Fork me on GitHub
#boot
<
2015-08-03
>
Petrus Theron10:08:09

What is the difference between boot's code reloader and figwheel?

juhoteperi10:08:17

@petrus: Figwheel is for Leiningen and Boot-reload for Boot

Petrus Theron10:08:59

@juhoteperi, I mean, is there a large difference I should be aware of? Like, how they reload namespaces? Are they functionally identical?

juhoteperi10:08:29

@petrus: They should be functionally identical for reloading

ragge12:08:59

Hi! Any updates on the next boot release? There are a few issues in 2.1.2 with the uber task that are preventing me from using it, that have since been fixed in master. Anything someone like me can do to help prepare for a release?

micha13:08:55

ragge: i've been remiss lately due to work and family stuff

micha13:08:53

@ragge: i'd like to go over the outstanding issues and the changes for this next release and make a nice changelog

micha13:08:20

any assistance with the changelog would be enormously appreciated

micha13:08:03

(i have been writing changelog in the message for the commit associated with the release build)

micha13:08:24

@ragge: what's your github id?

ragge14:08:14

@micha thanks, i def appreciate boot is an open source project with all the implications so I feel a bit bad about asking again... but as I said, anything I can do to help it along

ragge14:08:21

@micha i'm ragnard on github

martinklepsch14:08:57

As far as I understood micha creating a proper changelog would already help a ton simple_smile

ragge14:08:06

@micha i can definitely look into helping with changelog

ragge14:08:42

@micha how do you currently prepare the release commit? is there a changelog file somehwere?

ragge14:08:31

ie linking to closed issues

micha14:08:39

i'm not sure if we should have a CHANGES file or not

onetom15:08:59

@micha or @alandipert we were talking about whether it would make sense to try to turn our homepage into a single page app. i read the related https://github.com/tailrecursion/hoplon/issues/22 again and also the canonical google doc (linked from there) about ajax crawling (https://developers.google.com/webmasters/ajax-crawling/docs/getting-started) i looked into the pre-rendered source of http://hoplon.io and i noticed 2 potential issues which makes the sub-pages not turning up in the search results: 1, there is no <meta name="fragment" content="!"> in it 2, the links are just # simply instead of #!. the bang is supposed to suggest crawlability to the google-bot

onetom15:08:01

i was wondering if i could help to fix these, but a, im not super confident updating the related gh-pages branch b, i got 400+ warnings when i did boot1 dev on the master branch c, dont know if the boot2 branch already in production use d, i dont have access to the google analytics / webmaster tools to trigger a crawl and eval the results

onetom15:08:31

e, the boot2 branch doesnt even compile for me (`No such namespace: io.hoplon.content.getting-started`)

micha16:08:11

onetom: there is an issue with boot-hoplon that affects the haml parsing

micha16:08:26

caused by the way the package was made

micha16:08:40

i need to redeploy that and the warnings should go away

micha16:08:49

i want to rebuild http://hoplon.io and move documentation into the docs/ folder in the hoplon repo, so we can have versioned documentation

micha16:08:03

and http://hoplon.io be a landing site with examples and demos

micha16:08:11

for a "brochure" type site, like http://hoplon.io or a homepage you probably don't need to have anything special for AJAX crawling by webservers because all content is in the HTML if you've compiled with prerendering on

micha16:08:25

but if you have content in a database or dynamic content you can use the wrap-escaped-query ring middleware included with castra to render in phantomjs

micha16:08:48

you would of course need to make links with the #!... convention

micha16:08:23

i didn't do it for http://hoplon.io because there isn't any content created dynamically on the backend

micha16:08:33

(there is no backend)

onetom16:08:54

i have just arrived to the "no backend" conclusion, however to make the sub-pages recognized as individual pages, i suspect u have to use the #! convention and provide responses for the ?escapedfragment requests. it cant really be done with ghp or s3, thats true... the best solution from SEO perspective is still the generation of multiple pages... 😕

micha16:08:04

@onetom: you mean separate pages in the google result?

onetom16:08:31

it's more of a hoplon question though, so im not sure why are we talking about it here

onetom16:08:55

but that's clearly not acceptable how http://hoplon.io turns up in the search results currently

micha16:08:00

@onetom: what's not acceptable about it? it's #1 in all searches related to "hoplon", which is pretty good i think

micha16:08:11

shows up above the wikipedia entry, even

micha16:08:09

i didn't do any specific SEO optimizations other than just compiling with prerendering

onetom16:08:41

we have different products, pricing pages, educational resources, etc if someone searches for something specific, it's not satisfactory to just provide them the homepage

micha16:08:22

there are too many keywords associated with the single page is the actual problem?

micha16:08:44

there must be ways to alleviate that

micha16:08:11

google indexes single-page HTML books, for instance

micha16:08:28

perhaps something along those lines might have benefits

micha16:08:08

i don't know much about SEO but i'm keen to learn

micha16:08:41

but it's totally reasonable to divide the application into separate sub-applications, too

micha16:08:36

especially since you can have shared cljs (hoplon custom elements, etc) in jar dependencies used by all the different pages

micha16:08:59

that's how i organize my projects anyway

micha16:08:20

separate the ui kit into a separate project so i can use it across multiple applications

onetom16:08:57

im not sure whats the source of the misunderstanding here... if i search for myself for example, then the http://www.exiconglobal.com/about-us/ page should come up, but if i search for "exicon press release" the http://www.exiconglobal.com/news/ page should be the result. in either case if it would just be the homepage - http://www.exiconglobal.com/ - it wouldn't be obvious how to navigate to these subpages, no? the result would not be very relevant to the question

ej22:08:30

having problems with boot, datomic and reader literals

ej22:08:33

can anyone help?

ej22:08:47

keep getting NullPointerExceptions when running my datomic transactions wrapped in a fucntion

ej22:08:58

even with d/tempid being used.

martinklepsch22:08:27

@ej: no clue if this will help but that’s what came to mind

ej22:08:15

@martinklepsch: got this weird thing where if i save the file and auto-reload the ns, the (boot.core/load-data-readers!) applies

ej22:08:31

and transactions/queries work subsequently in the running application

ej22:08:39

but not before saving a file?

ej22:08:41

😞

martinklepsch23:08:07

@ej: I’m afraid I won’t be able to help more than pasting random links 😄

ej23:08:20

😞 been struggling with this for a few hours now

ej23:08:19

tried 😞 even ran (boot.core/load-data-readers!) in a server repl before running a datomic transaction function

ej23:08:28

still NullPointerException.