3 ms·

Hey, so this would be my job, I guess. The reason I don't bother writing a separate change log is that I'd end up duplicating what the 5 changes on that page (

by enneff 13y ago

Hey, so this would be my job, I guess. The reason I don't bother writing a separate change log is that I'd end up duplicating what the 5 changes on that page (above 1.2, the previous release) already say. All are subtle things that are of little meaning to most people, so I figure those that really care are going to read the changes anyway. Maybe I should revisit this next minor release. I should note that there is WAY more in that Firefox point release than in any Go point release. We strictly include only critical fixes for issues for which there is no workaround. For Go point releases, most people should just install them and move on. For 99.9% of Go programmers there is "nothing to see here". As has been already mentioned, we do provide a detailed change log for major releases: http://golang.org/doc/go1.2 http://golang.org/doc/go1.2 http://golang.org/doc/go1.1 http://golang.org/doc/go1.1

simonsarris 13y ago
I suppose that's a fair response, but at the risk of being rude please consider some advice from a marketing perspective. > All are subtle things that are of little meaning to most people... If the changes do have little to no meaning in themselves, it's important to consider: Why did you submit it, and why does it sit at the top of HN? HN didn't upvote it because of the changelog itself. There's nothing there. HN upvoted it because HN wants to know more about what's going on with Go. So what's really happened, is that HN is affording you a great opportunity here, but the contents of the post are sorta clutching defeat from the jaws of victory. I think we could both agree that Go is a great language that we want people to be excited about, and releases can make good front-of-mind advertising at the very least. If making the release announcements is part of your job then I think you should take it on yourself to get the prospective audience excited. Tell us everything that's happened since 1.2, including reminding us of any major advances surrounding go. Here's what people want to hear: The Go team is plugging along. Go is getting more dependable than ever. Here's what we're doing with Go, by the way. Here's the latest projects. Here's the latest tools. Don't just say the changes since 1.2, tell us everything in your ecosystem that's changed. Teach us to long for the immensity of the sea, as it were[1]. [1] https://gist.github.com/simonsarris/9318374 https://gist.github.com/simonsarris/9318374
- enneff 13y ago
  My goal with the mailing list post and the submission was to let as many Go users as possible that a new release is out and they should upgrade. Should I use a minor release announcement as an occasion to get people more excited about Go? I'm not sure. I like the short and sweet nature of these point releases, and I'm wary of turning them into a marketing exercise. Otherwise, I do spend a lot of time telling people what's going on with Go. For instance, my presentation on what's new in Go 1.3 was at the top of HN for a day. Appreciate the feedback, though.
  - bsdetector 13y ago
    So how is this different from any other spammer? You're get paid to use Hacker News as a Go mouthpiece to promote your product every chance you can take. The reason why you, a Google employee working on Go, shouldn't be submitting these stories is because you have a clear conflict of interest. You think that 5 changes totaling a few dozen lines of code is worth everybody's attention because you're in the thick of it. But really, a story like this being front page I think calls into question whether there is a voting ring or voting bot; I can't imagine more than a person or two actually looked at the changes and said this is important.
    - enneff 13y ago
      No need to be so cynical. You may not be interested in Go, but many programmers are, and many of them read HN (and vote). You underestimate the size and enthusiasm of the Go community.
      - endersshadow 13y ago
        Right, so please add some summary of changes/content to this. I don't have time to go through SVN commit logs, but I'm interested in Go and use it in some side projects. I want to quickly be able to say, "What changed? Can I apply it to my projects?" Not having that is wildly frustrating when I see these "version releases" where I have no clue what's worth my time.
    - 13y ago
hrjet 13y ago
> We strictly include only critical fixes for issues for which there is no workaround. For Go point releases, most people should just install them and move on. For 99.9% of Go programmers there is "nothing to see here". Isn't that a little contradictory? You are saying all the changes in this release are critical, yet there is "nothing to see here". As I understand it, these changes are critical corner cases but you believe that they are not experienced by 99.9% of the users. However, you should consider two things: 1. Such critical issues might be useful information for users who are not facing those issue currently (but might hit them in the future). 2. Your estimation of 99.9% might be wrong, because not all users might have complained loudly or even realized that there is an issue.
- enneff 13y ago
  It seems contradictory in this case, but not when you consider all the previous releases. With previous releases there has been what I call a "point-release-triggering fix". That is, a fix that simply cannot wait until the next release (for example, one that fixes a security hole). In those cases there is some advisory warning in the announcement. In this case, it had been 3 months since the last release and a few point-release-worthy (but not point-release-triggering) fixes had accumulated, so we though it's probably worth getting them out there before 1.3 (in another 3 months' time). I have considered your points. That's why (in response to point 1) I have linked to the changes directly from the release announcement and acknowledged in my message that perhaps I should write the changes up in future, and (in response to point 2) I issued the point release in the first place and submitted the announcement post here so that Go users would know to upgrade.
- jussij 13y ago
  > Isn't that a little contradictory? You are saying all the changes in this release are critical, yet there is "nothing to see here". To quote T. B. Lance 'if it ain't broke don't fix it'
jerf 13y ago
An automated process to take https://code.google.com/p/go/source/list?name=release-branch.go1.2&r=7ada9e760ce34e78aee5b476c9621556d0fa5d31 https://code.google.com/p/go/source/list?name=release-branch... and output it into a textual list that has the full title of the bug fix would be a huge improvement. There's a lot of noise in that table, along with some things being cut off even on my desktop monitor (to say nothing of when I was trying to read it on my phone this morning), and it's hard to skim visually for interesting things.
jksmith 13y ago
Yeah I was hoping for some blogging goodness but for a point release: >For 99.9% of Go programmers there is "nothing to see here". That's really correct I bet. I scanned the changes for stuff I'm interested in, got what I needed. Back to work and thanks Go team for your work.

by enneff 13y ago

simonsarris 13y ago

I suppose that's a fair response, but at the risk of being rude please consider some advice from a marketing perspective. > All are subtle things that are of little meaning to most people... If the changes do have little to no meaning in themselves, it's important to consider: Why did you submit it, and why does it sit at the top of HN? HN didn't upvote it because of the changelog itself. There's nothing there. HN upvoted it because HN wants to know more about what's going on with Go. So what's really happened, is that HN is affording you a great opportunity here, but the contents of the post are sorta clutching defeat from the jaws of victory. I think we could both agree that Go is a great language that we want people to be excited about, and releases can make good front-of-mind advertising at the very least. If making the release announcements is part of your job then I think you should take it on yourself to get the prospective audience excited. Tell us everything that's happened since 1.2, including reminding us of any major advances surrounding go. Here's what people want to hear: The Go team is plugging along. Go is getting more dependable than ever. Here's what we're doing with Go, by the way. Here's the latest projects. Here's the latest tools. Don't just say the changes since 1.2, tell us everything in your ecosystem that's changed. Teach us to long for the immensity of the sea, as it were[1]. [1] https://gist.github.com/simonsarris/9318374 https://gist.github.com/simonsarris/9318374

enneff 13y ago
My goal with the mailing list post and the submission was to let as many Go users as possible that a new release is out and they should upgrade. Should I use a minor release announcement as an occasion to get people more excited about Go? I'm not sure. I like the short and sweet nature of these point releases, and I'm wary of turning them into a marketing exercise. Otherwise, I do spend a lot of time telling people what's going on with Go. For instance, my presentation on what's new in Go 1.3 was at the top of HN for a day. Appreciate the feedback, though.
- bsdetector 13y ago
  So how is this different from any other spammer? You're get paid to use Hacker News as a Go mouthpiece to promote your product every chance you can take. The reason why you, a Google employee working on Go, shouldn't be submitting these stories is because you have a clear conflict of interest. You think that 5 changes totaling a few dozen lines of code is worth everybody's attention because you're in the thick of it. But really, a story like this being front page I think calls into question whether there is a voting ring or voting bot; I can't imagine more than a person or two actually looked at the changes and said this is important.
  - enneff 13y ago
    No need to be so cynical. You may not be interested in Go, but many programmers are, and many of them read HN (and vote). You underestimate the size and enthusiasm of the Go community.
    - endersshadow 13y ago
      Right, so please add some summary of changes/content to this. I don't have time to go through SVN commit logs, but I'm interested in Go and use it in some side projects. I want to quickly be able to say, "What changed? Can I apply it to my projects?" Not having that is wildly frustrating when I see these "version releases" where I have no clue what's worth my time.
  - 13y ago

hrjet 13y ago

> We strictly include only critical fixes for issues for which there is no workaround. For Go point releases, most people should just install them and move on. For 99.9% of Go programmers there is "nothing to see here". Isn't that a little contradictory? You are saying all the changes in this release are critical, yet there is "nothing to see here". As I understand it, these changes are critical corner cases but you believe that they are not experienced by 99.9% of the users. However, you should consider two things: 1. Such critical issues might be useful information for users who are not facing those issue currently (but might hit them in the future). 2. Your estimation of 99.9% might be wrong, because not all users might have complained loudly or even realized that there is an issue.

enneff 13y ago
It seems contradictory in this case, but not when you consider all the previous releases. With previous releases there has been what I call a "point-release-triggering fix". That is, a fix that simply cannot wait until the next release (for example, one that fixes a security hole). In those cases there is some advisory warning in the announcement. In this case, it had been 3 months since the last release and a few point-release-worthy (but not point-release-triggering) fixes had accumulated, so we though it's probably worth getting them out there before 1.3 (in another 3 months' time). I have considered your points. That's why (in response to point 1) I have linked to the changes directly from the release announcement and acknowledged in my message that perhaps I should write the changes up in future, and (in response to point 2) I issued the point release in the first place and submitted the announcement post here so that Go users would know to upgrade.
jussij 13y ago
> Isn't that a little contradictory? You are saying all the changes in this release are critical, yet there is "nothing to see here". To quote T. B. Lance 'if it ain't broke don't fix it'

jerf 13y ago

An automated process to take https://code.google.com/p/go/source/list?name=release-branch.go1.2&r=7ada9e760ce34e78aee5b476c9621556d0fa5d31 https://code.google.com/p/go/source/list?name=release-branch... and output it into a textual list that has the full title of the bug fix would be a huge improvement. There's a lot of noise in that table, along with some things being cut off even on my desktop monitor (to say nothing of when I was trying to read it on my phone this morning), and it's hard to skim visually for interesting things.

jksmith 13y ago

Yeah I was hoping for some blogging goodness but for a point release: >For 99.9% of Go programmers there is "nothing to see here". That's really correct I bet. I scanned the changes for stuff I'm interested in, got what I needed. Back to work and thanks Go team for your work.