TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

MRI Developers Don't Use RubySpec and It's Hurting Ruby

291 点作者 jc00ke超过 10 年前

21 条评论

nateberkopec超过 10 年前
Rubyspec is a great project for all the reasons Brian outlines.<p>However, it&#x27;s also a failure - in part, due to Brian and his attitude towards contributors. See this twitter conversation: <a href="https://twitter.com/the_zenspider/status/547527644535726080" rel="nofollow">https:&#x2F;&#x2F;twitter.com&#x2F;the_zenspider&#x2F;status&#x2F;547527644535726080</a> He&#x27;s been grinding on Rubyspec for years, bless him, but I think there&#x27;s a reason why he was unable to rally the community behind his effort, both in terms of gathering more contributors and in making it &quot;official&quot; in terms of the language spec.<p>JRuby, currently the only non-MRI ruby implementation that you can seriously consider for production use, runs the MRI test suite against JRuby. RubySpec is far from The Only Solution, though Brian would like you to think it is.
评论 #8821610 未加载
评论 #8821543 未加载
评论 #8821781 未加载
评论 #8821665 未加载
vorg超过 10 年前
It&#x27;s disappointing to hear about such problems with Ruby. I hit the same problem with the Groovy Language spec when I first came across Groovy. Its creator, James Strachan, initiated an implementation, test kit, and spec all within 6 months of each other (impl beta-1 in Dec 2003, and spec JSR-241 in May 2004). The project managers who took over from him, Graeme Rocher and Guillaume Laforge, changed direction by stopping work on the spec and refocusing the Groovy reference implementation to be the scripting language behind Grails. (Of course, <i>Groovy &#x27;n&#x27; Grails</i> was intended to chisel away at some of the market share of <i>Ruby on Rails</i> but that&#x27;s another story.) Strachan often wrote that the spec was to enable anyone to make their own implementation of Groovy if they want to, and right up to his very last posting ever <a href="http://groovy.329449.n5.nabble.com/Paris-write-up-tt395560.html#a395571" rel="nofollow">http:&#x2F;&#x2F;groovy.329449.n5.nabble.com&#x2F;Paris-write-up-tt395560.h...</a> on the Groovy mailing list on 5 Dec 2005, he maintained that what they were building was the reference implementation.<p>If Rocher and Laforge had come clean about how they turned the RI into the language itself, the backlash might have blown over quickly, but instead they led developers along for many years afterwards, not changing the spec to dormant until April 2012. Projects other than Grails who&#x27;ve tried to build atop Groovy have had to risk the ref impl changing in breaking ways between versions. The most spectacular incident was when Groovy++, an experimental static compiler built by Alex Tkachman that hooked via annotations into Groovy&#x27;s AST, had to drop back down from Groovy 1.8 to 1.7 in 2011, and my own side project was also affected by the change. It turned out Rocher and Laforge had secretly employed a mate to extend Groovy with the exact same static type-checking and compilation functionality as Groovy++ and were obviously trying to shake us off.<p>Unlike Ruby, Groovy only has one other implementation, GrooScript, built by Jorge Franco, which generates JavaScript from Groovy syntax. When the developers of the most used implementation of a language want to protect their control, it certainly does hurt the ecosystem, turning it into an &quot;echo system&quot;.
评论 #8824797 未加载
评论 #8823121 未加载
lgleason超过 10 年前
Both Charles Nutter and Matz are smart guys. They are also both great people. I say this having spent time with both of them. This whole, my implementation is better than theirs coming from Brian is crazy.<p>Matz and Charles have helped to set the tone for the community. While I appreciate Brian&#x27;s passion, it sounds like he needs to check his ego. No matter how smart we are, we can always learn things from other people. Without the collaboration of others neither JRuby or Ruby would be what they are today which is why they are successful.
评论 #8821647 未加载
评论 #8821687 未加载
sams99超过 10 年前
The thing I find depressing and frustrating about this kind of discussion is the fatalism and non-constructiveness.<p>I really wish it was:<p>&quot;I set up a server that runs ruby spec on ruby-head daily and automatically reports spec failures to ruby-bugs&quot;<p>So many companies are making big bucks off Ruby, yet so little are willing to fork out a bit of money and time to make Ruby better.
评论 #8821224 未加载
评论 #8822008 未加载
评论 #8821606 未加载
bhrgunatha超过 10 年前
&gt; Later that year, at RubyConf 2008, I gave a talk titled, What Does My Ruby Do about RubySpec. Matz and several other MRI developers attended. Immediately after my talk, contributors to Rubinius sat down with Matz and other MRI developers to discuss their effort to create an ISO specification for Ruby. We asked whether RubySpec could be part of the specification but were told that it was not appropriate to include it.<p>This seems telling.<p>Does anyone have any concrete information about why Matz and the Ruby team are opposed to using RubySpec then?<p>Has there been any progress on creating an ISO spec?<p>EDIT: It seems Ruby has a published ISO spec since April 2012 - <a href="http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=59579" rel="nofollow">http:&#x2F;&#x2F;www.iso.org&#x2F;iso&#x2F;iso_catalogue&#x2F;catalogue_tc&#x2F;catalogue_...</a>
评论 #8822215 未加载
tessierashpool超过 10 年前
I don&#x27;t want to get in the middle of drama, but if you take anything away from this, read the actual code.<p>Just read the actual code of the MRI tests.
评论 #8821755 未加载
lnanek2超过 10 年前
Matz is the creator of Ruby. If this guy really wanted to play ball, why didn&#x27;t he just submit PRs for improving the real Ruby&#x27;s test suite? Instead he just made his own thing, well duh it didn&#x27;t end up a part of the real Ruby. OK, a hosting provider wanted their own Ruby implementation to fix concurrency issues, but that doesn&#x27;t suddenly make the creator of the language have to do things their way. If you fork or reimplement a project and do whatever you want, well you took control so you gained something, but the creator has no obligation to change to your fork&#x2F;reimplementation.
评论 #8823617 未加载
Terr_超过 10 年前
Regardless of other warts with the language, I really love how the Java Language Specification lays down the law.
评论 #8821378 未加载
hartator超过 10 年前
Has Matz stated the reasons they are not using RubySpec? Knowing him, I can&#x27;t really believe he has chosen to ignore RubySpec without good reasons.
评论 #8821280 未加载
jes5199超过 10 年前
I almost didn&#x27;t catch the epilogue - he&#x27;s shutting down the RubySpec project entirely, because it hasn&#x27;t accomplished what he hoped.
评论 #8821553 未加载
senthilnayagam超过 10 年前
@senthilnayagam: . @yukihiro_matz can you respond <a href="http://t.co/qeeVAluOhJ" rel="nofollow">http:&#x2F;&#x2F;t.co&#x2F;qeeVAluOhJ</a><p>@yukihiro_matz: @senthilnayagam I am not in charge of testing. But as far as I understand it has been communication problems. Blaming no use.<p>@senthilnayagam: . @yukihiro_matz when merb could merge with rails, rubyspec shpuld merge with MRI &amp; become official reference for all Ruby implementations
joevandyk超过 10 年前
Why don&#x27;t MRI developers use RubySpec?
评论 #8821392 未加载
msie超过 10 年前
What of the ISO standard that mRuby is based on?<p><a href="http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=59579" rel="nofollow">http:&#x2F;&#x2F;www.iso.org&#x2F;iso&#x2F;iso_catalogue&#x2F;catalogue_tc&#x2F;catalogue_...</a><p>Aw crap, I have to buy this doc to see it?
评论 #8822277 未加载
评论 #8822273 未加载
dyeje超过 10 年前
Some of the discussions he links to in the article are disturbing. Many posts refusing to implement process because they don&#x27;t want any process at all. Doesn&#x27;t seem like a healthy approach.
评论 #8821562 未加载
issaria超过 10 年前
THIS Brian again, nobody uses rubinius and it doesn&#x27;t hurt anyone. This guy has a long history of promoting rubinius by attacking MRI, first time I saw him was on Baruco 2013, if you watch the video, his attitude is so annoying. Evan Phoenix&#x27;s Ruby implementation was an interesting project, but now totally ruined by this childish guy. Brian, my advise is to stop bitching about MRI and seriously improve the documentation of rubinius, the are many interesting topics on rubini.us, but most of them are WIP since forever.
serve_yay超过 10 年前
Jeez, what a mess.
danielweber超过 10 年前
Why <i>should</i> they have been using RubySpec?
评论 #8821186 未加载
crb002超过 10 年前
MRI unit tests are the spec.<p><a href="https://github.com/ruby/ruby/tree/trunk/test/ruby" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ruby&#x2F;ruby&#x2F;tree&#x2F;trunk&#x2F;test&#x2F;ruby</a>
jc00ke超过 10 年前
I mistakenly cut off the full title of the post: &quot;Matz&#x27;s Ruby Developers Don&#x27;t Use RubySpec and It&#x27;s Hurting Ruby&quot;
grover_hartmann超过 10 年前
Fuck, this idiot is now being pendatic in #ruby as well.<p>2015-01-02 14:45:30 brixen raise your hand if you have implemented Ruby and used MRI&#x27;s tests<p>Glad he didn&#x27;t get things to go his way. He cries too much. If you don&#x27;t like some projects go write your fucking own.
jagawhowho超过 10 年前
Ruby is about to die permanently. A language that copied emacs-lisp but with syntax? Lol that gives it features to impress the ignoramus majority but real Haskell or Lisp programmers know better.
评论 #8823192 未加载