These are chat archives for reactioncommerce/reaction

5th
Apr 2015
Aaron Judd
@aaronjudd
Apr 05 2015 15:58
@faddat @Gouthamve see #357 for overview. that’s ‘reaction-drive’, which is currently the private project running reactioncommerce. It connects to launchdock and, yes, uses docker to launch sites. initially that was going to be a public project as well, but - I haven’t had time to make it public, - well- because it needs some cleanup
Jacob Gadikian
@faddat
Apr 05 2015 16:05
oh I understand about that — no worries there. I had a feeling there was some launchdock magic in there :)
Right now, we are working on containerizing (not necessairly dockerizing) these apps, probably for the platform apache brooklyn:
Ghost
Discourse
Reaction
Our Web Frontend (it is going to visualize organizational activity)
Taiga
Shit, little too tired…. Rest of the list is not flowing atm...
top-tier open source webapps
anyway— and then when we sell a subscription, we’re going to kick a percent back to the author— I’ve done some research and it seems few others do that— since you write one— what is a good/fair %?
The end goal of all of that
Aaron Judd
@aaronjudd
Apr 05 2015 16:10
re: reaction-drive & launchdock: we wrote launchdock pre-meteor 1.0, pre-docker 1.0 - there are better solutions, ways to do things,etc so I’m currently trying to decide the future of launchdock and the general direction we take there.
I mean, that code is like a couple of months old, so of course it’s really outdated already, lol
Jacob Gadikian
@faddat
Apr 05 2015 16:11
understood— we did a similar thing, too
yep
I think everyone got a little burned by docker because of doing that….
Aaron Judd
@aaronjudd
Apr 05 2015 16:12
looks like brooklyn is YaDM - yet another docker management tool
with docker-compose, swarm,etc not sure we need it - particulary with kubernetes etc...
re: subscriptions - not sure I follow -> are you talking about package rev shares?
Jacob Gadikian
@faddat
Apr 05 2015 16:14
yeah as in:
we want to sell subscriptions…. brooklyn is an awesome hybrid AND YaDM
when we sell subs to open apps
Aaron Judd
@aaronjudd
Apr 05 2015 16:15
we don’t have a firm number, but we’ve discussed 75/25 split
Jacob Gadikian
@faddat
Apr 05 2015 16:16
Seems fair to us. Basically we wan to make all those apps— about 20 or so
available and themed to businesses…. boom you’ve cloud.
Aaron Judd
@aaronjudd
Apr 05 2015 16:17
I hope I don’t see that on my headstone “ Boom.. you’ve cloud!"
but yes, that’s the general range I think would be acceptable, not sure what others think - but similar to other marketplaces
Jacob Gadikian
@faddat
Apr 05 2015 16:18
I wasn’t aware others were yet in the habit of paying— can you let me in a little on who and how much?
Aaron Judd
@aaronjudd
Apr 05 2015 16:19
I mean in general for app marketplaces, like Apple, google play, shopify, wordpress
Jacob Gadikian
@faddat
Apr 05 2015 16:20
ahhh gotcha
Aaron Judd
@aaronjudd
Apr 05 2015 16:20
where 75% goes to the developer...
Jacob Gadikian
@faddat
Apr 05 2015 16:21
yes, I had a bit of an inversion there— now, I can continue to talk about that model— you going to handle hosting?
Aaron Judd
@aaronjudd
Apr 05 2015 16:21
yes, that’s the goal, where you can push custom or off the shelf releases
Jacob Gadikian
@faddat
Apr 05 2015 16:23
okay may we game-theory it for a second? I think we just hit an incompatiblity of sorts but I’m pretty game to negotiate. I think that today what a lot of hosts do is akin to theft. It’s not theft, becuse GPL & the commons, but certainly a tragedy of the commons at least
so we want to as you know, it takes more thought and work to do something open than to do something closed. ANyway we want to do the hosting. Why, you ask? Well, we want to do the hosting because we want to make MANY OSS apps commonly themable
and sell them, with a kickback to the author. Under that model, what’s fair?
Aaron Judd
@aaronjudd
Apr 05 2015 16:27
well, not sure how that differs from the hosting companies like dreamhost, that for years have had projects like wordpress and magento included as one button installs on their hosting account - without support or paying the underlying projects
Jacob Gadikian
@faddat
Apr 05 2015 16:28
right yes— well the differences are two:
Aaron Judd
@aaronjudd
Apr 05 2015 16:28
at least from my view, that’s acceptable use, and helps to evangelize and expand the user base
if you choose to support the projects by sharing some revenue, I think that’s awesome
Jacob Gadikian
@faddat
Apr 05 2015 16:28
1) We value open code ,we want to pay to keep the ecosystem afloat
$=development
Aaron Judd
@aaronjudd
Apr 05 2015 16:29
as long as $=development support, doesn’t equate to direct control development - ala node.js
Jacob Gadikian
@faddat
Apr 05 2015 16:30
and yep it’s acceptable, but in my mind they should kick back. We publish our source, too. open souce is a grand human challenge, and it should be recognized that developing for the commons is far more beneficial to all than plain old development.
oh hells no
that’s like the first rule of metadev: Creative control to creators
We see it like this: If these people (open developers) got paid even a fraction of the value they created, they’d be able to create far more value
that is what makes brooklyn not just YaDM
(screenshot coming)
brooklyn.png
It has a heroku-esque bit, and it’s designed to— have you heard this term yet— I love it: “chase the sun"
Aaron Judd
@aaronjudd
Apr 05 2015 16:34
compose.yml in https://docs.docker.com/compose/ is doing that “compose model".. I’m actually testing this with tutum.co (today!)
so one problem I see - if you are going to pay the oss projects, who get’s the funding? not sure that would always be clear cut
Jacob Gadikian
@faddat
Apr 05 2015 16:37
That’s going to be harder and I think the project will need to do somethign akin to github.com (search: valnet )
Aesop7
@Aesop7
Apr 05 2015 16:37
Hi very quick noob question here... I installed the reactioncommerce:core package but didn't get the admin info in the console. Is it retrievable somewhere in the source docs?
Spencer Norman
@spencern
Apr 05 2015 16:38
easiest way I've found it to reset the server
and then run meteor again
Aaron Judd
@aaronjudd
Apr 05 2015 16:39
@faddat generally a major problem we hope to address - we’d like packages that are part of the reaction distribution to be tested, core reviewed,licensing approved, and have a maintenance plan in place. if there are some acceptance criteria, it would be easier to identify someone to compensate.
@Aesop7 meteor reset then meteor as @spencern suggests - you should see the login info
Aesop7
@Aesop7
Apr 05 2015 16:40
thanks @spencern & @aaronjudd !
Aaron Judd
@aaronjudd
Apr 05 2015 16:42
@Aesop7 are you trying to run reaction-core as a standalone package, or using reaction app? just ask because it’s been a long while since I’ve tested it as a standalone meteor package, and probably would require a few extra steps.
@faddat in any case, curious to see what you come up with
Jacob Gadikian
@faddat
Apr 05 2015 16:45
@aaronjudd well hey man— you’re a part of the journey :)
Aaron Judd
@aaronjudd
Apr 05 2015 16:50
I’ve got a dilemma for the project that I’d love to get some feedback on, even though I think I know the answer. Should we focus on #354 as part of the “beta” goals, or should we focuse on the order workflow fulfillment management first and come back to it. It’s either refactor now for that workflow and not worry about it in the future… or refactor later when all the order workflow is all settled down. I’m leaning towards just doing it now, as there seems to be a lot of ‘marketplace’ interest, but it’s a possible delay on other “use reaction for real now” blockers
certainly would be easier to do now, but I’m worried that it’s a less important use case to delivering a beta. I had previously chosen to focus on “single tenant”, just to keep scope down.
Spencer Norman
@spencern
Apr 05 2015 16:52
I've been thinking about this a bit too, but haven't looked too much into what is required to refactor the base/core to make the marketplace be a more natural add-on
I think that adding other marketplace components adds a significant amount of work pre-beta, but it does seem like the marketplace is becoming a more common use case
Aaron Judd
@aaronjudd
Apr 05 2015 16:55
well we started with marketplace in the back of our minds as a goal, so it’s not like it’s a total change - most of the structure should be compatible already -and I think the majority of the work is going to be permissions/roles/subscriptions.. but regardless, it’s a whole bunch of new things to test
Spencer Norman
@spencern
Apr 05 2015 16:55
Workflow fulfillment is obviously a necessary step to polish before even a beta launch, but if building that before a refactor makes the refactor much tougher, then it seems like a better order would be to get the core ready for a marketplace and then move forward
There's definitely some tension between launching earlier, or launching w/ marketplace and refactoring
I'd personally lean towards single-tenant and refactor later to keep scope down, but that also makes it easier to keep pushing the refactor back and adding on other components that will also have to be refactored.
Do you see the marketplace refactor as being a breaking change for third-party packages?
Aaron Judd
@aaronjudd
Apr 05 2015 16:58
some images of Easter sunday flagellation and open source project management are popping up in my thoughts now
Spencer Norman
@spencern
Apr 05 2015 16:58
haha :/
at least github isn't closed on Sundays
Aaron Judd
@aaronjudd
Apr 05 2015 17:01
No, I think we can keep the core from breaking (much). I think the schema etc shouldn’t require much, likely pub/sub and roles - and some minor updates to the copyToOrder workflow -> most of the work that would break things would be in the admin order work - but that’s where some contributions/feedback might avert trouble (not sure how much others have done in the order workflow)
Spencer Norman
@spencern
Apr 05 2015 17:02
In that case, my lean would be towards launching w/ single tenant. I have a feeling that launching anything in a usable state will bring on significantly more contributors which would be helpful either way
though might complicate a few things :circus_tent:
what are your thoughts on it?
Aesop7
@Aesop7
Apr 05 2015 17:04
@aaronjudd I'm trying it out as a standalone package
Aaron Judd
@aaronjudd
Apr 05 2015 17:11
@spencern well - one side of me says- we’ve been sorta staying under the radar - as we really want the core to be “solid" before we get too much momentum - if that makes sense. Completely locking down single tenant, would be a good launching point. The other side of me says - spend a little more time on marketplace will really help for the larger vision be more intuitive and eventually should grow us quicker. maybe. I guess I’m 50/50 - the swaying factor might be contributions from @boboci9 and @evliu and other contributors who are working marketplace-esque approaches (I don’t know the specifics of their goals though) (and probably save us some major refactors and breaking changes later)
Spencer Norman
@spencern
Apr 05 2015 17:14
Yeah, totally get staying under the radar. Only get to launch for the fist time once too, would hate to spoil the opportunity by launching something you feel is not the best representation of the platform
Aaron Judd
@aaronjudd
Apr 05 2015 17:18
realistic, small goals has always been a problem with my visions though.lol. thus, the question.
it’s like a big canvas painting… I’m starting with the numbers in the middle - but can’t wait, so I like painting the edges as well
(ok, weird analogy)
Spencer Norman
@spencern
Apr 05 2015 17:22
I think there is immense value and momentum that comes with actually pushing something out the door, but I'm also very interested in the marketplace side of things as well (and plan to contribute on that front as well)
So, I'm not being super helpful and kinda agreeing with your 50/50 I guess
Aaron Judd
@aaronjudd
Apr 05 2015 17:24
haha, well - that’s sort of a vote on the marketplace, even if logically we agree that pushing single tenant out the door makes more sense - but I think we should build what is most requested
Aaron Judd
@aaronjudd
Apr 05 2015 17:29
there is some self-centered thinking on my part going on as well - I think we’ll build a stronger solution if we acknowledge what our current community you wants/needs now.. and focus our efforts that way. If we pull together, instead of each building our own marketplace solution, then (we as a company) might have a better shot at launching something that community doesn’t necessarily feel like has to be done solo. I know everyone has their business plans, but ours relies on all you guys succeeding - and delivering something (the PaaS, prediction, marketplace analytics), that you’ll all want to use in addition to your own stuff
Jacob Gadikian
@faddat
Apr 05 2015 17:30
So:
AHA
YES!
Okay
Whom here is trying to advance next-gen e-commerce?
Spencer Norman
@spencern
Apr 05 2015 17:30
I think we all are :)
Jacob Gadikian
@faddat
Apr 05 2015 17:31
Okay, well you’re all invited
I’m so not bullshitting when I say that I’ve got a factory with 20ish offices…. and we use 2 of them right now.
We’re a development platform, please excuse our site
Spencer Norman
@spencern
Apr 05 2015 17:32
blob
Jacob Gadikian
@faddat
Apr 05 2015 17:32
I’d like to give anyone interested access to our quip folder with plans and what have you— just toss me your e-mail address
\?!
Aaron Judd
@aaronjudd
Apr 05 2015 17:32
@faddat yes - broken here as well
Jacob Gadikian
@faddat
Apr 05 2015 17:32
sorry
Will have to talk to Dan about what’s up with www in the morning.
We want to be able to do BGA/SMT boards downstairs, program them upstairs and ship them globally.
Can’t do it in China— why? No freaking Internet
Aaron Judd
@aaronjudd
Apr 05 2015 17:34
but all the cool developers are hanging out in Thailand ;-) lol
Jacob Gadikian
@faddat
Apr 05 2015 17:36
no hardware scene worth mentioning there
and logistics are even worse
Aaron Judd
@aaronjudd
Apr 05 2015 17:36
@DenisGorbachev still coding in Thailand?
@faddat I’m interested in see how you progress, for sure. (just kidding about Thailand - although if you find me there coding one day, you can say it was premonition)
Jacob Gadikian
@faddat
Apr 05 2015 17:39
HW is booming here in Hanoi…. though it could be the “all around wrong” kinda boom at the moment— knowledge isn’t diffuse enough surrounding next-generation techs for mfg
so we’re going to diffuse it
China, it’s kinda a situation where everyone and their cousin and their aunt can do BGA with just a heat gun…. (Which, I’ll be the very first to admit— is damned impressive)
Aaron Judd
@aaronjudd
Apr 05 2015 17:40
the only thing we can do with a heat gun in the US is bedazzle iphone cases.
ok - I’m about to mute my channels and get back to work..
Jacob Gadikian
@faddat
Apr 05 2015 17:41
kk
:)
Aesop7
@Aesop7
Apr 05 2015 17:46
Hi all, if anyone has some time, I'd love to see an example of how to change the index ("/") page yield using iron router (and reaction-core). Thanks in advance!
Aesop7
@Aesop7
Apr 05 2015 18:25
nevermind, i figured it out :smile:
Aaron Judd
@aaronjudd
Apr 05 2015 18:34
:thumbsup: feel free to make suggest some changes to the docs
Goutham Veeramachaneni
@gouthamve
Apr 05 2015 18:50
So, is any body working on the COD package?
I am assuming that COD would require some changes to the core. But if we can fake some things on the lines of paypal package (like authorise and capture automatically), it should work.
Goutham Veeramachaneni
@gouthamve
Apr 05 2015 18:59
What do you think?
Aaron Judd
@aaronjudd
Apr 05 2015 20:42
@Gouthamve see #129 (I'm not but somebody could / should )
Goutham Veeramachaneni
@gouthamve
Apr 05 2015 20:44
I think I will give it a shot. But it is extremely important for 3rd world countries. In india 80% of my sales are COD
Aaron Judd
@aaronjudd
Apr 05 2015 21:10
agree! it would be great to have someone who has some good ideas on how to approach this.