One of the main questions on the part of users, the use of cloud-based solutions is definitely the secrecy and confidentiality of the data that are stored in the storage cloud, many fear that their data (documents, photos, code, etc..) Can first eventually be violated, and in some cases compromised, thus losing days and days of work.
Stable version of Google's mod_pagespeed
Among the open source software offered by Google in recent times, mod_pagespeed is probably the most interesting, and is certainly intended to save a huge amount of time for web developers.
For about a month is online version 1.0, after two years of beta versions, and the next is now ready as beta.
Let's see what it is. Mod_pagespeed is a module for Apache, the most popular web server, to optimize the way in which Apache serving pages, so as to achieve a much faster loading client side.
The software is easy to install (it is already available from Google packaged for Debian, Ubuntu, Red Hat and Fedora, and users have already made packages for much more, including Windows) consists on a series of filters that the server applies to web pages before serve them, automatically optimizing what would typically optimize the developer of the site, and allowing the visitor's browser a more efficient use of the cache.
One of the great problems of web designers, for example, is the modularization of CSS to use: the more are used, the more are comfortable to use. At the same time, however, many CSS likely to slow down the loading of the page. Usually, you lose a lot of time to create systems that server-side, dealing combine CSS with each other. This will be done automatically by mod_pagespeed, joining all the sheets into one, and entering the name of the virtual sheet so created its MD5 sum and set the caching time for one year. In this way, the visitor's browser will download the CSS only once a year, unless the developer of the site does not change, thus changing the MD5 and making it clear the browser should re-download everything again.
Another example of a filter is that which applies to images. Anyone who had to deal with websites that contain photographs or similar, knows well how boring it is every time create different sized images for thumbnail previews and various. A filter mod_pagespeed takes care of reading the image should be displayed as large, CSS and HTML, and automatically generate the small version. In addition, the image is optimized in terms of definition, often too high for a monitor, is relieved of unnecessary metadata, and is (if the requesting browser supports the format) converted into WebP, an open source format for the transmission of images web, which compresses a lot. This allows you to create websites that use WebP without discriminating the browser "by hand" by analyzing the request, and without having to make more versions of the same image. Also in the file names of the images you set a MD5, which allows you to keep in cache for a year if they are changed.
There are many other filters, well documented and testable on the website of mod_pagespeed. The filters are activated one by one, and Google has grouped all those basic and extremely stable in order to activate them by default (even though it is possible to disable all and enable only what you want). The activation is done from the server configuration, but you can also request via http header to be activated in order to test specific filters before you configure everything.
The software will soon see improved caching system server through the use of memcached. Memcached caching system known Open Source, is already on the beta branch of mod_pagespeed, and allows those who manage sites that reside side by side on most servers using a single area of cache: in this way, if a server creates such thumbnail for an image, all the other servers will quickly available.
Mod_pagespeed is actually nothing more than a wrapper that can be used to make Apache a set of libraries of Google, called PageSpeed Optimization Libraries, which are in fact independent of Apache (have been created in parallel to a proprietary server Apache and Google). These libraries have well-documented APIs that allow anyone interested to write modules for web servers other than Apache, or using other memcached cache systems, such as those existing in the architecture on which you want to run the libraries.
Already several hoster, particularly U.S., integrate the module and allow you to activate it via the control panel.
I mean, you go to a web always faster and less bandwidth-hungry for those who surf the Internet, while developers began an era in which they can finally address the logic and content of a website, without having to think about how to optimize to make it quickly downloaded and at the same time easily modifiable.
A beautiful thanks to Google, which has once again made freely available technology designed for its own systems. Anyone who wants to work has to have a mailing list, and I see that Google has been very active also to consider the demands made on StackOverflow.