on being and zealousness

On Fri, Jan 15, 2010 at 10:15 AM, Paul Phillips <paulp@improving.org> wrote:

I would like to comment on this commit message

 "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email, and that's why I'm writing this: I don't
like the idea that the visible historical record might give someone the
impression that I'm some kind of VCS loose cannon.

I gurarantee nobody tests their patches more thoroughly and nobody
throws away more code.  You can come look through my trash if you don't
believe me! Whenever I delete a branch I first print the diffs and run
them through the shredder so as to experience the loss more viscerally.

Summary: I am zealous, no doubt.  Zeal is my constant companion.  But I
keep him on a tight leash.

<my_2_cents>
As an external observer, the quality of Scala (both the visceral feel of looking at the code and the my actual experience running the code) has materially improved as a result of Paul's efforts.  Plus, for every bug Paul fixes, it's time that folks at EPFL can spend doing not bug fixing, so keeping Paul on the job helps move Scala the language forward.  Finally, for every cycle of waste heat for undoing something not-obviously-wrong that Paul's done turns into 10 lost cycles of work for the team as a whole.
</my_2_cents>
 

--
Paul Phillips      | We must respect the other fellow's religion, but only
Stickler           | in the sense and to the extent that we respect his
Empiricist         | theory that his wife is beautiful and his children smart.
i'll ship a pulp   |     -- H. L. Mencken

--
Lift, the simply functional web framework http://liftweb.net
Beginning Scala http://www.apress.com/book/view/1430219890
Follow me: http://twitter.com/dpp
Surf the harmonics

On 01/15/2010 10:15 AM, Paul Phillips wrote:
> I would like to comment on this commit message
>
> "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."
>
> and I should preface by saying I am not offended. However it seems to
> me that I may have an undeserved reptutation among at least a few people
> for making inadequately considered changes. In the majority of these
> situations where a commit of mine is later reverted, I had martin's
> specific approval beforehand. Very often (as in this case) I obtained
> the approval via private email,

Aside from your reputation, which I have in high regards, the core issue
is having private conversations. If it were public, people would see
there was consensus on this idea. I would just cc scala-internals in
the future on any threads.

As Karl Fogel says who wrote "Producing Open Source Software" and one of
the founders of the Subversion project said: "Don't do it."

http://producingoss.com/en/setting-tone.html#avoid-private-discussions

I've stated this before to the Scala developers, but I'm a believer in
having as much open as possible and the more closed conversations there
are the more issues arises.

Also, it's fine to use IRC to work stuff out and to make decisions, but
if its a bigger design question or anything that needs to be recorded or
discussed in a bigger group, it is moved to the mailing list.

Blair

On Fri, January 15, 2010 7:15 pm, Paul Phillips wrote:
> I don't like the idea that the visible historical record
> might give someone the impression
>

History can be changed, at least in SVN. I will change the
log entry on Monday.

http://subversion.tigris.org/faq.html#change-log-msg

+1

On Fri, Jan 15, 2010 at 1:33 PM, Blair Zajac <blair@orcaware.com> wrote:

On 01/15/2010 10:15 AM, Paul Phillips wrote:

I would like to comment on this commit message

  "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email,

Aside from your reputation, which I have in high regards, the core issue is having private conversations.  If it were public, people would see there was consensus on this idea.  I would just cc scala-internals in the future on any threads.

As Karl Fogel says who wrote "Producing Open Source Software" and one of the founders of the Subversion project said: "Don't do it."

http://producingoss.com/en/setting-tone.html#avoid-private-discussions

I've stated this before to the Scala developers, but I'm a believer in having as much open as possible and the more closed conversations there are the more issues arises.

Also, it's fine to use IRC to work stuff out and to make decisions, but if its a bigger design question or anything that needs to be recorded or discussed in a bigger group, it is moved to the mailing list.

Blair

--
http://erikengbrecht.blogspot.com/

On 15 Ιαν 2010, at 20:35, David Pollak <feeder.of.the.bears@gmail.com> wrote:

On Fri, Jan 15, 2010 at 10:15 AM, Paul Phillips < (paulp [at] improving [dot] org> wrote:

I would like to comment on this commit message

 "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email, and that's why I'm writing this: I don't
like the idea that the visible historical record might give someone the
impression that I'm some kind of VCS loose cannon.

I gurarantee nobody tests their patches more thoroughly and nobody
throws away more code.  You can come look through my trash if you don't
believe me! Whenever I delete a branch I first print the diffs and run
them through the shredder so as to experience the loss more viscerally.

Summary: I am zealous, no doubt.  Zeal is my constant companion.  But I
keep him on a tight leash.

<my_2_cents>
As an external observer, the quality of Scala (both the visceral feel of looking at the code and the my actual experience running the code) has materially improved as a result of Paul's efforts.  Plus, for every bug Paul fixes, it's time that folks at EPFL can spend doing not bug fixing, so keeping Paul on the job helps move Scala the language forward.  Finally, for every cycle of waste heat for undoing something not-obviously-wrong that Paul's done turns into 10 lost cycles of work for the team as a whole.
</my_2_cents>

"well said" is an understatement

 

--
Paul Phillips      | We must respect the other fellow's religion, but only
Stickler           | in the sense and to the extent that we respect his
Empiricist         | theory that his wife is beautiful and his children smart.
i'll ship a pulp   |     -- H. L. Mencken

--
Lift, the simply functional web framework http://liftweb.net
Beginning Scala http://www.apress.com/book/view/1430219890
Follow me: http://twitter.com/dpp
Surf the harmonics

--Christos KK Loverdos

On Fri, Jan 15, 2010 at 2:31 PM, Erik Engbrecht <erik.engbrecht@gmail.com> wrote:

+1

-6 (because some people's votes should count more than others... although this is no reflection on how I value Erik's voteworthiness.)

The whole "everything should be out in plain sight, everybody's opinion" stuff is crap.

Martin is a better language designer than all but 10 people who have ever lived.  Giving him some latitude to practice his excellence by talking privately to others in the community without have to write a report is critical.  Giving him some opportunities to make decisions that are not "consensus" is absolutely necessary.  Martin's also a pretty busy guy and requiring that he "send a note" to the list after a private communication with Paul or Philipp or one of the other key Scala guys is not a good use of anyone's time.

When you try to democratize the pushing of creative and research boundaries (which is what is happing in this and satellite communities) will fail.  You'll wind up with skilled political players dominating rather than people who are skilled with designing systems and writing code.  You'll wind up with lowest common denominator decisions that bug fewer people rather than bold decisions that might not seem correct to people who don't understand the whole picture.  You wind up with Subversion rather than Git.  You wind up with Apache2 rather than Lighhtd or Nginx.  In both cases, the former was a result of committee and the latter was the result of someone who actually understood the problem applying themselves to addressing it (and even with Lighttd's support issues, it's worlds better than Apache2 has ever been.)

I am very happy with the level of transparency we get from EPFL.  I am very happy with the amount that EPFL listens to and takes into account what the community's needs are.  No, I don't get everything I ask for and sometimes I go off and build it myself (e.g., QBase -- using the type system to define compiler enforced immutable classes, Option vs. Box, Actor vs. Actor) and sometimes it just doesn't happen (def! vs. override def), but never ever have I felt that EPFL has turned a deaf ear to my requests.  In my 3+ years in Scala-land, I have not seen Martin or the EPFL team turn a deaf ear to coherent suggestions from the community.  Through the 2.8 redesign process, Martin and EPFL have done an admirable job communicating with the community.

So, just so I'm not misconstrued, I am not advocating for private code or siloed development.  In Lift-land, the ratio of public : committer : private communication is about 1600 : 25 : 2.  For each 1,600 messages on the main Lift list, there are about 25 messages on the private committers list and about 2 private messages between me and one or two committers.  That ratio seems about right for Lift-land and other ratios may work for other communities.  Sometimes Marius will run a design by me before he puts it out on the list... he and I can have a more direct conversation in private.  I've also occasionally coached committers as to how to present ideas in a way that will win community support in private.

So, Paul went public asked to be cut some slack.  Let's cut him some.  Let's not pile onto Martin with the "you're not doing open source the one true divine way."  If there's a problem, an actual problem that someone is experiencing themselves that they can trace to a lack of transparency, let's all talk about it.  Otherwise, let's let EPFL focus on the business of getting 2.8 out the door.

Thanks,

David

 

On Fri, Jan 15, 2010 at 1:33 PM, Blair Zajac <blair@orcaware.com> wrote:

On 01/15/2010 10:15 AM, Paul Phillips wrote:

I would like to comment on this commit message

  "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email,

Aside from your reputation, which I have in high regards, the core issue is having private conversations.  If it were public, people would see there was consensus on this idea.  I would just cc scala-internals in the future on any threads.

As Karl Fogel says who wrote "Producing Open Source Software" and one of the founders of the Subversion project said: "Don't do it."

http://producingoss.com/en/setting-tone.html#avoid-private-discussions

I've stated this before to the Scala developers, but I'm a believer in having as much open as possible and the more closed conversations there are the more issues arises.

Also, it's fine to use IRC to work stuff out and to make decisions, but if its a bigger design question or anything that needs to be recorded or discussed in a bigger group, it is moved to the mailing list.

Blair

--
http://erikengbrecht.blogspot.com/

--
Lift, the simply functional web framework http://liftweb.net
Beginning Scala http://www.apress.com/book/view/1430219890
Follow me: http://twitter.com/dpp
Surf the harmonics

Ooooh, I like =>?


And back on topic, I'm with you on this one Paul.

On Fri, Jan 15, 2010 at 7:15 PM, Paul Phillips <paulp@improving.org> wrote:

I would like to comment on this commit message

 "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email, and that's why I'm writing this: I don't
like the idea that the visible historical record might give someone the
impression that I'm some kind of VCS loose cannon.

I gurarantee nobody tests their patches more thoroughly and nobody
throws away more code.  You can come look through my trash if you don't
believe me! Whenever I delete a branch I first print the diffs and run
them through the shredder so as to experience the loss more viscerally.

Summary: I am zealous, no doubt.  Zeal is my constant companion.  But I
keep him on a tight leash.

--
Paul Phillips      | We must respect the other fellow's religion, but only
Stickler           | in the sense and to the extent that we respect his
Empiricist         | theory that his wife is beautiful and his children smart.
i'll ship a pulp   |     -- H. L. Mencken

--
Viktor Klang
| "A complex system that works is invariably
| found to have evolved from a simple system
| that worked." - John Gall

Blog: klangism.blogspot.com
Twttr: twitter.com/viktorklang
Code: github.com/viktorklang

I so like it that I used it the day I saw it in a commit message. Sad that it got reverted. PartialFunction really is verbose.

On Fri, Jan 15, 2010 at 10:39 PM, Viktor Klang <viktor.klang@gmail.com> wrote:

Ooooh, I like =>?


And back on topic, I'm with you on this one Paul.

On Fri, Jan 15, 2010 at 7:15 PM, Paul Phillips <paulp@improving.org> wrote:

I would like to comment on this commit message

 "Reverted over-zealous replacement of 'PartialFunction' with '=>?'."

and I should preface by saying I am not offended.  However it seems to
me that I may have an undeserved reptutation among at least a few people
for making inadequately considered changes.  In the majority of these
situations where a commit of mine is later reverted, I had martin's
specific approval beforehand.  Very often (as in this case) I obtained
the approval via private email, and that's why I'm writing this: I don't
like the idea that the visible historical record might give someone the
impression that I'm some kind of VCS loose cannon.

I gurarantee nobody tests their patches more thoroughly and nobody
throws away more code.  You can come look through my trash if you don't
believe me! Whenever I delete a branch I first print the diffs and run
them through the shredder so as to experience the loss more viscerally.

Summary: I am zealous, no doubt.  Zeal is my constant companion.  But I
keep him on a tight leash.

--
Paul Phillips      | We must respect the other fellow's religion, but only
Stickler           | in the sense and to the extent that we respect his
Empiricist         | theory that his wife is beautiful and his children smart.
i'll ship a pulp   |     -- H. L. Mencken

--
Viktor Klang
| "A complex system that works is invariably
| found to have evolved from a simple system
| that worked." - John Gall

Blog: klangism.blogspot.com
Twttr: twitter.com/viktorklang
Code: github.com/viktorklang

--
Daniel C. Sobral

I travel to the future all the time.

Viktor Klang wrote:
> Ooooh, I like =>?
>
Please see #2860.

>
> And back on topic, I'm with you on this one Paul.
>

Now, I really didn't want to get involved in this matter, but is there
really a confrontation between sides somewhere? Because if I happen to be
one, I swear I didn't know it.

It may have been a poor choice of wording; I could have used a different
word, or written nothing at all. There was no intent whatsoever to diss
Paul. Frankly this whole discussion seems to me utterly ridiculous.

There is nothing to see here. Please move along.

Toni

Indeed, I didn't intend to make a thing of it and I'm sorry if it is
even in the ballpark of thinghood. I was not out to criticize anyone:
like I said in the opener, I took no offense. Only wanted to go on
record that I'm not a complete nut. ("Let the record show extempore is
not a complete nut.")

On Sat, Jan 16, 2010 at 2:15 AM, Antonio Cunei <antonio.cunei@epfl.ch> wrote:

Viktor Klang wrote:

Ooooh, I like =>?

Please see #2860.

And back on topic, I'm with you on this one Paul.

Now, I really didn't want to get involved in this matter, but is there really a confrontation between sides somewhere? Because if I happen to be one, I swear I didn't know it.

It was more like "I understand you Paul, and I think you have a good point here."
 

It may have been a poor choice of wording; I could have used a different word, or written nothing at all. There was no intent whatsoever to diss Paul. Frankly this whole discussion seems to me utterly ridiculous.

There is nothing to see here. Please move along.

Toni

--
Viktor Klang
| "A complex system that works is invariably
| found to have evolved from a simple system
| that worked." - John Gall

Blog: klangism.blogspot.com
Twttr: twitter.com/viktorklang
Code: github.com/viktorklang

Hey all,

On Fri, 2010-01-15 at 16:24 -0800, David Pollak wrote:
> Martin's also a pretty busy guy and requiring that he "send a note" to
> the list after a private communication with Paul or Philipp or one of
> the other key Scala guys is not a good use of anyone's time.

I am late to this conversation, but I would like to add a minor comment.
I agree that we should not _require_ Martin or anyone else to send a
note to the list after a smallish[1] decision has been made during a
private conversation (that happened for whatever reason). Having said
that, I would suggest that the fact that an approval was given is
important and one should aim to record it in the commit message, as it
prevents potential misunderstandings (like Paul being a nut ;)). I
should note that I don't remember if Paul did this for the commits in
question or not.

> So, just so I'm not misconstrued, I am not advocating for private code
> or siloed development. In Lift-land, the ratio of public :
> committer : private communication is about 1600 : 25 : 2. For each
> 1,600 messages on the main Lift list, there are about 25 messages on
> the private committers list and about 2 private messages between me
> and one or two committers. That ratio seems about right for Lift-land
> and other ratios may work for other communities.

That seems like a good ratio and it would fit my own opinion of what
"avoiding private conversations" should mean.

Ismael

[1] I would hope that we all agree that decisions that have a medium or
large impact should be publicly documented. I think a good job has been
done on that with the various SIDs.