Chuqui 3.0

Web companies organize massive effort to patch fundamental Internet flaw

This is seriously bad. What I'm curious about, and we'll only find out after the fact, I guess, is what the impact of unpatched systems and servers is. Given I spent some time yesterday tracking down a web statistic that was bogus that turned out to be a Windows 2000 box running IE 5 in china that downloaded a software package 10,000 times in 36 hours yesterday, I think we all understand just how likely it is that everything that needs to be patched will be...

Web companies organize massive effort to patch fundamental Internet flaw » VentureBeat:

The entire underpinnings of the Internet are vulnerable to a major bug in how Internet addresses are managed, security researchers announced today. The problem is so big that dozens of companies and government organizations have secretly synchronized an effort to fix the problem. The companies — from Microsoft to Cisco to AT&T — are all releasing patches today or in the next few days to eliminate the major vulnerability, which was discovered early this year by security researcher Dan Kaminsky, who kept it a secret until a conference call with the press today.

Twitter is not a chat room... or does it just feel that way some times

Michael Gartenberg - Twitter is not a chat room... or does it just feel that way some times:

"Twitter is not a chatroom." And yet, it sure feels like that at times. I've been twittering now for a few weeks and being part of that conversation has been a good experience and i plan to continue.

That's because, Dave Winer's opinion notwithstanding, Twitter is a chat room. What makes Twitter an interesting hack is that it's an infinitely large chatroom, which could easily become unlivable -- except you choose which subset of users you want to listen to. Which means it's an infinitely complex weaving of conversations.

to me, it's like a huge cocktail party, with a large room of people. And while you gravitate towards those groups with people you already know, just like in a non-virtual party, it's those half-heard snippets of interesting chatter that draw you into new groups and get you to meet new people.

That's the hack that makes Twitter interesting beyond just another chat room; discovery is built into the system, not something that depends on "hey, we meet in this room at 10PM on wednesdays" word of mouth...

The trick to keeping Twitter usable is twofold: one, don't grow your following list too large or make it too tangental (and don't be afraid to unfollow someone, including me. maybe especially me), and get over the mentality that you need to closely follow everything. The conversation is 24x7, it's value is allowing you to wander in and wander out as you want, just like a real party. Try to follow everything all the time, and you'll drown in the firehose.

Yahoo Layoffs For Real—But What’s the Real Number?

Yahoo Layoffs For Real—But What’s the Real Number?:

Nobody likes layoffs. But if you are going to bother, you might as well make it financially meaningful in Wall Street’s eyes. Doing some quick back-of-the-envelope calculations, let’s assume for arguments sake that each terminated employee costs an average of $100,000 in salary and benefits. Then cutting 500 people would only save Yahoo $50 million, whereas cutting 1,500 would bring $150 million to its bottom line.

Not necessarily. the first goal should be to solve the problems within the company, which means identifying the people who are making Yahoo better and separating them out from those who aren't. And keep the ones who are....

Setting layoffs by some generic number is a recipe for exacerbating the problem, not solving it. We had way too many of those "10% layoffs" at Apple in the good old days, and if you don't do it intelligently, all you do is lengthen the death spiral.

Are you top-heavy in management? Are there business units not holding their own? Groups where expenses are well out of line with revenues? Find the fat, find the rot. Fix those. Make sure you keep enough (and the right!) people to actually run the business well. Invest in improving the business, don't just cut expenses.

And if current management isn't capable of doing that, the board better find new, better management fast....

Coding Dilettantes

Burningbird » Coding Dilettantes:

Sivers discovered that moving to Rails because the community supporting the language was passionate, wasn't the wisest move to make.

heck, if all you need is passion, the Ruby and Rails people have nothing on the group that supported and evangelized LISP -- 25 years ago.

And now that I think about it, I find the parallels between the two languages interesting. both seem to have solid and passionate followers, especially among some aspects of the geek elite; both seem to solve some classes of problems well, and both view themselves as the future of computing, while the mainstream looks on and adapts the interesting parts out of them without adapting the language itself... And in both cases, there's a lot of smoke (interest and passion), but limited fire (jobs).

7 reasons I switched back to PHP after 2 years on Rails - O'Reilly Ruby

7 reasons I switched back to PHP after 2 years on Rails - O'Reilly Ruby:

The answer is no.

I threw away 2 years of Rails code, and opened a new empty Subversion respository.

Then in a mere TWO MONTHS, by myself, not even telling anyone I was doing this, using nothing but vi, and no frameworks, I rewrote CD Baby from scratch in PHP. Done! Launched! And it works amazingly well.

It’s the most beautiful PHP I’ve ever written,

I'm glad someone else feels this way. I've looked at Rails a few times, and I certainly don't see any reason to criticize rails -- but I simply couldn't see why people were so hot on it. It left me a bit cold, honestly. I could never generate much enthusiasm for it.

I look around the net, and when you look at large, scaled sites based on Rails, you find, well, Twitter. And it's had its scaling challenges. After twitter? it's hard to think of a site based on rails that's "grown up" and thrived.

And since I'm watching the job market, I've been watching the job market for Rails. lots of "hot, stealthy startups (come be employee 3)", not a lot of Rails seeming to have actually shipped and gone into production.

I keep getting the feeling that Rails is a great prototyping tool, something to get you to market fast something the geeks really like to play with; but I also get the feeling that for sites that need to grow up, they'll end up re-coding in something else. Basically a fast prototyper for future J2EE work.

And, honestly, you can do all that in PHP, and in most cases avoid the re-coding.

So, having spent some time getting a first serious look at Java, I'm going back to PHP and Perl. And frankly, lots more jobs there than in RAILS land. It may not be sexy or new, but it works....

I do like the layer separation of MVC, and I'm going ot have to experiment with that in my next project, but that's not RAILS, that's something you can do pretty much anywhere...

The Last Word (for Now) on Our RSS Feed

The Last Word (for Now) on Our RSS Feed: An Excruciatingly Long and Boring Post That Will Please Exactly No One - Freakonomics - Opinion - New York Times Blog:

1. Over the past two years, we have built up a substantial feed readership, and from the sound of these readers’ comments, many of them will be unsubscribing. To all of you, we say: sorry we’ve disappointed you; we’re sad you’re leaving; thanks for reading in the past; and I hope you’ll come back if we ever resume full feed. In the meantime, we hope you’ll visit the site, but as some of you have made brutally clear, that option doesn’t work for you.

This is cool -- a high visibilty site with large readership who's finally going to TRY this and see what happens. I do hope (and I encourage both the Freaks and the Times to use this as an experience gathering opportunity!) that they'll track metrics and then publish the results back out to the net so we can see what happened.

What percentage of subscribers actually leave? Do they stay away, or do they get replaced? Where those users good or bad users for the advertisers?

There's been a lot of noise over this issue for the last couple of years; now we have a chance to get some real data. Here's hoping those who CAN make that happen do.

What'll happen? God knows -- that's why we need numbers. But if experiences in other parts of the net can be used as a basis for guessing, I'd say the following is likely:

1) the % of people leaving will be small.

2) A good percentage of those threatening to leave won't. Or will leave -- and come back quietly soon thereafter, perhaps from a different location.

3) Many of those leaving are probably bad matches for the advertisers anyway, and/or are hostile to online advertising in general; I wonder if it's possible to correlate the unsubscribers in some way back to people who also use ad blockers or some sort?

4) these people are what I call the "squeaky wheels", and not representative of the user base as a whole.

A great example from my past: the old "reply-to" fight on mailing lists. Trust me, you get a mailing list larger than about 50 people, you'll likely fight this one every year or so, because there's a small but noisy group that believes that there's only one possible right answer, and it's that reply-to should be set to the list (for reasons why many of us who've run email systems and lists over the years don't like it, see reply-to considered harmful. Or don't, it's increasingly irrelevant these days...)

I used to do a nasty thing to these folks -- I'd ask the entire list, instead of just listening to the ones complaining.  And the funny thing is, having surveyed mailing lists probably a dozen times over the years, the same basic facts kept cropping up: 80% of the users don't care, and will take the default. Of those that do care, reply-to vs "no reply-to" basically split in half.  Certainly no mandate here.

The other interesting fact of these fights was that these folks weren't interested in doing what the majority wanted, or what was right for the community -- they wanted what they wanted, and it wasn't open to discussion. They knew better than everyone else, including the administrators and experts. And the sound of threats of leaving and etc in this situation sounds almost exactly the same as what happened here with reply-to.

The reality from my experience? The people who leave won't be missed. I rarely lost a "significant" contributor over this fight -- and I expect the same here.

But we won't know until we study this, and now we have the chance to create some data instead of people arguing with opinions. And isn't that really  what Freakanomics is about?

Here's one quick and easy way to generate a data point -- count up how many people sent complaints, and compare to the size of the feed. How big is it? 25%? 10%? 2% 1/2 of a %?

Chances are, the pool of people threatening to leave is a tiny percentage of the subscribers. And even if you assume each complainer represents ten who will leave silently (and my experience is that the number's MUCH smaller than that; happy people watch quietly, folks like this pushing an agenda tend to do so very forcefully, so the ratio is more like 2:1 or 3:1, most likely), I'd be amazed if they can push the subscriber numbers by 2-3%; frankly, moving the blog to the Times with the visilbilty that'll give it will probably wash away any impact of the "squeaky wheel" crowd's leaving. If they actually leave.

Isn't it funny how some people consider the format more important than what's actually said? That a well-written blog will be ignored if it's not following the right dogma on feed formats? Fortunately -- people that enamoured with the politics of the blogging technologies are rare. Noisy, but rare.

I think this is all a tempest in a teapot. but I sure hope the Times tracks the numbers and reports them so we have some real data to examine the next time this fight happens....

There is no Web Operating System (or WebOS) (by Jeremy Zawodny)

There is no Web Operating System (or WebOS) (by Jeremy Zawodny):

But when you stop and look at how technology has changed and opened in response to networks, the Internet, simple/open protocols, view source, and self-serve business systems, there's a clear pattern emerging.  Strong forces are at work here--strong enough that you're better taking advantage of them rather than fighting them.  You will lose.  I don't care if you're Microsoft, Google, or even Facebook (the latest golden child).

Open beats closed.  Simple beats complex.  Freedom of choice beats being told what technology to use.

The idea of starting now (or recently) to build up a full blown "stack" of services for the next generation of web-based applications (the kind that run on the Internet) and deciding to market it that way seems rather insane to me.  You're using strategies and tactics from the previous war.  It's last year's thinking (where "last year" is about 2001).

But there are also challenges and risks here.

Think about this -- your facebook profile loads only as slowly as the slowest widget you add to it (ditto your blog with all those widgets you've added to them, etc ad nauseum); worse,  how do you debug something like that? Answer: you probably don't, you blame facebook -- even though, at worst, they're proxying in data from some other service.

AS we continue to distribute this stuff, these problems will simply get worse -- and more common -- and that may be the biggest reliability and performance problem facing sites like Facebook as they scale.

Worst case? Imagine being on duty in the NOC at Skype this last week when suddenly your traffic volume drops and your network shuts down. Alarms go off, people start screaming -- and nobody can find a problem, because, since it's a P2P system and so heavily reliant on all of those P2P systems sharing resources -- the problem isn't in any of their machines they operate and control.

How do you build reliable operations around environments where you don't control the reliability of the systems you're relying on? Interesting challenge, and it's just starting. Assuming you believe Skype's answer for their outage, that's pretty much a worst case scenario for this kind of setup -- but it's one all systems that front a P2P or distributed environment are going to face...

Why Full Text Feeds Actually Increase Page Views

Previous discussion is here...)

Techdirt: Why Full Text Feeds Actually Increase Page Views (The Freakonomics Explanation):

However, in our experience, full text feeds actually does lead to more page views, though understanding why is a little more involved.  Full text feeds makes the reading process much easier.  It means it's that much more likely that someone reads the full piece and actually understands what's being said -- which makes it much, much, much more likely that they'll then forward it on to someone else, or blog about it themselves, or post it to Digg or Reddit or Slashdot or Fark or any other such thing -- and that generates more traffic and interest and page views from new readers, who we hope subscribe to the RSS feed and become regular readers as well.  The whole idea is that by making it easier and easier for anyone to read and fully grasp our content, the more likely they are to spread it via word of mouth, and that tends to lead to much greater adoption than by limiting what we give to our readers and begging them to come to our site if they want to read more than a sentence or two.

Interesting piece following up on a discussion I've been involved in. I can see the logic, too. Consier me, if not fully convinced, heavily swayed.

Followup 2: Truncated RSS Feeds Kill Conversations

Chuqui 3.0.1 Beta: followup: Truncated RSS Feeds Kill Conversations and Long Term Traffic: Technology Evangelist:
The answer seems to be there's no really good answer, but it seems clear people (at least those that care enough to discuss the issue) would rather see ads in the feeds than truncated feeds.

well, the Freakinomics folks moved to truncated feeds with their move to the times, so we now have a living experiment, if we get real data out of it:

Online Spin » Blog Archive » Freakonomics Sparks Debate Over Partial RSS Feeds:

Not surprisingly, it was Dubner’s recent post announcing the move to the NYTimes that fostered especially vibrant discussion, both congratulatory and protesting. On one hand, many cheered Freakonomics’ upgrade into the esteemed NYTimes stable. However, roughly six hours after Dubner published his announcement, roughly 100 of the 120 comments were in protest of Freaknomics’ move from full to partial RSS subscription feeds, forcing feed subscribers to begin their flow and then abruptly transfer attention to the NYTimes site via a separate browser window!

A sampling of the comments in Dubner’s post underscores the dilemma that online publishers face as they adapt to a more open Web, heightened competition, attention scarcity, savvier readers, rising expectations and RSS syndication:

What I want to know is not how many people complain -- but how many people drop the feed. This is complicated by the fact that they swapped sites as part of the transition and that'll muddle the waters.

But what matters here, and hopefully, after some time the numbers will be made available, is who votes with their wallet. And, frankly, if the people voting with their wallet weren't generating revenue anyway, or some other kind of positive contribution. It might actually give us some data to go with all of the opinions. Scary thought.

MySQL Community split officially a failure

jcole’s weblog: Jeremy Cole’s take on life. » Blog Archive » MySQL Community split officially a failure:

The “MySQL Community” concept has failed

Jeremy has a good take on the current situation within MySQL land. I know a bunch of folks who've moved to Postgresql in the last 18 months for various reasons, and much as I like MySQL, I've wondered, too, what makes sense moving forward.

Jeremy's nailed the essential conflict between the commercial MySQL part of the world and the open community aspect; that was a big aspect of what attracted people to MySQL initially, and MySQL AB seems to be screwing it up...