June 2009 was a huge month for www.christian-sauve.com, probably its single most important month since the site became a dot-com in 2002. In a manic blaze born out of boredom, I completely redesigned the site, not simply updating the sorely-outdated visual look, but crammed about nine megabytes of flat HTML content into a WordPress-powered dynamic architecture. (While, in the process, going from about 250 static pages to 2,700 dynamically-generated ones.) The more curious among you will find a summary of the process in the last section of this report, but in the meantime do enjoy the last Web Site Report of the non-dynamic era, since I’m completely updating the methodology of those reports next month to reflect the new architecture of the site: Goodbye Urchin, hello Google Analytics!
1. Mmm. Numbers…
One last time, my prickly “Urchin” web stats engine tells me that…
Report for: christian-sauve.com, June 2009
Total Visitors 7,147
Total Pageviews 25,812
(Corrected total 12,155)
Total Hits 32,295
Total Bytes Transferred 477.8MB
Average Visitors Per Day 238.23
Average Pageviews Per Day 860
(Corrected average 405)
Average Hits Per Day 1,076
The “corrected” numbers take out the CSS, robots.txt, PDFs, mis-filed graphic files (ICO, GIF, JPG) and other non-public files mistakenly considered “pages” by the statistics pre-digestion engine. All numbers are up from last month, but a good chunk of that is a consequence of the architecture switch and my own testing of the site.
Meanwhile, Google Analytics is a bit savvier in telling me that I only had 650 visits and 1,471 pageviews. Visits down; pageviews up, in keeping with the expected consequences of the redesign.
According to Urchin, our top ten most popular pages are
/index.html 966
/texts/free-movie-tickets.htm 316
/reviews.html 241
/about.html 225
/writings.html 208
/texts/solaris-explanation.htm 204
/links.html 199
/search.html 170
/contact.html 146
/francais/index.html 138
Little change here, just some re-ordering. (My redesign took place on June 27, too late to have an impact on the rankings.) Meanwhile, Google Analytics says…
| 1. |
/index.html |
192 |
| 2. |
/reviews.html |
139 |
| 3. |
/texts/solaris-explanation.htm |
126 |
| 4. |
/francais/index.html |
38 |
| 5. |
/the-contact/index.html |
32 |
| 6. |
/essays/index.html |
29 |
| 7. |
/the-reviews/index.html |
28 |
| 8. |
/the-about/index.html |
20 |
| 9. |
/writings.html |
19 |
| 10. |
/reviews/2003/books03k.htm |
17 |
This is a bit different from the usual results, and I can already see some of the new-template files sneaking onto the top-10, not so much as evidence that new visitors are flooding in, but as a consequence of the top-level site testing.
If you care about such things, (and it is fascinating stuff!), here’s a look at browser statistics for the month (by visitors), as provided by the clever gerbils at Google Analytics:
|
Browser |
This Month |
Last Month |
| 1 |
Firefox (all) |
212 |
230 |
| 2 |
IE 7.0 |
127 |
212 |
| 3 |
IE 8.0 |
65 |
(not counted) |
| 3 |
IE 6.0 |
63 |
67 |
| 4 |
Safari (all) |
68 |
47 |
The sudden jump in IE8 numbers is heart-warming, but I won’t rest until IE6 disappears from this list. (Since the new site looks really really weird in IE6, the process may accelerate next month.)
I should mention once again that as of next month, I’m going to use Google Analytics are the primary source of data, and maybe look at the Urchin stuff for fun: With the new dynamic architecture, Urchin is becoming more technically useful than analytically valuable, and trying to extract useful data out of it is going to be just too time-consuming. Fortunately, I have used Google Analytics for long enough that I won’t be too disturbed by what I’ll be seeing.
2. Where do these people come from?
According to Urchin, our top five sources of referrals (in visitors) were
google.com/search 474 (486)
live.com/results.aspx 109 (94)
www.google.ca/search 104 (80)
google.com/books 49 (51)
google.co.uk/search 47 (64)
As you may expect by now, Google Analytics has a slightly different view of the situation:
|
Source |
This Month |
Last Month |
| 1. |
google / organic |
410 |
430 |
| 2. |
yahoo / organic |
12 |
26 |
| 3. |
bing / organic |
10 |
11 |
| 4. |
books.google.com / referral |
9 |
8 |
| 5. |
en.wikipedia.org / referral |
9 |
11 |
(Lingo key: “Organic” is Google’s way of saying that no one has paid for links leading back to christian-sauve.com on those search engines. “Referral” is supposed to be a direct link to this site.)
Google now lists about 2980 links for “Christian Sauvé”, down from last month. A look at the top-100 results showed no new links of significance.
3. Ohh! Visitor comments!
Plenty of activity in the mailbox this month. Let’s have a look:
1. Chris from Boston wrote to say…
About 2002 Solaris- a new interpretation which also reveals Andrei Tarkovsky s Solaris (1972) to be genius, and not so self-indulgent! The faux Rheya, and the other “visitors”, are presented as projections of the Crew’s unconscious– but as such they symbolize film itself! These non-human creatures who do not die; who are imperfect copies of actual memories; who are sometimes rejected violently, sometimes loved irrationally; artificial products of advanced technology (moviemaking). Soderbergh, like Tarkovsky before him, is writing about our relationships to film itself (and by extension to the actors, etc, etc) I’m not sure if Lem’s book has the same self-conscious self-references to creative writing and reading- but this idea really opened up a new appreciation for me. I’ve also discovered a ’68 Soviet made-for-TV version- great stuff!
Interesting!
2. Sofie from Belgium says…
Hi! My name is Sofie, I’m 15 years old and I’m from Belgium. We watched the movie “Solaris” during English class, but I didn’t understand the movie well, especially the end. I was wondering if maybe you could help me to understand the several possible endings, because I already read a lot of different ones. And I have to explain them on my final, but the teacher doesn’t want to explain it to us. My final is the 18th of Juin. Thank you! Sofie
and then, moments later:
It’s sofie again…
Where can I see your answer?
I really wanted to help, Sofie, but you didn’t include a return email address. Given the date of your final, I doubt you’re reading this, but that’s just as well: Everything I’ve thought about the film is in my Solaris Explained article, and I’m not holding anything back. To anyone else: please re-read the article, it’s all there.
3. An anonymous drive-by:
Your interpretations are not scholarly.
Well, as a way of saying “You suck!”, that’s more polite than usual. Poking around at the server logs to figure out what had so provoked my anonymous correspondent, I discovered (without any surprise) that a disappointed review of a right-wing political thriller (Vince Flynn’s Term Limits) had once again earned me some nameless name-calling. For the record, I note that whenever I get a nasty anonymous comment, it’s almost always about a not-entirely-positive review of a right-wing thriller. What really makes me laugh is that the review essentially says “someone smarter than me will have to review the political implications of this.” Anonymous commentator’s reading comprehension: Poor.
4. Finally, a chance real-time encounter with a former local SMOF earned me the first real message handled by my new technical infrastructure…
For your SF reading, I suggest that you add Hal Clement and Harry Harrison. Harry’s Stainless Rat series and “Star Smashers of the Galaxy Rangers” constantly bring tears (of laughter) to my eyes. And for complete obsure – look for Charles R. Saunders…
They may not have been reviewed at length, but I have read plenty of Hal Clement (Mission of Gravity and others) and Harry Harrison (both works mentioned, and more). On the other hand, Charles R. Saunders was a new name I should have known!
4. Search Queries Oddities
According to Google Analytics, here are the month’s most popular search keywords:
|
Keywords |
Visits |
| 1 |
solaris explanation |
17 |
| 2 |
solaris ending |
13 |
| 3 |
christian sauvé |
9 |
| 4 |
solaris explained |
9 |
| 5 |
the runaway jury john grisham thesis synopsis |
8 |
| 6 |
christian sauve |
6 |
| 7 |
explain solaris |
6 |
| 8 |
solaris movie explanation |
6 |
| 9 |
dale brown american holocaust |
5 |
| 10 |
solaris 2002 plot |
4 |
Much Solaris again…
Other odd, special or amusing search keywords:
- “plenty of cleavage” cleavage movie review
- a movie about a bartender who becomes a hero and is hunted by a past
- how edible is the book from the christian writer mary backster to hell and back?
- review of symbolic, technical and audio codes used in the devil wears prada
- when you trust a person’s word only, is this wrong
Until next time, my name is Christian Sauvé and I remain… obsessed by web statistics.
X. Special section: Site Redesign Notes
I woke up in mid-June with a few new priorities: I was pretty much caught-up in what I had to do, was giving a speech to a few dozen people at the end of the month, and couldn’t stand looking at my current web site. Taken together, those three elements forced me to do something I had thought about since at least January 2008: Not only redesign the site visually, but convert the content to a database-driven Content Management System.
Picking WordPress was easy given the amount of experience I had with the system, and the just-right capabilities it offered. But converting 250+ files of HTML code totaling about nine megabytes of data isn’t the find of thing you do in a lazy afternoon. It’s pathetic to do elementary project management for a personal site redesign, but that’s what I ended up doing, with a real schedule on paper, deliverables, day-by-day objectives and critical paths. All the HTML data had to be concatenated, then meta-tagged, corrected, normalized and exported to a format that could be automatically imported in a WordPress database.
I ended up doing things I wasn’t planning on doing, in ways I hadn’t even imagined. I never could have finished the project without the industrial-strength text editor NotePad++ and the latest version of Microsoft Excel: Both applications covered for the other, but it’s Excel’s string-manipulation capabilities and ability to combine data in an XML template that really saved the day. I ended up reverse-engineering the WordPress XML format and then generating files to import 2,500+ items back in the system. The really dull stuff? Adding the date meta-data that was until then implicit in the HTML filenames, and converting my MOVIE TITLES to semantically-rich Movie Titles. There weren’t many programmatic ways of solving those issues. Also annoying? Finding out that being 90% accurate in entering meta-data over a dozen years is far from being good enough when that means hundreds of corrections over a thousand-items database.
It’s fair to say that my ambitions for my visual design were much, much higher than what I ended up applying. But the bulk of the redesign time budget went to the data formatting, with the “fun” visual design ending up “good enough” early on. There was a lot of tweaking (save for the commenting module code blatantly taken from the default Kubrick template, most of the template for this site was written from scratch), but it sort of came together. I only tested with last-generation browsers: this is my personal site, and I don’t really care if IE6 blows up for visitors sad enough to be stuck with it.
Overall, the project took about forty-to-fifty hours, spread over four-hours weekdays and twelve-hour weekend days. As an occasional web designer, I would have turned down that particular commission if it had come from an outside client. On the other hand, there’s little in WordPress that scares me now that I’m done.
I have opened comments on nearly all items: I wonder how long it’s going to take until I regret it.
Very little content didn’t make it from the old to the new architecture. What disappeared were usually essays I now strongly disagreed with (including one explaining why this site wasn’t a blog!), or collections of mini-reviews I couldn’t fit in the new site architecture, necessarily more rigid than the old one. I really wanted to review my older reviews to remove the more embarrassing ones, but simply didn’t have the time.
I felt pretty proud of myself for finding a way to automatically integrate posters and book covers to my reviews, but plans to provide scans of all book covers and movie posters quickly disappeared once I realized how many of them there were (2,500+) and how long it would take to generate every graphical item even if I was on a fully-automated process. (Basically: five minutes per book cover, two minutes per movie poster. If everything went well.) I compromised by including full imagery for everything reviewed since January 2008… and will provide images from now on. I hope, though, that my server space and bandwidth will be able to accommodate all of that new data.
What else should I note? Well, XAMPP worked superbly as a local development web server. Facebook was useful to send a “new site” announcement to my usual network. I’m still learning The GIMP –now that’s another full-time project not to attempt in the middle of a major redesign! I had some trouble with my host in installing WordPress, but nothing we weren’t able to handle after a few emails back and forth. I had a lot of trouble sleeping during the twelve days of the project: Hot temperatures, heightened stress, lots of sugar and a head boiling with ideas. But the site was more or less ready on-time, the speech went well and the project management worked more or less as planned. Hopefully, future redesigns won’t be so difficult!
Expect a lot of changes in the statistics next month: More pages, but smaller pages: I can’t really predict what will be the impact on Google referrals, especially since I won’t remove the majority of the old static pages until August 1st. We’ll see…
SiteReport | No Comments »