CommenTag | official blog

Last Tuesday I was invited by Andy Bargery to present commentag at the London bloggers meetup. It was the second time I attended such event. The first time was one month ago with @gandalfar from Zemanta and it was great fun.So this time I had to present. As usual I started preparing my slides the day before. But hopefully I could reuse some slides that I’ve done for the London Minibar but which I couldn’t show there because Mike Butcher (@mbites) only allowed me 5mn to present (which is pretty fast).But most of the presentation was all new stuff. I get inspired by this reading about the 20 ways to get more blog comments.I found them a bit repetitive and some a bit weird (like don’t be afraid to get off topic… I think a the main difference of a forum and a discussion on a blog is that the discussion refers to the topic of the post… hummmm…), so I remixed them in a highly visual keynote and here is the result. 

| View

Btw, thank you guys @slideshare for having featured this presentation on your homepage :-)The next London Bloggers Meetup will be held on the 29th July at 7.30pmFor a more extensive review of this London Bloggers meetup, check this post.

Sorry to have been so long.But here it finally is. It is just a summary of interesting thoughts that came up from this meeting and which might be interesting for anyone who is in the process of bootstrapping a startup. We talked about many things and describing everything would be a bit too long so let’s focus on the essential. 

First there is that funny quote about French people saying that because of our French culture we feel like we have to release fully polished products (and therefore we wait too much). 

About the concept of commentag he highlighted the fact that our real contribution (ie. our killer feature) was our concept of “Nested Tags”. The fact that when you click on one tag, you get directly the other related tags allowing you to narrow down your selection. This is our real innovation and we should more communicate about that.

Concerning my slides, he said that we should more highlight the market.The number of potential users, converted users.How many users will we be able to reach?That’s the very question. It is first all about how to grab the market, then in second how to turn this market of users into a revenue model.

Funny enough, this is almost contradictory with what we are used to in Europe. We just met Patrick Crasson, responsible of the Startup Essentials program of SUN in Belgium. His very first question was “What is your business model?”. “How will you make money” comes here first and not “how will you gain users”. So in Europe, you first have to focus on your revenues, then on the user experience. While VC in the States would allow you to first focus on the user experience (as long as there is a real market).

Both approaches can be good or bad depending on the business. The very question is in which game should we play? Should we see big and focus on user experience, or should we play safe and focus on a smaller market but on which we will be able to make money quite quickly?

Talking about the future applications of commentag he said “it’s not about a piece of software that could be applied to many different situations. It is all about a distribution channel”.  “If you start targeting public content like comments, then you can’t apply your product to manage emails or contacts. Because the distribution is not the same. The former is public whereas the latter is private.
However, you could apply it to different type of public content.So the idea here is to focus on developping a distribution on which a piece of software will be applied. Not the contrary.” 

This is a very interesting thought.Because at a first glance we would all think in terms of software (or product) and in its capabilities.At best, we will think in terms of market pain and try to address the issue (what we did), but never we would have thought in terms of distribution.So it is very important to focus on one distribution and then leverage it.

About the current downturn he said that it only gonna hurt companies selling advertises in the range $4 to 20$ cpm (cost per mille). ie. all companis playnig the role of useless intermediaries.However, in the funding stage the situation will probably be more difficult as investors would invest in less projects (which is already the case in Silicon Valley, see doom and gloom hits Silicon Valley).

Thanks again Reshma and Scott. Your advises are like always very helpful and much appreciated.

That was a great, great time we spent in Paris last week-end, while attending the WordPress BarCamp at La Cantine !

There, we were able to meet really interesting people, had edifying talks and debated about really hot topics such as WP & video, collective translation, OpenID, Social Nets, and even… our big brother Semantic Web !

Not only almost every attendee was able to give us his critics about Commentag’s concept and features, but we also discovered more about their own work, projects or blogs :

• Seesmic
• Travailleurs du Web
• Gobz
• L’Observatoire des Medias
• Agendeo
• Cinod
• Lense
• …

Thank you all for your remarks ! Looking forward to see you soon in a Brussels WordCamp… ^^

The main critic about Video comments is that we can’t scan through them as we can do with text comments.

Coupling Seesmic with Commentag would address that issue in a brilliant and innovative way. People could easily tag their video and visitors would be able to filter Video comments based on their topics.
That way, visitors would be able to only play video comments which really interest them.

The website Francofous-seesmic has installed both Wordpress plugins.
So we did. Give it a try!

It all started with Philippe Jeudy who interviewed me at yesterday Paris WordCamp and uploaded live the video to his seesmic account.

I’ve just submitted the idea to Seesmic team. Let’s see what will happen!

The Prototype library in its version 1.5 was quite buggy.
The problem is when you have a fancy old Wordpress plugin (like Lightbox 0.5), you might load Prototype 1.5_pre1 without notice, and it may broke other plugins (like the Commentag one).

If you are in that case, don’t panic.
Just update your Lightbox plugin to its latest version to date (0.6.3, available here)

If you are using another fancy plugin using an old version of Prototype, make sure to update it (Prototype is now at version 1.6.0.2)

Strangely enough, the Commentag plugin was not working on Nellio.com.
It produced some strange javascript outputs.

Digging into the problem showed that jQuery is not a friend of Prototype (two Javascript librairies).
So, I’m now running jQuery in noConflict() mode.

Besides, I also had to change
for (t in tags[])
to something like
for (t=0;t < tags.length;t++)

Because it looks like Prototype is adding some helpers to manipulate arrays which causes a for each loop to actually call these helpers. Which is really bad.

So now, it is all fixed.
Oh, btw, no one has got to update their commentag plugin
If you had the problem and turned off the commentag plugin, just re active it and it will work like a charm.

Don't forget to give us your feedback

Commentag.com - sorted thinking (teaser)Vidéo envoyée par xdammanNobody reads comments? That may change!Commentag is mixing Web semantic and UGT (User Generated Tags) to cope with information overload.Give it a try on commentag.com.

It took a bunch of hours but now it is done.
How come can  people still use IE? It is pure rubbish and a real nightmarre for web developpers.

Anyway. I’ve borrowed an old and noisy Dell Inspiron 5100  which runs IE6 and I spent my last evening focusing on making IE6 understands standards.

The most annoying thing is its lack of support of transparent PNG files. I’ve used the excellent script of 24ways which takes advantage of the IE-only functionality of post processing a page using an AlphaFilter.
More info here.

Amongst other  hacks I had to go through, there is to mention that there is a way to specifiy CSS properties which will be only interpreted by IE6.

For example, if you start your CSS property with an underscore (eg. _background: …) then only IE6 will interpret it.

In the same way, if you start with a double slash (eg. // background: …) then all browsers but IEx will interpret it (note: the standard way to put comments in a CSS file is only to use the delimiters /*  …   */ ).

Last but not least, Firefox and Safari both consider a defaut left margin of 0px for <ul> while IE6 does not.
So don’t forget to state
ul {
margin-left:0;
}

It also looks like floating elements do not always have a good presentation under IE6. Sometimes it’s better to put your object with a _position:absolute; (what I did for the commentag logo to make it overlap the main part of the site).

Other useful links to deal with compatibility issues between IEx, Firefox and Safari:

• CSS Contents and compatibility
• [ws] The Underscore Hack | Podtržítkový hack 

I noticed something strange on my parent’s Senseo. First, it seems to take longer to boot up than mine. I should investigate this and do some benchmarks. And second, the “auto shut-off” function doesn’t work. The Senseos have differents versions but mine is older. I really don’t have a good explication; I can only suppose that my parents’ machine is more used. I will keep you updated on this.

Besides this, I have been looking at Amazon’s S3 and EC2. While I think EC2 is more adapted for computing than for web services, it can help to manage the scalability.

EC2 is a service to run virtual machines (”instances”). Sure, you can build an (almost) complete virtual web infrastructure if you want. But if you want reverse proxy and load balancing, it’s up to you. You have to configure everything yourself. While this is interesting (and fun) to do, it will take a lot of time to build and maintain. Or you can use solutions like scalr or weoceo. The first is free and open source but very young and the second costs and is not so fine-tunable. Those solutions integrate load-balancing and managing tools.

S3 is the storage solution. It stores juste simple files (”buckets”) and the virtual machine images. The most important is that EC2/S3 lacks database support. Sure you can use S3 to backup the database but that’s just a hack and I wouldn’t recommend it for production use.

I think of EC2/S3 is not going to replace a real server for a web service but I still think about it for the scalability. Or to answer traffic spikes, like what Judd Vinet did (see http://hostingfu.com/article/setting-up-part-time-web-cluster-with-amazons-ec2).

I will continue to investigate Amazon’s services’ use for commentag. Stay tuned.

Good NLP is very difficult, because of real-world ambiguity inherent to nearly every real-world situation. By its nature, it truly is a NP-complete problem. Trying to solve it brute force is pointless.

“He knew what he had to do. It was, of course, an impossible task. But he was used to them. Dragging a rat all the way from the wood to the hole had been an impossible task. But it wasn’t impossible to drag it a little way, so you did that, and then you had a rest, and then you dragged it a little way again… The way to deal with an impossible task was to chop it down into a number of merely very difficult tasks, and break each one of them into a group of horribly hard tasks, and each one of them into tricky jobs, and each one of them…” — quoting Hamish Cunningham’s quote from Terry Pratchett, Truckers, p. 119.

Here comes the strategy of Commentag. True NLP is impossible. Right. So what ? No big deal ! Let’s just move on and make at least a few things work. And then some better POS-tagging. And then some more powerful semantem extraction. Until we finally obtain as good results as our big brothers.

So we are on a concrete detection of local and specific information extraction, such as dealing with nicknames, recognizing several specific subjects, defining a minimistic set of microgeneric sems or taxems, etc.

Let’s travel a bit lighter this week. Do you feel that fresh air ?