Archive Page 2
Twitter Blog: We Learned A Lot
OK, this is actually pretty sweet: Twitter Blog: We Learned A Lot.
- A response (a second response, in fact)
- More details of the explanation (scalability, confusion)
- A temporary compromise (more details below)
- A statement of intent to provide a more permanent solution
- And some words to make it clear they understand
Really, how many calls would you have to make to, say, your cable company, to get that kind of satisfaction? I dunno about you, but I have no idea how many that would be, because, let’s face it, I never ever got anything even vaguely close to that good a response from my (former) cable company. Hence, the dish on my roof. And, of course, I haven’t paid a red cent to Twitter. But I digress.
If the Twitter blog entry above doesn’t make clear to you what was, what was done, what’s been done to it, and what’s coming, then there are several other explanations on the web; this one’s nice: (Marshall Kirkpatric / ReadWriteWeb).
Here’s my chop at it.
What Twitter have done is expand the taxonomy of tweets.
- Originally, there were just tweets: anything you said could and would be heard by your followers.
- Then, there were mentions: twitterers added “@username”, just as a convention to catch the eye of “username”. The tweets were still seen by the same folks, the sender’s list of followers, but “username” would perk up and take notice
- Then, there were replies: Twitter volunteered to forward mentions to “username” even if s/he wasn’t actually following the speaker. At some point or other, this was applied only to mentions where the mention was at the beginning of the tweet. “@username right on” would get special treatment; “right on, @username” would not. Then, fairly recently, this was changed so that the mention anywhere in the tweet would get reply treatment, and even several usernames “Let’s meet at @fred ’s house, @joe” would go to both, even if they didn’t follow.
- Somewhere along the way, and this seems to be the development Twitter didn’t actually notice, people started noticing their friends talking to folks they didn’t know, and using that to expand their own community. A lot of us think this is the fundamental juju of Twitter.
- The new thing, as of some time today, is this: while above, “mentions” and “replies” were the same thing (identical messages), now Twitter has a formal notion of a difference: a “mention” is made by typing the “@username” thing; a “reply” is made by clicking a button labelled “reply”. In these terms, “mentions” get wide dissemination, as always, but “replies” only go to the addressee.
This is a bit odd, as everyone acknowledges. Look at it this way: if you seem a message appear in your Twitter timeline, “@you did you see this?”, there may be no way for you to guess whether anyone else has seen it. Also odd, the explanation is something about which button you pushed, but what’s that mean? Are they changing the code in their web UI to support this? And if so, what’s the impact on the 3rd-party clients?
I’m only guessing, but I’m guessing that they actually are using something that’s already common. You may not have noticed it, because not all clients show it, but the system already supports a notion of linking tweets and replies. One way to see it, no matter what client you start with, is to go to the “status” page for the individual tweet. This (and many-but-not-all other presentations) sometimes says “in reply to <username>”. I suspect this is what Twitter refers to when they talk about messages sent by clicking a Reply button. Which would be good, in that all the clients I know of already support that (they may not show it, but the provide it).
So, I suspect: any message with “in reply to” info will only be shown to the reply-ee; other kinds of mentions will be shown to all followers of the sender (there’s still no personal option available).
Long term (again, see the blog post linked at the top of this one), Twitter promises to “ive folks far more control” via “a per-user setting,” so expect some check boxes or something, soon.
Filed under: Toys | 4 Comments
Twitter replies
Well, at least Twitter hears: Twitter Blog: Whoa, Feedback!. As you can see over there, they’ve heard loud and clear the message coming from everywhere (including my previous posting, though I’m only a snowflake in the avalanche).
What’s very weird, though, is that Twitter still has not offered any explanation as to why they made the change. It appears to me that they really hadn’t realized the value of the feature they removed (discovery of “interesting people”). That seems odd, but then again they’ve always seemed to have a very guarded attitude toward the feature. Even on the configuration page, I seem to recall it was described very negatively. (I’d go get an exact quote, but it’s gone now!)
What they’ve said in explanation seems to be confined to that one blog post, and what it says is (1) “technical reasons” and (2) “scalable.” The first contains no details. The second seems, if anything, counter-intuitive: the behavior that used to be default, and is now mandatory, seems to me far less “scalable” than the behavior they now do not provide (for which I will now coin the name “my way”
My way is: “show me anything said by this person.” Very simple.
The other way is: “find everything said by this person. Consider whether it’s a reply. If it’s a reply, check whether I’m also following the addressee. If so, show me.” That’s two dips into my follow list, or in database terms two joins to the same table. It happens my company runs a highly scalable web service, and those words “two joins to the same table” scare me to death!
Maybe they’re concerned about the number of messages? I have no doubt at all that this change reduced the number of messages they send out. To judge from my personal stream, I’d guess traffic was cut roughly in half. We all know the scalability challenges Twitter has had in the past. It’s been pretty stable, lately–remarkably so, for those of us who twittered through the invention of the Fail Whale–and I’m sure we all want it to stay stable. But 2x is really peanuts, in scalability land. Their traffic is probably growing 2x per week, so what, they’ve delayed apocalypse by a week?
Filed under: Toys | Leave a Comment
Twitter loses it
The folks at Twitter have lost their minds: Twitter Blog: Small Settings Update. #twitterfail #fixreplies
I suspect I’m speaking to three kinds of people:
- Those still trying to figure out what this Twitter business is all about
- Those who like it, but don’t quite understand why it’s so cool
- Those who know exactly what I’m about to say
That makes the explanation a little tricky, but let’s start at this end:
The world is drowning in ways to chat. Twitter has exploded because it’s different. It really has exactly one difference: its novel way of building your community. Twitter’s community-building dynamics are unique, which makes it interesting. Twitter’s community-building dynamics are powerful, which makes it popular. Twitter’s community-building dynamics are easy, which makes it fun.
Well … “were.” Because the people who run Twitter have decided to toss out their one distinguishing feature, the one thing that makes Twitter unique, and popular, and fun. In other words, Twitter is now … pointless.
What am I talking about?
This is where it matters which “kind of person” you are, and I’m going to have to explain something remarkably subtle, which I know is annoying to those who already understand, and confusing to those who don’t yet. Sorry about that.
The Twitter community-building process is … oops, sorry: “was” … built on two pillars: asymmetric following (just ’cause you listen to what I say doesn’t mean I have to listen to you), and overheard fragments. You can see how those two work hand in hand: you “follow” (as Twitter calls it) someone who you think might turn out to be interesting, usually by looking over their past tweets. But how do you learn about this someone in the first place? Most often, by overhearing some conversation they’re having with someone you already follow: it’s viral, organic, natural, and obvious once someone thought it up.
This is a complete contrast to most other Social apps, where friendship is mutual: I can’t hear what you say unless you also hear what I say, and in fact typically “befriending” requires some sort of bidirectional protocol; both have to express willingness before any connection happens.
The Twitter model means that particularly interesting people get a very large followership. It’s entirely up to you who you find interesting. You’ve heard, of course, about celebrity Twitter heros; I personally prefer to follow people like @timoreilly, of O’Reilly publications, who is an absolute fountain of provocative and insightful comments, links, and conversations. But make your own choices, follow whom you like. The point is, that whoever you follow is a window for you into a larger world, an introducer of new, interesting people.
Or, “was.”
Because as of yesterday morning, your current friends (those you follow) no longer introduce you to their friends. Twitter is now blocking that. As of yesterday morning, you’ll see far less on Twitter (which, of course, is not without its good aspects), so you can follow more people … but you no longer have any way to discover them.
What Twitter has done, actually is sold out. They were a community building service, they’re now only a community mirroring service.
OK, I admit it, I’m old: over a hundred years ago, a writer named G. K. Chesterton pointed out that special-interest clubs and societies (such as the “Gentlemen’s Clubs” of London, which were becoming popular at that time) were often described by their fans as a “larger world” than the home and family they were replacing, but they were in fact far smaller worlds: the self-selection of like-minded souls means that you never bump into someone too very different from yourself. You pick a club like you pick a mirror, for how flattering it is to your own self, and how well it insulates you from surprises, change, difference — life.
Twitter was a true Chestertonian “larger world.”
Now, it’s not.
Filed under: Toys | Leave a Comment
Michael Coté is stalking me!
You’ve heard that all this “social media” stuff is getting out of hand. But when “That Guy” on Flickr and Twitter and blogs sneaks out of your hand and onto the walls of your neighborhood coffee shop, it’s really too much. Especially early in the morning. Before you’ve had your coffee.
Which, of course, often happens in your neighborhood coffee shop.
Filed under: TMI | 1 Comment
Which generation are you?
It’s very odd. I used to be on the younger end of the “Baby Boom” generation (born 1944-1955, and there I was in 1953). It was uncomfortable, actually: I always felt slightly left out. And after some struggle, the generation following was dubbed “the Me Generation,” with heavy emphasis on the unsavory selfishness that implies. I was glad, eager, to be at least a little over the line away from that!
But gradually, people have stopped talking about that “Me” generation. Shame, really … check the ages of the rogues of the economic melt-down; lotta “Me” going around there, it seems!
But now, I seem to be mathematically almost dead center Boomer. Take The Technology Generation Gap at Work is Oh So Wide . (I hasten to point out that I’m far more hip than the Boomers discussed in this article!) (“More hip”? Did I just say that?) The article, or perhaps it’s the LexisNexis survey they’re reporting, defines Boomers as “ages 44-60.” Boy, at 54, no “long tail” for me, buddy: right in the center. A bit old for the center, in fact. (That can’t be right!)
Filed under: TMI | 2 Comments
Is there some reason, do you suppose, why Amazon only wants to sell their Kindle to people who don’t read very much? I’d think it would be just the opposite. Silly me. So far as I can see, if you actually love books, you’re expected to stay away from the Kindle.
Why do I say this? Because it doesn’t seem to provide any mechanism for keeping track of more than a dozen or so books. It (or their service) can store them, sure, but how many titles can you scroll through before you go nuts?
I love books. Every room in my house has at least one book case. I sort them endlessly: the Science Fiction, the history, the sailing stories, the Bibles, the detective stories … oops: Science Fiction detective stories, what to do? Some other organization now needed! Rearranging my books is itself a part of the love.
My electronic books, same story. I have just a bit over 200 volumes of electronic literature that I’ve collected in the ten or so years I’ve been reading ebooks out of my hand. They’re organized in a fairly rich folder hierarchy—and reorganized into some other hierarchy fairly often. I’ve been through three or four reader programs, written conversion utilities on occasion, and every one of them has come out initially unable to list these volumes in anything but a simple linear list. I’ve worked with each developer to help them understand the need for organizing tools, and each one has added the feature … only to be supplanted by the next wave of reader technology or platform, sadly, and I have to start all over again.
Now, Kindle. Once more around the barn.
And due to format wars (another blog post, for a time when my teeth can unclench), also Stanza, an excellent reader, but also without organizing capabilities.
For the iPhone, there appears to be only one reader with organizing capabilities, BookShelf. It doesn’t actually organize the documents on your iPhone, but it works in concert with a free helper on your laptop or desktop: the documents can be organized there, and can be browsed from the iPhone, and then downloaded onto the handheld. Almost close enough to suffice: the dox on the device are still just listed linearly, but the other connection’s good enough you don’t care. It’s like the laptop is just external storage for the iPhone.
But BookShelf is payware, and startlingly high priced for an iPhone app at that. I fully support the developer’s right to make a living, and I’m rather baffled at how LexCycle can provide Stanza for free … but they do. So I have to decide whether my need to organize my books is worth the whole price of BookShelf. Tricky question … not yet, anyway.
Filed under: Toys | 2 Comments
My brother’s a snob. It’s OK, we’re all snobs about something or other, right? (Don’t even try to talk to me about coffee or operating systems, you’ll just get nowhere.)
But my brother, now, he’s a snob about music. I don’t mean one band or another, one genre or another. He has “taste” in these areas, some of it good, but he’s not a snob about it. What he’s a snob about is “real music.” And by “real,” I mean “Not Guitar Hero. Not Rock Band.” Music that you actually, you, yourself produce.
Which is why I immediately called my former sister-in-law, to commiserate and make funeral arrangements, when I saw “Why Rock Band Is Better Than Actual Music.” I knew his head had exploded, messily and extravagantly. Imagine my surprise, when even as I was dialing, I received a tweet from … my brother … actually soliciting opinions on the whole “Better” business! And hardly even incoherent! I am so proud!
So here’s what I think: I think “the fake plastic Fisher Price musical instrument fad” is a bad thing. It’s not very much worse than what came immediately before (which would be: recorded music), but it’s a teensy step down “the path of least resistance, which leads forever downward.” What these toys should be compared to is something actually outside most people’s experience, these days: once upon a time, people learned to actually play music so they could play for each other at parties. There weren’t records or radio or TVs or iPods, there was only us. And when you heard music, it was either hirelings (and you were one of the unimaginably few elite), or it was your actual friends, who had spent their actual time actually learning to play actual music. For you.
Real musicians still have that experience (though they tell their wives they’re “practicing,” but domestic tranquility management is a topic for another post). Hardly anyone else does, any more. Which is why hardly anyone understands when real musicians deplore Rock Band: the one-eyed man may be king in the kingdom of the blind … but no one beholds his glory.
Filed under: Toys | 2 Comments
The best thing about the iPhone
I’ve finally decided. I love my iPhone in so many ways, it was hard to pick. But after three or four months’ constant use and play, I think I can pick a winner:
With the iPhone, I get the crappy AT&T service, with fade outs, dropped calls, and even “calls that never ring the phone,” a feature I’d never even heard of until I started using the service.
I never had any of these experiences with my former carrier, Sprint. Seven years with Sprint, and not one dropped call. But now, with AT&T, I feel special just to be allowed to finish a conversation without dialing back! I used to feel so left out: all those commercials about dropped calls, all my friends dancing like crazed monkeys trying to find a phone-sized ether-bubble with signal. The camaraderie! The exercise! I was jealous. But no longer! I’m One Of Them!
Filed under: Hello | 1 Comment
I’ve just pushed out a new build of SCPlugin, the Mac OS X “Finder” plug-in for Subversion. This one includes the new release of Subversion 1.6.0. You can learn more about the new stuff in Subversion 1.6.0 from their release notes. Good stuff!
You can also get an easy-to-install command-line build of Subversion 1.6.0 for your Mac from the CollabNet community. Which brings up an interesting question that might deserve some extra words: why are there so many versions available?
As you may know, the Subversion community jealously guards compatibility between clients and servers. As a result, every Subversion client ever released works with every Subversion server ever released. That’s pretty cool! Granted, there are some features, and some performance enhancements, that you only get if both ends are upgraded, but you always get at least the full feature set of whichever end is at the lowest rev. And remarkably often, you get something better even than that. This is seriously sweet work, and demonstrates just how dedicated the Subversion developer community is to you, the user of SVN.
None the less, and despite best efforts, and really not entirely their fault, and a couple other disclaimers, the compatibility story between mixed versions at the client end is not quite so good. While some versions of client tools can be intermixed freely, the only real guarantee is that you can always upgrade your client software without losing the use of your working copies. There’s no corresponding guarantee that you can downgrade. Which sounds reasonable enough–why would you downgrade???–until you stop to think about having multiple clients on the same machine. Such as, say,
- SCPlugin
- CollabNet Subversion
- Xcode
- the built-in SVN (available in OS X 10.5)
Now we see a problem: each of those is upgraded in a different way, so keeping them in sync might mean upgrading some for no reason other than to match the others.
This gets more complicated. There are a lot of Subversion clients for OS X. A lot. Some, like SCPlugin and Xcode, provide their own built-in copy of Subversion. Some, like CollabNet Subversion, fink, and MacPorts, offer full replacements for the Apple-supplied one. Others, like svnX, use the Apple-provided one (or, can be configured to use one of the full replacements). In such a world, it’s far, far too easy to get in over your head, and suddenly your favorite tool may stop working, just because you were fiddling with something you don’t really care about!
The Subversion community understands the problem. Among their efforts to keep things under control is a script they provide, called change-svn-wc-format.py. Basically, whenever you find that one of your tools complains that it’s “too old,” you can use the script to down-rev your working copy for its benefit. Understand that the moment you use a newer tool, it will silently up-rev the working copy again. If you’re mixing tools, you may have to get real real friendly with this script!
To help a bit more, SCPlugin has done several things:
First, we bundle this script with every SCPlugin install. If you have SCPlugin, you have the script (in /usr/local/bin/). And of course, if you have SCPlugin, you have a Mac, and if you have a Mac, you have Python, which you need in order to run the script.
Second, when SCPlugin happens to bump into that “too old” problem, it doesn’t just complain about its aching joints, it also tells you about this script, where to find it, how to use it, and what voluntary organizations exist to help you recover afterwards. All in one handy alert.
Third, we have begun a wiki page to track the various combinations and configurations. Go check it out, preferably before you get into trouble. It includes guidance on which versions of what go with which versions of what else, and how to avoid shooting your own foot off. In the nature of things, the page can’t guarantee to be complete, nor to remain at all times current, but it’s a wiki: we can all help expand and update it. So hopefully, it’ll always be at least useful.
So, check it out! Download from our Documents and files area. Discuss in our users discussion and mail list. Join the community!
Filed under: Subversion | 1 Comment
… psssst! there are no memes!
Stormy Peters tagged me into the “7 Things” meme, a kind of collaborative exploration of the conscious unconscious, and apparently the only force known that moves faster than the speed of Wikipedia. How the “7 Things” meme works:
- Link to your original tagger(s) and list these rules in your post.
- Share seven facts about yourself in the post.
- Tag seven people at the end of your post by leaving their names and the links to their blogs.
- Let them know they’ve been tagged.
The Things:
- It makes perfect sense to me when my iPod goes from Janis Joplin to Stryper to Coldplay to Chopin to (any) Marsalis. But I never knew Myley Cyrus recorded “7 Things” until I sat down to do this post.
- I once led a team who implemented a major software product (component of a desktop tool suite) that had zero customer-reported defects during the first year of sales, and was eventually the only part of its suite to show a profit (when it was sold to another company). It was a bug tracking system. Which no one uses any more.
- If the U. S. Navy hadn’t forbad me to play my trumpet on a submarine, and taught me to type 60 wpm instead (your tax dollars at work!), I might have been a musician today, instead of a geek. Well, “instead of a software geek.”
- I mentor a great high school student, Snyder, (wish I had time to do this more). We hang out, talk about what’s happening, and help each other sort out what’s important.
- I once bought three Macintoshes, two printers, two iPods, and a load of software in a single day. In general, I see the last 20 years of computer evolution divided equally between “figuring out what the Mac did right” and “figuring out what the Mac did wrong.”
- I make a pretty fair cappuccino (even though I can never remember how to spell it). I get about as much practice at this as your average barista, but have only myself to judge the results. Maybe it could be a new career….
- I am happiest when working a speck of feathers, hook, and string over an icy mountain stream. This has very little indeed to do with catching fish.
So, who do I know who deserves to be brought into this meme? Here are some people who I’ve found all the more interesting, the more widely I knew them and their writings:
- John Mark Walker Who manages his multiple personalities in more blogs than I can count
- Ben Collins-Sussman Who does it all in One Single Blog (truly amazing)
- Karl Fogel Activist, peacemaker, musician, coder, author, advocate, and friend
- Guy Martin Witty, opinionated, provocative, savvy–and sometimes even not wrong!
- Michael Cote Likewise. And I don’t even work with him
- Matt Asay Pretty much my inspiration for getting into blogging
- Jon Swanson Pretty much my inspiration for managing my blogging / life balance
Filed under: Hello | Leave a Comment