Slashdot Mirror
Yahoo's YSlow Plug-in Tells You Why Your Site is Slow
Stoyan writes "Steve Souders, performance architect at Yahoo, announced today the public release of YSlow — a Firefox extension that adds a new panel to Firebug and reports page's performance score in addition to other performance-related features. Here is a review plus helpful tips how to make the scoring system match your needs.
Sites are only as fast as the slowest path through the site.
If your site has 10 different affiliate links/sponsors, all hosted on different providers, your site will be slow.
Similarly, if your site has 100 different java/javascript crapplets,widgets, your site will be even slower.
Here is a simple guide for site creators:
1. Don't overload on ads, I'm not going to view them anyway
2. Put some actual content I'm interested in on your site
3. Don't overload me with java/javascript crap, I don't care what my mouse pointer looks like, just let me click
4. Not everything needs a php/mysql front/back end.
Feel free to use this as a guide, and I might just visit those sites.
My own site also got a 'D', so that seems to be the standard grade. Everything that matters, it got an 'A' for, except for using non-inlined CSS which it got a 'B' for the test that said you shouldn't (to reduce HTTP requests), and an N/A for the test that says you should (to take advantage of caching). Then there were a whole lot of irrelevant things that it got an 'F' for. The fact that none of my site is hosted on a distributed network, the fact that I leave the browser cache to make its own decision about expiring pages since I don't know in advance when I'm going to next change it, and something about ETags, I'm not sure whether it is saying I should have more of them, or I should get rid of the ones I've got.
I guess I flunk!
-mcgrew
It depends on the headers (server), browser, and method, actually. Under some circumstances, for instance under SSL, full copies of all files will be downloaded for every request. As HTTP headers get more complex (some browsers with toolbars, etc, plus a plethora of cookies), the HTTP request/response cycle expands. It may not seem like a lot, but a .5kb request header multiplied by dozens of elements and you can quickly use up a lot of bandwidth. Firefox does a much better job than Internet Explorer under SSL, but not by much unless you enable disk-based caching.
Something I would love to see are some of the headers condensed by the browser and server. For instance, on first request the browser sends the full headers. In the reply headers, the server would set a X-SLIM-REQUEST header with a unique ID that represents that browser configuration's set of optional headers (Accept, Accept-language, Accept-encoding, Accept-charset, User-agent, and other static headers). Further requests from that browser would then simply send the X-SLIM-REQUEST header and unique ID and the server would handle unpacking it -- if the headers are even needed. Servers that don't supply the header would continue to receive full requests, preserving full backward and forward compatibility.
There are a few things to reduce request sizes for web applications. MOD_ASIS is one of the best ones. We use it as one of the last steps of our deployment process. All images are read in via script, compressed if they are over a certain threshold, and minimal headers are added. Apache then delivers them as-is -- reducing load on Apache as well as the network (the only thing Apache adds is the Server: and Date: lines). ETags and last-modified dates are calculated in advance. Also certain responses such as simple HTTP Moved (Location:) responses, GZip isn't used -- GZiping the response actually *adds* to the size due to their very small document size.
I've killed some time on this since it's a pretty interesting idea. It turns out there are plenty outside the D and F range. It does seem to like pages with a single Flash object and not much else, so that's bad. It also makes some pretty arbitrary decisions which don't mean squat to many sites. There are some sites that get enough traffic that speed is a factor but not so much that a content delivery network is really necessary, for example.
I skipped the actual link and score on sites that are pretty much just representative of the sites around them. I wanted to include them by name, though, to show where they fall. I've stuck mostly to main index pages, and I've noted where I've gone deeper.
A: Google (99%), Altavista main page (98%), Altavista Babelfish (90%) (including upon doing a translation from English to French), Craigslist (96%), Pricewatch (93%), Slackware Linux, OpenBSD, Led Zeppelin site at Atlantic (100%), supremecommander.com, w3m web browser site (96%)
B: Apache.org (87%), the lighttpd web server (84%), Google Maps, which also got a C once (84% in most cases), Perlmonks (84%), Dragonfly BSD (85%), Butthole Surfers band page (81%), 37 Signals
C: One Laptop Per Child,, ESR's homepage, the Open Source Initiative (78%), Google News (73%), Lucid CMS (74%), Perl.org (75%), lucasfilm.com, Charred Dirt game
D: gnu.org, The Register, A9 (66%), kernel.org, Akamai (64%), kuro5hin.org, freshmeat.net, linuxcd.org, Movable Type (61%), Postnuke, blogster.com, Joel on Software (67%), Fog Creek Software, metallica.com, gaspowered.com, Scorched 3D (68%), id software (64%), ISBN.nu book search
F: MS IIS (49%), microsoft.com, msn.com, linux.com, fsf.org, discovery.com, newegg.com, rackspace.com, the Simtel archive (26%), CNet Download (29%), Adobe (58%), savvis.com, mtv.com, sun.com, pclinuxos.com, freebsd.org, phpnuke.org, use.perl.org, ruby-lang.org, python.org, java.com, Rolling Stones band page (56%), powellsbooks.com, amazon.com, barnesandnoble.com, getfirefox.com
My site for my company (96%) gets an A (no, I'm not going to get it slashdotted) which is pretty simple but has a pic and some Javascript on it. Several sites I have done or have helped design with someone else get C or D ratings.