text
stringlengths
0
1.17k
[85.02 --> 93.94] setup takes only three minutes no excuses to have untested code in production head to codeship.io and
[93.94 --> 101.48] also check out their blog which i love by the way blog.codeship.io one more thing to mention for our
[101.48 --> 109.86] members um you can save between 294 and 2994 on your first year with codeship so make sure you take
[109.86 --> 115.18] advantage of that the changelog.com benefits if you're not a member what are you waiting for
[115.18 --> 120.82] membership is just 20 bucks a year and you support us to support open source once again
[120.82 --> 129.68] codeship.io sign up three minutes that's all it takes do it today and now on to the show
[129.68 --> 139.76] we're joined today by Amos Aviv and Roe Schwaber Cohen talking about Mean.js it's a product oriented
[139.76 --> 145.30] full stack javascript boilerplate so why don't one of you guys give us a uh well first of all why don't
[145.30 --> 153.28] you guys introduce yourself i guess Amos you can go first um so my name is Amos Aviv and i'm a developer
[153.28 --> 164.68] a web developer for about uh 12 years now um i've been through the days of i5 and quicksmod uh through
[164.68 --> 174.80] uh the birth of the uh new generation um browsers and finally to where we are today uh in this exciting
[174.80 --> 182.40] new world of javascript uh in in the server and stuff like that um i do a lot of um
[182.40 --> 192.48] side projects i guess um because uh you have to do something for your soul um
[192.48 --> 200.10] and one of the latest side projects i had was Mean.io um which evolved to be Mean.js
[200.10 --> 207.56] um yeah so there's there's a little bit of drama around that which we'll kind of touch on a little
[207.56 --> 212.34] bit um later in the show but we you know we'll kind of avoid that as much as possible just to kind
[212.34 --> 217.34] of shine a light on the good of Mean.js um so that's cool so you've been around you've been doing
[217.34 --> 223.68] this for a while yeah uh Roe yeah Roe why don't you so i've been a developer in israel um for the past
[223.68 --> 231.44] uh nine years i would say um yeah i come from a more of a php background but recently also found
[231.44 --> 238.76] the light and node um very interested in like all of what this framework has to offer um and then
[238.76 --> 244.60] angular is again one of the newer tools that i've started using in the past uh maybe a year or two
[244.60 --> 253.52] um and i'm currently working for a company called leafly here in seattle um yeah that's about it
[253.52 --> 259.36] awesome so why don't you guys give us a uh introduction to what Mean.js is um in its current
[259.36 --> 270.34] in its current form okay so um Mean.js is a full stack javascript boilerplate um it was born out of
[270.34 --> 279.02] uh our attempt at um a better flow for this kind of stack so the term was coined um in the MongoDB
[279.02 --> 287.32] blog actually um and we basically found out that it was super um efficient um almost basically wrote
[287.32 --> 294.74] the stack um i think it was uh six or seven months ago no no it was am i making a mistake yeah it's
[294.74 --> 304.62] about uh a year ago um the blog post was uh written by a developer called valery karpov he uh on the
[304.62 --> 313.02] mongodb blog um i used to do uh i'm a freelance developer so i i'm involved in several projects
[313.02 --> 320.82] and i recognized a certain pattern in projects where where developers used mongodb as the database
[320.82 --> 330.62] uh node as the web server and uh angular js as the as the client web framework um and decided uh
[330.62 --> 337.38] uh those projects could look better be better organized if uh they put in some sort of order
[337.38 --> 346.48] so i created uh a Mean stack uh which i used uh uh for about uh two months before we released it as
[346.48 --> 355.22] an open source um i i i thought the name Mean was cool and when i looked it up i found out uh
[355.22 --> 364.98] several people are already using it uh so i just released it um as Mean um to to address the namespace
[364.98 --> 378.02] i guess um it can it it basically um was constructed to offer uh MVC whatever um structure to both the
[378.02 --> 386.82] server side and the server side um in a way that um represents your entities properly um
[386.82 --> 396.66] i think what what i i tried to do uh was help developers from coming from strict type languages
[396.66 --> 407.76] with background in in a java um uh sp.net uh and stuff like that um step into the world of node.js
[407.76 --> 416.42] web applications and feel a little more comfortable um and so the point of Mean uh just sorry to interrupt
[416.42 --> 421.70] the point of Mean so it's it's it's opinionated right it's mongo express angular node and this is
[421.70 --> 428.02] something that you guys basically noticed was like a really common trend amongst uh for a node stack
[428.02 --> 434.56] right which was this thing called Mean uh what is it about Mean.js now i've used all of these things
[434.56 --> 441.00] and all of them are relatively easy to uh get started with on their own what is it about Mean.js that that
[441.00 --> 445.34] kind of makes it easier to get started with the whole stack rather than trying to do each one individually
[445.34 --> 452.08] so i think uh the most important uh stuff we wanted to address the most important issue we wanted to
[452.08 --> 459.42] address was um the interface between the different parts um you have your angular js application uh
[459.42 --> 466.58] running and you want to communicate with your node server so your node server um it should present
[466.58 --> 477.64] some sort of a rest api for angular js to use um and we wanted that um uh interface point uh to be
[477.64 --> 488.24] properly organized so when you download the stack you get um a folder structure uh and a couple of config
[488.24 --> 494.56] files that help you configure the different parts of the application like um the connection point between
[494.56 --> 503.62] node and uh MongoDB uh where we use the Mongoose module or uh a user authentication layer uh which uses the
[503.62 --> 511.62] the popular passport module and we wanted to give this all out of the box uh to feel to let uh developers just
[511.62 --> 520.96] uh begin writing their code uh instead of uh i don't know trying to figure out how to um build their project
[520.96 --> 531.46] um and just concentrate on building uh what they want instead of um the infrastructure which which i used to see
[531.46 --> 538.70] uh which i used to see developers uh taking a lot of time uh dealing with the infrastructure and the way
[538.70 --> 545.98] the proper way to uh connect those parts uh so right means we have a very i think we have a very
[545.98 --> 552.40] important um concept that we're trying to relay um with mean which is first of all we chose um
[552.40 --> 562.18] only if not um like mostly if not only popular um components so um we intentionally chose to go with
[562.18 --> 568.24] Mongo instead of couch we intentionally chose to go with express instead of some other um um framework
[568.24 --> 575.06] uh that does routing we chose to reduce uh our jade imprint eventually because all of these things
[575.06 --> 580.82] were were were coming from the community so we're trying to pick um again components that are really
[580.82 --> 586.60] popular from the one hand and from the other we really wanted to not abstract away the simplicity of
[586.60 --> 591.98] of all these frameworks so we don't want to create some sort of like a layer that takes all this
[591.98 --> 596.62] complexity away from you as a developer we wanted it we wanted you to still be hands-on and really
[596.62 --> 603.36] understand how the parts work but um that doesn't necessitate um you learning each of the components
[603.36 --> 609.02] and like really knowing how to integrate them properly um and we found that it was really easy
[609.02 --> 615.30] to start um creating vertical stacks uh which means like all the stacks in the server all the all the
[615.30 --> 622.26] parts of the server and the client for for a given entity um it made that whole process a lot easier and a lot
[622.26 --> 628.92] faster right so this is different from a lot of our listeners are rubyists and this is different from
[628.92 --> 634.46] rails and that rails is its own thing right it builds all the layers into its own thing and and
[634.46 --> 638.60] this would be similar if you guys were to say build your own express that had its own templating and
[638.60 --> 644.00] its own database and its own you know all that mean it takes the other tools so you still have
[644.00 --> 649.54] complete control over the tools individually and it allows you to but it makes it easier to kind of
[649.54 --> 654.68] connect it all together yeah i think one of the more the more uh prominent patterns we saw with uh
[654.68 --> 661.74] with mean is the amount of forks um that were kind of unusual to yeah about um 1000 forks already
[661.74 --> 667.38] yeah so like yeah the total the total is about 1000 forks and the reason for that is people
[667.38 --> 672.12] like different flavors of mean and they like different flavors of full stack javascript really
[672.12 --> 677.74] um so if we look at this entire picture we're we're of the mind that um all these flavors are
[677.74 --> 683.94] totally valid and we're not necessarily opinionated um in a way that says like our our flavor is better
[683.94 --> 688.82] than any other flavor it's just that we found these tools to be super popular and super helpful
[688.82 --> 694.54] and powerful and we thought that they were our our best choice for the scenarios where we were
[694.54 --> 701.04] at the time so i was um um developing for a startup called go mango and i had to rewrite my website
[701.04 --> 706.14] really quickly and it was sort of built with a lot of fragmented pieces of jquery and angular
[706.14 --> 712.08] and node in the back and some php and some dotnet and all all these systems were working together
[712.08 --> 717.78] but not very well um and my first experience with mean.io was when i basically converted my website
[717.78 --> 722.94] that was built before that um it took about eight months to build and i converted it in about two
[722.94 --> 732.30] weeks so i feel i really felt a very significant and real um uh change in in the speed uh of my
[732.30 --> 737.78] development and i think it really it really has a lot to do with a the fact that we're doing only
[737.78 --> 743.94] javascript um and not like switching between languages which is immensely helpful and it's
[743.94 --> 750.94] sort of easier to keep on thinking sort of the same way um with with thinking and and i and i specifically
[750.94 --> 756.96] mean about thinking about modularity and thinking about you know um asynchronous um uh workflow instead
[756.96 --> 761.70] of a synchronous workflow all of that together creates this effect of everything is so much easier
[761.70 --> 768.22] and faster gotcha so the the idea that you're using like mongo instead of couch or react and express
[768.22 --> 773.74] instead of happy or getty or you know different different things uh that's that's your choice but
[773.74 --> 778.78] but what happens when like somebody forks this and it's reen right with all the same stack except
[778.78 --> 783.24] react instead of mongo and that one kind of takes off as the popular fork like that's that's a real
[783.24 --> 789.44] problem that this open source project faces um and so what happens there we're actually not yeah we're
[789.44 --> 794.62] not seeing it as a problem yeah we're not this is a great yeah how else you can go so we're not
[794.62 --> 801.52] viewing it as a uh we're not viewing it as a problem um we actually find this as a great opportunity