I would like to get a sense of when the next Gerrit hacking session should be held. In the past few years we have had one in the spring, and one in the fall, both at Google's Mountain View campus.
Venue is something of an issue for me, for personal reasons I can only attend in person if it is at our Mountain View campus.
I would like to target the group size to be around 10 contributors, I think this has worked well for us in the past.
Since I don't have a date in mind, I would like some input from existing contributors on when would be good (or not good) times to attempt to schedule this.
For my own personal reasons but also to give distant folks plenty of
time to make travel reservations in advance at reasonable ticket
prices, we can't plan anything before say mid-April. :-)
> Do you have a duration in mind - a couple
> days vs a full week?
I'm open. We could do 2 days, we could do 4/5. I know some of the
folks from the EU prefer a longer time, it makes it easier to justify
on the company expense reports and it gives them some time to get out
of the sealed tin tube and stretch their legs. Its a long flight from
the EU to here. :-\
Replying for Edwin and myself:
We prefer 5 days to justify the long flight and starting from Monday
to have some time to adjust to 9 hours timezone difference.
For April-May we have only two possible time slots where both of
us would be available:
30.April - 4.May
or
7.May - 11.May
We still need budget approval from SAP which should be decided soon.
>
> --
> To unsubscribe, email repo-discuss...@googlegroups.com
> More info at http://groups.google.com/group/repo-discuss?hl=en
4-5 days would be nice. As previously mentioned in the mail thread, a longer Hackaton would make it a bit easier (at least for us living on the other side of the Atlantic) to motivate the travel costs.
Best regards
Gustaf
I'll look for space on campus this week and see if we can get
something arranged. Don't book travel just yet. :-)
-Martin
--
Employee of Qualcomm Innovation Center, Inc. which is a
member of Code Aurora Forum
Is it feasible or in next plan to carry out the linguist feature like that github has now ?
Reference to https://github.com/github/linguist
most guys around me like github, linguist feature is a reason . if Gerrit will have this feature that would be very fine.
Is it feasible or in next plan to carry out the linguist feature like that github has now ?
Reference to https://github.com/github/linguist
/Bruce
This uses Pygments to format, which is pretty darn slow to do from
Java. I had hoped to be able to eventually switch to Jygments [1],
which is a port of Pygments to Java using a mostly automated lexer
conversion. Unfortunately nobody has had time to work on it. The
current syntax highlighting used by Gerrit Code Review is done on the
client side... the server only sends the raw text and the client marks
it up as HTML. Switching to something like Pygments or Jygments means
the server has to do the markup.
> How do folks feel about the 7.May - 11.May dates proposed by Sasa?
>
> I'll look for space on campus this week and see if we can get
> something arranged. Don't book travel just yet. :-)
If there's room for me I'd be happy to join.
--
Magnus Bäck
ba...@google.com
OK, so May 7-11 seems like the dates. I'll start hunting for space on
Google's campus.
So I managed to find space in my building. I basically stole the
largest conference room (Dr. Porkchop) and booked it solid for the
week. I may look for other space on campus, but right now this room is
mine. :-)
The room is sized for 14. 14 is actually a tight fight in a room for
14, especially all day. So I would like to keep the hacking session
around 10-12 people, including myself and my team. I suspect Dave
Borowitz may want to join us, so this leaves room for about 8 others.
Sasa, Edwin, Gustaf are coming so that drops us to about 5 slots
available.
I want to give preference to existing contributors, especially Martin,
before letting other folks attend since this is a working session to
hack on Gerrit source code.
Ok, Martin and I are approved. Luca, Brad, Magnus, and Deniz have all also either expressed interest or have travel approved. That leaves us 1 over the "about 5 slots". Think that'll work?
Nasser
>
> --
> To unsubscribe, email repo-discuss...@googlegroups.com
> More info at http://groups.google.com/group/repo-discuss?hl=en
--
On Mar 30, 2012, at 12:29 PM, Shawn Pearce wrote:> On Tue, Mar 27, 2012 at 09:49, Shawn Pearce <s...@google.com> wrote:
>> On Tue, Mar 27, 2012 at 08:45, Lundh, Gustaf
>> <Gustaf...@sonymobile.com> wrote:
>>> 7-11 May seems great. I would absolutely like to join this year's Gerrit Hackaton.
>>
>> OK, so May 7-11 seems like the dates. I'll start hunting for space on
>> Google's campus.
>
> So I managed to find space in my building. I basically stole the
> largest conference room (Dr. Porkchop) and booked it solid for the
> week. I may look for other space on campus, but right now this room is
> mine. :-)
>
> The room is sized for 14. 14 is actually a tight fight in a room for
> 14, especially all day. So I would like to keep the hacking session
> around 10-12 people, including myself and my team. I suspect Dave
> Borowitz may want to join us, so this leaves room for about 8 others.
> Sasa, Edwin, Gustaf are coming so that drops us to about 5 slots
> available.
>
> I want to give preference to existing contributors, especially Martin,
> before letting other folks attend since this is a working session to
> hack on Gerrit source code.Ok, Martin and I are approved. Luca, Brad, Magnus, and Deniz have all also either expressed interest or have travel approved. That leaves us 1 over the "about 5 slots". Think that'll work?
Nasser
>
> --
> To unsubscribe, email repo-discuss+unsubscribe@googlegroups.com
> More info at http://groups.google.com/group/repo-discuss?hl=en
I'm interested as well... but couldn't get an approval, unfortunately.
--
To unsubscribe, email repo-discuss...@googlegroups.com
I'm not sure I understood this.
We have limited space right now for two reasons:
* I want this to be a working session. The more people that attend,
the less likely we are to write code that improves Gerrit, and the
more likely we are to just chat about our various problems with Gerrit
and not actually improve the software itself. Getting together to
share experiences can be very informative and productive, but its a
different kind of productive from the kind where we sit down and write
code together as a group for a week.
* Google has limited space on our campus to host events, and limited
budget to buy everyone breakfast, lunch and snacks throughout the day.
Yes even Google has its limits as to what it can spend on things.
Getting space/budget to host for a 10ish people for a week is
different from getting space/budget to host a single day event for a
larger number of people.
If you are suggesting that a larger Gerrit Code Review "user
conference" might be well attended, I think you are sort of right. At
GitTogether '2011 last fall we had a turnout of around 50 people, and
maybe at least half of those where there primarily because their
company uses Gerrit internally and wanted to learn more about how to
use it, and hopefully gain insight to problems they were having with
the software. Its a very different group of people. Gerrit developers
who write code for Gerrit vs. Gerrit server administrators who are
tasked with keeping it running for their organization.
As a community we may explore the idea of running a Gerrit user
conference in the fall. But this is still a pie-in-the-sky concept.
I would like to enhance:
- Security on password encryptions (AES on secure.config with master password and SHA-1 on HTTP password in DB)
- REST-Ful API extension for in-DB account management (and possibly a new "local" realm to Gerrit)
- Auditing API
- A new layer draft of REST-Ful API more stateless and easy to be used
... or anything else the other would like me to change / provide / enhance :-)
Being face-to-face with all the other would ease the interactions, but a couple of remote sessions, not desirable, but could work as well for me.
Luca.
Remotely joining may be possible with Google+ Hangout... for those who
can't be there in person. :-)
> I would like to enhance:
> - Security on password encryptions (AES on secure.config with master password and SHA-1 on HTTP password in DB)
Seems reasonable. I guess you also want to put a passphrase onto the
server's private key blob, or also AES encrypt that too with some sort
of master key?
I would suggest using KeyCzar here. It works pretty well. Google might
want to have a bit of a hand in helping to define how this loading of
secure values happens. We currently store secure values out of process
using something based on KeyCzar and currently we wedge those values
into a fake secure.config in memory before feeding them into Gerrit.
It would be nice if we could get Gerrit to just talk directly to our
key storage API by defining a simple bridge API.
> - REST-Ful API extension for in-DB account management (and possibly a new "local" realm to Gerrit)
The whole user management code needs to be seriously overhauled. Colby
is starting to do some work on that with the group membership
refactoring, but there is a lot remaining.
> - Auditing API
It seems to me that plugin hooks at the right points throughout the
code would make an auditing plugin, and a lot of other stuff,
possible.
> - A new layer draft of REST-Ful API more stateless and easy to be used
Yes. I have tried to get Conley going in this direction with being
able to post review comments and approval scores using JSON over HTTP.
It turns out selecting a single change is really hard, the "correct"
long term "key" is a triplet of project name, branch name, and
change-id string. And the first two of those accept / as characters.
On Sun, Apr 1, 2012 at 17:51, Luca Milanesio <luca.mi...@gmail.com> wrote:I could join remotely if needed: my goal is really to focus on a joint code-session to provide improvements in Gerrit we've discussed so far in the mailing list.
Remotely joining may be possible with Google+ Hangout... for those who
can't be there in person. :-)
I would like to enhance:- Security on password encryptions (AES on secure.config with master password and SHA-1 on HTTP password in DB)
Seems reasonable. I guess you also want to put a passphrase onto the
server's private key blob, or also AES encrypt that too with some sort
of master key?
I would suggest using KeyCzar here. It works pretty well. Google might
want to have a bit of a hand in helping to define how this loading of
secure values happens. We currently store secure values out of process
using something based on KeyCzar and currently we wedge those values
into a fake secure.config in memory before feeding them into Gerrit.
It would be nice if we could get Gerrit to just talk directly to our
key storage API by defining a simple bridge API.
- REST-Ful API extension for in-DB account management (and possibly a new "local" realm to Gerrit)
The whole user management code needs to be seriously overhauled. Colby
is starting to do some work on that with the group membership
refactoring, but there is a lot remaining.
- Auditing API
It seems to me that plugin hooks at the right points throughout the
code would make an auditing plugin, and a lot of other stuff,
possible.
- A new layer draft of REST-Ful API more stateless and easy to be used
Yes. I have tried to get Conley going in this direction with being
able to post review comments and approval scores using JSON over HTTP.
It turns out selecting a single change is really hard, the "correct"
long term "key" is a triplet of project name, branch name, and
change-id string. And the first two of those accept / as characters.
-----Original Message-----
From: repo-d...@googlegroups.com [mailto:repo-d...@googlegroups.com] On Behalf Of Shawn Pearce
Sent: Monday, April 02, 2012 5:40 AM
To: Bruce Zu
Cc: repo-d...@googlegroups.com; Edwin Kempin; Saša Živkov; Martin Fick; Lundh, Gustaf; David Borowitz
Subject: Re: Planning next Gerrit hacking session
--
To unsubscribe, email repo-discuss...@googlegroups.com
I found a different room that is "rated" for 24. 24 is what the fire
marshal will let us seat in the room. I think 24 is a lot of people
for an all-day hacking session. 12 sounds better.
I think the attendee list is currently:
1) Shawn Pearce
2) Dave Borowitz
3) Magnus Bäck
4) Saša Živkov
5) Edwin Kempin
6) Gustaf Lundh
7) Martin Fick
8) Nasser Grainawi
9) Brad Larson
10) Deniz Türkoglu
11) Luca Milanesio (unclear if travel is possible?)
Have I missed anyone?
It sounded like Bruce Zu was interested also. He shows up
on page 3 here (7 commits):
http://www.ohloh.net/p/gerrit/contributors?page=3
and I think he has several (many?) more changes currently
for review. He also regularly reviews any changes which I
add him to, so I would consider him a fairly active
contributor and coder,
-Martin
--
Employee of Qualcomm Innovation Center, Inc. which is a
-EA
We might have space. I'm at ~12 folks right now.
|
a short update: I will be able to participate on-site :-)
[Nr. 11 :-) ... is the seat still available for me ?]
Will start pushing some code for review before then anyway :-)
Luca.
Yes.
> Will start pushing some code for review before then anyway :-)
OK. :-)
Should you have the same problem ... I found space in the Best Western Silicon Valley @Sunnyvale (http://bit.ly/HWvOzS)
I booked a car in case anybody would need a daily lift to Google :-)
See you there !
Luca.