Archive for the ‘Internet’ Category

Facebook PHP compiler

Tuesday, February 15th, 2011

More and more php compilers are emerging, and it now finally looks like there are a number of decent ones out there.

Just a couple of years ago, there only were bcompiler, eaccelerator and perhaps some others, but none really good. Eaccelerator would be sensitive to server environment and only work under certain conditions, and even have special problems with different php programs, and others had similair problems.

But now it seems like there are 5-10 different compilers, and all seem to work really nice. An especially interesting one is HipHop for PHP – a php compiler written by facebook to run in their own servers. The ultimate goal of all compilers is the SPEED (and, hwell, perhaps, obsfucation too), and I can only imagine what kind of speed problems the facebook experiences, with all their traffic, so if hiphop is good for them, then it should be too for the rest of the normal mortal aplications out there.

One important thing to know though is that hiphop for PHP is not an opcode cacher or a JIT. It is actually a compiler, and produces binary code as such. Probably because of that it has problems with eval() and some other dynamic features of PHP. If that is the case with your application, then it’s probably not for you, but how often does one really use eval?

Definetely check it out if you are looking to speed up your php application without changing a line of code. http://developers.facebook.com/blog/post/358/

Github: https://github.com/facebook/hiphop-php/wiki/



Anonymity in 2010

Tuesday, November 30th, 2010

A great podcast/presentation that has started to circulate around the blogosphere, about how the internet and the progress is affecting our anonymity.

Steve Rambam (who is/was a private investigator, and ironically looks very much like one, thou also like an italian mafia boss) takes on many interesting things in this 3 hour long talk, including
– What information we are putting in “The Database” ourselves and how the way people are disclosing information about themselves changed over the past 10 years.
– Just how much information does a mobile phone broadcast about its owner.
– How easy it is nowadays for a detective to find out anything about you without leaving his or her desk.
– What kind of databases are being put together, sold and used, and what kind of information can be found there.
– A couple of interesting sites for information hunting.

Yahoo Pipes

Monday, August 30th, 2010

*** The pipes didn’t come out yesterday, so those of you who have already heard about them, can simply skip this post.

But those who didn’t, are about to find out something wonderful!

The pipes are a web service from Yahoo and can be accessed by going to http://pipes.yahoo.com/pipes/

As an example of what it can do, let’s take today’s featured pipe: eBay price watch. Apparently, it gives you information about certain items from ebay, found by certain keywords that lie in a certain price range.

Without any tools at all, finding this information would require going to eBay, searching for items, sorting, finding the ones you want, and repeating that over and over again. If you are decent programmer, you could easily make a PHP script or something that does it for you: goes to eBay, fetches a page using a search request, parses it and so on. It is doable, but is it worth the trouble?

Well with yahoo pipes a script like that takes no more than 10 minutes to create! It is all because instead of writing code, you are combining structural blocks that process information in certain ways, which can include search/replace, regexps, combining functions, well basically any kind of aggregation you could possibly imagine.

Yahoo pipes are based on RSS, so the most used output method from the pipe is an RSS channel, which, if we are talking about the eBay example, could be imported into your favorite RSS reader.

Basically any data from any webpage, RSS channel or any url can be inserted into a yahoo pipe and converted to the format you want. Several sources of data can be used, records can be combined, removed, parsed, and so on, with limitless possibilities. And, what is more important, all the above is done simply, fast and visually, which, besides the development speed, also gives you a lot of room for experimentation and creativity.

(And no, I don’t work for Yahoo, it’s just that this service is really a great contribution for human kind :P)

And I haven’t even told you the best part: Yahoo pipes support information io in JSON and if I remember correctly, even PHP-serialized, which, combined with its possibility to fetch anything that has a url, effectively means that you can create your own modules for pipes that can do anything that pipes cannot do already, making the service omnipotent.

If nothing else, it is a great way to test new ideas without having to write a single line of code.