I remember asking my brother a while ago what programming language people would still use in 10 years. I was suggesting Ruby, and since he’s into .NET, I thought he’d say C# or asp. To my surprise, Javascript was his immediate answer.

That was before words like AJAX or javascript libraries were mainstream. Javascript is now used in all sorts of flavour on the web, from dynamizing user interfaces, up to powering whole websites like gmail or 280Slides. In 2008 alone, we’ve seen the apparition of stand-alone javascript web frameworks like SproutCore or Objective-J.

A new service called the Reasonable Smart Platform aims to push further boundaries, by offering a platform for web developers that will host their application and store the data for them.

In practice, you open an account by using OpenId. You can then start a new host, or clone existing ones, with files hosted on RSP through git. Other programmers can also clone your host, and then push changes.

There is currently a minimal set of features. HTTP requests are taken care by bootstrap.js; you can also access the filesystem, store, retrieve and search data, and manage sessions. All of this is done with javascript; which is executed on the RSP server upon user requests.

If you go through the website, it seems raw; but the service is already functional. The website in itself uses RSP. Here is for instance a sample application done today, a demo made just for this post. It’s a board where visitors can leave a message, like what they are doing currently. If you have a RSP account, the code is hosted at http://mtw-demo.reasonablysmart.com:81/ if you want to play with it.

I’d like to believe that this is a strong contender to other cloud computing companies. They are actually looking for investors to fund and develop the project; it’s for instance a direct competitor to 10gen. Of course, there are many shortcomings (such as lack of documentation and examples); but I know how talented they are, seeing them working at blitzweekend so this shouldn’t be a problem.

Mini … minimalism

Marc-André Cournoyer has just published Invisible (link to the GitHub repository), a ruby web framework that aims to be as minimalist and compact as possible, with the core code taking 2k lines.

Yes, you’ve read it well, it’s a web framework and it’s under 2000 lines of code. In comparison, in other web frameworks like Ruby On Rails, there are base files that are well beyond 2000 lines (here or here for instance)

Thanks to its simplicity, I imagine the web framework could be used to get a simple website in a minimal amount of time, such as a website that has only a dozen or so (static) pages. It might not be production ready though for web developers, since it has minimal no support for stylesheets, javascripts, or any other fancy web2.0 effects.

In related news, the prolific Marc-André Cournoyer also released a while ago Thin Turbo, aiming to speed up Thin.

There are no clues if the author aims to release one day a web server or a framework coded in one line, or in assembly

QuebecTorrent, a quebec-based torrent tracker, was ordered by the Superior Court of Quebec to close the website late last week.

To my knowledge, it’s a first in Canada, where file-sharing is still legal (well, until bill C-61 comes). Here, the legal team from media groups Musicor, Zone 3 Musique, TVA Groupe Inc, succeeded into making the administrator of the website, M. Sébastien Brûlotte, admit that the website was damaging the entertainment industry. They were in fact threatning M.Brûlotte of $200.000 in damages if he won’t “play nicely”.

QuebecTorrent has had 100.000 members during its lifespan, with torrents tracking in total 2700To. Compared to other websites like isohunt or thepiratebay, it is a relatively small torrent tracker though; and it’s likely that the owner of the website didn’t want to go through all these legal hassles for a website which is obviously a side-project.

The court judgement is interesting though because it sets a precedent in Montreal and in Canada for future similar cases, for companies or individuals considering p2p.

For the anecdote, the judge ordered the M.Brûlotte to display the court’s decision on the website till july 2009. The defendant complied; but the notice is drowned by advertising, plus a liberal dose of pop-ups and pop-unders.

(just for fun)

In a poll which run here for the past 3 months, Amazon AWS and Google App Engine were choosen by MTW readers as the most promising technologies, well ahead of other new technologies with 34% of votes. I’ve grabbed then the opportunity to delve into virtualized and on-demand web hosting, aka cloud computing.

We sat down with Alistair Croll from Bitcurrent, MC of GigaOm’s cloud-focused Structure08 event and track chair for Interop’s Cloud Computing track, to find out about the cloud.

mtw: Back from last week’s cloudcamp, would you say that cloud computing is hype or it’s here to stay?

Alistair Croll: Cloud is here to stay. The economics are too good to pass up. But there’s a ton of hype surrounding it, because everyone is jumping on the cloud bandwagon. Duncan Hill said it best: Cloud computing is self-service IT outsourcing. That’s a good definition, because it recognizes that what’s new is the ability to self-provision.

I just finished a joint Bitcurrent/GigaOm report on clouds ($250 at http://briefings.gigaom.com), and I was surprised at just how much innovation there is within the field.

Amazon web services and also Google App Engine are the most known providers, with mostly similar features and pricing. But do you see a space somewhere that can by occupied by another startup, by offering for instance different service level agreements, prices, technology, or custom services? Or would are we going towards a market driven by scale, and dominated by Amazon and Google?

It’s still open territory. There are two dimensions of cloud: Operational clouds are about letting someone else run things, and here, Amazon is king (though there are lots of others — Joyent, XCalibre, Gigaspaces, Rackspace.com’s Mosso, for example. On the other hand, Development clouds are about making it easier to build things. The extreme example is Ning, which lets you build a social network.

Generally speaking, development clouds lock you in; operational clouds are “open” and portable, but you have to do a lot of the heavy lifting. Because of that, there are companies like Rightscale and Platespin and enomalism that try to make it easier to manage a cloud.

There’s also a class of companies and open source projects that let you build your own clouds. 3Tera is commonly mentioned here, but there’s also the Eucalyptus project and others.

There are still popular web hosting providers such as dreamhost who offer commoditized web hosting plans. Do they still have a place here? How unique exactly is the cloud computing technology? Would it be possible for web hosters to transform their data centers and offer competing services?

Shared hosting means you’re on the same machine. If someone else kills the machine, you die too. Admittedly, it’s a blurry world — I would say, at a minimum, that a cloud has to have virtualization (to put the customer in a “sandbox”) and self-provisioning. Generally, the virtualization runs atop a grid of commodity hardware so it can expand and shrink as needed.

Many companies (Enki is one) have taken 3Tera’s software and made a cloud. But simply offering a Plesk server and a virtual IP on a load-balancer does not make someone a cloud provider.

How wise is it for a company to host their software to a company like Google or Amazon? Isn’t there a problem if those web giants rolls out a competing product?

Some have claimed that Google Apps is a “farm team” for Google, since you can’t really run your app somewhere else if you want to.  There’s always a tension between the OS and the apps — think about Microsoft introducing disc compression (which competed with Stac) or remote desktop (which competed with Citrix Winframe.) Heck, once upon a time someone competed with MS Paint!

In general, the market will sort out which things belong to virtualization companies like Citrix Xen, VMWare, and Microsoft; which belong to clouds like Amazon, Mosso, and Joyent; and which belong to the apps on top of them.

But consider this: If you build something and run it yourself, you’ll spend a ton of time running it instead of innovating. If you let someone else run it, at least you get to keep developing new features to stay ahead of them. I’d say for many startups, not using the cloud is a huge mistake.

Are cloud computing meant for temporary computing needs? such as handling traffic spikes or temporary data analytics? or are we going towards complete transfer of server infrastructure?

“Bare metal” clouds like Amazon can spin up a server quickly. Enterprises are using them tentatively, for just the kinds of tasks you mention. Internet companies are building their entire businesses on them (think Smugmug.) Some clouds are even built on clouds (like Rails cloud Heroku, which runs on AWS.)

Eventually, cloud computing will be taken for granted. When you come right down to it, you’re not going to build a data center next to a dam, or construce a nuclear power plant. But Google or Amazon might.

There’s seldom any differentiation from running your own hardware. And many services coders normally need to build — like authentication, or storage, or a friend feed, or a mailing system, or a message queue — are already built, and connected to millions of users. So why would you build your own?

To finish, do you have any data on cloud computing adoption? or pointers to best practices for web startups?

The number of developers on Amazon is approaching 400,000. The data from Amazon Web Services already outstrips Amazon.com. This is a fundamental transformation. Companies are building new applications in weeks, and refusing VC money because they don’t have to buy servers in advance.

Some clouds (Joyent is a good example) offer “accelerators” specifically aimed for development of a type of application, such as Facebook. So if you’re a new startup, you’d better have a good reason to not use the cloud. And even if you have one, the cloud should still be factored in for things like testing, bursty capacity, and backups.

For companies wanting to be a cloud, the answer is to specialize. “Vertical” clouds like Salesforce.com’s focus on a specific industry or application. Salesforce is for CRM; Webex is for communications; and so on. I think we’ll see cloud platforms for video (Nirvanix is doing some of this), or for HIPPA compliance, or for travel, and so on. There are big advantages to tailoring your cloud to a specific industry.

Thanks to Alistair Croll for the opportunity!

I would also like to take the opportunity to invite readers to share what their web Infrastructure is. Going around events and blogs, I know this is one of the least talked about subjects, even though it’s a crucial point for a technology company. So, don’t hesitate to leave a comment on how and where you host your software, and your experience with web hosting providers, be it aws or other services. 

Thanks!

Datalicious is a relatively new company in Montreal started by Sébastien Pierre. As stated in the company’s homepage, it focuses on creating optimized interfaces for data visualization.

Sébastien Pierre is a newcomer from Belgium and France, having started there a small software R&D company named ivy.fr. He is known for the revealicious project (demo also avalaible on datalicious), which grabbed a lot of buzz a few years ago.

Now, MontrealTechWatch don’t usually cover service providers; however since MTW’s main focus is “Technology and Innovation”, I think the company fits here. Datalicious uses a wide range of open source technologies, as well as public APIs to create new user interfaces, and seems to excel on both fields. It’s an approach I haven’t yet seen in Montréal, and brings fresh ideas that I hope will initiate and inspire local great projects.

I especially like the craftsmanship & attention to detail that can be seen in the website and projects like revealicious.

PS: ok, I just re-read the post, and it seems like blatant advertising. No, I was not paid for this post :-)  But still, I like datalicious’ approach.

Control Yourself, Inc. has launched officially today Identi.ca, an open microbloggling service that takes a radical approach on how new social web services should be built and used.

Identi.ca is Evan Prodromou’s take on current SaaS and web applications practices. He writes in great detail about identi.ca’s inception in a blog post today. He writes for instance about walled gardens, mentionning how web companies such as Google or Yahoo are accustomed to suck in your data and history, and how most of them don’t allow data portability between different services.

Identi.ca in itself is a service similar to Twitter, in the sense that you use it to tell your network of “friends” about your current activity. The beauty of the system though lies with laconi.ca, a AGPL‘d software that anyone can install on their own servers. Once setup, you can exchange data friend updates with other laconi.ca servers.

If you sit back a while and think about the concept, it solves many problems. It’s for instance an immediate solution to Twitter’s scalability problems — the flow of activity streams are now handled by the users’ servers and not by one single and monolithic web company. It also allows users to customize, add, improve their own version of laconi.ca, opening new ways to use the service, for those who want more security, more flexibility or maybe an in-house closed version.

I foresee laconi.ca as the wordpress of microblogging platforms but in many ways, better. It’s also the only service I know of that has implemented OAuth with a new microblogging standards named OpenedMicroBlogging, used by laconi.ca to authenticate and publish updates to other connected servers. I haven’t myself tried out laconi.ca; as I am still trying to understand the ins and outs of OAuth. Anyways, it’s great to see such innovation, and congrats to Evan!

Codefest, a weekend hackfest for developers supported by phpquebec, was this weekend, and it begun Friday at ETS’s Pub sans génie.

There is something hardcore about spending a whole weekend with other developers with the objective to code on projects. The majority of Montréalers would want to go outdoors, and do something at the opposite end of programming. Here, attendees gladly came for the event; and many were there to code … just for the pleasure of coding.

There was for instance a Msc. student from Université de Sherbrooke, who was doing a research on Québecers involved in open source projects; and I think he went around for 30 min, unsure on how to approach the group (!!), maybe thinking that we were from another world.

I also thought about the lifestyle a great deal of time when coming there. However, I just thought that I also spent a lot of time developing new tools and architecturing new ideas and applications, be it weekdays or weekends; so it might as well be in a hackfest event.

Like the other CodeFest, attendees were free to choose any project they want to work on. There was a couple of guys who worked on drupal, Jonathan Bond (!) from GoldenEye Solutions worked on openmv; other projects were about the MediaWiki engine, TikiWiki, java and unit tests, selenium, etc. There was an actionscript geek from tribalnova, Evan Prodromou was there working on a new secret project. We also had a girl from New Zealand who came over to participate.

There was also Philippe Gamache, who talked extensively about web application security. He is very knowledgeable, and I learned a lot — by overhearing the topics he was discussing with Marc Laporte and a sysadmin from iweb, such as xss, sql injecting, css, overflow, playing with character encoding, social engineering, playing with processes, firewalls, etc. I told him it’s a good thing he was on the “good side” since he was capable of wreaking havoc in any québec website; and he replied he used to be on the other side… I also discovered he co-authored a book on security. Anyway, I’ve since then shielded and upped the level of security on servers I am working on.

Philippe Gamache currently works at mobivox, although I am not sure why a voip startup would need his skills.

In retrospective, this was a great event, with great potential. I wished there were more people coming and/or maybe organize them in regular intervals, in a pre-defined place. Yes, I am thinking about MetaLab. Quote:

The Metalab is a hacker space in Vienna, Austria — an open center for people who do creative things with technology.

Wouldn’t that be great in Montréal?

More photos from the event.

Akoha has announced this morning $1.9 million in funding from various angel investors. The list of investors include David Chamandy (Lavalife), John Bromley, JS Cournoyer, Ron Dembo (Zerofootprint.net), Jake Eberts (film producer), Alan Gershenfeld (E-Line Ventures, Games for Change), John Meeks, Reg Weiser, Jonathan Wener, Rober Montgomery (Achilles Media), Chris Emergui (BAM Strategy), and Montreal Start Up.

The funds are supporting ungoing development and launch of Akoha.

The news has been picked by many blogs, in Techcrunch, StartupNorth, Techvibes, and mashable

What’s curious is that the startup got featured on those high-profile blogs even though noone was able to tell what exactly Akoha was doing — it probably has a lot to do more with Austin Hill’s profile than the product. The funniest on the blog coverage was Fagstein’s post:

… It could be a social-networking site for fundraisers or it could be a giant multiplayer Pong game with the Sesame Street theme playing all the time. They’re still kind of being coy about it …

Of course, it’s due to the fact that Akoha is still under development. I can’t myself tell you more what Akoha is doing - other than they are playing with social gifting, both on and offline. There are couple of things to get though from the announcement:

1. if you read the announcement well, the funding was not made recently but earlier this year, with the first tranch provided late last year. The Akoha team is also going to CIX this week. This is not fortuitous and I believe has been planned meticously. Frequent media apparitions means Akoha is looking for impact, either for an upcoming launch/beta, or either for anohter funding round.
2. the list of investors is quite interesting; apart from the financial investors, there are media entrepreneurs, professionals in games, and leaders in social initiatives in the provided list, which is smart for Akoha’s founders. As Austin Hill said in a previous talk, it’s important to meet investors in the same space, as they can also be an advisor to your business.

In lieu of news and articles, I thought it would be interesting to know what were your computing environment.

The following poll accepts multiple answers

This one is for *developers*

Feel free to comment below about your choice.

The proposal for a Montréal Mobile developer meetup received an unexpected interest, and I have now nearly 20+ people interested in showing up.

I am excited to announce that we have now a venue, which is Station c, 5369 bvd St-Laurent, and a date, which is Friday 21st of March, at 5.30pm

Station C is Montréal’s new coworking space, and I arranged with Patrick Tanguay so that we can use exceptionally the place for this meetup. Thanks, Patrick!

The meetup is open to hackers, developers, user interface designers, or anyone interested in mobile applications and its usage. The format of the event will be barcamp’s — we will have an open schedule/live wiki, and anyone wishing to talk about a special topic can just write an entry. If you have been to RococoCamp, then we are heading to exactly the same format.

As always, be prepared to share, and also come with your mobile devices/laptops/tools.