Go 1.7 Is Trading Much Slower Compile Times For Better Generated Code

[jntesteves]

Go 1.7 Is Trading Much Slower Compiler Times For Better Generated Code

Should read: "is trading short compile times for better code". The title, as it is, describes a win-win situation, which is not the case here. And, sci-fi apart, time can't be faster nor slower. What you mean there is a shorter/longer comparison.

There, I said it! Hopefully in a way that won't trigger any heated discussion about understanding basic English this time

[Luke_Wolf]

Should read: "is trading short compile times for better code". The title, as it is, describes a win-win situation, which is not the case here. And, sci-fi apart, time can't be faster nor slower. What you mean there is a shorter/longer comparison.

There, I said it! Hopefully in a way that won't trigger any heated discussion about understanding basic English this time

Well... Since you're intent on asking for it... when dealing with perfectly valid English which apparently you don't and in fact refuse to understand...

In English when we have a word followed by the word time, for example: "compile time", or "running time" this is considered a span of time. Now the terms "slower" and "faster" in the context of such a phrase refer to expansion or contraction of this span, or are used to describe the relation between two spans in the same fashion. "Slower" and "faster" are also relative to the span in question whereas "short" vs "long" is relative to the speaker, and are absolute statements as far as the reader is concerned.

Now let's take some example sentences to contrast what you're trying to push with how English is actually used

* "Bob's run time has become faster since he started working out."

"Bob's run time became short when he started working out."

* "Jane's run time slowed down after she was involved in a car accident."

"Jane's run time became long after she was involved in a car accident."

The one's with the *s are how it's actually written in English in case you didn't catch that. Also where precisely do you get off telling native speakers how to write? Hm? Particularly given that your grasp is obviously fairly weak, else you wouldn't be whining at Michael like this. This is supposed to be a tech site and a tech forum, not a non-native speakers whine at Michael for their misunderstanding of English website or forum. If you have something tech related to discuss, do so.... otherwise please redirect your forum entries to /dev/null.

There, I said it!

[Guest]

Should read: "is trading short compile times for better code". The title, as it is, describes a win-win situation, which is not the case here. And, sci-fi apart, time can't be faster nor slower. What you mean there is a shorter/longer comparison.

Just some notes:

- The article title indirectly contains the word "tradeoff", which contradicts the idea that the title mentions a win-win situation.

- In the case of some people's Go codes, Go 1.7 will actually be a true win-win situation, because for them the following equation will be true:

(compile_time_go_1.6 + run_time_of_my_app_go_1.6) > (compile_time_go_1.7 + run_time_of_my_app_go_1.7)

- The interaction of compiler speed, speed of the code generated by the compiler, and programming time is a complex problem.

[compiler speed + speed of code generated by the compiler + programer time] => [complex problem]

See also: https://en.wikipedia.org/wiki/Complexity

[jntesteves]

Luke_Wolf: I didn't address you directly on the other thread when I thought you were trying to deliberately offend some other people, because it was not directed to me and also, I thought you were just having a bad day too, or maybe I misunderstood your intention. Then, I said a bunch of BS to try to stand my point, which I admit and already apologized. But now you are addressing me directly and doing the same thing I did there (going out of your way to stand your point) and again I have the feeling that you sound somewhat rude. Could be just me. Maybe I sound rude to you too and it's just a cultural difference, in which case I apologize for that too, as I don't mean to be rude towards anyone in this community.

About the contents of what you just wrote in this thread, though, I hope no one learning the English language ever read what you just wrote and think that is appropriate use of the language for anything but an informal conversation.

Being a native speaker is not a free pass to publish wrongly phrased statements without deserving at least a little criticizing now and then. I don't usually do that but also don't see anything wrong with it. It's just your opinion that that should never be done. You can try to enforce your opinions if you want, that doesn't mean anyone will obey you.

The title given to this article states the opposite of it's contents, so I thought it was well deserving of criticism. You did not deny it, so I'll assume we agree on at least that.

[signals]

Just some notes:

- The article title indirectly contains the word "tradeoff", which contradicts the idea that the title mentions a win-win situation.

jntesteves is probably concerned about the misuse of the word "trading." If I trade X for Y, I give up X and get Y. Go is actually trading fast compile times for better code generation. The way the title is worded makes it sound like they got better code by giving up slow compile times.

I'm not sure why I decided to join this discussion. Maybe I just like abuse? But, since I'm pretty sure jntesteves is correct, it irks me to see people jump all over him. I guess I didn't see the other thread, and don't know the history of this conflict.

I am a little confused about his objection to fast and slow here. Yes, technically "short" and "long" is probably more correct. But I think fast and slow is commonly used in this situation.

[Guest]

Should read: "is trading short compile times for better code". The title, as it is, describes a win-win situation, which is not the case here. And, sci-fi apart, time can't be faster nor slower. What you mean there is a shorter/longer comparison.

There, I said it! Hopefully in a way that won't trigger any heated discussion about understanding basic English this time

I'm not sure why I decided to join this discussion. Maybe I just like abuse? But, since I'm pretty sure jntesteves is correct, it irks me to see people jump all over him. I guess I didn't see the other thread, and don't know the history of this conflict.

In general, it is about which rule to choose to get the most accurate meaning when interpreting article titles in the news.

Ideally, an article title should be an abstraction of the article content and it should not contradict the content, but unfortunately finding/formulating such a title is impossible in many cases because of how human language is structured.

[Luke_Wolf]

Luke_Wolf: I didn't address you directly on the other thread when I thought you were trying to deliberately offend some other people, because it was not directed to me and also, I thought you were just having a bad day too, or maybe I misunderstood your intention. Then, I said a bunch of BS to try to stand my point, which I admit and already apologized. But now you are addressing me directly and doing the same thing I did there (going out of your way to stand your point) and again I have the feeling that you sound somewhat rude. Could be just me. Maybe I sound rude to you too and it's just a cultural difference, in which case I apologize for that too, as I don't mean to be rude towards anyone in this community.

My rudeness scales to just how much certain groups of people have drawn my ire, and I am quite done with people coming in and whining at Michael about his English, particularly when they always focus on things where they're in the wrong instead of Michael instead of pointing out even any of his legitimate errors. Michael very often writes typos or or otherwise screws things up but you people latch onto the most inane things such as title casing, and think it's somehow wrong while completely ignoring his legitimate issues.

About the contents of what you just wrote in this thread, though, I hope no one learning the English language ever read what you just wrote and think that is appropriate use of the language for anything but an informal conversation.

And again, you're showing your lack of knowledge of English and in fact exposing the Dunning-Krueger effect in action by thinking what I wrote is wrong, it's entirely correct in any context because those are what those terms are and how they are used, regardless of whether you as a non native speaker think they're correct or not. The entire difference between slow/fast vs short/long is Object Relative vs Speaker Relative and Object Absolute respectively. We (the reader) do not know that Go's compile times are short, all we know is that it takes much longer to compile code than previous.

Being a native speaker is not a free pass to publish wrongly phrased statements without deserving at least a little criticizing now and then. I don't usually do that but also don't see anything wrong with it. It's just your opinion that that should never be done. You can try to enforce your opinions if you want, that doesn't mean anyone will obey you.

Being a native speaker means that 99 times out of 100 he will be using correct grammar, structure, and phraseology, and that if you think as a non-native speaker that he's wrong, that you need to check your own understanding. What a native speaker will mess up on frequently is spelling and punctuation, which primarily boils down to carelessness, although commas for example aren't really standardized.

Once again however, this is a tech blog (which Michael has readily stated in the past when he's done things like post articles about beer, and certainly he cannot be considered a journalist, because this site rates as a tabloid at best), we are here to discuss tech. Michael is not going to correct articles or titles because it's not worth his time doing so. Hell... he barely participates in this forum at all. So as a result all you are doing is derailing threads.

The title given to this article states the opposite of it's contents, so I thought it was well deserving of criticism. You did not deny it, so I'll assume we agree on at least that.

The title is ambiguous and not entirely clear so I saw how you could derive such an idea. However your "fix" does nothing at all to correct the underlying issue, and what should be done instead is simply to not use that ambiguous speech pattern.

The correct interpretation is: You are getting better code for slower compile time.
Your interpretation was: You are getting better code and faster compile time.

A much better way to write this would be:
Go 1.7 is generating better code at the cost of much longer compile times.
or
Go 1.7 compiles programs much slower, but generates better code

[jntesteves]

A much better way to write this would be:
Go 1.7 is generating better code at the cost of much longer compile times.
or
Go 1.7 compiles programs much slower, but generates better code

There! You finally got it right! TY! No need to argue here anymore as you just showed the right way to write it, and still you wrote a whole essay earlier trying to deny it. Get a life Luke, stop following me around in this forum and please don't answer my posts anymore.