Blog
Juno Records Launch
Posted in Blog, December 9, 2004 | 2 comments
Well, today we launched the site that's taken up the last few months of my life. Working in London has been a pain, and working as a team of one has had its ups and downs. However, it's been an interesting project, and I think (hope) I've learned a lot about everything from load balancing and database replication to tracking how users use sites and how subtle changes can affect that, as well as project management and diplomacy - but not nearly enough about the last two. A proper writeup is on the way. However, for now, you'll have to settle for a couple of URLS:
New Juno Website: www.juno.co.uk
Old Juno Website: classic.juno.co.uk
New Juno Website: www.juno.co.uk
Old Juno Website: classic.juno.co.uk
Mime Type Fun Revisited
On November 3rd, I posted Mime Type Fun, a quick note explaining that I was now serving this site, written in XHTML 1.1, with the correct "application/xhtml+xml" mime type.
It was a useful exercise. Apart from anything else, it gave me a good opportunity to clean up much of the code on the site. However, it also highlighted problems with XHTML 1.1 and the "application/xhtml+xml" mime type, that are the reason I have now switched back to plain ol' HTML.
Now, don't get me wrong. I heartily support XHTML and believe it is the future of web design. However, right now, it just isn't nearly ready for primetime.
I did consider moving back to XHTML 1.0 or 1.1, but serving it with the "text/html" mime type. However I do believe that if you are going to bother writing a site in XHTML, you should at least serve it correctly to those user agents that can understand it.
If this site was completely static, without the LiveSearch, AdSense or the other small pieces of JavaScript dotted around the place, things may be different. As a tool for marking up static pages, XHTML works well, even with content negotiation techiques. But for anything involving interactivity, I don't believe it's going to be worth the time and effort required to change until the majority of users can experience the benefits.
Once enough users can receive XHTML properly, this site will be changing straight back over. The lessons learned from this experience will be invaluable. Until then, it's back to HTML - which at least means the LiveSearch (on the right) is working again.
It was a useful exercise. Apart from anything else, it gave me a good opportunity to clean up much of the code on the site. However, it also highlighted problems with XHTML 1.1 and the "application/xhtml+xml" mime type, that are the reason I have now switched back to plain ol' HTML.
Now, don't get me wrong. I heartily support XHTML and believe it is the future of web design. However, right now, it just isn't nearly ready for primetime.
I did consider moving back to XHTML 1.0 or 1.1, but serving it with the "text/html" mime type. However I do believe that if you are going to bother writing a site in XHTML, you should at least serve it correctly to those user agents that can understand it.
If this site was completely static, without the LiveSearch, AdSense or the other small pieces of JavaScript dotted around the place, things may be different. As a tool for marking up static pages, XHTML works well, even with content negotiation techiques. But for anything involving interactivity, I don't believe it's going to be worth the time and effort required to change until the majority of users can experience the benefits.
Once enough users can receive XHTML properly, this site will be changing straight back over. The lessons learned from this experience will be invaluable. Until then, it's back to HTML - which at least means the LiveSearch (on the right) is working again.
Lycos Spam Screensaver Doomed To Fail
Posted in Blog, November 30, 2004 | 2 comments
Lycos has announced the release of a screensaver called "Make Love Not Spam". This, they say, will help in the fight against spam, by helping increase the costs to spammers of sending spam out.
Spam is only currently a practical business tool because the cost of sending email is so miniscule that a tiny response covers the costs of sending it - so you can see the logic behind Lycos's tool - increasing the costs of sending spam render spam impractical.
The system itself is fairly simple. A list will be maintained of URLs promoted through spam. Anyone who installs the screensaver from Lycos will then automatically visit these URLs when their computer is idle.
The idea is that with tens of thousands of people using the tool, the sites promoted through spam will end up with a large amount of extra - and completely useless - traffic. This will cost whoever runs the site money - so promoting a site or service through spam will be seen to be too costly to do so again.
However, though the idea on the face of it sounds reasonable, there is no way this can succeed at reducing spam on the internet (or even just irritating spammers).
The problems are with how the system functions. Essentially what Lycos are doing are organising a Distributed Denial of Service (DDoS) attack on URLs they select themselves. They have said their intention is not to take sites down and they will limit the level of traffic sent to a domain - but that doesn't alter the fact that it is a DDoS attack, even if it fails (which is illegal in many countries, I believe - though I am not sure - perhaps someone could clarify this?). Not to mention, they cannot possibly know the capabilities of a server they are attacking - so how they can limit their attacks so they will not crash a site (and thereby execute a successful DDoS attack), and still generate enough traffic to cause a problem, I don't know.
There is a human element to the process, which introduces a risk. URLs are selected from a generated list, from sources such as SpamCop - but they are selected by Lycos employees. I see no reason to believe it will be impossible for a Lycos employee, if they wished, to add a URL to the list to be spammed that just takes their fancy.
The URL list, even if URLs are checked well and the list is well maintained, can not be trusted. If I was feeling malicious, I could sent out a few thousand emails advertising a URL. Once it was added to the Lycos system, I could then redirect traffic from that URL to another one. In no time at all, Lycos would be costing my competitors a substantial sum of money.
Assuming all URLs in the system are ones that deserve to be attacked, there is no logical indication that this kind of attack will be effective at stopping, slowing, or affecting the cost of spam. Few spammers promote their own URLs. The vast majority of spam is sent by a very small group of spammers, and they are paid vast quantities by companies to send it.
They stand to lose nothing, directly, through Lycos. They will not be paying for their clients' traffic - that would be stupid. And contrary to popular belief, spammers are not stupid. They make millions by irritating people daily, and spend huge amounts of time working on bypassing sophisticated spam filters, after all.
Some have suggested that perhaps this will have the effect of stopping a company from sending out spam email twice. Perhaps. However, why would they be spamming in the first place? To increase traffic.
Will the traffic from the "Make Love Not Spam" screensaver be in any way distinguishable from normal traffic? If it is, it can be filtered and will have no effect on any site on a half-decent host. If it is not distinguishable from normal traffic, then the spam campaign will look like it has been successful - and the company will likely repeat the exercise.
True, this traffic increase will not yield a corresponding sales increase, but any salesman capable of convincing a company to send out hundreds of thousands of commercial emails to the unsuspecting public is going to be perfectly capable of convincing the same company that they just need to word the next email a little better, or that they need to make some changes to their site, to improve their conversion rate.
It can cost some people money, though. Anyone on a shared hosting environment, where they pay for extra bandwidth by the gig, will be hit hard by this. They could end up losing hundreds, maybe thousands, of pounds. However, a very tiny percentage of people hosting sites on that scale can afford to, or want to, send out spam. Stopping them, if it worked, would not make any noticable dent in overall spam levels.
My honest belief at this point is that the Lycos "Make Love Not Spam" screensaver is nothing more than a cheap publicity stunt, aimed to raise Lycos's profile. It cannot actually stop or hinder spam, and may even increase it. It will certainly increase the level of noise on the internet (noise being basically useless traffic), and that will cost the ISPs money as they nee to upgrade machines and connections (guess where those costs are passed on to).
Spam is only currently a practical business tool because the cost of sending email is so miniscule that a tiny response covers the costs of sending it - so you can see the logic behind Lycos's tool - increasing the costs of sending spam render spam impractical.
The system itself is fairly simple. A list will be maintained of URLs promoted through spam. Anyone who installs the screensaver from Lycos will then automatically visit these URLs when their computer is idle.
The idea is that with tens of thousands of people using the tool, the sites promoted through spam will end up with a large amount of extra - and completely useless - traffic. This will cost whoever runs the site money - so promoting a site or service through spam will be seen to be too costly to do so again.
However, though the idea on the face of it sounds reasonable, there is no way this can succeed at reducing spam on the internet (or even just irritating spammers).
The problems are with how the system functions. Essentially what Lycos are doing are organising a Distributed Denial of Service (DDoS) attack on URLs they select themselves. They have said their intention is not to take sites down and they will limit the level of traffic sent to a domain - but that doesn't alter the fact that it is a DDoS attack, even if it fails (which is illegal in many countries, I believe - though I am not sure - perhaps someone could clarify this?). Not to mention, they cannot possibly know the capabilities of a server they are attacking - so how they can limit their attacks so they will not crash a site (and thereby execute a successful DDoS attack), and still generate enough traffic to cause a problem, I don't know.
There is a human element to the process, which introduces a risk. URLs are selected from a generated list, from sources such as SpamCop - but they are selected by Lycos employees. I see no reason to believe it will be impossible for a Lycos employee, if they wished, to add a URL to the list to be spammed that just takes their fancy.
The URL list, even if URLs are checked well and the list is well maintained, can not be trusted. If I was feeling malicious, I could sent out a few thousand emails advertising a URL. Once it was added to the Lycos system, I could then redirect traffic from that URL to another one. In no time at all, Lycos would be costing my competitors a substantial sum of money.
Assuming all URLs in the system are ones that deserve to be attacked, there is no logical indication that this kind of attack will be effective at stopping, slowing, or affecting the cost of spam. Few spammers promote their own URLs. The vast majority of spam is sent by a very small group of spammers, and they are paid vast quantities by companies to send it.
They stand to lose nothing, directly, through Lycos. They will not be paying for their clients' traffic - that would be stupid. And contrary to popular belief, spammers are not stupid. They make millions by irritating people daily, and spend huge amounts of time working on bypassing sophisticated spam filters, after all.
Some have suggested that perhaps this will have the effect of stopping a company from sending out spam email twice. Perhaps. However, why would they be spamming in the first place? To increase traffic.
Will the traffic from the "Make Love Not Spam" screensaver be in any way distinguishable from normal traffic? If it is, it can be filtered and will have no effect on any site on a half-decent host. If it is not distinguishable from normal traffic, then the spam campaign will look like it has been successful - and the company will likely repeat the exercise.
True, this traffic increase will not yield a corresponding sales increase, but any salesman capable of convincing a company to send out hundreds of thousands of commercial emails to the unsuspecting public is going to be perfectly capable of convincing the same company that they just need to word the next email a little better, or that they need to make some changes to their site, to improve their conversion rate.
It can cost some people money, though. Anyone on a shared hosting environment, where they pay for extra bandwidth by the gig, will be hit hard by this. They could end up losing hundreds, maybe thousands, of pounds. However, a very tiny percentage of people hosting sites on that scale can afford to, or want to, send out spam. Stopping them, if it worked, would not make any noticable dent in overall spam levels.
My honest belief at this point is that the Lycos "Make Love Not Spam" screensaver is nothing more than a cheap publicity stunt, aimed to raise Lycos's profile. It cannot actually stop or hinder spam, and may even increase it. It will certainly increase the level of noise on the internet (noise being basically useless traffic), and that will cost the ISPs money as they nee to upgrade machines and connections (guess where those costs are passed on to).
Satellite TV
Posted in Blog, November 30, 2004 | One comment
Having just moved to a nice new flat, I'm now in the process of getting everything organised and set up. Decent TV would be nice, but is proving harder than it should be.
My road's a fairly nice one, and I'd need planning permission to put up a satellite dish. NTL, in their infinite wisdom, have decided that a road packed with people who can't get Sky isn't worth cabling. Maybe they don't like money much.
Apparently, indoor dishes are possible, and freeview is a possibility (though rather poor choice of channels, even with TopupTV). But if anyone has any ideas of other options available, I'd love to hear them!
My road's a fairly nice one, and I'd need planning permission to put up a satellite dish. NTL, in their infinite wisdom, have decided that a road packed with people who can't get Sky isn't worth cabling. Maybe they don't like money much.
Apparently, indoor dishes are possible, and freeview is a possibility (though rather poor choice of channels, even with TopupTV). But if anyone has any ideas of other options available, I'd love to hear them!
Mum vs Firefox
Posted in Blog, November 26, 2004 | 20 comments
I've tried for a while now to persuade my mum to take basic precautions to ensure she is as secure as possible when using the internet. Weening her off AOL took the promise of faster internet and better reliability with a Pipex broadband connection (1). Small things, like installing a firewall and antivirus for her, took many months of reasoning before permission was given.
In the end though, I am happy enough for my mum to use the net without major problems. And until recently, she was using Internet Explorer. ActiveX was disabled, though she had no idea what that actually was - or that it was disabled. Same with third party cookies and several plugins.
Occasionally I had to explain why a site wouldn't work, and she'd be happy enough finding and using a different one. She referred to Internet Explorer as "Google" and associated (and still does to a degree) the blue 'e' on her taskbar with Google, because it was her homepage.
Eventually, though, the only things left to do were to help her switch to a better browser and email client (Thunderbird's not quite ready, but that switch is coming soon).
Opera was something of a disaster. She liked the way the email client worked, and picked that up in no time, but as a whole she found there was too much to take in. I removed just about every button and bar, till we were left with basically a pretty-looking replica of the IE interface, but by then the damage was done. She just doesn't like Opera. At least she's tried it though, which is something.
After Opera, of course, came Firefox. Now it's actually been released, I'm happy to recommend it to her. And the funny thing is, she loves it. She says the top bar is prettier and easier to use than IE (actually, she says it's prettier than Google, but we all know what she means). She loves the popup blocker, though I'm fairly sure she's forgotten it's there.
Best of all though is the stuff she doesn't know is there. She has noticed that some sites look better. Not many, but a few. She doesn't know it, but she is more secure. She's even started using favourites - something she, until recently, avoided like the plague.
We were having a chat about it today, though, and that's when it hit me - most of the reasons I gave her for changing are completely and utterly worthless. She doesn't care if it renders pages like the W3C say they should be rendered. She doesn't know, or want to know, who the W3C are or that CSS is what controls the look of pages.
She doesn't care about tabbed browsing. She uses one web page at a time. When she's finished with it she goes back to Google and starts over again. She might, one day, start to use tabbed browsing, but I doubt it. She doesn't care about RSS. She doesn't know what that is. She wouldn't use it if she did.
At the end of the day, she likes - and uses - Firefox because it's faster, better looking, and just plain easier to use. She likes the fact she feels safer using the net with it. That's it.
Perhaps that's how we should start pitching it to the masses.
1 Before anyone complains that I've done something wrong by getting her to switch to broadband, she's very happy she's changed. She loves the speed, and it's saving her money - before, she had a second line and was paying for an unmetered connection.
In the end though, I am happy enough for my mum to use the net without major problems. And until recently, she was using Internet Explorer. ActiveX was disabled, though she had no idea what that actually was - or that it was disabled. Same with third party cookies and several plugins.
Occasionally I had to explain why a site wouldn't work, and she'd be happy enough finding and using a different one. She referred to Internet Explorer as "Google" and associated (and still does to a degree) the blue 'e' on her taskbar with Google, because it was her homepage.
Eventually, though, the only things left to do were to help her switch to a better browser and email client (Thunderbird's not quite ready, but that switch is coming soon).
Opera was something of a disaster. She liked the way the email client worked, and picked that up in no time, but as a whole she found there was too much to take in. I removed just about every button and bar, till we were left with basically a pretty-looking replica of the IE interface, but by then the damage was done. She just doesn't like Opera. At least she's tried it though, which is something.
After Opera, of course, came Firefox. Now it's actually been released, I'm happy to recommend it to her. And the funny thing is, she loves it. She says the top bar is prettier and easier to use than IE (actually, she says it's prettier than Google, but we all know what she means). She loves the popup blocker, though I'm fairly sure she's forgotten it's there.
Best of all though is the stuff she doesn't know is there. She has noticed that some sites look better. Not many, but a few. She doesn't know it, but she is more secure. She's even started using favourites - something she, until recently, avoided like the plague.
We were having a chat about it today, though, and that's when it hit me - most of the reasons I gave her for changing are completely and utterly worthless. She doesn't care if it renders pages like the W3C say they should be rendered. She doesn't know, or want to know, who the W3C are or that CSS is what controls the look of pages.
She doesn't care about tabbed browsing. She uses one web page at a time. When she's finished with it she goes back to Google and starts over again. She might, one day, start to use tabbed browsing, but I doubt it. She doesn't care about RSS. She doesn't know what that is. She wouldn't use it if she did.
At the end of the day, she likes - and uses - Firefox because it's faster, better looking, and just plain easier to use. She likes the fact she feels safer using the net with it. That's it.
Perhaps that's how we should start pitching it to the masses.
1 Before anyone complains that I've done something wrong by getting her to switch to broadband, she's very happy she's changed. She loves the speed, and it's saving her money - before, she had a second line and was paying for an unmetered connection.
Missing: One Web Developer
The site's a little quiet this month, and for that I apologise. Plenty more to come over the next few months, but life has decided to get in the way of regular updates for now.
I'm moving to Hove on Saturday, and changing jobs in a week. It's been great working for Juno Records (the new site will be out of beta next week), but it's time to move on. I'll still be involved with the Juno site for some time to come, but in a smaller capacity - and I'm looking forward to seeing how it develops over time - there's some very cool stuff planned for the site!
I'll be starting work for Propellernet on the 6th December, though, so it is that that is taking my free time at the moment. It's an interesting role, certainly, especially given that their focus is marketing rather than development. Another huge bonus is that this gives me the opportunity to work with, and learn from, Ammon Johns.
Finally, I'd like to wish Kim Krause, the driving force behind Cre8asite Forums, many congratulations on her imminent wedding!
I'm moving to Hove on Saturday, and changing jobs in a week. It's been great working for Juno Records (the new site will be out of beta next week), but it's time to move on. I'll still be involved with the Juno site for some time to come, but in a smaller capacity - and I'm looking forward to seeing how it develops over time - there's some very cool stuff planned for the site!
I'll be starting work for Propellernet on the 6th December, though, so it is that that is taking my free time at the moment. It's an interesting role, certainly, especially given that their focus is marketing rather than development. Another huge bonus is that this gives me the opportunity to work with, and learn from, Ammon Johns.
Finally, I'd like to wish Kim Krause, the driving force behind Cre8asite Forums, many congratulations on her imminent wedding!
Which Irritating "Which ... Are You?" Quiz Are You?
Posted in Blog, November 12, 2004 | 5 comments
Is there anything more annoying than a "Which ... Are You?" Quiz? Possibly just one. A Which "Which ... Are You?" Quiz Are You Quiz. Fill out the following survey if you have too much time on your hands:
One Month Smoke-Free
Posted in Blog, November 11, 2004 | 10 comments
Woohoo, one month without ciggies! A good start, I think. I've avoided the constant snacking, and though I have put on a little weight, it's less than I feared. In fact, next to the mound of blubber I started with it's nothing. I've been a little snappier, and went through a short phase of replacing my nicotine intake with caffeine. No sleep for a couple of weeks was fun. I seem to be back to normal now though. I just hope I can keep it up this time!
iPod, meet Winamp
Posted in Blog, November 10, 2004 | 6 comments
I have an iPod. I like Winamp. So until recently, my life was trickier than need be. The iPod relies on iTunes to update its music, and only iTunes can make use of ratings and so on from the iPod. This means that playlists on your iPod, ratings, play counts and so on, cannot be carried acrodd to Winamp - my player of choice.
Until recently, that is. I've been using ml_iPod for a few days now and have found it very useful. It basically operates the same was as iTunes, allowing you to sync your media library, create smart playlists, and save ratings and play counts in both directions. Very very cool.
Until recently, that is. I've been using ml_iPod for a few days now and have found it very useful. It basically operates the same was as iTunes, allowing you to sync your media library, create smart playlists, and save ratings and play counts in both directions. Very very cool.
Firefox Launches
Posted in Blog, November 9, 2004 | One comment
Firefox, one of the best of the current browsers on offer, was today finally released (up till now we have been using preview and beta versions). If you haven't got it yet, I suggest you go and download Firefox now - you will find using the internet easier, faster and more secure.
Lewes Bonfire Night 2004
Posted in Blog, November 6, 2004 | 5 comments
Lewes is a sleepy little town nestled in the Sussex Downs. For 364 days a year, nothing noteworthy happens. On November the 5th, though, things are a little out of the ordinary ...
Mime Type Fun
Posted in Blog, November 3, 2004 | 2 comments
I've made the switch, finally! This site is now served with the mime type "application/xhtml+xml" to those user agents that can understand it, and "text/html" to those that can't. However, this means that there is a fairly good chance that some things won't work perfectly. I'll be fixing bugs as I find them, but if you do spot any glitches (unencoded ampersands and so on) please let me know!
One problem I am aware of and will be fixing as soon as humanly possible is that the JavaScript in the site is, for the most part, now not working. The LiveSearch, for example, is currently inactive.
Notes
Adsense is broken. There are solutions, but none pretty. (fixed)
Caching system stuffed. (fixed) (no it isn't) (oh yes it is) (web developer pantomime mode fixed)
When serving "text/html", should really be using a different DTD and rewrite the code slightly. (fixed)
JavaScript problems, mainly LiveSearch (looks like Opera has problems executing a script appended with appendChild when page is served as "application/xhtml+xml")
Comment posting bug. (error fixed but "remember me" not working) (fixed)
One problem I am aware of and will be fixing as soon as humanly possible is that the JavaScript in the site is, for the most part, now not working. The LiveSearch, for example, is currently inactive.
Notes
Adsense is broken. There are solutions, but none pretty. (fixed)
Caching system stuffed. (fixed) (no it isn't) (oh yes it is) (web developer pantomime mode fixed)
When serving "text/html", should really be using a different DTD and rewrite the code slightly. (fixed)
JavaScript problems, mainly LiveSearch (looks like Opera has problems executing a script appended with appendChild when page is served as "application/xhtml+xml")
Comment posting bug. (error fixed but "remember me" not working) (fixed)
Page: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
ILoveJackDaniels.com is the online playground of