4 ms·

Bram Cohen: Comments on Go

65 points by jeff18 17y ago

andrewljohnson 17y ago
I really wish people wouldn't include the name of the author in the title. I think this should be banned and cleaned up by the moderators in the same way using numbers in the titles of submissions is, and the same way that moderators removed "Ohhhh snapppp" from the Go post about the author of Go! Invariably, Paul Bucheit, Paul Graham, Bram Cohen, Guido van Rossum, Joel Spolsky, and other famous hackers get voted to the very top of the list. If these guys are great writers and great thinkers, then their articles will speak for themselves, and they usually do (but not always). For the few people who have never heard of Bram Cohen or Joe Hewitt (currently 1 and 2 on HN), then their names are just noise. For those who have heard of them, it just biases their judgment in both clicking and voting. In this case, Bram's blog even carries his name... so why make his name the first two words in the headline? Do we really need to make to use this kind of name-dropping to highlight content? That sounds like the opposite of content democratization to me.
- bootload 17y ago
  "... really wish people wouldn't include the name of the author in the title. ..." I'd agree but only because the name is usually redundant when the domain name is shown.
- rms 17y ago
  Why is biased judgement of voting and clicking bad? That should give a list of links that more closely reflects the likes of the readers of the site. In this case, I agree it is unnecessary and redundant because the blog already carries the author's name.
- jeff18 17y ago
  Nice rant, however, I submitted this using the YC bookmarklet. :) No subtle manipulation of Hacker News intended. Hopefully democracy will still be intact after this serious infraction.
  - andrewljohnson 17y ago
    I didn't mean to impugn your motives... just saying I don't think it's a good thing to do in general. Also, I'm not familiar with the bookmarklet. Does that auto-generate the titles or something?
    - vorador 17y ago
      No, you can edit the title of the submission like for a regular submission.
  - jacquesm 17y ago
    You can still edit the title, it's easy.
- efaith 17y ago
  And should publishers also remove the names of the authors from the covers of their books, on the theory that they don't want to "bias" potential readers? How do you imagine this working out? In a bookstore, if books don't have the authors on the covers, then if I read a book I like, I will have trouble finding other books by the same author. This will doubtlessly counteract my "bias" in favor of that author, but is that a good thing? It's a good thing if your greatest desire is to dethrone Stephen King and Dan Brown and J. K. Rowling by making it hard for their fans to find their books, forcing them to read books by authors who may or may not be enjoyable for them to read. But if you care at all about maximizing the ability of people to find stuff they like to read, then the name of the author is an important bit of evidence that they can use for that purpose and you would be depriving them of it.
- mullr 17y ago
  In general I agree, but in this case it's really helpful for disambiguation from the dozens of other "what I think about Go" posts that people have been writing. Incidentally, I agree with him w.r.t. lack of exceptions being a big barrier. Everything else is cool / live-withable.
  - andrewljohnson 17y ago
    That's just the point... it's not helpful for disambiguation. If the article was great, then it would be disambiguated by votes!
    - jacquesm 17y ago
      Well, by that standard it seems that the majority of HN readers think it is great, it is currently #1 on the homepage. If only the number of votes would be a true measure of 'greatness', then I could stop reading HN on the 'new' page. That's where the gems come by and plenty of them are lost.
      - andrewljohnson 17y ago
        Jeez, I feel like I'm talking to a wall here. I'm arguing that the reason that it has so many votes could be because his name is in the title. I bet if you showed 200 people the article, and told 100 of them Bram wrote it, then the 100 who knew would score the article higher. Is that a good thing? Leave the names out, and let the content decide.
        jacquesm 17y ago
        > Jeez, I feel like I'm talking to a wall here. Sorry for being dense, it's early :) > I'm arguing that the reason that it has so many votes could be because his name is in the title. Yes, I got that, it's just that because it is also in the domain name in this case I'd expect it to be just as high right now if the name had not been in the title. > I bet if you showed 200 people the article, and told 100 of them Bram wrote it, then the 100 who knew would score the article higher. Is that a good thing? No, it isn't. But that would mean that you'd also have to drop the domain name from behind the title. Personally I don't mind, it helps me to avoid some of the more overexposed sites here.
- jacquesm 17y ago
  I'm ok with it if the name of the author is not already in the domain name. It helps to spend your time useful. Otherwise (like in this case) it is redundant.
- mjs 17y ago
  I think the name was meaningful in this case, and that this is true in most cases. I'm not especially interested in a random person's thoughts on Go. I may be interested in their thoughts on Go if their post happens to make it to the top of YC: this probably indicates that they have something interesting to say. And I'm also interested in their thoughts on Go if they're someone who is a recognised as an authority in the field, as is the case with Bram Cohen. Or: I might be interested if they have a history of writing good posts. So names in titles do usually convey useful information for me. (Even if I don't recognise it, the fact that the submitter used a name is informative.)
- scott_s 17y ago
  It makes sense as the title; it provides context. On Bram's site, using the title "Bram Cohen: Comments on Go" would be redundant. If you're already reading his blog, then you know every entry is by him. When his entry is mixed in with the rest of the internet, his title has no context. Putting his name in the title provides the context.
jacquesm 17y ago
On topic: I find it funny that someone of Bram Cohens' standing would miss the point of fast compilation. If you're working in plenty of other languages compile time does not go up linear with code size, but much faster. When you start out it zips right along, but as soon as you cross that 10,000 line mark the first bits of irritation set it in, and not much later than that you can go and get a cup of coffee and drink it while your code rebuilds. For really large projects an hour isn't much at all! Fast compile time = I won't lose my train of thought while waiting for the compiler to finish. It's a 'tool' thing, and I think that such details make for much better tools.
- andrewljohnson 17y ago
  Well, I guess I'll also jump on topic here :) I used to primarily write Javascript and Python code, but then I moved into the world of Objective-C and compilation. I also think that compiling is a productivity and focus killer, and I think addressing this issue is a worthy goal.
- klipt 17y ago
  "Go's speed of compilation is very nice, although I'm afraid I view that not so much as a strength of Go but as an awfulness of C++." I think his point is that Go isn't special in this regard since many languages already compile much faster than C++. For example, Delphi/Free Pascal.
  - jacquesm 17y ago
    Sure, but Go is aiming for the space that C/C++ occupy, not for the space that Delphi and Free Pascal occupy. And neither Delphi nor Free Pascal look like they are going to be threatening C/C++ any time soon, this new language just might.
    - klipt 17y ago
      Free Pascal is almost identical to C++ apart from syntax; they're both compiled, statically typed Algol descendants with object orientation and generics bolted on. Which is partly why Free Pascal use is still marginal - the only significant advantage is compilation speed, while lack of popularity is a major disadvantage in terms of libraries, quality of optimizing compilers, etc.

65 points by jeff18 17y ago

I really wish people wouldn't include the name of the author in the title. I think this should be banned and cleaned up by the moderators in the same way using numbers in the titles of submissions is, and the same way that moderators removed "Ohhhh snapppp" from the Go post about the author of Go! Invariably, Paul Bucheit, Paul Graham, Bram Cohen, Guido van Rossum, Joel Spolsky, and other famous hackers get voted to the very top of the list. If these guys are great writers and great thinkers, then their articles will speak for themselves, and they usually do (but not always). For the few people who have never heard of Bram Cohen or Joe Hewitt (currently 1 and 2 on HN), then their names are just noise. For those who have heard of them, it just biases their judgment in both clicking and voting. In this case, Bram's blog even carries his name... so why make his name the first two words in the headline? Do we really need to make to use this kind of name-dropping to highlight content? That sounds like the opposite of content democratization to me.

bootload 17y ago
"... really wish people wouldn't include the name of the author in the title. ..." I'd agree but only because the name is usually redundant when the domain name is shown.
rms 17y ago
Why is biased judgement of voting and clicking bad? That should give a list of links that more closely reflects the likes of the readers of the site. In this case, I agree it is unnecessary and redundant because the blog already carries the author's name.
jeff18 17y ago
Nice rant, however, I submitted this using the YC bookmarklet. :) No subtle manipulation of Hacker News intended. Hopefully democracy will still be intact after this serious infraction.
- andrewljohnson 17y ago
  I didn't mean to impugn your motives... just saying I don't think it's a good thing to do in general. Also, I'm not familiar with the bookmarklet. Does that auto-generate the titles or something?
  - vorador 17y ago
    No, you can edit the title of the submission like for a regular submission.
- jacquesm 17y ago
  You can still edit the title, it's easy.
efaith 17y ago
And should publishers also remove the names of the authors from the covers of their books, on the theory that they don't want to "bias" potential readers? How do you imagine this working out? In a bookstore, if books don't have the authors on the covers, then if I read a book I like, I will have trouble finding other books by the same author. This will doubtlessly counteract my "bias" in favor of that author, but is that a good thing? It's a good thing if your greatest desire is to dethrone Stephen King and Dan Brown and J. K. Rowling by making it hard for their fans to find their books, forcing them to read books by authors who may or may not be enjoyable for them to read. But if you care at all about maximizing the ability of people to find stuff they like to read, then the name of the author is an important bit of evidence that they can use for that purpose and you would be depriving them of it.
mullr 17y ago
In general I agree, but in this case it's really helpful for disambiguation from the dozens of other "what I think about Go" posts that people have been writing. Incidentally, I agree with him w.r.t. lack of exceptions being a big barrier. Everything else is cool / live-withable.
- andrewljohnson 17y ago
  That's just the point... it's not helpful for disambiguation. If the article was great, then it would be disambiguated by votes!
  - jacquesm 17y ago
    Well, by that standard it seems that the majority of HN readers think it is great, it is currently #1 on the homepage. If only the number of votes would be a true measure of 'greatness', then I could stop reading HN on the 'new' page. That's where the gems come by and plenty of them are lost.
    - andrewljohnson 17y ago
      Jeez, I feel like I'm talking to a wall here. I'm arguing that the reason that it has so many votes could be because his name is in the title. I bet if you showed 200 people the article, and told 100 of them Bram wrote it, then the 100 who knew would score the article higher. Is that a good thing? Leave the names out, and let the content decide.
      - jacquesm 17y ago
        > Jeez, I feel like I'm talking to a wall here. Sorry for being dense, it's early :) > I'm arguing that the reason that it has so many votes could be because his name is in the title. Yes, I got that, it's just that because it is also in the domain name in this case I'd expect it to be just as high right now if the name had not been in the title. > I bet if you showed 200 people the article, and told 100 of them Bram wrote it, then the 100 who knew would score the article higher. Is that a good thing? No, it isn't. But that would mean that you'd also have to drop the domain name from behind the title. Personally I don't mind, it helps me to avoid some of the more overexposed sites here.
jacquesm 17y ago
I'm ok with it if the name of the author is not already in the domain name. It helps to spend your time useful. Otherwise (like in this case) it is redundant.
mjs 17y ago
I think the name was meaningful in this case, and that this is true in most cases. I'm not especially interested in a random person's thoughts on Go. I may be interested in their thoughts on Go if their post happens to make it to the top of YC: this probably indicates that they have something interesting to say. And I'm also interested in their thoughts on Go if they're someone who is a recognised as an authority in the field, as is the case with Bram Cohen. Or: I might be interested if they have a history of writing good posts. So names in titles do usually convey useful information for me. (Even if I don't recognise it, the fact that the submitter used a name is informative.)
scott_s 17y ago
It makes sense as the title; it provides context. On Bram's site, using the title "Bram Cohen: Comments on Go" would be redundant. If you're already reading his blog, then you know every entry is by him. When his entry is mixed in with the rest of the internet, his title has no context. Putting his name in the title provides the context.

jacquesm 17y ago

On topic: I find it funny that someone of Bram Cohens' standing would miss the point of fast compilation. If you're working in plenty of other languages compile time does not go up linear with code size, but much faster. When you start out it zips right along, but as soon as you cross that 10,000 line mark the first bits of irritation set it in, and not much later than that you can go and get a cup of coffee and drink it while your code rebuilds. For really large projects an hour isn't much at all! Fast compile time = I won't lose my train of thought while waiting for the compiler to finish. It's a 'tool' thing, and I think that such details make for much better tools.

andrewljohnson 17y ago
Well, I guess I'll also jump on topic here :) I used to primarily write Javascript and Python code, but then I moved into the world of Objective-C and compilation. I also think that compiling is a productivity and focus killer, and I think addressing this issue is a worthy goal.
klipt 17y ago
"Go's speed of compilation is very nice, although I'm afraid I view that not so much as a strength of Go but as an awfulness of C++." I think his point is that Go isn't special in this regard since many languages already compile much faster than C++. For example, Delphi/Free Pascal.
- jacquesm 17y ago
  Sure, but Go is aiming for the space that C/C++ occupy, not for the space that Delphi and Free Pascal occupy. And neither Delphi nor Free Pascal look like they are going to be threatening C/C++ any time soon, this new language just might.
  - klipt 17y ago
    Free Pascal is almost identical to C++ apart from syntax; they're both compiled, statically typed Algol descendants with object orientation and generics bolted on. Which is partly why Free Pascal use is still marginal - the only significant advantage is compilation speed, while lack of popularity is a major disadvantage in terms of libraries, quality of optimizing compilers, etc.