blurry. I read
about the Metro style as I did not understand why it was used for a
news web site. I am not sure whether it is an appropriate design in this case.
At 17:04 11-11-2015, Nadim Bundhoo wrote:
>Do you remember the several animated GIFs on webpages in the early
>2000's ? Every animations distracted us from viewing the one piece
>of information on the page that was supposed to be the main "content."
Yes, that was distracting.
>On this new version of lexpress.mu, I feel the same. Which piece of
>info should I focus on?
You are supposed to focus on the ads. :-)
>I do not have an idea on the size of the old version, but I did a
>test on this new version. I found that the homepage weighs around
>23MB. The website is responsive, which means that on mobiles and
>tablets, dark matter are added to the pages. By dark matter I mean
>that assets are downloaded though not visible (most probably because
>of display:none in the CSS) [1].
I tested the (data) size for
http://www.elandsys.com/~sm/cellular-data-usage-201508.html The web
site was slow when I did tested it on a mobile phone. It might be
because of the "dark matter" which you mentioned.
>The cost for 1GB of mobile data is Rs299 [2] or Rs 0.299 per MB of
>data. If everytime I visit lexpress.mu on my mobile (with this 1GB
>data plan), I'll have to pay Rs 6.877 (it may be slightly less,
>depending on the amount of data downloaded).
It would be cheaper to buy the newspaper instead.
>Anyway, it is not encouraging at all to visit lexpress.mu on a mobile.
Yes.
>I don't know about lexpress.mu's analytics, but having 2 different
>versions of the website (desktop & mdot) would have been better,
>instead of one responsive version. Imagine the bandwidth wasted when
>the website is accessed on a mobile phone.
I didn't look into how the web site implemented responsive web
design. I would not expect such an approach to cause unneeded use of
bandwidth.
>The website has also been designed to cater for IE7 and IE8. Say if
>1% of the traffic comes from surfers on IE7 & IE8 - which makes
>around 10000 surfers according to "News web sites traffic in
>Mauritius" [3], why do we offer them a responsive website? Surfers
>on these browsers may still be on Windows XP (perhaps government
>officials) with a slow computer. Why should they have to download
>all the CSS & JS then?
1% of web traffic is a significant number. It gets complicated when
part of that 1% are readers you want to reach. This is where I would
start looking into the details (CSS, JS, etc.) of what is being proposed.
>I am on MyT 10M [4]. Defimedia's homepage weighs around 14MB and
>loads in around 20s whereas Lexpress.mu weighs around 23MB and loads
>in 37s. (number of CSS & JS files, images/ads/videos/audio?)
>
>Defimedia's page starts to render after 4.6s whereas Lexpress page
>starts to render after 6.2s. (CSS & JS ordering?)
I'll leave both of you to discuss this with Sandeep. :-)
>The browser makes 227 requests if I surf Defimedia's page for the
>first time and 22 requests subsequently. 338 requests are made if I
>surf on Lexpress homepage for the first time, and 163 requests
>subsequently. (caching of assets?)
338 requests is a very high number. A lot of the discussion relates
to how user experience is assessed.
>The only area where I could find L'express better than Defimedia is
>the first-time-to-byte (i.e. time to wait for server to send first
>piece of contents, usually the HTML). Lexpress.mu may be behind a
>cache server (Varnish?) and Defimedia not.
Defimedia.info is using nginx. www.lexpress.mu is using
Varnish. The time to first byte (not a local test) for
defimedia.info is 1.763 seconds, and 0.586 seconds for www.lexpress.mu.
At 00:01 12-11-2015, Nadim Bundhoo wrote:
>One thing that I need to be tested. If people reading this ML can
>send their result, it would be great. I want to know whether the
>batteries of mobiles get drained fast when surfing on L'Express'
>website. I surfed 4 times on the homepage. My battery level:
>
>1st: from 48% to 30%
>2nd: from 30% to 19%
>3rd: from 15% to 8%
>4th: from 100% to 92%
>
>I was only on the homepage, and scrolling up and down, without
>opening other pages.
The above behaviour is unusual. I'll try and do a test as I would
like to find out what is causing this.
Regards,
S. Moonesamy
Received on Thu Nov 12 2015 - 11:38:37 PST