All quiet on the Ansearch front

So I dropped a none-too-flattering allu­sion to Ansearch into a post about the search engine indus­try in Aus­tralia a bit over a week ago now, and their CEO Dean Jones popped in to say he’d picked up the post via Tech­no­rati (I was suit­ably impressed) and that much had changed since my first “review”.

I took up his invi­ta­tion to con­tact him via email with ques­tions and sug­ges­tions, but have yet to hear any­thing back. This is per­haps under­stand­able given the length of the email: but I did at least expect a reply acknowl­edg­ing receipt of this mes­sage. At any rate, at present any oppor­tu­nity Ansearch may have had to clar­ify mis­con­cep­tions pre­sented here hasn’t been taken up. I com­mu­ni­cated in the email that I was more than happy to pub­lish not only his responses to explicit ques­tions, but also “any­thing else you wish to add that I haven’t asked about”.

If any­one from Ansearch picks up this post via Tech­no­rati, the offer remains open. If you’d pre­fer not to use email, I’m happy to do a tele­phone inter­view for pub­li­ca­tion here (both in arti­cle form and as a sup­ple­men­tary transcript/podcast).

Update: Dean Jones picked up the post, and sent me an email say­ing he’ll reply shortly. Watch this space!

# by Josh on September 8th, 2005 Tags: , , , ,
| No Comments »

Something scary about HSC timetables

Despite more than 60,000 stu­dents in NSW doing their HSC this year, the first three per­sonal timeta­bles listed when a Google search for “HSC timetable” is con­ducted are those of peo­ple known to me. My timetable is rank­ing high­est, prob­a­bly because of incom­ing links and the rel­a­tive fresh­ness of infor­ma­tion, then Dale’s from last year (ranked not far below, prob­a­bly because of rel­a­tive age of the doc­u­ment), then, on the next page, Ben’s timetable posted a few days ago — I’d imag­ine because of incom­ing links, though the fact it’s not in a table may have some­thing to do with it.

Does that strike any­one else as scary? I’m so sure we can’t be the only peo­ple in the state (of approx­i­mately 6.6 mil­lion peo­ple) to have posted their timeta­bles online! And, if we are, that we all know each other directly — it’s a scary extrap­o­la­tion of the idea of a “digerati” into a broader con­text! Have we formed our own élite? Does that make us rel­a­tively “heard” voices in this global community?

Call this an over­re­ac­tion to search engine opti­mi­sa­tion if you will, but I think it’s sig­nif­i­cant that three stu­dents (or ex-students) in the state who all know each other can rank so highly on a term hardly com­mon to them alone — there is a can­di­da­ture of approx­i­mately 60,000 stu­dents, and then there are the rel­a­tives and friends of these peo­ple to con­sider. Don’t laugh at this com­ment too hard, but where are the Live­Jour­nal users on the first two pages? Cer­tainly, some­one must have posted such infor­ma­tion, or, at the very least, a string refer­ring to the same, on their own website.

“Democ­racy” on the web takes on a new tone when viewed like this. Cre­ation of some com­pelling con­tent in a seman­tic frame­work appar­ently enti­tles one to a greater voice on any issue. There are tech­ni­cal lim­i­ta­tions to doing it any other way, of course — but per­haps this sug­gests that “the web” isn’t quite so demo­c­ra­tic as it has often been heralded.

# by Josh on June 15th, 2005 Tags: , ,
| 1 Comment »

Statistics for April

This month past has regret­tably been one in which I haven’t spent enough time cre­at­ing con­tent, although March was fairly good, but the traf­fic didn’t really fal­ter for it.

Unique: 2091
Vis­its: 4281
Pages: 14158
Hits: 44419
Band­width: 484.76 MB

Unique impres­sions are up by over 600 from last month, sug­gest­ing increased dri­ven traf­fic (from exter­nal sites, search engines, etc., which accounted for 13.1% of total traf­fic, as opposed to inci­den­tal traf­fic — that is, reg­u­lar vis­i­tors, syn­di­ca­tion, etc.). Band­width is slightly up on March, exceed­ing 500MB if non-viewed sta­tis­tics are included (there was 192.48 MB of non-viewed traf­fic, which means search engine spi­der­ing as well as cer­tain syn­di­ca­tion ser­vices I think), which is get­ting fairly size­able I think, espe­cially com­pared to last year’s sta­tis­tics for this month (which aren’t really valid, because there was a holder page up then, but it’s fun to point out) — 2.38MB of traf­fic and 30 unique visitors!

The most pop­u­lar post on this web­site remains the orig­i­nal Dash­Lite announce­ment, although the updated ver­sion doesn’t really get a look in… which is okay, because it was released more out of social respon­si­bil­ity than any new need, and peo­ple can choose for them­selves what they want.

See also:
Feb­ru­ary 2005 sta­tis­tics
September/October 2004 sta­tis­tics
June 2004 statistics

# by Josh on May 1st, 2005 Tags: , , , ,
| No Comments »

Clearing form element value text

I just realised I hadn’t imple­mented this on my web­site yet, so I just reme­died that sit­u­a­tion. Due to work­loads at the minute I’m not get­ting time to write lots here, so I thought I’d just share this snip­pet quickly.

It’s a good prac­tice for search forms to include some text (the value attribute) so that users can know what they’re for, even if it’s appar­ently self explan­tory. On this web­site, the search box at time of writ­ing lacks any label at all except for the text denot­ing “search” in the box. This is all well and good, up until the time some­one wants to actu­ally search for some­thing. They can’t just click the field and type the query and bang return; they need to ensure they select the con­tents of the field prop­erly, so that they can delete it and it isn’t included in the query.

We can do this with event han­dlers — in this case, onclick. Essen­tially, the logic flows some­thing like this:
If a per­son clicks the field, the con­tents are wiped.

Any­thing wrong with that?

Yes, of course there is. That’s the sim­plest way to do it, but it’s wrong. What if the user types a query in and uses the mouse to click a point in the text to edit it? Their query would be wiped, because onclick clears the field when­ever it occurs, regard­less of content.

So what do we want? Using the text on this site as an exam­ple, we want to wipe the con­tents of the field when a per­son clicks it only if the con­tents are exactly “search”. Admit­tedly, this still has some prob­lems — if some­one were search­ing for a search engine or had any other rea­son to include “search” in their string and went to edit, there’d be poten­tial for issues, but that, by my think­ing, is tolerable.

The sim­pler ver­sion, which doesn’t check to see if the value is “search” before eras­ing, is simply

onclick="this.value=''"

The mar­gin­ally more com­pli­cated ver­sion, which is far more usable, goes like this

onclick="if(this.value=='search'){this.value=''}"

That means our new input field code looks like this:

<input type="text" class="search" value="search" name="s" id="s" size="13" onclick="if(this.value=='search'){this.value=''}" />

Now go use it!

p.s. No need to say that this isn’t par­tic­u­larly ground­break­ing. I’d just for­got­ten to imple­ment it here, and thought it could be help­ful to other peo­ple write about.

# by Josh on April 27th, 2005 Tags: ,
| 4 Comments »

A response to DashLite criticism

Navid Azimi posted the fol­low­ing in a com­ment on the ini­tial Dash­lite announce­ment post:

This seems like a good imple­men­ta­tion and def­i­nitely has it’s uses but for most admin­is­tra­tors this could actu­ally be more detri­men­tal in the long run than ini­tially expected. The pri­mary idea behind the Dash­board was to allow all Word­Press Admin­is­tra­tors to stay informed regard­ing devel­op­ments in the community.

Many WP users (or any com­mu­nity for that mat­ter) install and sit. Often times being obliv­i­ous to new ver­sions and (most impor­tantly) secu­rity updates. This sort of unpatched soft­ware can be detri­men­tal not only to your web­host, and your web­site, but also to the entire web com­mu­nity itself.

For exam­ple, when phpBB was exploited with a major secu­rity flaw — there was a major defi­ciency in con­tact­ing all admin­is­tra­tors regard­ing the secu­rity hole. The prob­lem is twofold. The more you pro­mote the secu­rity hole, the eas­ier it becomes for mali­cious users to exploit unpatched instal­la­tions. You see where I am going here.

Of course — right now — in the prime heat of your blog you feel that you are check­ing wordpress.org every­day and you’re prob­a­bly skim­ming the forums daily too. There is no way you’ll miss any updates. But as time goes on and you have tweaked, retweaked and redesigned your web­site five times you’ll real­ize that its time for your blog to push bet­ter con­tent and not just look pret­tier. And its then when you sim­ply stop keep­ing up with every nightly or read­ing the forums daily.

Then again, I could be com­pletely wrong.

I kind of felt that this required a response more pub­licly than the con­tin­u­a­tion of the com­ment thread would per­mit, hence this post­ing. Read the rest of this entry »

# by Josh on April 8th, 2005 Tags: , , , , , , , , , , ,
| No Comments »

Something about backwards search engines

No, I’m not talk­ing about elgooG.

The Syd­ney Morn­ing Her­ald pub­lished an arti­cle enti­tled “New Aus­tralian search engine launched” today, the first para­graph of which reads “Australia’s newest search engine Ansearch opens for busi­ness today with a novel twist, demo­graphic search­ing.” It’s not a par­tic­u­larly well writ­ten arti­cle, but the arti­cle ven­dor is AAP, not the SMH itself, so we’ll leave that alone, at least for the minute.

It goes on to laud the search engine for their inno­va­tion, both in this fea­ture of demo­graphic search­ing, and in other areas:

Ansearch says it cuts down search clut­ter by dis­play­ing the main search results as sin­gle web­sites and not the indi­vid­ual pages of websites.

What, like the Google [More results from domain­name] fea­ture? You know, the one that actu­ally works prop­erly? I say “works prop­erly”, because a quick search of Ansearch reveals that their “cut­ting search clut­ter” fea­ture is a tad bro­ken — not to men­tion their char­ac­ter encoding.

Proof that it's broken, demonstrated by duplicate entries and incorrectly encoded characters

Read the rest of this entry »

# by Josh on April 4th, 2005 Tags: , , , , , , , , , , , , , ,
| 8 Comments »

Graveyard retired

Some more atten­tive reg­u­lars (who don’t just peruse this web­site by means of syn­di­ca­tion) may have noticed the dis­ap­pear­ance of a link in the top bar in the last sev­eral hours. This is because I’ve finally got all the old con­tent into Word­Press, with no small amount of assis­tance from Michael, under a cat­e­gory called “Before Word­Press” (this post is cat­e­gorised sim­i­larly, and shall likely be the last ever entry into that category).

Prac­ti­cally, this means that that con­tent is using seman­ti­cally bet­ter markup, has bet­ter meta infor­ma­tion for search engines, and is inter­nally search­able, using the Word­Press search func­tion (it wasn’t before).

For most reg­u­lars, this prob­a­bly doesn’t mean much, but the old arti­cles attract the most search engine traf­fic, so this’ll be of ben­e­fit to peo­ple find­ing rel­e­vant con­tent, at least, because the old script could be some­what retarded in the way it was indexed, as there was no for­mal perma­link struc­ture, just a bunch of loose query strings, which search engines didn’t like.

</geek off>

# by Josh on April 3rd, 2005 Tags: , , ,
| No Comments »
« Older Entries
Newer Entries »