Similarly argued, icon fonts are cool but have x-browser rendering issues, dependence on webkit for accurate anti-aliasing (currently), require sticking to a 32x32 grid, and altogether lead to a massive headache for nimble development. In most cases, especially for new products and low traffic, it's premature optimization. Github's speakerdeck gives a great rundown of their experience:
SVGs are XML, which can be compressed transparently by the web server using GZip, which from my reading on the web typically gets around 70-80% compression on XML files, thus reducing the deficit fairly heavily.
Also, that first comparison chart? "Most browsers don't support the full SVG featureset" - Firefox 3.1, Chrome 1.0, Safari 3.2?
Comments
Great dissection, thanks for sharing!
Similarly argued, icon fonts are cool but have x-browser rendering issues, dependence on webkit for accurate anti-aliasing (currently), require sticking to a 32x32 grid, and altogether lead to a massive headache for nimble development. In most cases, especially for new products and low traffic, it's premature optimization. Github's speakerdeck gives a great rundown of their experience:
https://speakerdeck.com/u/jonrohan/p/say-hello-to-octicons
And note, this is Github, who actually has reason to optimize for >2MM users, saying the process was is too complex and tedious.
thank you for your comment, and the link, really interesting points!
SVGs are XML, which can be compressed transparently by the web server using GZip, which from my reading on the web typically gets around 70-80% compression on XML files, thus reducing the deficit fairly heavily.
Also, that first comparison chart? "Most browsers don't support the full SVG featureset" - Firefox 3.1, Chrome 1.0, Safari 3.2?
here is a followup post comparing gzipped versions: http://blog.pictonic.co/post/33163844823/icon-fonts-vs-svgs-...
this is true, although even when gzipped svgs are typically about double the size of ttfs
re browsers, yea none of them passed 100% of the tests, but admittedly that chart is about 10 months old
The first graph has the wrong label for the x axis.