status.wordpress.net

Today, we are mostly getting request timeouts and ‘server maintenance’ pages, and half the time the forums aren’t accessible either. Joy. I’m starting to understand why we don’t have an export feature; it’s not to make Six Apart look good, it’s because if they hadn’t locked us in there’d be nobody left.

I made two related points on the .com forums yesterday: one, that we need a status page like what everyone else has. Except that there’s no point because no-one would ever bother to update it. (Have I posted this before? Wouldn’t surprise me if I had, given its obviousness.)

And two: when the traffic got too much and the servers started going wonky and it became evident we needed new ones, why not shut down sign-ups temporarily? Or reinstate the invite system? Prospective users might get annoyed; but they’re also likely to be reassured that you’re looking after the users you already have and ensuring that their service isn’t compromised. Right now, my service feels pretty compromised. And while it probably wouldn’t make a vast amount of difference whether new accounts were being made or not, it would look as if the developers were doing something about the situation now, rather than just saying ‘yeah, we’re getting new hardware, w0ot!’

Except oh, I forgot, you can’t close signups because then your standalone wordpress users wouldn’t be able to get hold of their Akismet key, and there is no mechanism for giving them an Akismet key without the useless spacewasting usernamehogging free blog to go with it. Consequently there is no way of ensuring the service grows at a sensible, sustainable rate. You can’t shut the doors at the entrance. And you won’t open them at the exit.

[slaps ‘idiocy’ tag on entry]

11 Comments »

  1. Hexxenn said

    And they expect to offer paid hosting in the near future? ROFL.

  2. Matt said

    The server issues had nothing to do with signup or traffic growth, they would have happened if we had 50 users. We’re in the middle of a transition period, doubling the number of servers we have and splitting the app between multiple datacenters. This’ll give better uptime and availability than 99.9% of web applications, but getting there might have some bumps.

  3. Pissed off Joe said

    Matt, why the hell do I need to sign up for a fucking wordpress.com account just so I can use Akismet on my WP blog eh? Is there any legitimate reason apart from “we want to have a nice big number of users to tell people about”?

  4. Alan said

    Pissed off Joe: Probably so they can track how many Akismet keys are distributed… Though why they couldn’t track that with a generic key generator system, I don’t know.

    I think a status page would placate a lot of people. From my experience at Last.fm, which has a fairly detailed status page, people like to know what’s happening in the system. In Last.fm’s case, if there’s a queue on the user’s cluster, they can find out straight away and not have to ask questions in the support forum as to why their tracks aren’t being updated or whatever.

    It sure beats being snapped at by forum goers who here the same question x amount of times.

  5. wank said

    The server issues had nothing to do with signup or traffic growth, they would have happened if we had 50 users

    We can all understand a server falling over if it’s overloaded; and if it was just a case of dodgy hardware, that’s the kind of bad luck which could happen to anyone. If, however, something went horribly wrong at the planning stage or with the software, the inability to cope with more than 49 users is a little bit worrying.

    getting there might have some bumps.

    There’s no ‘might’ about it. Shifting data around from city to city and server to server involves downtime and general glitchiness. Most users are reasonable people who will be patient and understanding about this, as long as they’re kept informed and the information they’re getting is accurate. (For example, please don’t tell them to go make a cup of coffee and come back in a few minutes unless you can 100% guarantee that their blog will be back by then. Oh, and comedy messages from the server are likely to grate on people from the second viewing on.)

    Is there any legitimate reason apart from “we want to have a nice big number of users to tell people about”?

    This is a perfectly legitimate reason, especially if you want to impress potential investors. Who doesn’t like having money thrown at them?

  6. Matt said

    People who sign up purely for an Akismet key are, assuming none of them use their WP.com account at all, no more than 10% of our numbers. Not worth padding for, and any investor doing due diligence would look at activity, not the raw signup numbers. Angry Joe, if getting a WP.com account bothers you that much, there’s no need to use Akismet. There are plenty of other plugins out there to protect you from spam.

    As we’ve said before, we plan to allow people to get accounts without creating a blog, hopefully pretty soon, but the way things are right now it would be tricky to decouple that process (which is why it wasn’t done in the first place). Empty blogs just use resources, and give us no benefit, so there’s no reason for us to leave them. I’ve looked into deleting old blogs with no posts, but that would at most give us 8-12 days before the number of new blogs created has brought us back to where we started, so it’s not worth it given the risk involved.

  7. wank said

    10% of 161 000 is still over 16 thousand, no? That’s quite a lot of usernames forever made unavailable. Nor of course would you have figures for the people who registered just to get Akismet, but made one or two posts just to check out what wordpress.com was like. And then of course some of them stayed and blogged. Denying that the API keys were put here to increase user numbers would be disingenuous to say the least. They hooked people in who had no other reason to sign up. If that wasn’t the intention, you’d have let people sign up at the akismet site.

    Pissed off Joe does have a point, in that bundling Akismet with the main wordpress download does encourage people to think ‘well, this must be the best anti-spam plugin because it has the official seal of approval’ and railroads them into signing up here. Unless they’re upgrading from 1.5 or they hang round the forums, they may not even be aware that alternatives exist.

  8. Matt said

    16 thousand = two weeks of signups. Not a big deal. A good username is the reward people get for getting in early, nothing else. There are still hundreds of thousands of good ones left. Normal good usernames usually are available until around 2 million or so.🙂

    To clarify, 10% is assuming none of them blog or otherwise use their WP.com account. (Lot’s of people use it for commenting but not posting, and I imagine that’ll become more common once we have more privacy features.)

    Having people register on Akismet directly may happen at some point, but our team is so small (especially Akismet, which only I work on) that often things which are “nice to have” are delayed for things which work today. Maintaining a separate user system, anti-junk-registration system, billing system, account abuse system, etc just for Akismet would be a waste of time right now.

  9. Pissed off Joe said

    Yo wank, what happened? Nothing new posted in several days.

  10. Status said

    RE:Hexxenndoes wordpress offer paid hosting??:))))

  11. […] of the situation by email, since I neither got one nor would expect one. (I’d settle for a status page, as opposed to a sticky post on a forum which is DOWN :roll:) The likeliest scenario is that they […]

RSS feed for comments on this post · TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s