Python Secret Weblog for 2008 February 1 (entry 0)

(5) Fri Feb 01 2008 22:32 Python 3.0 is incompatible with Python 2.0:

Jesse Noller wrote:

I hate slashdot: Python 3000 is incompatible?!!!

and then says:

A sensationalist title, summary and normal slashdot commenters: "Python 3.0 To Be Backwards Incompatible"

I don't see how the title is sensationalist. Python 3.0 is backwards incompatible with Python 2.0. Conversion scripts are supplied, but that fundamental reality stands. The summary on Slashdot is this:

Organizations using Python will be affected in a major way by changes in store for the language over the course of the next twelve months, Linux.conf.au attendees were told this morning. The Python development community is working towards a new, backwards-incompatible version of the language, version 3.0, which is slated for release in early 2009. Anthony Baxter, the release manager for Python and a senior software engineer at Google Australia, said "We are going to break pretty much all the code. Pretty much every program will need changes." Baxter also added another tidbit for attendees, saying that Python accounts for around 15 percent of Google's code base."

I don't see anything sensationalist in this either. It's true: organizations and communities using Python are going to be impacted in a major way by the release of Python 3.0, and the reason they are going to be impacted is the non-backwards incompatibility. There are significant consequences for organizations that need to deal with this transition - for the organizations that go ahead and make it, and also for the organizations that wait and see and continue to use the Python 2.x version.

I haven't read any of the Slashdot comments. I'm sure some of them are very wrong, and very annoying; that's normal. It's just that I object to spreading the idea that this message is sensationalist or somehow wrong. It's not. It's the reality that one way or another the Python community will have to deal with. Objecting to an expression of this reality spreads the wrong message.

Recognizing that Python 3.0 is backwards incompatible makes us aware that we are heading for something we will have to deal with one way or another. It's a useful wakeup call for people who might not be aware of this.

The language developers are working on a migration script and are adjusting the Python 2.x line of Python so that we can prepare our code bases. By all means, spread that news too in addition to the backwards incompatibility news. But don't object if news sources forgo such subtleties in spreading the main message to take home about Python 3.0: without special action on your part, your code will break when you try to run it with the new interpreter. Adopting Python 3.0 as a community will mean an awful lot of code to transition. It'll be a process of years, and it won't be easy.

(as an aside: I think I wrote a reasonable posting about Python 3.0 just now. Of course that's what I thought when I posted about this last time, and then I had people land on my head like a ton of bricks. It left a very unpleasant taste in my mouth. Let's hope this time will be better. I think having an open discourse about this is important.)

---

Comments:

Posted by Paul Boddie at Sat Feb 02 2008 02:43

Here's a good response:

"See, the funny thing is that the changes going into Python 3 are fixes for much of what people have complained about in Python 2.x and prior."

Not very consistent with the reality, though:

http://wiki.python.org/moin/PythonWarts

Anyway, I think I've already said enough on the subject: there are some useful changes in 3.0 and some gratuitous breakage. Perhaps 2.7 or 2.8 will be the "good enough" compromise release.

Posted by Jesse Noller at Sat Feb 02 2008 03:02

It *is* sensationalist, as are most - if not all of Slashdot lately. It makes no mention that the 2.x line will be maintained, or that conversion tools will be maintained, or the fact this has been planned for some time. The person(s) who submitted, and approved the article did little (no) footwork and instead puked up a link to an equally detail-free article. Slashdot stopped being a news site long ago.

I am allowed to think a half-assed pseudo-story omitting any real information about a talk in Australia is sensationalist. The headline was silly (and that was my beef), and I'd expect more from an outlet I had respect for at one time (although that time passed long ago). I have seen bloggers and other people deride slashdot for equally sensationalist and information-devoid articles in the past, of which this is simply a new one. Most of the commenters on this article within the first few hours were equally devoid of anything coming close to value.

Listen - I know you have your bone with Py3k, it's not perfect, and you disagree with many of the things that have gone on, objecting to revolution versus evolution is OK. I don't agree with every aspect of py3k either, but encouraging people to stand in the town square and scream "the sky is falling" is silly.

People will not *have* to change their code when py3k comes out. There will be a gradual transition in which those who wish to use Py3k will, and those who don't, won't. Most linux distros will probably ship both 2.x and 3.x. The 2.x line will be around for some time to come. Yes it's going cause disruption in the community - no one is arguing that.

I think calling it sensationalist is fair: I want to expect more, not less from my "news". They could have at least made mention of the effort many people are putting in to transition tools, and back porting features into the 2.x line to make the gradual transition easier.

Posted by Martijn Faassen at Sat Feb 02 2008 14:38

Jesse, Slashdot is sensationalist *lately*? How long have you been reading Slashdot? :)

Slashdot is giving short summaries of articles with little room of subtleties. Subtlety is not what Slashdot is all about. Mentioning effort in transition tools and backporting features to the Python 2.x line might have gotten the summary over twice the length of what it was - too long for Slashdot.

I just still don't get why you complain about this Slashdot entry. They've been like this for as long as I've known them, and this entry itself did not contain any inaccuracies.

I did not read the story it linked to. You were not complaining about the story originally - you were complaining about Slashdot itself and the message it was giving out. If think it should have linked to a better story, then why didn't you say that?

Posted by Jesse Noller at Sat Feb 02 2008 14:53

Agreed on the lately point: I should have said "it's been getting worse and worse" - especially ever since they started trying to compete with Digg. I mean, there's low content - and then there is *no* content.

The post they chose to put up was low content: the Headline they used was sensationalist, and the summary was low content. Again: You're right, the point at which I could have expected more from slashdot is long past, but I can still be frustrated when things are misrepresented.

I frustrated when other things are misrepresented on the site as well - this just happened to be something I not only thought was sad: but sadly amusing.

Posted by Leandro Lameiro at Tue Feb 05 2008 18:19

Martijn: I agree with Jesse that Slashdot title and content is sensationalist. It is true that py3k will break everything, but the story misses the other side of the release. It just doesn't cover any of the efforts being done to help porting 2.x code.

To be a neutral news source they don't just have to say the truth. They need to write about bad and good things of the change.

I stopped reading slashdot exactly because of exaggerated stories like this. And they do it all the time.