Techie :: Techno ::: July 2005 Archives

There seem to be an ever growing number of Irish blogs judging by the two main aggregators; POTB and IrishBlogs. But does anyone know how popular the blogs are?

Technorati tagging your posts in Wordpress can be achieved in a number of ways. The key is doing it in such a way as to minimise effort on the bloggers part.

BOTD can bring up some very quirky snippets. I especially liked this one A very funky use of CSS which only a standards compliant browser can view. I wonder has anybody tried doing a tricolour.........

Over the last few months blogs and blogging have received a lot of media attention. In some cases the media provides useful incites via the "new" medium, such as the Guardian's news blogs.

Eric Davis has released a wordpress plugin for feedburner.

I've noticed that quite a few bloggers have started putting Creative Commons licenses on their blogs. Considering the nature of blogs, with their links both in and out and cross-fertilisation I can understand why the CC license is attractive, but why do people make a conscious decision to include it? Do you think it makes any difference?

Site has been moved over to new server and seems to be working correctly. If anybody spots anything odd please let me know ASAP For the curious this new server is running Ubuntu stable instead of Debian. MySQL 4.1 and all the other "bits and bobs"

Judging by what I'm seeing the Google PR update still hasn't settled down. A couple of days ago I was seeing a clear PR of 5 on this blog, but now I see 4, while other people see 5. The only way to see what it is across all Google data centres is to use a tool like this one You can change the URL to your own :)

John had posted some interesting statistics based on his crawling of the IE namespace:

Websites With Title, Keywords and Description: 10460

That's out of a corpus of 36198 domains!! So, if John's stats are true, and I have no reason to doubt him, approximately 66% of Irish websites have no meta content. Some SEO experts may say that meta content is not as important as it used to be, however they would surely agree that TITLE tags are a necessity. Others would also opt in favour of using all available meta content, as it can help with your listing in some cases. From reviewing sites submitted to browse.ie, search.ie and translationresearch.com I would have to agree. In the case of the two Irish directories the quality of submissions is quite poor. If they were being submitted to Skaffe , where I am also an editor, they would either be rejected or pushed between editors until the titles and descriptions had been beaten into an acceptable shape. A lot of the Irish sites submitted are completely lacking in title tags, or all pages share the same ones, which is almost as bad. As John points out, there is a definite need for SEO professionals in the Irish market. I wonder are people willing to pay them though....

Not so many moons ago I mentioned Google Sitemaps which may help with indexing your site. Google are now providing a handly list of plugins, produced by 3rd parties, that hook into their API. For bloggers there is a plugin for Wordpress, while there are others for popular applications such as oscommerce, vbulletin, phpbb and plone (to mention but a few) I've noticed some fellow bloggers are relying on third-party services to generate their sitemaps and google has listed a number of those also The only one I have any experience of is the Wordpress plugin, so I am only providing links to the others

I've been trying to get the runphp plugin with plugins used working. I was hoping to get a simple static page in Wordpress to do this, but for some silly reason it's refusing to work :( I have, however, managed to get it to work for an external page, so you can now see which plugins I'm using.

The launch of the 2005 Netvisionary Awards is this Wednesday evening in Clontarf Castle, Dublin. Free for IIA members. EUR25 for non-members. I'd love to meet up with people who are going :)

While doing some browsing I came across another useful SEO site: seomoz In order to use the site you need:

• Google API key
• Yahoo API key
• Amazon API key

As its tools query a number of online resources using your tag, not theirs, which makes a lot of sense, especially as it's a suite of free tools. The tools available include ones for valuing links, checking keyword popularity and a range of other things that any budding DIY SEO expert should find useful.

According to a post on a Hispanic blog, the latest Harry Potter is already available online, both the book and an audio version of it (read by a human no less) As the original post is in Spanish I'm providing my own translation below:

It took 12 hours for the fans of Harry Potter to publish an illegal electronic version of Harry Potter and the Half-Blood Prince, the latest and much awaited book about the young wizard written by JK Rowling. The process consists of scanning in the book's pages and digitising them using OCR software (Optical Character Recognition). After they proceed to the page by page revision (which is shared out) to ensure there are no errors before publishing it online. If this wasn't enough there is also an audio version of the book available on filesharing networks, read by a human not software. Why does this happen? the problem, for both the publisher and the author, is that being part of the information society is that if you don't make your work available and provide the required facilities, then someone else will. Not only will it happen,. but groups organise themselves in order to get them out as quickly as possible, as happened with the latest Harry Potter

David pointed out a very disturbing piece of tech news. It looks like DRM (digital rights management) is moving towards adding restrictions based on your monitor. Worrying? I would think so.

The Organic SEO wiki seems like a good starting place for DIY SEO. It also has a useful glossary of terminology that you will encounter.

It's almost amusing to see so many bloggers talking about the latest Google updates. We all remember the famous "Florida" update. I posted about it at the time, as I'm sure did many others. Prior to November 2003 there was almost a predictable factor in SEO. If you played by the rules and did certain things you knew more or less where you stood. Page rank (PR) was important, but more importantly you could rely on your position in the search results. With the Florida update all that changed. In the case of our company site it was pretty catastrophic, but we were not alone. Not by a long shot. Clients who had had good rankings saw their sites vanish almost overnight. A lot of theories were put forward as to what was happening, why and how to overcome it (this site contains a summary of some of the more popular theories). It didn't matter. The theories were simply that, theories. What was extremely frustrating was that the quality of the search results after the update seemed to have regressed. If Google had been trying to combat the SEO experts who used dubious methods to boost their clients' rankings, then a lot of innocent bystanders were caught in the crossfire. Almost 2 years later the updates still excite the online community and we all wait to see what will happen. Will our PR increase? Will our search terms drop? Will we still be on the first page for that search query that we cherish so much? In some ways none of this matters as much as before. A lot of us have resorted to using Google's contextual advertising to boost our presence. We still want to appear in the search results, but even if we don't we know that we can hope to catch people's eyes in other ways. Unfortunately the reality is this. Google is still king. Maybe this is a good thing. Maybe it's not. You cannot avoid the facts though. If I look at the referrer logs for any site that I manage I can expect to see the bulk of referrals coming from Google. What does that mean? Does it mean that Google is the most popular search engine? Or does it mean that the people who are most likely to want the kind of information on those websites are more likely to use Google to find the information? It's a hard one to call. Back in the early days of the web there was no Google. There was no MSN. Do you remember the virtual tourist? Can you recall the early days of Yahoo? Anybody remember Altavista (mk. 1)? Or webcrawler? Most of these sites are still around in some form or another, but the way we use them has changed dramatically, as has the face of those who choose to use them. Even the most technically inept 'net user ends up using Google these days, either directly or through one of the many sites that relies on their search technology. Of course I still use Yahoo, but not for searching. I might go to Yahoo for news or other information, but not to actually search.

An interesting article on XML.com on about encrypting and decrypting RSS feeds using blowfish and greasemonkey. I don't know too much about crypto, but it sounds like an intriguing idea, although I am not so sure how secure it really is.

Market research is an important exercise. You should know:

• What people have
• What people would like to have
• What people need (not the same as above)
• What is on offer (in the market)
• How it is offered

There's probably a more scientific way of presenting those ideas, but you should get the general idea. In the hosting market it is not uncommon to contact competitors to seek quotes for various services simply to see how other people are presenting their services and products. After all, when it boils down to it, a domain is still a domain. You can call it what you like, but it's not going to change what it is. I was rather amused, but slightly bemused, when we received two almost identical RFPs in the space of 12 hours. One arrived. I read it and was dealing with it. When the second one arrived I was hit with a sense of deja vu. Hadn't I read the same request only a couple of hours earlier? Maybe I was imagining things. No. I wasn't. Putting them side by side the only differences were minor punctuation and stylistic differences (3 or 4 elements to be precise). Smelling a rat I checked the headers, as did a couple of our technical staff. You would think that they'd have at least tried to hide their origin. Nope. The IP in the header was the competitor's office mail server, with customised SMTP flag and all!! Moral of the story. Don't send emails to competitors from your office IP.

Human edited directories take time to manage. If you insist on breaking the rules you will end up being banned. A few things to avoid:

• Submit the same URL more than once
• Request link exchanges
• Spam
• Argue with the editors

I've also made a point of posting some of this over on the new browse.ie forums

According to the Wordpress download counter close to 400,000 downloads have been recorded of the 1.5* release. So how many installs are there? If you take into consideration the number of blogging platforms based on WP or MU the number grows exponentially and this is without even considering the number of hosting providers offering it preinstalled.

Spamcop has a handy little tool to convert SPF record entries to IP address and/or ranges. Simply feed it a domain name and it will read in the SPF record if it exists and convert it to IPs. If no SPF is present then it will use the MX record. I suppose you *could* use this to check your own SPF records ....

Although I managed to get an internal cosmos and tagging thing working last week I seem to have run into a rewrite rule conflict due to my choice of wordpress plugins. The feedburner plugin I installed a few weeks ago uses mod_rewrite to pass all requests for RSS feeds, regardless of the format, over to the feedburner URI. As the tagging plugin is also rewriting the URLs this causes problems. I'm not sure how I can fix this one, but I may be able to come up with some kind of hack. I think I need to spend some time reading up on mod_rewrite and .htaccess :(

According to an article on The Register, UK ISPs have been asked to retain data prior to last Thursday's attack as it is felt that the bombers may have coordinated their efforts electronically (a not unreasonable assumption). Although this request is reasonable it is a bit impractical:

• Does a common format for logs exist? - No
• What volume of data is involved?- an unknown, but the amount of data involved would be huge
• Will any of the data in space controlled by UK ISPs be of any use to investigators?

If an sms was sent from a non-UK phone on roaming in the UK, for example, which network would hold the logs? If you start throwing around scenarios with regard to the movement of traffic it can cause severe headaches. Although I'm sure ISPs will be more than happy to cooperate in any investigations any logs provided could take months to decipher..

According to reports scam emails purporting to seek funds for victims of last weeks bombing have already appeared. Similar emails appeared shortly after the Asian tsunami.

I woke up this morning to find that MySQL had died during the night. Restarting it wasn't a problem, but my feeds were broken until I got Feedburner to resync.

There seems to be a cottage industry in domain scams. Every few months these scams come to the fore and we hear about them either directly from our clients or from other industry sources.

Sitevista was officially launched earlier this evening. The site should be popular with web designers and developers. I gave a brief overview of its features a couple of weeks ago

Google have released an official version of the Google toolbar for Firefox 1.*+ The site seems to be up and down quite a bit (ie. you may not be able to access it), but I was able to grab a copy of it. You can try accessing the firefox page here Another link someone posted is to a page full of Google extensions for firefox

Finally some sense amidst all the chaos. The vote was rejected 648 to 14 with 18 abstentions. More information: jmcc The Register RTE

The EU votes on software patents today. More information available: IFSO Silicon Republic The Guardian

Ask any *nix user group which distro you should use and you'll probably start a minor religious war :) Padraig Brady provided a very handy link to make it all a lot more painless Based on my answers it suggested Kubuntu for my laptop, which was a sane choice, as I like apt Edit: I installed Kubuntu on my laptop this evening via apt.

Ireland is supposedly a Catholic country and I suppose it has managed to retain some of its "moral values". Possibly more than other countries, although I doubt if all Irish people would subscribe as heartily to this ethos in the privacy of their own homes as they might like people to believe in public. A couple of weeks ago a new TLD specifically for the adult industry was approved by ICANN - .xxx (hardly the most inventive of TLD denominations, but at least its intended audience is clearly defined). Irish legislation is heavily influenced by Catholic and Christian values and anybody who denies that need only look to the Constitution to realise the error of their beliefs. That this is the case is not surprising, nor is it an exception. Other states have similar influences on their legislature. It is only countries, such as France, that have managed to truly separate Church and State (or more correctly Church from State). For a number of reasons the topic of adult content in the Irish context has arisen a few times in the last few weeks, both in my conversations with industry colleagues and in more public fora. Fergus Cassidy's most recent column makes reference to one of the IEDR's more bizarre naming policies: 3.4: The proposed domain name must not be offensive or contrary to public policy or generally accepted principles of morality. (see http://www.iedr.ie/NamingPolicy.php) Irish law is based on common law principles, which means that the law is more than the statute books, but a combination of statutes, case law, precedent and interpretation. It is not uncommon, therefore, to find tracts of "law" that leave a lot to be desired in terms of precision. However, laws governing the state should not be confused with rules governing the registration of domain names. Ambiguity and subjectivity are problematic and should be avoided. Up until the late 90s it was not possible to buy a copy of Playboy magazine legally in Ireland. In early 1992 there was a major scandal surrounding the distribution of contraceptives on Irish university campuses. It is now 2005. What has changed? A lot. Most newsagents stock Playboy and similar magazines in plain view, albeit on the top shelf. Subscribers to Sky and other services can receive premium adult channels that leave little to the imagination. Yet you cannot register porn.ie or hope to find an Irish hardcore website hosted in Ireland. (Playboy.ie is registered by the way, though hustler.ie isn't) Why is that? I know that we as a company would have problems if we were to host adult content on our network, but probably not from a legal point of view. It would be more likely to cause issues with some of our other clients. What exactly is the legislation covering this area? It's not too clear, and when I say that I am probably being generous. This lack of clarity is problematic to say the least. Although as a company I would not be in favour of us hosting hardcore porn or anything that could be interpreted as such, it would be preferable if the actual boundaries were defined clearly. There have been discussions of this issue in the past, but they have all led nowhere, as the actual legislation itself is hard to follow. For example, there are a couple of articles online that may act as starting points: http://www.siliconweb.ie/webtalk/020899.shtml http://www.ispai.ie/press.htm However the articles are very old, and are not exactly comprehensive. If anybody has any more recent legal references I would love to see them.

Link rot is the bane of directory editors. If you run a niche directory site, such as I do, one of the things you want to make sure of is that your links are current ie. that they do not give 404s or other errors. There is no point in fooling yourself into thinking that you can take on the likes of Google, but you can offer good resources for your chosen subject. The backend of Browse.ie is powered by Gossamer Threads Links SQL (now known as Glinks) which has a link checker builtin. Unfortunately, as I discovered, the UserAgent string it was reporting was generic and had been blocked by some website owners. With the help of Niall O'Broin I finally fixed the UserAgent to be something a little more descriptive. If you see the following string in your logs, then you'll know that it has visited your site: "Browse.ie Link Checker - See http://www.browse.ie" An earlier version reported itself as: "Browse-ie Link Checker - See http://www.browse.ie" The robot will simply check that the link is valid. Nothing more. Nothing less. As long as it gets a valid response, such as a 200, it will mark your link as being "ok". I'm not sure how often the robot will do the check, but I expect to run it once a week or so to ensure that there are no dead links ie. avoid link rot