FeedsSo, Firefox 1.5 came out yesterday. I highly recommend it for those not already using it 
.
I am also very impressed by the new-look mozilla.com . Clean, easy to navigate, and yet incredibly stylish.
Both the web team and the product team should feel very proud their work.
Congratulations to everyone involved, on the massive milestone.
If you haven’t downloaded it yet – why not?
I’m pleased to announce that when AWStats 6.5 gets released, it will feature a contribution from me.
Specifically, revision 1.23 of awstats_misc_tracker.js, is mine.
Woot!
It seems that on October 8th, the site went through the 100,000 visit barrier.
Wow.
Named in Planet tradition, and not through sheer ego, I’m releasing Planet Doug to the world.
What is Planet Doug I hear you ask?
Planet is a feed aggregator (i.e. it takes blog postings from numerous sources, and makes them accessible at a single point). Planet Doug simply takes the most recent posts of the people featured in the sidebar (in my brand spanking new blogroll), and makes a single page from them for easy perusal. Mainly a tool for my own convenience, it also lets anyone nosy enough gain an insight into my psyche through the presence/absence of feeds on a certain topic.
That’s if anyone cares.
Since (a) I get more Safari visitors than visitors using Opera, Netscape or Mozilla and (b) I can actually test on different versions, rather than ‘hope for the best’, I’ve hacked together another AWStats modification.
This one was more difficult than those I put together a month ago, since Safari doesn’t identify its version to a website – only it’s build – that is – Safari/312 is really Safari 1.3, and Safari/85 is really Safari 1.0. So after extracting the all-important-number, a comparison then has to be made against a list containing build number->version information (thanks Wikipedia!). This will also mean that whenever Apple release a new version, I’ll have to update the list*. The code for Internet Explorer, Firefox, Opera et. al will continue to work as long as the information is kept in the same format e.g. Firefox/xxx isn’t changed at some point to Firefox (xxx).
* I do have fallback code, so if an unknown build is found, it will report the build number instead. I’d love to know why Apple chose to identify the Safari version in use, but in such a non-obvious way…
Took Apple 3 days to make it and ship it from China to the Czech Republic. It then took TNT 7 days to ship it to the UK. But it’s here now.
I haven’t used it much, and this is the first time I’ve used a Mac, so these are only 1st impressions.
Good
——-
1) It’s tiny. Really tiny. I hadn’t quite grasped how small 16.5×16.5x5cm was.
2) Lot’s of visual flourishes in the OS – things like the icons in the Dock shrinking as more needs to fit.
3) WebCore appears to do an excellent job.
4) Camino, iCab and Omniweb all seem to be well-designed web browsers.
Bad
—–
1) It asked if I wanted a US or UK keyboard layout, but seems to be using the US version anyway – shift+2 gives me @, not ” and vice-versa.
2) The Home/End keys don’t seem to do anything – how do I skip to the beginning/end of a line?
3) Closing the last window of a program, doesn’t appear to actually quit the program.
4) Hate Safari
5) Weird sound when I turn it on. Very tinny speakers.
6) What the hell is the green button in the top-left of windows supposed to do? It certainly doesn’t maximise – it just seems to make the window slightly bigger (but how much bigger depends on the individual application.
7) 250Mb of software updates needed on a machine that only came out of the factory 10 days ago.
(posted from my Mac)
Building on the security features released at beta 1, upcoming new features will include ActiveX Opt-in: To reduce the attack surface and give users more control over the security of their PC, most ActiveX controls (even those already installed on the machine) will be disabled by default for users browsing the Internet.
…
IE 7 implements a native XMLHTTPRequest object for Javascript applications, instead of requiring an ActiveXObject to be created. This also means XMLHTTPRequest will function on machines that have ActiveX disabled.
…
Finally, in recognition of the need for great web developer tools, we are just about to beta a Web Developer Toolbar that provides web developers with rich object model and visual tools which will help them design standards-based HTML and CSS web pages. This feature will be delivered as an add-on for IE6+.
Chris Wilson (on the IEBlog)
Microsoft are worried
Not only does the official IVS Knowledge Base suck for ‘knowledge’, the incredibly bad forum search makes looking elsewhere extremely difficult.
So I’ve added one to the resource centre. At the moment, it sucks too – default mediawiki theme, and only a single article. Hopefully, it’ll take off.
The other day I blogged about getting visits from ‘Opera/9′, one of which originated from the headquarters of Opera Software, leading me to speculate that the next version was going to get a major bump to it’s version number. I’m now received a visit from ‘Opera/8.50′ – again from Opera’s headquarters in Norway.
So what is the next version? 8.50 or 9? Does anyone know?
Without my upgrade to AWStats, I’d never have noticed these 2 gems revealed here!
There are still people using IE 5 and 6 betas. People using 5.0 and 5.5, I can put down to user apathy / company policies. But leaving a beta version installed, when the full product has been out 4 or more years, seems odd to me. Oh well – it’s sub-0.1% of visitors.
The next version of Opera appears to be version 9. The current version is 8.02, so that’s quite a version leap, given the small version bumps of recent history. Normally when I see a version number that isn’t ‘right’ (e.g. ‘Firefox/reset’), I’ll go into the raw logs, and do a manual examination. If it looks suspicious (e.g. a spammer), I then exclude the visitor from the summaries, in the same way I exclude myself. 2 things stand out about this one though that make me think it’s genuine. The logs reveal 2 visits from Opera/9…
The first occurence was a visit directed at a testcase I created for one of their QA staff, regarding a bug in Opera.
This visit is listed in my logs as
84.48.74.200 – - [22/Aug/2005:18:45:03 -0400] “GET /wp/?page_id=121 HTTP/1.1″ 200 3761 “-” “Opera/9.00 (X11; Linux i686; U; en)”
84.48.74.200 is an IP address registered to an ISP in Oslo, Norway where Opera Software is located. An Opera employee would certainly have access to pre-release versions…Suspicions reduced, I moved on to the 2nd visit.
Visit 2 also confined itself to the aforementioned testcase, and is listed in my logs as
213.236.208.22 – - [26/Aug/2005:06:14:01 -0400] “GET /wp/?page_id=121 HTTP/1.1″ 200 3765 “-” “Opera/9.0 (Windows NT 5.1; U; en)”
213.236.208.22 is an IP address registered to Opera Software ASA. Good enough for me!
I’ve completed my modifications to the way the stats are generated, and the stats are slowly coming back up. The only thing holding them back now is time – I have to reprocess each configuration/month combo individually. 16 Resource Centre configurations (1/resource) x 7 months since February = a while.
Thanks to Microsoft (and Netscape, but mostly Microsoft), 90% of browsers identify themselves as something else. This makes life hard.
In the early days of the web, everything was fine. Browsers identified themselves in a sane manner. The most well known browser was Mosaic.
Along come Netscape with their ‘Mosaic-Killer’ browser. To web servers it identified itself as Mozilla (mosaic-killer->mozilla, geddit?). At this time, Netscape was inventing extensions to HTML, that no other browser understood, so often authors wrote 2 different versions of pages. Standard and ‘Netscape Enhanced’.
Microsoft introduced Internet Explorer, but people obviously wouldn’t switch if it couldn’t display all their favourite Netscape-enhanced pages properly. Because of this Internet Explorer identified itself (and still does) in the form Mozilla (compatible; MSIE), so that it received the same pages as Netscape.
To get over the same problem Opera identifies itself by default in the form Mozilla/4.0 (compatible;MSIE;Opera). So to get accurate statistics, you’ve got to look for Opera before IE, and IE before Netscape.
The modern Mozilla browser can’t identify itself as Mozilla/version, because then it receives pages saying ‘You’ve got a very old Netscape, we won’t let you in until you upgrade. Mozilla therefore does something like this – Mozilla/5.0 (rv:1.7) . Netscape 6 and above, are based on this competely different Mozilla, and do something similar to Mozilla/5.0 (rv:1.7) Netscape/6. Other Mozilla-based browsers do the same e.g. Mozilla/5.0 (rv:1.7) Firefox/1.0.
So…
The number after Mozilla/ is the version of Netscape in versions 1-4. But you’ve got to check for Internet Explorer & Opera first. For Netscapes 6 and up, you need to look for Netscape/. This needs to be done before checking for plain Mozilla, as does checking for Firefox etc…as you amass a list of browsers, the amount of ‘check this before that’ requirements for accurate identification of the browser is staggering.
I’ve rewritten parts of this multiple times to make more sense, apologies if it still doesn’t.
Stats are unavailable for a couple of days, while I try to expand the ability of AWStats to break down various browser versions. It currently handles IE/Firefox/Netscape, and I want to expand that to include Opera and Mozilla. I also want details of developmental versions – e.g. developmental versions of Firefox in the 9 months since 1.0 are versioned as 1.0+, but are being listed simply as 1.0 currently.
Since I’ve never written any perl before, and AWStats is a bloated mess of code (according to those who do know perl, it’ll be interesting. I’m hoping that 90% of it will be copy/paste.
Tweaking the look of these pages in the past has always been hard, because I’ve had to change some code, upload it to the server, refresh page, write more code, upload it, refresh page….and anything that broke would have also broken for someone trying to visit the page. Because of this I took a very cautious approach.
That’s been because all my previous attempts to get Apache (web server), PHP (programming language) and MySQL (database) working together on my machine at home ended in failure. Thanks to my discovery of XAMPP[1], this is no longer a problem, so I’ve been able to make more extensive changes to things without fear of inconviencing anyone else. My 404 error page is snazzier as well, including a reference not only the page that you tried to access, but a time-wasting narrative as well (try it!). It now works for the blog as well.
I’ve also managed to beat into submission most of the remaining Internet Explorer bugs on the site – the only one I’m aware that remains is the fuzzy header on the sidebar
[1]XAMPP is combined installer for all of those web tools and more. The best part being that all of the necessary configuration is perfomed automatically.
Over the years I’ve seen many pictures of Barbie dolls played with in ways never imagined by Mattel. One-armed Barbie. Barbie and Ken making babies. Lesbian Barbie. Bondage Barbie. Today I’ve seen something which I don’t think can be topped — Menstruating Barbie.
Why would you
a) Decide that Barbie needed this bodily function?
b) Carry out the neccessary modification?
c) Photograph your work?
d) Post the pictures to the Internet?
(If you know the answers, I’d rather not be told)
Microsoft -> Windows, a ‘shiny’ look, and a yellow ring instead of a blue one. I’m also fairly sure that on the taskbar at 16x16px, that piece of yellow will look very odd indeed.
Hopefully, by now you’ve seen from our posts that there are a lot of new features and work going into IE 7. As part of this update, we’re refreshing our icon and logotype. We considered more radical departures from our current logo, but blue “e” with the ring is very recognizable and familiar to users, so we elected instead to make more subtle changes.
As you can see, the new “e” has more modern look, and the edges are a bit darker so the icon stands out better against different backgrounds. We liked the gold ring too since it brings in new energy and helps the icon pop a bit more than the old one.
This icon and text treatment will be used on the versions of IE 7 for Windows XPSP2, Server 2003 SP1, and x64 versions. We’ll have a slightly different look for the Windows Vista™ version, one that’s more consistent with the icons there. We’re not quite ready with that one yet, but we’ll share it here once we do.
Tony Chor (on the IEBlog)
As of 4pm today, DougWeb is hosted by Site5 rather than Lunarpages. Not only are they cheaper ($6.95/month) – I get more hosting for my money, as well as a zippier site. And Site5 are willing to pay off old hosting contracts (within limits!), so I haven’t lost out by switching.
It’s not that I was upset with Lunarpages – they’ve been fantastic. But Site5 are offering a fantastic deal that anyone would be hard-pressed to beat. Thanks to Doron Rosenberg for the recommendation.
Everything should be fully transferred over, but please do let me know if you spot anything is missing!
The overwhelming negative response to beta 1 has forced Microsoft to reveal exact details of some of the CSS improvements that will be in beta 2.
The IEBlog has the details – the listed bugfixes/enhancements are a very encouraging sign, although MS have confirmed that Acid2 will have to wait for a further release. In the words of the lead IE developer.
We fully recognize that IE is behind the game today in CSS support.
Chris Wilson (on the IEBlog)
Remember that the first stage to overcoming a problem is to admit the problem exists. And it looks like MS have the next few stages well in hand.
