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。

My PhD advisor rewrote himself in bash (2010)

257 点作者 vismit20005 个月前

24 条评论

ergotux5 个月前
Two resources which helped me improving my writing, when I was writing my thesis were &quot;How to Write Mathematics&quot; by Paul R. Halmos and &quot;Mathematical Writing&quot; by Donald E. Knuth et al. I would always start with Halmos to get into the spirit of perusing clear and precise communication. The &quot;Bad&#x2F;Better&#x2F;OK&quot; suggestions especially reminded me of the discussions in the lecture notes from Knuth et al. And at a third step a linter such as the proposed one is probably helpful, if something slips through.<p>I think these resources are essential for anyone who writes on any subject which at least involves definitions here and there.
评论 #42415177 未加载
评论 #42414281 未加载
nazka5 个月前
If anybody likes this article and wants to know more about the process of writing effective PhD papers they should watch [1]. In fact, anybody who desires to improve their communication skills should watch it. It is so good that I would have paid to have access to this video!<p>[1] <a href="https:&#x2F;&#x2F;youtu.be&#x2F;vtIzMaLkCaM" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;vtIzMaLkCaM</a>
评论 #42418132 未加载
quickgist5 个月前
I find myself disagreeing with many of the examples. E.g. according to the article:<p>Bad: It is quite difficult to find untainted samples. Better: It is difficult to find untainted samples.<p>Bad: We used various methods to isolate four samples. Better: We isolated four samples.<p>Something being <i>quite</i> difficult reads significantly differently than just being difficult. You haven&#x27;t made the sentence better, you&#x27;ve changed the meaning.<p>And the fact that you used various methods instead of a single method is information missing from the second sentence.
评论 #42415388 未加载
评论 #42415210 未加载
评论 #42416327 未加载
评论 #42415866 未加载
mooreds5 个月前
If you are looking for more of this, check out vale.sh. It lets you add all kinds of style guidelines, including avoiding weasel words.<p>Here&#x27;s the GH action we use to run vale on our website at PR time: <a href="https:&#x2F;&#x2F;github.com&#x2F;FusionAuth&#x2F;fusionauth-site&#x2F;blob&#x2F;main&#x2F;.github&#x2F;workflows&#x2F;vale.yml">https:&#x2F;&#x2F;github.com&#x2F;FusionAuth&#x2F;fusionauth-site&#x2F;blob&#x2F;main&#x2F;.git...</a><p>and our config: <a href="https:&#x2F;&#x2F;github.com&#x2F;FusionAuth&#x2F;fusionauth-site&#x2F;tree&#x2F;main&#x2F;config&#x2F;vale&#x2F;styles">https:&#x2F;&#x2F;github.com&#x2F;FusionAuth&#x2F;fusionauth-site&#x2F;tree&#x2F;main&#x2F;conf...</a><p>We&#x27;ve found it helpful to enforce style but probably aren&#x27;t using it to the full extent.
评论 #42414323 未加载
simonw5 个月前
I got Claude to port those Bash scripts to a web UI so I could paste code directly into it and see what came out: <a href="https:&#x2F;&#x2F;tools.simonwillison.net&#x2F;writing-style" rel="nofollow">https:&#x2F;&#x2F;tools.simonwillison.net&#x2F;writing-style</a><p>Claude transcript here: <a href="https:&#x2F;&#x2F;gist.github.com&#x2F;simonw&#x2F;e9902ed1cbda30f90db8d0d22caa06d2" rel="nofollow">https:&#x2F;&#x2F;gist.github.com&#x2F;simonw&#x2F;e9902ed1cbda30f90db8d0d22caa0...</a>
评论 #42417657 未加载
评论 #42418363 未加载
emmelaich5 个月前
<i>diction</i> and <i>style</i> commands were present in early Unix, which did similar jobs.<p>You can get them (or versions of them) from GNU. They&#x27;re in homebrew.<p><a href="https:&#x2F;&#x2F;www.gnu.org&#x2F;software&#x2F;diction&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.gnu.org&#x2F;software&#x2F;diction&#x2F;</a><p>From what I remember, they weren&#x27;t great; this bunch of programs probably does just as well.
评论 #42414190 未加载
Gimpei5 个月前
In my field, writing quality was on the very lowest rung of importance, below even teaching evaluations. As much as I value clear, concise prose, I’d say a grad student would be better served working on public speaking especially when faced with hostile questioning, and, sadly, with brown-nosing. Yes I am bitter :p
评论 #42414733 未加载
评论 #42414148 未加载
liontwist5 个月前
&gt; To market a paper, the author must make a compelling case for why her idea deserves access to that resource.<p>In other words, journals are filled with papers that were sold the best, not the most important ideas. And as the author also says, superficial things like hard to detect typos are often a deciding factor because the reviewers can detect them.<p>we should stop pretending there is objectivity and embrace journals that reflect taste and opinion of the editor.<p>Or have standard places like arxiv for publishing everything. There is no scarce resource for uploading pdfs.
评论 #42414724 未加载
评论 #42417670 未加载
nonrandomstring5 个月前
Useful for those who, for some unfathomable reason, don&#x27;t use emacs.<p>Might be hard to get into a good workflow as running them and then re-editing seems tedious.<p>Author gives credit to emacs &quot;writegood&quot;, but my all-time fave style-nazi plugin is &quot;artbollocks-mode&quot;.
dang5 个月前
Related:<p><i>Shell scripts to improve your writing</i> - <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=13295530">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=13295530</a> - Jan 2017 (55 comments)<p><i>Shell scripts to improve your writing, or &quot;My advisor rewrote himself in bash.&quot;</i> - <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=1529166">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=1529166</a> - July 2010 (31 comments)
tgraf_805 个月前
Good intentions, indeed. Creating lots of steering committee slides, I know about the wish from the audience of a simpler language. But ‘very close’ is different from ‘close’. It’s not just salt and pepper but trying to articulate a complex and nuanced reality. And yes, research papers then sound a bit less solid and complete- sorry, but often this is the reality you should not hide.
评论 #42417449 未加载
gmac5 个月前
Can confirm the value of checking for repeated words such as ‘the the’. This is the final example I give my Econ PhD students in a session on RegExps: <a href="https:&#x2F;&#x2F;users.sussex.ac.uk&#x2F;~gm268&#x2F;iphd&#x2F;regexps&#x2F;regexps.pdf" rel="nofollow">https:&#x2F;&#x2F;users.sussex.ac.uk&#x2F;~gm268&#x2F;iphd&#x2F;regexps&#x2F;regexps.pdf</a>
评论 #42415912 未加载
评论 #42416503 未加载
ofalkaed5 个月前
As a humanities sort I appreciate this but the scripts sort of go against the general thrust of the text since the scripts can not understand context or semantics, it feels like they would push many towards blindly following prescription instead of what he is advocating for. I think elaboration would have served better than bad, better and good examples which do not explain the issues and assume the reader will intuitively understand. We get some elaboration but not enough.<p>Gardner&#x27;s Modern English Grammar should also be the primary recommendation for further reading, Gardner has a gift for explaining the nuances of these things. Style guides are guides for the style of a given publication or writing within a discipline, not guides on writing well or with style.
评论 #42417457 未加载
0xbadcafebee5 个月前
Put &quot;AI&quot; in the name and sell it for $5M?
anonymousDan5 个月前
I found this an interesting read: <a href="https:&#x2F;&#x2F;www.cambridge.org&#x2F;us&#x2F;universitypress&#x2F;readers-brain" rel="nofollow">https:&#x2F;&#x2F;www.cambridge.org&#x2F;us&#x2F;universitypress&#x2F;readers-brain</a><p>It gives a neuroscience perspective on what makes certain writing styles clearer.
655 个月前
<p><pre><code> I&#x27;d even go so far as to say that the removal of all adverbs from any technical writing would be a net positive for my newest graduate students. </code></pre> I&#x27;ve heard this before. Stephen King hates adverbs. However, it can be very difficult to remove all adverbs from your writing. &quot;randomly&quot; for example is an adverb, and if your sentence uses it, it can be difficult to rewrite the sentence without &quot;randomly&quot; that isn&#x27;t long and complicated. Many adverbs are emphasis words that don&#x27;t need to be in the sentence (&quot;extremely&quot; for example), but other adverbs are critical to the sentence.
评论 #42418477 未加载
teekert5 个月前
This what Claude&#x2F;Chat do when you ask them to make text more concise.<p>Have to say, Chat does produce what I am looking for in one go more often. Claude always makes lists or makes things way to concise. Maybe I need other prompts.
tetris115 个月前
duplicate words:<p><pre><code> sed -z -r &#x27;s|(\b[^ ]+\b)(\1)+|~Dupe!\1~|g&#x27;</code></pre>
lupire5 个月前
The script appears to be missing a check for &quot;long series of single-sentencr paragraphs&quot;, which quite harms readability.
barrettondricka5 个月前
I use a version of these scripts all the time.<p>though all-in-one spell-checkers like vale.sh are more convenient.
cm21875 个月前
Next: create a bash script to find out if the PhD advisor reads HN
评论 #42414741 未加载
dogboat5 个月前
See also <a href="https:&#x2F;&#x2F;vale.sh" rel="nofollow">https:&#x2F;&#x2F;vale.sh</a>
morpheuskafka5 个月前
&gt; I&#x27;d even go so far as to say that the removal of all adverbs from any technical writing would be a net positive . . .<p>&gt; Bad: We offer a <i>completely</i> different formulation of CFA.<p>&gt; Better: We offer a different formulation of CFA.<p>Getting rid of all adverbs (e.g. &quot;<i>vigorously</i> stir the solution,&quot; &quot;<i>monotonically</i> increasing function,&quot; &quot;<i>spontaneously</i> combustible&quot;) hardly seems wise, whether technical writing or not.
评论 #42413948 未加载
评论 #42414021 未加载
bsder5 个月前
My first reaction was &quot;This person isn&#x27;t in a hard science like physics or engineering. Maybe medicine?&quot; Ayup--medicine.<p>Style can be annoying but <i>lack of context</i> is deadly.<p>&quot;To our surprise, false positives were low (3%).&quot; is not much of an improvement. &quot;To our surprise, false positives were low (3%) as we expected closer to 10%.&quot; <i>IS</i> an improvement.<p>However, that statement in a hard science is going to cause people to start asking some questions: Is 3% actually low? Why was the expectation so high to start? Did you screw up your error bars? Is it <i>really</i> 3%? etc.<p>&quot;It is difficult to find untainted samples.&quot; <i>Because</i> is missing. In addition, this comment is rarely germane. I, the reader, don&#x27;t really care. That&#x27;s <i>your</i> problem. Either find the samples to get to statistical power or get a different method. As the reader, I only care if you did something weird in order to sidestep the problem--<i>that</i> I want to know about.<p>All of the recommended things still retain &quot;weaseliness&quot; while trying to sound like they don&#x27;t.<p>Finally, I would love to have his problem of merely cleaning up some weasel words when most students I know (even at the PhD level) still have trouble stringing together coherent arguments and understanding where the holes and weaknesses are.
评论 #42414352 未加载