[ruby-core:21039] Happy new year and... moving Ruby development to Git?

On Sun, Jan 4, 2009 at 12:14, Joel VanderWerf <vjoel@...> wrote:

And now we can compare ruby with perl instead of linux:

Yes, we should definitely take advice on working practices from the people developing Perl 6...


mathew

[evolving_jerk]

On 04.01.2009, at 21:53, mathew wrote:

> Yes, we should definitely take advice on working practices from the  
> people developing Perl 6...


If you read the post, it was Perl 5 that moved to Git.

MK

[Robert Dober]

On Fri, Jan 2, 2009 at 4:18 PM, Florian Gilcher <flo@...> wrote:

> In addition to that, I want to mention that Windows support of GIT is
> still sub-standard. I'm using the cygwin-build on a daily basis[1] and can
> speak out of personal experience.
I never had any issues with Mercurial between my Windoze and Linux
boxes. I do not know, however how Mercurial acts under high load.
Cheers
Robert



--
It is change, continuing change, inevitable change, that is the
dominant factor in society today. No sensible decision can be made any
longer without taking into account not only the world as it is, but
the world as it will be ... ~ Isaac Asimov

[Nikolai Weibull-11]

On Sun, Jan 4, 2009 at 20:08, Michael Klishin
<michael.s.klishin@...> wrote:

> On 04.01.2009, at 21:53, mathew wrote:

>> Yes, we should definitely take advice on working practices from the people
>> developing Perl 6...

> If you read the post, it was Perl 5 that moved to Git.

I'd say, with the same conviction that everyone else posting on this
topic seems to have, that this is proof positive that no one actually
reads what everyone else is writing on this topic and everyone is
simply spewing out their own opinion on the matter and claim that
anyone who doesn't agree is wrong and is therefore a moron (sometimes
cleverly hiding this statement behind sarcasm).

["Martin J. Dürst"]

First, happy New Year to everybody.

Second, despite what some of the posters have said on this thread,
at least I have read it carefully, and learned quite a bit about git
and other systems.

Third, I hope that we can end this thread soon, because it doesn't
produce much more enlightment anymore. Matz already said in
[ruby-core:21068] that Yugui is looking at the issue of how to
port the necessary tools. Even if Ruby doesn't do as complete a
job as Perl with regards to integrating history
(see http://developers.slashdot.org/article.pl?sid=09/01/04/1610256),
the change may also take quite a bit of time (it was one year for
Perl, according to the above article).

Forth, from my current understanding of git, I wouldn't mind using
git for Ruby development. For the way I work for Ruby, the careful
staging facilities in git would provide some help. (I have to admit
that I have zero actual experience with anything except cvs and svn).
However, I can easily live with subversion for the time being.

Fifth, I'm personally interested in how all these systems perform
for 'projects' that combine very, very loosly related stuff
(e.g. a huge Web site with high independence for each subpart,
or a 'project' that just serves as a backup repository for
the students in a lab or research group). Because this is
not relevant for Ruby, please send me replies off-list
(to duerst@...). If requested, I can send a summary
to the list.

Regards,    Martin.


#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst@...    

[Florian Gilcher]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On Jan 2, 2009, at 5:59 PM, Luis Lavena wrote:

> On Fri, Jan 2, 2009 at 12:18 PM, Florian Gilcher  
> <flo@...> wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>>
>> On Jan 2, 2009, at 2:32 PM, Yukihiro Matsumoto wrote:
>>
>>> Hi,
>>>
>>> We already set up quite a few tools around Subversion, so we need to
>>> port them to support git.  Yugui (who uses git herself) is in charge
>>> of the issue.  You can contact (and offer help to) her.
>>>
>>>                                                      matz.
>>>
>>
>> In addition to that, I want to mention that Windows support of GIT is
>> still sub-standard. I'm using the cygwin-build on a daily basis[1]  
>> and can
>> speak out of personal experience.
>> Configuring everything beyond the most simple things gets arcane  
>> and hard.
>> If "ease of use" is the argument, it doesn't hold true on win.
>>
>> In the light of this, I would suggest an "official" mirror.
>>
>
> I would disagree on that topic.
>
> I use on a daily basis msysGit (1.5.5 and 1.6.0) without issues. Uses
> notepad as editor or bundled vim too.
>
> Also, for the record, configured to use ssh and plink (PuTTY) and both
> work without issues.
>
> One important thing: disable autocrlf (false) since is the worse
> default msysGit developers could have introduced.

Well, thanks for the proposed help. At the moment, it works good for  
my uses
(as a better perforce client). My point is: all this is too much.  
Subversion is:
install => use.

git takes a lot of thought to set up on windows. If it runs, it runs,  
thats true.
But it's not a tool that i would suggest any non-command-line-lover to  
use.
And Windows users usally are just that.

Also, it sometimes has the tendency of biting you in the back. In my  
first
installation, the default hook to deny commits with trailing  
whitespace was on.
Combined with the fact that it treated CRLF-endings as trailing  
whitespace,
this really messed my windows development up. Sure, thats treatable.  
But it
raises a lot of questions for the uninitiated.

Thats why I said: sub-standard. Its not entirely broken or unusable.  
It just
feels like an afterthought (might be because of the fact that the whole
windows shell feels like an afterthought).

>>
>> [1]: Yes, I also tried the msysgit, which is even worse.
>
> Again: which part is worse? It lacks things like git-serve, but I can
> live without the need to "host" git repositories in my own computer.

The constant crashes. I don't know whether it works now. I stuck with
the cygwin version.

Regards,
Florian
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAklhzXoACgkQyLKU2FMxSOIjMACgrHhz7XOlnQE2J64AuP/VV+lg
8YgAnA8bHJU2E15Gg2tw3sQSTMUTqtDV
=aE1W
- -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAklhzaMACgkQyLKU2FMxSOJObQCdH5z952HwYiHA5e6GjQus+3Dv
Q20AoI+yd7qR0DUS1Uu6bGrx+TK7RtTU
=K4h/
-----END PGP SIGNATURE-----

On Mon, Jan 5, 2009 at 2:28 AM, Martin Duerst <duerst@...> wrote:
> First, happy New Year to everybody.

And to you too.

> Fifth, I'm personally interested in how all these systems perform
> for 'projects' that combine very, very loosly related stuff
> (e.g. a huge Web site with high independence for each subpart,
> or a 'project' that just serves as a backup repository for
> the students in a lab or research group). Because this is
> not relevant for Ruby, please send me replies off-list
> (to duerst@...). If requested, I can send a summary
> to the list.

The main difference between svn and git wrt to loosely related stuff
is that in git you cannot checkout a part of a repository: you always
grab the whole thing when cloning a repo, and you always check it all
out. So, for loosely related stuff you would usually have a separate
git repo for each component of this large project, and then maybe use
submodules to bind them together.

--
Giuseppe "Oblomov" Bilotta

[Meinrad Recheis]

On Mon, Jan 5, 2009 at 10:06 AM, Florian Gilcher <flo@...> wrote:

[...]

[1]: Yes, I also tried the msysgit, which is even worse.

Again: which part is worse? It lacks things like git-serve, but I can
live without the need to "host" git repositories in my own computer.

The constant crashes. I don't know whether it works now. I stuck with
the cygwin version.

I played around with msysGit 1.5 and 1.6 yesterday on winXP and had a very bad experience. 'git pull' ends up in an endless loop asking for the passphrase which makes it completely unusable.

The cygwin git packages work like a charm. 

Both msysGit and cygwinGit come with a tk gui which is helpful for visualization of branches and all but it is not quite user friendly IMO.

-- henon

[Luis Lavena]

On Mon, Jan 5, 2009 at 8:08 AM, Meinrad Recheis
<meinrad.recheis@...> wrote:

> On Mon, Jan 5, 2009 at 10:06 AM, Florian Gilcher <flo@...>
> wrote:
>
> [...]
>
>>>> [1]: Yes, I also tried the msysgit, which is even worse.
>>>
>>> Again: which part is worse? It lacks things like git-serve, but I can
>>> live without the need to "host" git repositories in my own computer.
>>
>> The constant crashes. I don't know whether it works now. I stuck with
>> the cygwin version.
>
> I played around with msysGit 1.5 and 1.6 yesterday on winXP and had a very
> bad experience. 'git pull' ends up in an endless loop asking for the
> passphrase which makes it completely unusable.

That's because you tried to checkout a private (ssh) repository and
lack the proper ssh authorization and keys?

I had that working with PuTTY (Pageant) and even with the bundled ssh
without issues.

> The cygwin git packages work like a charm.

Yes, a beautiful and charming snail.

> Both msysGit and cygwinGit come with a tk gui which is helpful for
> visualization of branches and all but it is not quite user friendly IMO.

gitk is great for quick looking, is not like TortoiseSVN or anything like that.

Anyhow, what is the point of this discussion? How many Windows users
will hack in the Ruby source code?

And those who does, I guess they are quite command-line savvy.

I know Git has a complex and less userfriendly command line interface,
but the whole discussion was not Git alone but the set of tools around
it.

--
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[evolving_jerk]

On 04.01.2009, at 21:14, Joel VanderWerf wrote:

> And now we can compare ruby with perl instead of linux:
>
> Perl Migrates To the Git Version Control System
> http://developers.slashdot.org/article.pl?sid=09/01/04/1610256


http://www.python.org/dev/peps/pep-0374/

MK

[Daniel Berger-2]

Michael Klishin wrote:

>
> On 04.01.2009, at 21:14, Joel VanderWerf wrote:
>
>> And now we can compare ruby with perl instead of linux:
>>
>> Perl Migrates To the Git Version Control System
>> http://developers.slashdot.org/article.pl?sid=09/01/04/1610256
>
>
> http://www.python.org/dev/peps/pep-0374/

I have this odd feeling they're going to pick Mercurial over Git.

Just a hunch. ;)

Regards,

Dan

[Radosław Bułat]

On Mon, Jan 5, 2009 at 12:29 PM, Michael Klishin
<michael.s.klishin@...> wrote:
> http://www.python.org/dev/peps/pep-0374/
>

Very nice approach to make a switch (look at the comparison!). I think
that there should be similar talk in ruby-core (very technical and
informative).

On Mon, Jan 5, 2009 at 12:40 PM, Daniel Berger <djberg96@...> wrote:
> I have this odd feeling they're going to pick Mercurial over Git.

It doesn't matter, they choose what they think is the best (even if
"the best" mean "must be written in python").

--
Pozdrawiam

Radoslaw Bulat

[evolving_jerk]

On 05.01.2009, at 16:10, Radosław Bułat wrote:

> Very nice approach to make a switch (look at the comparison!). I think
> that there should be similar talk in ruby-core (very technical and
> informative).


Ruby needs PEP like process badly for virtually anything, but it is a  
whole different story.

MK

[Meinrad Recheis]

On Mon, Jan 5, 2009 at 12:09 PM, Luis Lavena <luislavena@...> wrote:

On Mon, Jan 5, 2009 at 8:08 AM, Meinrad Recheis
<meinrad.recheis@...> wrote:
> On Mon, Jan 5, 2009 at 10:06 AM, Florian Gilcher <flo@...>
> wrote:
>
> [...]
>
>>>> [1]: Yes, I also tried the msysgit, which is even worse.
>>>
>>> Again: which part is worse? It lacks things like git-serve, but I can
>>> live without the need to "host" git repositories in my own computer.
>>
>> The constant crashes. I don't know whether it works now. I stuck with
>> the cygwin version.
>
> I played around with msysGit 1.5 and 1.6 yesterday on winXP and had a very
> bad experience. 'git pull' ends up in an endless loop asking for the
> passphrase which makes it completely unusable.

That's because you tried to checkout a private (ssh) repository and
lack the proper ssh authorization and keys?

No, it floods the screen with prompts without waiting for input. Its a bug that happens in certain environments and I saw on the net that others had experienced the same thing.

-- henon

[Tommy Morgan]

On Mon, Jan 05, 2009 at 10:22:49PM +0900, Michael Klishin wrote:
> Ruby needs PEP like process badly for virtually anything, but it is a  
> whole different story.
>
> MK

Time for a question from the community-newb:

What is it that's prevented an 'REP' from being established? I've seen the above
sentiment tossed around a lot, so if it were a matter of "we don't have anywhere
to post them" then I'm surprised someone hasn't thrown together a quick little
webapp to take care of that. Am I just hearing a vocal minority, or is there
something else going on there?

--Tommy M.
http://www.duwanis.com

[Florian Gilcher]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On Jan 5, 2009, at 2:36 PM, Tommy Morgan wrote:

> On Mon, Jan 05, 2009 at 10:22:49PM +0900, Michael Klishin wrote:
>> Ruby needs PEP like process badly for virtually anything, but it is a
>> whole different story.
>>
>> MK
>
> Time for a question from the community-newb:
>
> What is it that's prevented an 'REP' from being established? I've  
> seen the above
> sentiment tossed around a lot, so if it were a matter of "we don't  
> have anywhere
> to post them" then I'm surprised someone hasn't thrown together a  
> quick little
> webapp to take care of that. Am I just hearing a vocal minority, or  
> is there
> something else going on there?
>
> --Tommy M.
> http://www.duwanis.com
>

Ruby had something called RCR (Ruby Change Request)[1]. I can't tell  
you why,
but it didn't seem to work out and ruby-core is now the place to speak  
about
changes.

If you are interested in that, you should fork the discussion :).

[1]: http://rcrchive.net/ 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAkliEA8ACgkQyLKU2FMxSOKY2ACfXJTk1Gb7tAt5nA3isTIOpO9+
sYIAnjcR1o8Fs1xjjIaT/GlLnWTZxwL3
=e0Eq
-----END PGP SIGNATURE-----

[evolving_jerk]

On 05.01.2009, at 16:36, Tommy Morgan wrote:

> What is it that's prevented an 'REP' from being established? I've  
> seen the above
> sentiment tossed around a lot, so if it were a matter of "we don't  
> have anywhere
> to post them" then I'm surprised someone hasn't thrown together a  
> quick little
> webapp to take care of that. Am I just hearing a vocal minority, or  
> is there
> something else going on there?


I think that if someone puts together a small merb app or rack app or  
whatever,
and helps Matz and co deploy it, nobody would object about having a  
more organized way
to propose language/process/stdlib changes.

It would require a shift in perspective about the process though. My  
observations so far (3+ years with Ruby)
are that Ruby people are a bit like children (in a good sense, and  
myself included). They like playing and not necessary like keeping  
things organized.

Python people are more "boring", and this is why — IMO — they have PEP  
and really well thought out process of 2.5 => 3.0 migration,
  and we don't have REP and any 1.8.x => 1.9.x migration plans. I am  
not bitching/complaining here, I just want to say, this would be a way  
more significant move for Ruby community that a switch from Subversion  
to Git. This is it.

MK

[Florian Gilcher]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On Jan 5, 2009, at 2:59 PM, Michael Klishin wrote:

>
> On 05.01.2009, at 16:36, Tommy Morgan wrote:
>
>> What is it that's prevented an 'REP' from being established? I've  
>> seen the above
>> sentiment tossed around a lot, so if it were a matter of "we don't  
>> have anywhere
>> to post them" then I'm surprised someone hasn't thrown together a  
>> quick little
>> webapp to take care of that. Am I just hearing a vocal minority, or  
>> is there
>> something else going on there?
>
>
> I think that if someone puts together a small merb app or rack app  
> or whatever,
> and helps Matz and co deploy it, nobody would object about having a  
> more organized way
> to propose language/process/stdlib changes.

Why so big? Ruby already uses Redmine (http://redmine.ruby-lang.org),  
which itself has a wiki that can directly be linked by the Ticket  
tracker. No need for a merb application, the infrastructure is there.  
Everyone can get an account and edit.

I think if someone seriously wants to garden an RCR-Wiki and propagate  
it, no one will hold him back.

Regards,
Florian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (Darwin)

iEYEARECAAYFAkliFVYACgkQyLKU2FMxSOJutACeOvGDI6i7osyftVLV4/849qa3
ec4An3ZO/NOY1mi7R1sAtTVg0N69Ziun
=vsSO
-----END PGP SIGNATURE-----

[M. Edward (Ed) Borasky]

Tommy Morgan wrote:

> On Mon, Jan 05, 2009 at 10:22:49PM +0900, Michael Klishin wrote:
>> Ruby needs PEP like process badly for virtually anything, but it is a  
>> whole different story.
>>
>> MK
>
> Time for a question from the community-newb:
>
> What is it that's prevented an 'REP' from being established? I've seen the above
> sentiment tossed around a lot, so if it were a matter of "we don't have anywhere
> to post them" then I'm surprised someone hasn't thrown together a quick little
> webapp to take care of that. Am I just hearing a vocal minority, or is there
> something else going on there?
>
> --Tommy M.
> http://www.duwanis.com
>
>

There have been a couple of attempts at something like this for the
language itself. There was a Ruby Change Request web site and a few
others that I've forgotten. But I don't recall anything on the support
infrastructure. Meanwhile, the "language itself" discussion seem to take
place (in English, anyhow) here on ruby-core. Maybe there should be a
"ruby-infrastructure" list for those who want to talk about workflows,
benchmark suites, test suites, version control systems, etc.

[Paul Brannan]

On Sat, Jan 03, 2009 at 12:48:09PM +0900, Michael Klishin wrote:
> Darcs has idea of "every branch is a separate directory" (or used to
> have it). I can't say it is a problem but you will end up having 3-5
> copies of the pretty much the same code, and if you repository is 600
> megs, not all movies you try to  download from iTunes may fit into
> your HD.

The bigger problem with keeping separate copies of each branch is that
it forces the user to do a full recompile when a new branch is made (*).
This discourages short-lived development branches.

Paul


(*) at least, this is true when using make -- a build system that uses
md5sums makes this less of an issue