Category Archives: Rant

When something needs to be said.

Infrastructure as Competitive Advantage

There’s an interesting post at GigaOM: Web 2.0, Please Meet Your Host, the Internet. It’s a good read, though could be shorter, but a few things struck me after reading it. I don’t disagree with him per se, I just think the emphasis is on the wrong thing. (Probably for effect.)

Infrastructure can be a competitive advantage today — the speed and reliability of WordPress.com has certainly put us in a favorable light with users, especially large customers — but that’s going to disappear over time. We’re very much at version 0.1 of things like Amazon’s web services and App Engine, but it’s not hard to read the writing on the wall and understand that level of abstraction is going to be the future foundation of web applications. I’m not counting on infrastructure to be a long-term competitive advantage for Automattic.

If you have a few minutes it’s worth reading On Grids, the Ambitions of Amazon and Joyent which has the real definition of a grid and Sunshine, which is worth it for the extended analogies to Greek mythology. (Both end in ads for Joyent.) Also check out Early notes on GoogleApps, Dave Winer groks where this has to go.

Second, Allan describes a case of a DDOS attack hurting a friend’s startup who had very little information about how to stop it:

Unfortunately, the poor site performance was not missed by the blogosphere. The application has suffered from a stream of bad publicity; itโ€™s also missed a major window of opportunity for user adoption, which has sloped significantly downward since the DDOS attack and shows no sign of recovering.

We can all name startups or sites that aren’t particularly known for their performance, but that flourished in spite of it. Twitter and MySpace comes to mind. If we dug a little deeper we could also find thousands of startups who were prepared for the world to show up to their door, and it never did. Building something people want is much harder than scaling it. (In most cases.) If you solve the what-people-want problem, they’ll use you no matter how bad your interface is, how slow your site is, just give them somewhere worth waiting for. I would suspect the friend here isn’t seeing their usage decline because on their Techcrunch day the site wasn’t responsive, it’s that they’re probably still in the before market fit stage.

Third, I am a huge believer in the importance of performance, but most people forget that on the web 80-95% of performance is on the front end not the page generation time. (I realize I’m saying this on a site with a 140kb header graphic. :)) Yahoo has fantastic resources on this. When a website “pops” it probably has very little to do with their underlying server infrastructure and a lot to do with the perceived performance largely driven by how it’s coded at the HTML, CSS, and Javascript level. This, incidentally, is one of the reasons Google Gears is going to change the web as we know it today – LocalServer will obsolete CDNs as we know them. (Look for this in WordPress soonish.)

Finally, for the next few years before we have true utility computing, there are some great “hardware as a service” providers like Layered Tech and Server Beach that essentially handle everything from the power to the network to hardware, and let you take over from the operating system up. This is what we use for WordPress.com, Akismet, WordPress.org, and it’s great. It’s allowed us to focus on what matters — our software and service. You still need a pro like Allan describes to handle things at the OS level (most performance problems I see are badly configured servers, not hardware limitations) but leave networking and hardware to people with economies of scale. This comment nails it.

Update: I’m in a video Rod Boothby did asking What is Cloud Computing, good timing.

Armchair Scaling Experts

random($foo): Internet Asshattery, Armchair Scaling Experts Edition. If you’re not the largest site using a given piece of software or framework and you’re having more trouble than someone who is, you’re doing it wrong.

With WordPress specifically, there are hundreds of sites I can point to that scale just fine to meaningful traffic levels with no caching, plugins, or anything. If your server is tuned for serving static files instead of dynamic requests, then a plugin to make WP output static files is a fine band-aid, but only if you don’t have the access or expertise to properly configure things in the first place. (In which case you should consider alternative hosting, help, or a hosted service like WordPress.com.) But people like to think that (1) they’re bigger or more special than anyone else or (2) that the 5-6 layers that sit under WordPress have nothing to do with its performance.

I don’t expect everyone to know about this, it’s very much a learning-by-doing thing and everyone’s situation is different. But at least operate with the assumption that if there’s someone bigger running without troubles that they (or sufficient Googling) might be able to help you out.

See also: the shockingly ignorant comments (over 200 at this writing) on this post. There are some smart people in there, but they’re drowned out by “wind0z sux!” and “that’s what you get for using (PHP|MySQL|WP|IIS|RDBMS)…”

Here’s a WordPress blog doing just fine:

WordPress is Open Source

Six Apart has recently decided that the best way to win back customers fleeing their platforms is to target WordPress, which is a new strategy they call competing. (What have they been doing the past 7 years?) A good example is this exchange between a commenter on Valleywag and Byrne Reese, the lead developer of Movable Type:

Sundown: “@anildash: what part of WordPress is not open source?”

byrnereese: “@Sunnduwn – I think that is a question better asked of Automattic. Anil, and certainly not Six Apart, has never been briefed, nor has anyone for that matter been presented with an accounting of what is open and closed source at Automattic.”

Okay, here’s some accounting:

WordPress is 100% open source, GPL.

All plugins in the official directory are GPL or compatible, 100% open source.

bbPress is 100% GPL.

WordPress MU is 100% open source, GPL, and if you wanted you could take it and build your own hosted platform like WordPress.com, like edublogs.org has with over 100,000 blogs.

There is more GPL stuff on the way, as well. ๐Ÿ™‚

Could you build Typepad or Vox with Movable Type? Probably not, especially since people with more than a few blogs or posts say it grinds to a halt, as Metblogs found before they switched to WordPress.

Automattic (and other people) can provide full support for GPL software, which is the single license everything we support is under. Movable Type has 8 different licenses and the “open source” one doesn’t allow any support. The community around WordPress is amazing and most people find it more than adequate for their support needs.

Movable Type, which is Six Apart’s only Open Source product line now that they’ve dumped Livejournal, doesn’t even have a public bug tracker, even though they announced it going OS over 9 months ago!

I had held off criticizing them after they went OS and before they decided to start an all-out confrontation because that’s not generally what OS projects do to each other.

For as long as I can remember the WordPress about page has linked and thanked Movable Type for ideas and inspiration.

Movable Type once led the market, it had over 90% marketshare in the self-hosted market. Now they call “pages” and “dynamic publishing”, features WordPress has had for 4+ years, innovation and you still can’t do basic things like click “next posts” at the bottom of home page.

For the record, I’m glad they’ve taken the license of MT in a positive direction that prevents them from betraying their customers like they did with MT3, but they have a long way to go before the project could be considered a community.

WordPress did 3 major releases last year, we’ll do 3 major releases this year. Along the way thousands of people will contribute, as well as every employee of Automattic. What we build will be greater than the sum of its parts because we’ve been a community and open source from the beginning, and always will be.

Percentage of Splogs

I’ve been indicated a few places saying a third of blogs are spam. Someone came up with this by me saying we’ve axed around 800,000 splogs on WordPress.com, and looking at our number of blogs, which is 2.5m.

As for percentage of the total blogosphere, reported by Technorati as north of 100 million, which are splogs, I’d say the number is much higher – probably 80%. This isn’t as bad as it sounds, I just think spammers are very effective at creating hundreds of thousands to millions of blogs, they tend to stick around, and I feel like Technorati’s number doesn’t doesn’t adequately scrub these out.

While I’m making data-less estimates, I’d say there are about 25-30 million non-spam blogs, and about 8-14 million of those are active in terms of getting traffic or new posts. You could cover a meaningful portion of the blogosphere by just indexing 4 or 5 million blogs.

Splogs and blogger attrition are two problems no one really talks about, but that’s okay because I don’t think either is hindering anyone’s growth as measured by metrics that matter, like pageviews or uniques. (Though many of the services supporting so many splogs must have an inordinate amount of resources devoted to them.)

See also: Blog Ping and Spam Statistics, WordPress.com February wrap-up.

Wither Dreamweaver

I’ve done my coding in Dreamweaver for 5+ years now. I think I’m the only one who does so at Automattic, but it’s a good fit for me with the network/SFTP integration, decent PHP highlight, regex search/replace, and good project support. It was a natural transition for me from Homesite. I know there are a thousand other editors that I could use, and I know I shouldn’t be on Windows most of the time, but that’s not what this post is about. I’m utterly appalled by how bad Dreamweaver CS3 is. I paid hundreds of dollars to upgrade to something that consistently crashes when I edit certain parts of PHP files and CTRL + F no longer opens a search box unless I have a document open.

Dear Dreamweaver team, I’ve been putting up with these bugs for close to a year now. I will come down to San Jose and show you the bugs personally. Just please do something, or feign the appearance of movement. For now, I’m switching from CS3 to version 8, which is just sad.

Socialtext Spam

(Rant warning.) What’s it called again when you can’t unsubscribe from a service that sends you email almost every day? This is broken:

  1. I get my daily email about changes in the LoicLeMeur Wiki that I signed up for sometime at Les Blogs.
  2. “To stop receiving or change how often you receive emails about changes
    to LoicLeMeur Wiki, go to: http://www.socialtext.net/loicwiki/emailprefs “
  3. Redirects me to this URL.
  4. It appears that I am logged in. I change my preferences to “Never”
  5. Of course when I go to my settings it shows me as “user78247”, so I didn’t actually cancel my subscription, it just showed a stupid subscription page to me even though it obviously doesn’t have my email address right now and I’m not logged in.
  6. Okay, click the “log in” link.
  7. Type in email and password.
  8. I am now mysteriously redirected to the FOAFnet wiki, which hasn’t been touched in like a year. NO IDEA WHY.
  9. Wait, the top still has a login button, but also a settings button?
  10. Settings takes me to the user page for “user19254”, even though I JUST LOGGED IN.
  11. Okay, click login again.
  12. Now I’m redirected back to the FOAFnet wiki again but this time it shows me as being logged in.
  13. Where the heck do I go now???
  14. I go back to the Loic wiki via the address bar.
  15. It says login in the top right corner again.
  16. I enter my information, and am redirected by to the FOAFnet wiki. I can sense the wiki mocking me.
  17. Okay, maybe if I go back to my settings page where it worked.
  18. It forces me to login to the FOAFnet wiki AGAIN even though I did just SECONDS AGO.
  19. I now am logged in, I click on settings and go to “My Workspaces.”
  20. It only shows the FOAFnet wiki, even though I know I’m a member of the Loic one because I GET EMAILS FROM IT EVERY DAY.
  21. Also now for some reason every Socialtext wiki I try to visit, like this one from Web 2.0 conference, redirects me back to the FOAFnet wiki. (cue blood-curdling scream)

Any suggestions? I would just dev/null them in my procmail, but I prefer not to do that to legitimate companies. At what points could the user experience be improved?

(And yes, I have reported this problem to them. I demonstrated it in person in August 2005 at BarCamp. It’s January.)

Update: Socialtext responds in the comments. “[T]his morning we disabled email notifications for all the public wikis on our site, due to the confusion people had when trying to turn them off.” A thoughtful, effective, and quick response.

Dave Matthews Stand Up

“Crash into me…” If you buy the new Dave Matthews Band Stand Up CD, do not put it in your Windows computer. I did and it popped up an annoying software installation and I said no to it just like I did 9 years ago when I first saw that happen with Fiona Apple’s Tidal. It then ejected the CD. First strike! I just wanted to listen to it. So I did the logical thing and pushed the CD drive back in, and the CD spun up and then the computer crashed and rebooted. Needless to say, this is a very bad thing.

Frustrating Amazon

Some days Amazon just drives me batty. Part of it may be the fact that I have 3 distinct accounts on Amazon, all with the exact same login, but which one I get depends on on what password I use when I log in. One of these accounts gives me the discount for using A9, the rest don’t. Depending on which one I logged into last, my A9 search history is hosed. The wishlists are in various states of upkeep, and all different. It’s confusing and frustrating and really turns me off from Amazon. I wish I could just delete all those accounts and start fresh. (If they let me export my wishlist first.)

Braindead Finder Behaviour

Because of what I consider totally braindead behivour in the OS X Finder I appear to have lost about 60 pictures from my trip. When I offload pictures from card I generally drag the 100PENTX folder from the card onto my desktop and I leave things in that folder until I have a chance to compress the pictures, divide them into days, and upload them. Well OS X does this crazy thing where when you drag the folder onto the desktop it asks you if you want to replace the folder with the same name. On Windows I always say yes and it just adds the new pictures to those already in the folder. In OS X it apparently means delete the folder that’s already there with no way to recover it and replace it with the one you’re dragging. This happened to me a few times and I couldn’t figure out what was happening, luckily though I had backups on my iPod. Unfortunately going back over the pictures from the trip it seems a day is missing. Fortunately it was a day of mostly travel so I’ll live, but still a bummer. Updates: John Gruber weighs in, and here’s the exact message in Windows.

New update: Robert Scoble put me in touch with Bob Day who had this to say:

If the question is just “Why do merge by default?”, there are lots of
answers.

1. Because it maps well to operations that users are likely trying to
accomplish (see the scenario of dropping a picture folder from a
camera).
2. Folder replace can be done by deleting the destination folder first,
and then copying. If you have replace be the primary method, then merge
becomes a very tedious process.
3. Because it is less destructive?

Please realize that having a camera that uniquely assigns picture
numbers until you reset them becomes very important with this merge
behavior. If your pictures are all uniquely named, the default of
replacing files with the same name will allow you to not lose any files.

Also realize that this is a complicated scenario for most users. Almost
any choice is going to be bad for some users.

And yes, the behavior is a concious choice. We had to implement this
feature in Windows 95.

I followed up that “So before that [Windows 95] folders were deleted and
overwritten?” Bob responded: “I need the source code to Windows 3.1 to confirm. Anybody remember “File
Manager”? wow, that is old.” And dug up:

Ok, archeological discovery over. (wipes the dust off his sleeves)

Win3.1 would say the same thing for folders as it did for files:
“Replace file with file

And if you said “yes” for a folder, it would try to delete the
folder first, which would error out if the directory
wasn’t empty. Not sure what the error message is there.

I would love to get similar background for the Mac OS X behaviour.

Out for the Count

Well, it’s happened again. I should have seen it coming, the signs were all there, but I hoped “Not this time, it’s different.” However here I sit with what is most likely strep throat, judging from my track record and the terrible pain every time I swallow. This was not entirely unexpected, as every year for the past five or so I have contracted some alphabet letter of strep combined with something else, the worst being last year when I actually ended up in the hospital, on Valentine’s day. This year has been my healthiest ever though, and I’ve hardly gotten a cold or a sniffle for almost a year, so I was hoping this wouldn’t happen. That said, the timing could have been worse; I’m glad it wasn’t at SxSW (like poor Jane or Ernie); I’m glad it wasn’t on Valentine’s day; I’m glad it wasn’t during an “important” school week; I’m glad that you’re still reading at this point. I’m going to try and get some more rest. Updates will either be light, or come at a Kathyesque rate—we’ll see.

Thoughts from An Old Hand

I was sitting next to Derek when he wrote this. Some interesting thoughts, and I’m as guilty as any of being weblog-centric in my thinking. I think WiFi, which I’m involved in through the Houston Wireless group, and weblogs, which you are experiencing now. I see a tremendous potential for enfranchisement, which I’m not even sure if that’s a word but it’s the best suited I can think of. My confession: I used to hate the word ‘weblog’, ‘blog’, and every variant thereof. At one point I spent hours scouring this site for every mention of it and removing on sight. I’ve gotten over myself.

Scum

There has been a lot of talk about referrer spamming in blogspace, and while browsing a PHP network query tool I stumbled across this, which looks terrible. What’s even worse is there is some genuinely neat looks scripts on their site, some things that I’ve been meaning to write myself, and it looks like I’m going to have to because I couldn’t support any that sells a product like ReferBomb.

This script will get up to 1000 Google search results for the search term of your choice (we’ve provided a few suggested default search terms), then automatically fetch all of the resulting web pages. What does this do for you? Think “public web stats” or think “blogs who reciprocate referring links!”