I'd like us to think a bit about events in the next 12 months.
I'm going to eventually put this into the project meeting agenda for recording some kind of consensus, but for now this is half-baked.
One of the backdrops is that going forward CloudBees will be organizing one JUC per year, at San Francisco, instead of doing 4 like we did the past 12 months. From what I've been told, the main issue is that the cost involved in running it is just too high (I saw one of the cost summary and it's in the order of tens of thousands of dollars!)
Given that, I think we need to change the model here.
First, I think we need to run events cheaply --- as a case in point, our friends in Japan organized Tokyo JUC a lot cheaply:
- we found the venue who's willing to host us for free
- we did video recording by ourselves and so that was free
- the conference was from noon to 5 so that there's no need to
cater food.
- the drink-up after the event was paid for by attendees who
opted in for that part.
Second, I find that it works better when the local people organizes the event, like in Israel and Tokyo.
In this new model, I think the community could act as a facilitator --- I think we want to put this model out there (so that we can find people like Lars who's interested in driving one in their home city), and we want to help those efforts so that they feel like a series of the same event.
Maybe it starts with a mailing list to hook up people interested in organizing one, with some Wiki pages to share the knowledge. Alyssa thinks she can continue to help with the actual work (of handling CfPs, handling money between sponsors and vendors, registrations, etc.)
I also think the Jenkins project should spend some of its budget on those events, and this is another reason we need some guidelines.
One way I think of this is as a parallel to Jenkins CIA, in which we set out some mechanism to encourage small local meet-ups and evangelism effort. This could be a bigger brother to CIA that just requires more commitment from those who are willing to do it.
I'm not sure how this fits into that FOSDEM thread I just started, or if it needs to fit into this.
I'm in complete agreement to what Kohsuke stated below. I've found that it
is much more cost effective when we have the support of the local
community. ROI would be much greater than costs, i.e JUC Tokyo, Herzelia.
JUC Antwerp was recently cancelled as costs outweighed ROI. However, if
there's great interest within the local community to lead and support JUC
Antwerp or any other locations, I would be happy to help out.
On Fri, Aug 3, 2012 at 5:00 PM, Kohsuke Kawaguchi <k...@kohsuke.org> wrote:
> I'd like us to think a bit about events in the next 12 months.
> I'm going to eventually put this into the project meeting agenda for
> recording some kind of consensus, but for now this is half-baked.
> One of the backdrops is that going forward CloudBees will be organizing
> one JUC per year, at San Francisco, instead of doing 4 like we did the past
> 12 months. From what I've been told, the main issue is that the cost
> involved in running it is just too high (I saw one of the cost summary and
> it's in the order of tens of thousands of dollars!)
> Given that, I think we need to change the model here.
> First, I think we need to run events cheaply --- as a case in point, our
> friends in Japan organized Tokyo JUC a lot cheaply:
> - we found the venue who's willing to host us for free
> - we did video recording by ourselves and so that was free
> - the conference was from noon to 5 so that there's no need to
> cater food.
> - the drink-up after the event was paid for by attendees who
> opted in for that part.
> Second, I find that it works better when the local people organizes the
> event, like in Israel and Tokyo.
> In this new model, I think the community could act as a facilitator --- I
> think we want to put this model out there (so that we can find people like
> Lars who's interested in driving one in their home city), and we want to
> help those efforts so that they feel like a series of the same event.
> Maybe it starts with a mailing list to hook up people interested in
> organizing one, with some Wiki pages to share the knowledge. Alyssa thinks
> she can continue to help with the actual work (of handling CfPs, handling
> money between sponsors and vendors, registrations, etc.)
> I also think the Jenkins project should spend some of its budget on those
> events, and this is another reason we need some guidelines.
> One way I think of this is as a parallel to Jenkins CIA, in which we set
> out some mechanism to encourage small local meet-ups and evangelism effort.
> This could be a bigger brother to CIA that just requires more commitment
> from those who are willing to do it.
> I'm not sure how this fits into that FOSDEM thread I just started, or if
> it needs to fit into this.
Hi Kohsuke and devs, thanks a lot for bringing up this topic.
By direct encouragement from Kohsuke I will contribute to this thread.
Just to set the context of my comment: Like CloudBees we (Praqma) have our
business model based pre-dominantly around Jenkins CI. Where CloudBees
focus on PaaS, we have our focus on business driven, sponsored Open Source
development and consultancy services related to tool support of software
development and verification processes (with Jenkins Ci as the natural
hub). We strive to support the Jenkins community in any way we possible
can. So besides our contributions to the pool of plugins we also found it
natural to help organize and fund some of the more costly community
activities - such as events.
CloudBees and Praqma entered a formal partnership a few month back. At
about the same time where we had our first vision of a Jenkins User
Conference in Copenhagen. We have debated the format of such a CPH user
conference together with CloudBees: Should it be an official JUC event
(just another city along with Paris, New York, San Fransisco and Antwerp)
and simply squeeze it into the program as with Herzelia and Tokyo or should
it be a CIA event. Or something in between?
As you may have noticed we had to settled on a different format than JUC
(and I realize now, that this is probably due to the relatively costly
format the JUC represent): So the CPH event in September is kind of a new
approach, and perhaps a trial of the new format.
- We use the term "User Event" instead of "User Conference" - Actually
we see the gathering very much as a conference, but we did this simply to
deliberately distinguish it from CloudBees' JUC format.
- We start at noon - so attendees are expected to have their own lunch
before they show up.
- We continue to late in the evening, but the conference meal will be
sandwiches - enough to fill everybody, but not in any way fancy (a note
hereto: The conference meal I had at JUC in Paris was the best conference
meal I had in my entire career ;-)
- When the event officially ends at the conference hotel (19:00'ish), we
have invited the participants to join us to a cafe or pub in the
neighborhood for more relaxed discussions and free drinks - This social
after-event has it's own sponsor (Programming Research in the CPH event).
- The token attendee give-away will be a mug instead of a T-shirt. The
sponsor value will be the same, but the production cost will be less than
half of what it cost to produce a T-shirt. And then we just hope, that the
attendees will like their mug as much a they would have liked a T-shirt.
In our case (CPH september) we didn't have a call for paper session - but
that is not necessarily part of the format we are suggesting, it was
simply because we had so many good speakers lined up in our network (Nokia,
Sony Mobile, Programming Research, Grundfos, CloudBees and Ourselves) that
this year with south short notice - we didn't need a CfP session.
The CPH user event is hosted and funded by Praqma and a generous sponsor
contribution is added from CloudBees (Thanks!!!) and as I mentioned - the
social after-event has it's own sponsor.
...And just to backup Kohsuke in the experiences he's sharing: Even though
we are trying to make this a cheap event (also for the attendees - signup
fee is only 25€) we are *still* going to loose a considerable amount on
money on hosting this event. And on top of that we are going to spend an
equally considerable amount of time on planning and executing it.
I'm not whining about it! Quite the opposite. We have a commercial interest
in Jenkins and we are happy to be able to 'pay-back' to the community. But
I think that two point are important for large events:
- Any event will need a commercial sponsorship to become a success:
Cheap is good, but moneyless as in 'driven solely be initiative' - is
probably not going to happen outside the CIA format.
- If a commercial sponsor should contribute - they would want the event
to be 'officially accepted and approved' by the community in order to shine
in the light of the Google doctrine: "Do No Evil"
As we in Praqma are developers ourselves we found that we had an urge to
fill in content that was developer oriented (as opposed to user oriented)
as well. But this represents a paradox; it's really difficult to mix such
different groups (users and developers) at the same conference unless it's
a big one with more tracks, so instead we have decided to host a separate
developer event along with the user event (a full day workshop on the day
before the user event).
One of the complaints or critics we often hear from developers who attended
hackatons (including ourselves) is that they are often not very well
organized: People meet with no specific agenda, they spend a day writing
code (often alone) on the same stuff as they would have worked on had they
stayed at home - at the end of the day they eat pizza together and break up.
We have internally worked on developing a good format for a well organized
hackaton:
It's a "Code Camp".
It's inspired by open space technology/discussions (the term 'camp' is also
used within the open space approach): Developers are encouraged to bring a
few topics or issues to the camp, which starts with a short introduction
where each develop gives a few words of explanation to the topics. Then
there is a vote among the attendees: What are the interesting topic that
should be worked on? The topics that gets the most votes win a time-slot on
the code camp. Then we break out into small groups - one per topic (no one
works alone, all groups are at least 3 participants).
Following up on the inspiration from the Open Space technology the good
topics are characterized by:
1) a real issue of concern
2) a high level of complexity
3) a high level of diversity
4) a real or potential conflict
5) a high level urgency.
- The camp has it's own wiki - each break out session will have an
appointed scribe, who captures findings, results, deliveries, discussions
etc on the wiki. This supports that the summary-sessions can be kept very
short, anybody who wants to know what the other groups worked with can find
it on the wiki - and so can people who didn't even attend the camp.
- The camp is well-organized although even though it's not specifically
pre-organized.
The scribes can be recruited among more novice developers, who would like
to attend the Code Camp, but potentially would be intimidated by the size
of the brain trust. Or scribes could be recruited among once-was-developer
managers who tend to have strong opinions on direction, but due to rusty
code-fingers they lack to ability to actually contribute to the code base
(I'm sad to admit, that the description fits myself ;-).
At Praqma we run Code Camps like this internally among our developers about
every second month and we find them quite efficient - we'd like to share.
It could potentially also be a format for a Jenkins devroom at FOSDEM?
In order to bring down the cost, the Code Camp is hosted at Praqma Plex -
and we simply serve pizzas from the local joint. It's very cheap! Seats are
limited to 20 developers, together with scribes and facilitators we'll be
around 25 in total. The registration fee is purely symbolic; 15€, just to
make sure that no-shows are encouraged to unregister as opposed to just not
showing up.
Like Kohsuke just announced that CloudBees is prepared to take
responsibility for hosting the San Fransisco JUC once every year. We at
Praqma are also prepared to commit ourselves the make both the "Jenkins CI
Copenhagen Code Camp" and the "Jenkins CI Copenhagen User Event" annual
events.
This year it's in September - but like JUC in San Fransisco is piggybacking
on Java ONE we are also considering on having our events piggyback on a
larger one - and we have our eyes set on the GOTO conference (
http://gotocon.com) which would imply that the 2013 event probably will be
in May - not September.
And like Alyssa @ CloudBees we, here at Praqma, also declare ourselves
ready to assist any community driven initiative or CIA event that need
assistance in our region (Europe) with what ever is needed: Wiki hosting,
Money handling, Event registration, CfP, design of stickers, logos,
give-aways etc.
On Sat, Aug 4, 2012 at 2:00 AM, Kohsuke Kawaguchi <k...@kohsuke.org> wrote:
> I'd like us to think a bit about events in the next 12 months.
> I'm going to eventually put this into the project meeting agenda for
> recording some kind of consensus, but for now this is half-baked.
> One of the backdrops is that going forward CloudBees will be organizing
> one JUC per year, at San Francisco, instead of doing 4 like we did the past
> 12 months. From what I've been told, the main issue is that the cost
> involved in running it is just too high (I saw one of the cost summary and
> it's in the order of tens of thousands of dollars!)
> Given that, I think we need to change the model here.
> First, I think we need to run events cheaply --- as a case in point, our
> friends in Japan organized Tokyo JUC a lot cheaply:
> - we found the venue who's willing to host us for free
> - we did video recording by ourselves and so that was free
> - the conference was from noon to 5 so that there's no need to
> cater food.
> - the drink-up after the event was paid for by attendees who
I think having a tried-and-tested format that can be replicated elsewhere would be nice. It will reduce the overhead involved in organizing an event.
If we can package this such that it can be held in a bigger conference room of a big company or an university classroom, I wonder if that opens up a lot of possibilities to replicate this elsewhere.
In Asia and Europe, food catering isn't necessary, and moving to a nearby pub after the event where attendees pay for themselves is quite acceptable. This is the proven model in Tokyo.
In US, we'll most likely need some delivered food, but I suspect the only reason it's so expensive is because the hotel robs you, and that will be a non-issue if we have the event in a company or an university.
One thing I do wonder is how important it is to keep the event free of charge (or keep it minimum just to make sure we won't end up a half-empty room with 50 people on the waiting list.) In Tokyo we never charged for an event and we just oversell the seats instead, and that was considered important.
In other parts of Japan, I've heard that it's mostly up to $5 or $10 even when we charge. That's a price someone can pay out of his own pocket, and I suspect that'd make it a lot easier for many (at least myself) to attend.
Just as a hypothesis, if we are to have this event in the Nokia campus in Copenhagen (I remember you guys had a pretty big room that fits 100) and in the above-mentioned format, would that drive down the cost enough that we can get away without spending many $1000s and charging people?
> Hi Kohsuke and devs, thanks a lot for bringing up this topic.
> By direct encouragement from Kohsuke I will contribute to this thread.
> Just to set the context of my comment: Like CloudBees we (Praqma) have
> our business model based pre-dominantly around Jenkins CI. Where
> CloudBees focus on PaaS, we have our focus on business driven, sponsored
> Open Source development and consultancy services related to tool support
> of software development and verification processes (with Jenkins Ci as
> the natural hub). We strive to support the Jenkins community in any way
> we possible can. So besides our contributions to the pool of plugins we
> also found it natural to help organize and fund some of the more costly
> community activities - such as events.
> CloudBees and Praqma entered a formal partnership a few month back. At
> about the same time where we had our first vision of a Jenkins User
> Conference in Copenhagen. We have debated the format of such a CPH user
> conference together with CloudBees: Should it be an official JUC event
> (just another city along with Paris, New York, San Fransisco and
> Antwerp) and simply squeeze it into the program as with Herzelia and
> Tokyo or should it be a CIA event. Or something in between?
> As you may have noticed we had to settled on a different format than JUC
> (and I realize now, that this is probably due to the relatively costly
> format the JUC represent): So the CPH event in September is kind of a
> new approach, and perhaps a trial of the new format.
> * We use the term "User Event" instead of "User Conference" - Actually
> we see the gathering very much as a conference, but we did this
> simply to deliberately distinguish it from CloudBees' JUC format.
> * We start at noon - so attendees are expected to have their own lunch
> before they show up.
> * We continue to late in the evening, but the conference meal will be
> sandwiches - enough to fill everybody, but not in any way fancy (a
> note hereto: The conference meal I had at JUC in Paris was the best
> conference meal I had in my entire career ;-)
> * When the event officially ends at the conference hotel (19:00'ish),
> we have invited the participants to join us to a cafe or pub in the
> neighborhood for more relaxed discussions and free drinks - This
> social after-event has it's own sponsor (Programming Research in the
> CPH event).
> * The token attendee give-away will be a mug instead of a T-shirt. The
> sponsor value will be the same, but the production cost will be less
> than half of what it cost to produce a T-shirt. And then we just
> hope, that the attendees will like their mug as much a they would
> have liked a T-shirt.
> In our case (CPH september) we didn't have a call for paper session -
> but that is not necessarily part of the format we are suggesting, it
> was simply because we had so many good speakers lined up in our network
> (Nokia, Sony Mobile, Programming Research, Grundfos, CloudBees and
> Ourselves) that this year with south short notice - we didn't need a CfP
> session.
> The CPH user event is hosted and funded by Praqma and a generous sponsor
> contribution is added from CloudBees (Thanks!!!) and as I mentioned -
> the social after-event has it's own sponsor.
> ...And just to backup Kohsuke in the experiences he's sharing: Even
> though we are trying to make this a cheap event (also for the attendees
> - signup fee is only 25 ) we are *still* going to loose a considerable
> amount on money on hosting this event. And on top of that we are going
> to spend an equally considerable amount of time on planning and
> executing it.
> I'm not whining about it! Quite the opposite. We have a commercial
> interest in Jenkins and we are happy to be able to 'pay-back' to the
> community. But I think that two point are important for large events:
> * Any event will need a commercial sponsorship to become a success:
> Cheap is good, but moneyless as in 'driven solely be initiative' -
> is probably not going to happen outside the CIA format.
> * If a commercial sponsor should contribute - they would want the
> event to be 'officially accepted and approved' by the community in
> order to shine in the light of the Google doctrine: "Do No Evil"
> As we in Praqma are developers ourselves we found that we had an urge to
> fill in content that was developer oriented (as opposed to user
> oriented) as well. But this represents a paradox; it's really difficult
> to mix such different groups (users and developers) at the same
> conference unless it's a big one with more tracks, so instead we have
> decided to host a separate developer event along with the user event (a
> full day workshop on the day before the user event).
> One of the complaints or critics we often hear from developers who
> attended hackatons (including ourselves) is that they are often not very
> well organized: People meet with no specific agenda, they spend a day
> writing code (often alone) on the same stuff as they would have worked
> on had they stayed at home - at the end of the day they eat pizza
> together and break up.
> We have internally worked on developing a good format for a well
> organized hackaton:
> It's a "Code Camp".
> It's inspired by open space technology/discussions (the term 'camp' is
> also used within the open space approach): Developers are encouraged to
> bring a few topics or issues to the camp, which starts with a short
> introduction where each develop gives a few words of explanation to the
> topics. Then there is a vote among the attendees: What are the
> interesting topic that should be worked on? The topics that gets the
> most votes win a time-slot on the code camp. Then we break out into
> small groups - one per topic (no one works alone, all groups are at
> least 3 participants).
> Following up on the inspiration from the Open Space technology the good
> topics are characterized by:
> 1) a real issue of concern
> 2) a high level of complexity
> 3) a high level of diversity
> 4) a real or potential conflict
> 5) a high level urgency.
> * The camp has it's own wiki - each break out session will have an
> appointed scribe, who captures findings, results, deliveries,
> discussions etc on the wiki. This supports that the summary-sessions
> can be kept very short, anybody who wants to know what the other
> groups worked with can find it on the wiki - and so can people who
> didn't even attend the camp.
> * The camp is well-organized although even though it's not
> specifically pre-organized.
> The scribes can be recruited among more novice developers, who would
> like to attend the Code Camp, but potentially would be intimidated by
> the size of the brain trust. Or scribes could be recruited among
> once-was-developer managers who tend to have strong opinions on
> direction, but due to rusty code-fingers they lack to ability to
> actually contribute to the code base (I'm sad to admit, that the
> description fits myself ;-).
> At Praqma we run Code Camps like this internally among our developers
> about every second month and we find them quite efficient - we'd like to
> share. It could potentially also be a format for a Jenkins devroom at
> FOSDEM?
> In order to bring down the cost, the Code Camp is hosted at Praqma Plex
> - and we simply serve pizzas from the local joint. It's very cheap!
> Seats are limited to 20 developers, together with scribes and
> facilitators we'll be around 25 in total. The registration fee is purely
> symbolic; 15 , just to make sure that no-shows are encouraged to
> unregister as opposed to just not showing up.
> Like Kohsuke just announced that CloudBees is prepared to take
> responsibility for hosting the San Fransisco JUC once every year. We at
> Praqma are also prepared to commit ourselves the make both the "Jenkins
> CI Copenhagen Code Camp" and the "Jenkins CI Copenhagen User Event"
> annual events.
On a related note, if you are in a big city (New York, Sao Paulo, Seoul, Chennai, Paris, Beijing, London, etc.) and has access to a venue, and if you /might/ be interested in hosting an event like this, please let us know.
Our usage stats should have a pretty accurate distribution of users around the world, so if the plan outlined below works, all we need in theory is some venues plus local speakers... (and then a lot of love from Alyssa, et al.)
> I think having a tried-and-tested format that can be replicated
> elsewhere would be nice. It will reduce the overhead involved in
> organizing an event.
> If we can package this such that it can be held in a bigger conference
> room of a big company or an university classroom, I wonder if that opens
> up a lot of possibilities to replicate this elsewhere.
> In Asia and Europe, food catering isn't necessary, and moving to a
> nearby pub after the event where attendees pay for themselves is quite
> acceptable. This is the proven model in Tokyo.
> In US, we'll most likely need some delivered food, but I suspect the
> only reason it's so expensive is because the hotel robs you, and that
> will be a non-issue if we have the event in a company or an university.
> One thing I do wonder is how important it is to keep the event free of
> charge (or keep it minimum just to make sure we won't end up a
> half-empty room with 50 people on the waiting list.) In Tokyo we never
> charged for an event and we just oversell the seats instead, and that
> was considered important.
> In other parts of Japan, I've heard that it's mostly up to $5 or $10
> even when we charge. That's a price someone can pay out of his own
> pocket, and I suspect that'd make it a lot easier for many (at least
> myself) to attend.
> Just as a hypothesis, if we are to have this event in the Nokia campus
> in Copenhagen (I remember you guys had a pretty big room that fits 100)
> and in the above-mentioned format, would that drive down the cost enough
> that we can get away without spending many $1000s and charging people?
> On 08/04/2012 06:31 AM, Lars Kruse wrote:
>> Hi Kohsuke and devs, thanks a lot for bringing up this topic.
>> By direct encouragement from Kohsuke I will contribute to this thread.
>> Just to set the context of my comment: Like CloudBees we (Praqma) have
>> our business model based pre-dominantly around Jenkins CI. Where
>> CloudBees focus on PaaS, we have our focus on business driven, sponsored
>> Open Source development and consultancy services related to tool support
>> of software development and verification processes (with Jenkins Ci as
>> the natural hub). We strive to support the Jenkins community in any way
>> we possible can. So besides our contributions to the pool of plugins we
>> also found it natural to help organize and fund some of the more costly
>> community activities - such as events.
>> CloudBees and Praqma entered a formal partnership a few month back. At
>> about the same time where we had our first vision of a Jenkins User
>> Conference in Copenhagen. We have debated the format of such a CPH user
>> conference together with CloudBees: Should it be an official JUC event
>> (just another city along with Paris, New York, San Fransisco and
>> Antwerp) and simply squeeze it into the program as with Herzelia and
>> Tokyo or should it be a CIA event. Or something in between?
>> As you may have noticed we had to settled on a different format than JUC
>> (and I realize now, that this is probably due to the relatively costly
>> format the JUC represent): So the CPH event in September is kind of a
>> new approach, and perhaps a trial of the new format.
>> * We use the term "User Event" instead of "User Conference" - Actually
>> we see the gathering very much as a conference, but we did this
>> simply to deliberately distinguish it from CloudBees' JUC format.
>> * We start at noon - so attendees are expected to have their own lunch
>> before they show up.
>> * We continue to late in the evening, but the conference meal will be
>> sandwiches - enough to fill everybody, but not in any way fancy (a
>> note hereto: The conference meal I had at JUC in Paris was the best
>> conference meal I had in my entire career ;-)
>> * When the event officially ends at the conference hotel (19:00'ish),
>> we have invited the participants to join us to a cafe or pub in the
>> neighborhood for more relaxed discussions and free drinks - This
>> social after-event has it's own sponsor (Programming Research in the
>> CPH event).
>> * The token attendee give-away will be a mug instead of a T-shirt. The
>> sponsor value will be the same, but the production cost will be less
>> than half of what it cost to produce a T-shirt. And then we just
>> hope, that the attendees will like their mug as much a they would
>> have liked a T-shirt.
>> In our case (CPH september) we didn't have a call for paper session -
>> but that is not necessarily part of the format we are suggesting, it
>> was simply because we had so many good speakers lined up in our network
>> (Nokia, Sony Mobile, Programming Research, Grundfos, CloudBees and
>> Ourselves) that this year with south short notice - we didn't need a CfP
>> session.
>> The CPH user event is hosted and funded by Praqma and a generous sponsor
>> contribution is added from CloudBees (Thanks!!!) and as I mentioned -
>> the social after-event has it's own sponsor.
>> ...And just to backup Kohsuke in the experiences he's sharing: Even
>> though we are trying to make this a cheap event (also for the attendees
>> - signup fee is only 25 ) we are *still* going to loose a considerable
>> amount on money on hosting this event. And on top of that we are going
>> to spend an equally considerable amount of time on planning and
>> executing it.
>> I'm not whining about it! Quite the opposite. We have a commercial
>> interest in Jenkins and we are happy to be able to 'pay-back' to the
>> community. But I think that two point are important for large events:
>> * Any event will need a commercial sponsorship to become a success:
>> Cheap is good, but moneyless as in 'driven solely be initiative' -
>> is probably not going to happen outside the CIA format.
>> * If a commercial sponsor should contribute - they would want the
>> event to be 'officially accepted and approved' by the community in
>> order to shine in the light of the Google doctrine: "Do No Evil"
>> As we in Praqma are developers ourselves we found that we had an urge to
>> fill in content that was developer oriented (as opposed to user
>> oriented) as well. But this represents a paradox; it's really difficult
>> to mix such different groups (users and developers) at the same
>> conference unless it's a big one with more tracks, so instead we have
>> decided to host a separate developer event along with the user event (a
>> full day workshop on the day before the user event).
>> One of the complaints or critics we often hear from developers who
>> attended hackatons (including ourselves) is that they are often not very
>> well organized: People meet with no specific agenda, they spend a day
>> writing code (often alone) on the same stuff as they would have worked
>> on had they stayed at home - at the end of the day they eat pizza
>> together and break up.
>> We have internally worked on developing a good format for a well
>> organized hackaton:
>> It's a "Code Camp".
>> It's inspired by open space technology/discussions (the term 'camp' is
>> also used within the open space approach): Developers are encouraged to
>> bring a few topics or issues to the camp, which starts with a short
>> introduction where each develop gives a few words of explanation to the
>> topics. Then there is a vote among the attendees: What are the
>> interesting topic that should be worked on? The topics that gets the
>> most votes win a time-slot on the code camp. Then we break out into
>> small groups - one per topic (no one works alone, all groups are at
>> least 3 participants).
>> Following up on the inspiration from the Open Space technology the good
>> topics are characterized by:
>> 1) a real issue of concern
>> 2) a high level of complexity
>> 3) a high level of diversity
>> 4) a real or potential conflict
>> 5) a high level urgency.
>> * The camp has it's own wiki - each break out session will have an
>> appointed scribe, who captures findings, results, deliveries,
>> discussions etc on the wiki. This supports that the summary-sessions
>> can be kept very short, anybody who wants to know what the other
>> groups worked with can find it on the wiki - and so can people who
>> didn't even attend the camp.
>> * The camp is well-organized although even though it's not
>> specifically pre-organized.
>> The scribes can be recruited among more novice developers, who would
>> like to attend the Code Camp, but potentially would be intimidated by
>> the size of the brain trust. Or scribes could be recruited among
>> once-was-developer managers who tend to have strong opinions on
>> direction, but due to rusty code-fingers they lack to ability to
>> actually contribute to the code base (I'm sad to admit, that the
>> description fits myself ;-).
>> At Praqma we run Code Camps like this internally among our developers
>> about every second month and we find them quite efficient - we'd like to
>> share. It could potentially also be a format for a Jenkins devroom at
>> FOSDEM?
>> In order to bring down the cost, the Code Camp is hosted at Praqma Plex
>> - and we simply serve pizzas from the local joint. It's very cheap!
I will check with some local developers to check their interest in a Jenkins event in Sao Paulo (anyone here? Algum desenvolvedor ou interessado por aqui? :-).
There are some universities that I believe would gives access to a venue for this event with no cost or at low cost. I liked the idea of starting after noon and having a pub where each attendee pays for themselves, and if needed TupiLabs could put some money on this event, and keep it at a very low price or even free.
>________________________________ > From: Kohsuke Kawaguchi <k...@kohsuke.org> >To: jenkinsci-dev@googlegroups.com >Cc: Lars Kruse <l...@praqma.net> >Sent: Monday, 6 August 2012 11:57 PM >Subject: Re: Agenda for project meeting: event planning
>On a related note, if you are in a big city (New York, Sao Paulo, Seoul, >Chennai, Paris, Beijing, London, etc.) and has access to a venue, and if >you /might/ be interested in hosting an event like this, please let us know.
>Our usage stats should have a pretty accurate distribution of users >around the world, so if the plan outlined below works, all we need in >theory is some venues plus local speakers... (and then a lot of love >from Alyssa, et al.)
>On 08/06/2012 07:36 PM, Kohsuke Kawaguchi wrote:
>> I think having a tried-and-tested format that can be replicated >> elsewhere would be nice. It will reduce the overhead involved in >> organizing an event.
>> If we can package this such that it can be held in a bigger conference >> room of a big company or an university classroom, I wonder if that opens >> up a lot of possibilities to replicate this elsewhere.
>> In Asia and Europe, food catering isn't necessary, and moving to a >> nearby pub after the event where attendees pay for themselves is quite >> acceptable. This is the proven model in Tokyo.
>> In US, we'll most likely need some delivered food, but I suspect the >> only reason it's so expensive is because the hotel robs you, and that >> will be a non-issue if we have the event in a company or an university.
>> One thing I do wonder is how important it is to keep the event free of >> charge (or keep it minimum just to make sure we won't end up a >> half-empty room with 50 people on the waiting list.) In Tokyo we never >> charged for an event and we just oversell the seats instead, and that >> was considered important.
>> In other parts of Japan, I've heard that it's mostly up to $5 or $10 >> even when we charge. That's a price someone can pay out of his own >> pocket, and I suspect that'd make it a lot easier for many (at least >> myself) to attend.
>> Just as a hypothesis, if we are to have this event in the Nokia campus >> in Copenhagen (I remember you guys had a pretty big room that fits 100) >> and in the above-mentioned format, would that drive down the cost enough >> that we can get away without spending many $1000s and charging people?
>> On 08/04/2012 06:31 AM, Lars Kruse wrote: >>> Hi Kohsuke and devs, thanks a lot for bringing up this topic.
>>> By direct encouragement from Kohsuke I will contribute to this thread.
>>> Just to set the context of my comment: Like CloudBees we (Praqma) have >>> our business model based pre-dominantly around Jenkins CI. Where >>> CloudBees focus on PaaS, we have our focus on business driven, sponsored >>> Open Source development and consultancy services related to tool support >>> of software development and verification processes (with Jenkins Ci as >>> the natural hub). We strive to support the Jenkins community in any way >>> we possible can. So besides our contributions to the pool of plugins we >>> also found it natural to help organize and fund some of the more costly >>> community activities - such as events.
>>> CloudBees and Praqma entered a formal partnership a few month back. At >>> about the same time where we had our first vision of a Jenkins User >>> Conference in Copenhagen. We have debated the format of such a CPH user >>> conference together with CloudBees: Should it be an official JUC event >>> (just another city along with Paris, New York, San Fransisco and >>> Antwerp) and simply squeeze it into the program as with Herzelia and >>> Tokyo or should it be a CIA event. Or something in between?
>>> As you may have noticed we had to settled on a different format than JUC >>> (and I realize now, that this is probably due to the relatively costly >>> format the JUC represent): So the CPH event in September is kind of a >>> new approach, and perhaps a trial of the new format.
>>> * We use the term "User Event" instead of "User Conference" - Actually >>> we see the gathering very much as a conference, but we did this >>> simply to deliberately distinguish it from CloudBees' JUC format. >>> * We start at noon - so attendees are expected to have their own lunch >>> before they show up. >>> * We continue to late in the evening, but the conference meal will be >>> sandwiches - enough to fill everybody, but not in any way fancy (a >>> note hereto: The conference meal I had at JUC in Paris was the best >>> conference meal I had in my entire career ;-) >>> * When the event officially ends at the conference hotel (19:00'ish), >>> we have invited the participants to join us to a cafe or pub in the >>> neighborhood for more relaxed discussions and free drinks - This >>> social after-event has it's own sponsor (Programming Research in the >>> CPH event). >>> * The token attendee give-away will be a mug instead of a T-shirt. The >>> sponsor value will be the same, but the production cost will be less >>> than half of what it cost to produce a T-shirt. And then we just >>> hope, that the attendees will like their mug as much a they would >>> have liked a T-shirt.
>>> In our case (CPH september) we didn't have a call for paper session - >>> but that is not necessarily part of the format we are suggesting, it >>> was simply because we had so many good speakers lined up in our network >>> (Nokia, Sony Mobile, Programming Research, Grundfos, CloudBees and >>> Ourselves) that this year with south short notice - we didn't need a CfP >>> session.
>>> The CPH user event is hosted and funded by Praqma and a generous sponsor >>> contribution is added from CloudBees (Thanks!!!) and as I mentioned - >>> the social after-event has it's own sponsor.
>>> ...And just to backup Kohsuke in the experiences he's sharing: Even >>> though we are trying to make this a cheap event (also for the attendees >>> - signup fee is only 25€) we are *still* going to loose a considerable >>> amount on money on hosting this event. And on top of that we are going >>> to spend an equally considerable amount of time on planning and >>> executing it.
>>> I'm not whining about it! Quite the opposite. We have a commercial >>> interest in Jenkins and we are happy to be able to 'pay-back' to the >>> community. But I think that two point are important for large events:
>>> * Any event will need a commercial sponsorship to become a success: >>> Cheap is good, but moneyless as in 'driven solely be initiative' - >>> is probably not going to happen outside the CIA format. >>> * If a commercial sponsor should contribute - they would want the >>> event to be 'officially accepted and approved' by the community in >>> order to shine in the light of the Google doctrine: "Do No Evil"
>>> As we in Praqma are developers ourselves we found that we had an urge to >>> fill in content that was developer oriented (as opposed to user >>> oriented) as well. But this represents a paradox; it's really difficult >>> to mix such different groups (users and developers) at the same >>> conference unless it's a big one with more tracks, so instead we have >>> decided to host a separate developer event along with the user event (a >>> full day workshop on the day before the user event).
>>> One of the complaints or critics we often hear from developers who >>> attended hackatons (including ourselves) is that they are often not very >>> well organized: People meet with no specific agenda, they spend a day >>> writing code (often alone) on the same stuff as they would have worked >>> on had they stayed at home - at the end of the day they eat pizza >>> together and break up.
>>> We have internally worked on developing a good format for a well >>> organized hackaton:
>>> It's a "Code Camp".
>>> It's inspired by open space technology/discussions (the term 'camp' is >>> also used within the open space approach): Developers are encouraged to >>> bring a few topics or issues to the camp, which starts with a short >>> introduction where each develop gives a few words of explanation to the >>> topics. Then there is a vote among the attendees: What are the >>> interesting topic that should be worked on? The topics that gets the >>> most votes win a time-slot on the code camp. Then we break out into >>> small groups - one per topic (no one works alone, all groups are at >>> least 3 participants).
>>> Following up on the inspiration from the Open Space technology the good >>> topics are characterized by: >>> 1) a real issue of concern >>> 2) a high level of complexity >>> 3) a high level of diversity >>> 4) a real or potential conflict >>> 5) a high level urgency.
>>> * The camp has it's own wiki - each break out session will have an >>> appointed scribe, who captures findings, results, deliveries, >>> discussions etc on the wiki. This supports that the summary-sessions >>> can be kept very short, anybody who wants to know what the other >>> groups worked with can find it on the wiki - and so can people who >>> didn't even attend the camp. >>> * The camp is well-organized although even though it's not >>> specifically
