I never understood this thing about "communities". It's not like someone marry a language and a community, and exclusively use only that language for everything, and then, when some better/newer language comes, divorce from the "community" not ever to touch it again.
Come, we are all adults. Use the right tool for the job, and belong to many (or none) communities. Your favorite programming language is not your identity.
What strikes me most about both Ian's history and Armin's is "Starting again". Rewrites, starting over having learnt more the first time. Yet the idea of that happening in a commercial setting is ludicrous for almost every company I have known. Yet perhaps look at Apple's development of the iPhone - a constant set of retries trying to get past a single Jobs.
> Rewrites, starting over having learnt more the first time. Yet the idea of that happening in a commercial setting is ludicrous for almost every company I have known.
I am not a commercial setting, I am a programmer. If I would not learn new things and adapt I would become rather pointless quickly. What's important for a company is not important for an individual.
The point I was trying to make was there is very little likelihood of a company saying "yes, go rewrite your code, start a new project, because you think it can be done better".
This is a much needed counter-balance to Ian's post. While I can understand the technology reasons for moving away from Python, the community remains the most important reason for hanging on to it. I believe, which ever open source tool one chooses to use, the community should be one of most, if not the most, important criteria.
Hmmm, I guess that just didn't make such a big impression on me. I mean, the PyCon people actually tried to do the right thing, even before the fact, and seemed to handle the situation correctly as it developed. The problems that resulted were probably due more to flaky personalities and poor judgment on the part of those personalities' management.
I think that the problems went deeper than "flaky personalities and poor judgment on the part of those personalities' management" since Adria continues to feel/be threatened even a year later:
Umm, yeah, I probably shouldn't touch that stuff above with a 39.5' pole. What I meant to say was that although I'm sensitive to everyone's emotions, it would have been a better result for everyone involved had the two CEOs imposed a two-week gag order on all parties and then after all us chattering morons forgot it, issued a joint press release acknowledging the incident and reporting that all had been resolved to everyone's satisfaction through private dialogue.
What I should have said instead was that I don't think the python "community" can really be blamed for it since they actually instituted an anti-harassment code of conduct before many other conferences and have actively, repeatedly sought and used input from advocates on this issue.
A bit off-topic, I would like to view the hype-trend of programming languages on Google Trends. The new Google Trends interface allows one to search for "Ruby" the programming language and not Ruby the mineral.
But I have troubles with Python, as the search suggestions show me Python molurus, Monty Python, Python reticulatus, Burmese python but sadly "Python" the language is missing. Maybe someone found a solution how to input Python (the language).
When I search for "Scala" the programming language in that interface, I still get the majority of the results from Italy, specifically the region of Milan (and the Teatro alla Scala). Might be, because "making a program" is an activity in music too, and google is unable to filter that?
I wonder where most people go from python? As has been mentioned in other threads (specifically the comments on Ian Bicking's recent blog post) many seem to be moving on. Where are you all going? Personally after having been a python dev for a long time, I now self identify as a Haskeller (although I'd certainly still use python for many tasks over Haskell). I find Haskell to be as mind expanding now as python was when I taught myself to code ~10 years ago.
I've seen others mention clojure and go as jumping off points. So, those of you who feel you have "moved on" from python, where would you say you hang your digital hat nowadays?
I think lots of people in the Python community are very good at what they are doing and just adjust to whatever the new situation requires. There are lots of Python people that made a move to the JVM after they outgrew their Python environments. There are many that are now writing lots of Go and JavaScript.
I don't think people "move on", they just also start using more than one programming language. A few years ago Python + C was all you needed. Now you need to deal with JavaScript if you want it or not, often there is Ruby involved because of deployment tools, there is more and more Go code, Java became popular in the same community etc.
I've moved on to Scala. I never liked the lack of typing in Python, but I thought it was the price one paid for an expressive, low-noise language; then I discovered it was possible to have it all.
> If the tool is just bad then there is no good reason to deal with the community.
Python is not that bad and while I, like Armin, have been moving slowly away from it due to a number of reasons, it is still one of the most versatile tools in my toolbelt.
It's not a bad technology choice. It's that people themselves evolve, get interested and learn new things.
Armin apparently is somewhat involved with Rust now, I doubt that's because he is going to write web application in it but it's because he has become interested in developing applications that require a low-level language.
Ian Bicking has been working for Mozilla for a while now and Mozilla is very focused on using Javascript...
Important people in the Python community becoming interested in new things, learning new things, is not a bad thing, it's a very good thing. You want people with influence to learn new things, get excited about them. That's what sparks evolution, new ideas and creativity.
Sometimes - not always. Especially if it's not merely mediocre, but outright bad as postulated. (I don't think Python is bad, but I don't have the experience with it to judge either.)
All the community in the world is unlikely to make up for tools that I have to reverse engineer from disassembly to file proper bug reports against (too frequently necessary to even figure out if it's their bug or mine in the first place), or create workarounds for.
If a tool is good, I'm probably not turning to the community much in the first place: It's all right there in the docs or obvious by design, right? (If the answer is no, is it actually a good tool?)
> If a tool is good, I'm probably not turning to the community much in the first place
Then there's no community, there's nothing shared and the tool ends up dying because there's only so much you can do on your own. See: the vast majority of lisps.
Communities survive my absence ;). But more importantly, the lack of a user community doesn't necessarily translate to the lack of an internal developer community outside the FOSS sphere, or lack of a feedback loop into that developer community.
Do you use your browser because of it's user community? Your media player? Your IDE? Your VM manager? Your task tracker? Your IRC client? System APIs? The standard library of your favorite Language?
For myself, I'd answer "none of the above". There are hundreds of devs behinds these tools, and I interact with next to none of them - and that works out great.
> there's nothing shared and the tool ends up dying because there's only so much you can do on your own.
I'm frequently entirely OK with this. Some of these tools will evolve without my input. Some of them will die (read: finally stabilize) when better alternatives surface (win/win!) Neither situation is generally going to be helped along significantly by throwing in my lot with an existing tool's user community.
On the developer side of the equation, when even possible, joining is expensive: Time spent familiarizing with the codebase and coding standards, agreeing on bikeshed colors, wrangling contracts to avoid ending up like Sergey Aleynikov...
> See: the vast majority of lisps.
This may be heretical, but I'm afraid I don't grant that "the vast majority of lisps" were good tools, although I'll certainly grant that community inertia wasn't enough to save them.
> Do you use your browser because of it's user community? Your media player? Your IDE? Your VM manager? Your task tracker? Your IRC client? System APIs? The standard library of your favorite Language?
I'd actually say yes to many of those. The main reason I switched away from my previous browser was the absence of certain extensions/userscripts; the reason I use the IDE I do is entirely down to the plugins.
Come, we are all adults. Use the right tool for the job, and belong to many (or none) communities. Your favorite programming language is not your identity.