LLVM 2.9 Released | Hacker News Reader

LLVM 2.9 Released

llvm.org

69 pointsDrJokepu19 comments

Does anyone have experience with using the C API from clang ? Does it improve from former versions (doc, stability)

I would really like to get a reliable C header parser for code generation, and gccxml is just damn awful to use.

I'm curious - why do you need the C API? Do you plan to call it from some dynamic language?

That said, the C API of clang should be very usable, and they even have working Python bindings for it AFAIK

For the C rationale: yes, I want to use clang to generate cython interface files from pxd (I have an horrible hack using gccxml ATM: https://github.com/cournape/cython-codegen), and I would prefer writing the tool in python from ctypes.

Do you have any experience with the C API ? Last time I checked (clang 2.7 ?), there was little to no documentation, and the few clang tutorials with C++ were out of date.

Personally I have no experience. But I'm routinely monitoring the clang development mailing list and sometimes discussions of the C API surface and it appears it's active and supposed to be operational.

Have you looked into using Boost::Python?

Flagged. Not likely to be of interested to non-LLVM users, LLVM users can subscribe to the LLVM announcement mailing list if they want release updates.

What's the value to the HN community?

Hacker News Guidelines
What to Submit

On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.

http://ycombinator.com/newsguidelines.html

This article is "On Topic", IMO, as per the guidelines.

It's a minor release announcement. Where's the intellectual curiosity ?

Here's a simple test: Are people commenting on the content of the story ? - I'd be curious to know how many of the people upvoting even read the release notes and how many just upvoted based on the title.

Please do realize that there are youngsters who come across new terms, like linux, unix, design patterns, kernel, unix, oop, functional programming, open source, SEO etc. and many others, every single day for the first time.

Just because you and I have already gained (some) knowledge about these things, does not mean they will not be intellectually simulating for someone else, especially someone new to the profession/field.

Also in the guidelines,

> Please don't submit comments complaining that a submission is inappropriate for the site. If you think something is spam or offtopic, flag it by going to its page and clicking on the "flag" link. (Not all users will see this; there is a karma threshold.) If you flag something, please don't also comment that you did.

I'm aware of that and I consciously chose to ignore it.

I felt it would be inappropriate for me to flag it without publicly saying I was doing so, as I know other people might disagree and stating that I was flagging it would allow for a discussion as to whether the link is on-topic or not.

If somethings wildly off-topic or just spam I'd flag it without commenting, but I think if something is borderline off-topic it's something that deserved to be discussed.

Why would you flag this? Its relevant to all of us with an interest in programming and development of systems. Dont you realise the significance of llvm?

As someone who used to hang out on LtU I'm well aware of the significance of LLVM, doesn't mean it's release announcement is on-topic on HN.

There are hundreds of packages which are of interest to programmers, if we posted minor releases of every single one of them to HN, we'd soon be drowning.

So we need to ask what's special about this release?

Are there features that are going to attract new users, are they doing something impressive, etc.

In this case the answer is no, it's just a standard release, so I don't think it's a good use of the first place postion on HN.

It spans beyond just a release announcement. It introduces llvm to people who are unaware of its existence. It brings llvm into the mainstream so that more people may start using it. It gives us an opportunity to discuss the project in both positive and negative lights, whether that be relevant to the release or the project as a whole.

I agree that we cannot post the release announcement for every project in existence and I believe everyone here on HN understands that, but when an interesting and significant project has an update and may not have been on the front page for some time, it seems appropriate to submit it.

That need would be much better served by posting a link which talks about a practical example of LLVM being used in the wild, or an interesting hack you can do using it.

I've no objection to LLVM stories being posted here, I do object to minor releases being posted here which push down stories more useful to the community.

Posting release notes means people who aren't familiar with it at best are just going to glance at it and hit back.

But the same logic applies to any other interesting piece of software...

And what's result of that on HN? In the past 7 days on the front page we had release notes for:

- jQuery 1.5.2,

- Redis 2.2.3,

- Portable C Compiler 1.0,

- Node.js 0.4.5,

- KeePass 1.19,

- GNOME 3,

- SproutCore 1.5 Release Candidate 1.

How many of those are really worth it? Maybe PCC and GNOME, since they are major releases.

Also

- Rails 3.0.6

- Palm WebOS 3.0 (probably ok as major release)

- AppEngine 1.4.3

I stay with my previous opinion of "when an interesting and significant project has an update and may not have been on the front page for some time, it seems appropriate to submit it".

Everyone is entitled to their opinion and there will be debate about this. I asked why it was flagged and I got a response, I'm happy being able to have conversation in which I've understood the other persons argument.

So, that's 10 things, of which 3 are major releases and uncontroversially appropriate. 7 in total. So, on average, either

1. these things will last at most (say) a day on the front page, in which case they take up ~ 1 slot there; or

2. they will attract enough interest from the HN community to stay on the front page for longer than a day.

I find it hard to see either of these as terrible outcomes, given that these submissions are of enough interest to get to the front page in the first place.

Reply on news.ycombinator.com