louquillio has quit [Remote host closed the connection]
louquillio has joined #nanoc
kitallis has joined #nanoc
kitallis has quit [Ping timeout: 252 seconds]
kitallis has joined #nanoc
kitallis has quit [Ping timeout: 255 seconds]
kitallis has joined #nanoc
<ddfreyne>
Keltia: posted
<Skyr_>
musicmatze: Uh, can't say :) Upgraded my hardware 18 months ago, bought something reasonably priced (not the latest sh*t)
Skyr_ is now known as Skyr
Guu has left #nanoc [#nanoc]
bghost has joined #nanoc
bghost has left #nanoc [#nanoc]
<bobthecow>
Keltia: did you get your thing sorted?
<bobthecow>
the "missing layout" bit?
<ddfreyne>
bobthecow: Maybe guard-nanoc should work and do the compilation in a child process so that loaded code doesn’t stick around. What do you think?
<bobthecow>
and let everyone else deal with the eventual insanity.
<ddfreyne>
Yeah, I was thinking about doing that…
vukini___ has quit [Read error: Connection reset by peer]
vukini_ has joined #nanoc
yogsototh has joined #nanoc
<musicmatze>
I have an Idea: I dealt around with datamapper some time ago. Maybe it is possible to write a datamapper backend for nanoc to store the current state of the compiler in a sqlite or mysql database on disk!?
<musicmatze>
Or even remote, if someone has a usecase for this. With this, it would be possible to set up a distributed-compile some time...
<ddfreyne>
musicmatze: That’s certainly an interesting area and something like that has been scheduled for 4.x
<musicmatze>
Nice.
<ddfreyne>
musicmatze: nanoc already caches a lot of state in tmp/ so that recompiles run faster (but not as fast as they could be, and the amount of in-memory caching is too big, which is why you’re seeing the out of mem errors)
<ddfreyne>
nanoc is already halfway there already, actually… but it can be improved
yogsototh has quit [Remote host closed the connection]
<musicmatze>
that's fine. What I think about is a module system for nanoc which offers the possibility to "plug in" different storage backends for example. .. say JSON for small sites, sqlite (datamapper) for bigger ones or even mysql for large ones.
<musicmatze>
Just an Idea!
<ddfreyne>
musicmatze: That is definitely also on the to-do list!
<bobthecow>
i have one valid use case for disabling pruning: assets with cachebusters in the URL might want to leave old versions around.
<bobthecow>
in fact, i think i want to implement #postprocess for 3.x for a similar reason :)
<bobthecow>
i'm cool with prune-by-default, but auto-prune is problematic.
<bobthecow>
s/auto-prune/force-auto-prune/
<ddfreyne>
bobthecow: I’m not sure whether cache busting is problematic here
<bobthecow>
i've got a javascript app that has a bunch of asset URIs it references.
<bobthecow>
so if you have the app open and are interacting with it, you're expecting the same asset URIs to be there as when you started interacting.
<bobthecow>
so ideal for me is pruning with a combination of age + no longer needed.
<bobthecow>
i could have three different copies of foo-HASH.js if i compiled three times within whatever window i decide to keep 'em around for.
<bobthecow>
i'm not saying nanoc needs to support such a scheme, just that it shouldn't prevent me from implementing it.
<ddfreyne>
So postprocess do ; prune ; end would be okay for you? Maybe with options to prune (like… block or regex that indicated which ones to skip?)
<bbommarito>
Is there some irony here: Perusing sites built using nanoc, and Github uses it for documentation yet they invented Jekyll
<bobthecow>
bbommarito: jekyll is a blog engine.
<bobthecow>
hardly the right tool for the job for a documentation site :)
<ddfreyne>
bobthecow: Any other remarks except for the #sort?
<bobthecow>
oh, no. sorry. i'll merge.
<bobthecow>
i got distracted.
<ddfreyne>
bobthecow: Merging that pull request == closing 4 other issues :D
<ddfreyne>
33 open issues → 28 open issues
<bbommarito>
Well, I'm again peeking at nanoc as I now have ideas how to do archives without...having archive pages or pagination, and while Jekyll would work, it's a pain in the butt to do multiple 'blogs' and I have to use CodeKit for everything if I want to build in say Sass...
<bbommarito>
Jekyll also makes it somewhat tricky to inject different head elements depending on if something is a 'post' or not...trying to see if nanoc will be any more forgiving on doing that.
<ddfreyne>
bbommarito: nanoc is pretty flexible
<ddfreyne>
bbommarito: For example, for stoneship.org I used to have different backgrounds for different articles (my Braid review had a wonderful Braid background, for instance)
<bbommarito>
There are certain things I want to achieve, and two of those are: Based on the 'section', have all categories listed in the sidebar, and all month/year listed in the sidebar (So may 2013)
<bbommarito>
What editor did you go from -> to ddfreyne?
<ddfreyne>
bbommarito: TextMate a long time ago, now vim
yogsototh has joined #nanoc
<bbommarito>
ddfreyne: Have you messed with Sublime at all?
<ddfreyne>
bbommarito: Nope, but I have heard quite a bit of praise
<ddfreyne>
I reworked the Stoneship site a few days ago and now I have different layouts based on the kind of content (normal pages and review pages)
<bbommarito>
It's not a bad editor, and you do get the vim movement keys and all that...but...the big issue is that with those vim movement keys, no matter how fast you put your key repeat up, you can't just hold down j for instance, and keep going down.
<ddfreyne>
I never hold down keys in vim
<bbommarito>
Can nanoc do something like yield(:header) by any chance?
<ddfreyne>
Copyright (C) 2009 tm2gtksw2(Alexandre da Silva)
<ddfreyne>
That’s my theme…
<bobthecow>
it seems to be automatically extracted with tm2gtksw2
<bobthecow>
which adds the copyright header.
<ddfreyne>
Yep, pretty naughty
<bbommarito>
Alright, it's official: I have to get out of this stupid office job and back into remoting.
<bbommarito>
Crohns disease + office does not always work well, plus...well the job sucks anyway.
<ddfreyne>
4 weeks and I’m quitting my job :)
<ddfreyne>
No more commuting…
<ddfreyne>
The highway from Ghent (where I live) to Brussels (where I work) has been shut down since yesterday morning
<ddfreyne>
Gas leak
vukini_ has quit [Remote host closed the connection]
<ddfreyne>
So I’m working at home today again
vukini_ has joined #nanoc
vukini__ has joined #nanoc
vukini_ has quit [Read error: Connection reset by peer]
<bbommarito>
ddfreyne: Congrats on quitting your job. I am assuming you have something lined up.
<bbommarito>
Or have become extremely wealthy.
<ddfreyne>
bbommarito: Yep. I’m moving to Berlin and starting at SoundCloud in 9 weeks
<bbommarito>
So, here's my job in a nutshell: Drive 30 miles to work, sit and stare at a computer for 6-7 hours (Because...yea) maybe make a tiny fix to some website and then drive 30 miles hom.
<bbommarito>
home
<bbommarito>
Everything could /easily/ be done remotely when my stomach has decided to be stupid but the owner of the company is "You must be here to actually be productive" type of person while...I am /far/ more productive remote
<ddfreyne>
Yeah. I’m working from home today and it is quite cool
yogsototh has quit [Remote host closed the connection]
<bbommarito>
I have, in a month and a half, probably worked 10 hours total, and the rest has been...not even trying to look busy, just sitting there.
vukini_ has joined #nanoc
<ddfreyne>
bbommarito: That sucks :(
<bbommarito>
So, 60 mile roundtrip to stare at a screen.
<bbommarito>
In fact, this job is costing me more, because I have animals on my farm, and I have to pay someone to care for them while I am gone.
yogsototh has joined #nanoc
vukini__ has quit [Read error: Connection reset by peer]
<ddfreyne>
bbommarito: My commute takes about two hours per day
<bbommarito>
Yep, about what mine takes.
<bbommarito>
Sad part: If I started applying in San Francisco, I would have a job very quickly...remote is not so simple.
* Skyr
currently spends 2 hrs/day on the highway, too
<Skyr>
This is simply not good
<Skyr>
From time to time, personal presence is important.
<bbommarito>
Skyr: Time to time, yes, every single day to stare at a screen, no...
<ddfreyne>
I found a neat appartment (but expensive) about 5 minutes on foot away from the SoundCloud offices :D
<Skyr>
But if not, telecommuting is beneficial for everyone. Less stress, lower costs, good for the environment, ...
<Skyr>
bbommarito: I once worked in a company where the boss required (beside presence every day) daily reports on what was done
<Skyr>
Written reports, that is.
<bbommarito>
Skyr: Oh, if someone ever said "We need written reports on what you did", I would tell them to go screw themselves and walk away.
<Skyr>
The same boss once complained about the cost for the company when a worker sent a non-business info mail to all colleagues (in this case it was the info on a new tennis club nearby with very low rates for children)
<Skyr>
bbommarito: Guess where I'm no longer working :)
<bbommarito>
Skyr: There are things I will tolerate, and things I will not...and that would be one of them. Second thing I will not tolerate: Stupid logic questions at an interview. No, I will not tell you how far a bowling ball can drop from a 100 story building...
<bbommarito>
And how do you add a task to my list saying to fix a spelling error, when it's a damn graphic that I cannot edit?....grrrr
<Skyr>
bbommarito: Logic questions in an interview seems to be an American habit ;) I never encountered such a thing in Germany...
<bbommarito>
Skyr: I think it is an American thing. Companies here think "Well, Google and Microsoft do these, so we need to do it" when neither Google nor Microsoft have done those types of questions for years.
<bbommarito>
Or, when in the Ruby on Rails realm, Pivotal told them they should ask those questions
<Skyr>
*gg*
yogsototh1 has joined #nanoc
<bbommarito>
And I didn't know Soundcloud was over in your neck of the woods ddfreyne, for some reason I thought they were a SF company.
yogsototh has quit [Ping timeout: 245 seconds]
<ddfreyne>
bbommarito: Nope, their HQ is in Berlin, but they do have operations in SF
<bbommarito>
I would love to move to Berlin...I love Germany.
<bbommarito>
Or, should I say the more proper Deutschland.
<bbommarito>
Speaking of: I think it's time for a little Der Rhythmus der Hitze
Segismond has quit [Quit: Leaving]
<bbommarito>
Okay, 6 AM, so now must shower, so I can eat at 7, so I can leave at 8...god I hate schedules.
Segismond has joined #nanoc
<bbommarito>
Anyway.
<bbommarito>
Looks like I may not be going in...roads out of town are flooded.
<ddfreyne>
Hehe :)
<ddfreyne>
Where do you live?
<darix>
so thumbnailize now also knows about darktable-cli ... galleries directly from raw files at some point
<bbommarito>
ddfreyne: City called Memphis, Michigan.
<bbommarito>
Population of 1183 humans...(Note humans, there are more animals than humans), 1 square mile, and I own 75 acres of that (So a tenth of the city)
<bbommarito>
Basically it's all farmville...pretty dull city but...meh...it's home
jugglinmike has joined #nanoc
louquillio_ has quit [Ping timeout: 264 seconds]
louquillio_ has joined #nanoc
<guardian>
ddfreyne: what's the modern way of capturing content for a sidebar then?
<guardian>
darix: the page explicitly says "Showing captured content in a sidebar the old, deprecated way (do not use or I will become very angry)"
<guardian>
hence my question :)
<darix>
guardian: you scroll down too much
<darix>
guardian: found it?
<darix>
ddfreyne: if i want to update the metadata for an item on the run. is that possible or discouraged?
<ddfreyne>
darix: Not possible
<darix>
hm k
<ddfreyne>
During compilation, all metadata is frozen
<ddfreyne>
You can modify it during preprocessing though
<darix>
nanoc add-token-for-hidden
<darix>
hmm ok
<darix>
will do that :)
<guardian>
found it
robnormal has joined #nanoc
<guardian>
ddfreyne: would that fly? http://pastebin.com/f5XQ0Wig I'm annoyed capturing captures raw content and you somehow have to duplicate the capturing logic for captured content (or be very rigid about captured content and impose it's always a specific format)
<ddfreyne>
guardian: Hmm, that won’t work. What you *could* do, however, is split up items like that in multiple items in the preprocessor
<ddfreyne>
guardian: The item is the basic data type in nanoc
<guardian>
I was suggesting nanoc may address that in a future version, without requiring splitting items into sub items as part of preprocess
<ddfreyne>
guardian: I've thought about that (Hakyll supports it) but I don’t think it’s quite worth supporting. It increases complexity for no apparent reason
<guardian>
is there a way to implement a capturing helper that would use the same filter chain as the item in which it's invoked?
<ddfreyne>
guardian: Capturing doesn’t really do any filtering at all, so I’m not sure I understand your question
vukini_ has quit [Ping timeout: 252 seconds]
<guardian>
I want the captured content to be filtered, not raw
<guardian>
and that would be the same filter chain as the item owning the captured content
<stbuehler>
perhaps you could add "magic" markers and have a filter capture it later in the chain
yogsototh1 has quit [Remote host closed the connection]
<guardian>
stbuehler: that or something based on naming: some_item.md and some_item_sidebar.md
<guardian>
having related content in different files might become cumbersome. having a magic marker that survives a whole filter chain may turn being fragile
yogsototh has quit [Ping timeout: 246 seconds]
vukini has quit [Remote host closed the connection]
vukini has joined #nanoc
<ddfreyne>
<ddfreyne>
oops
robnormal has left #nanoc ["Leaving"]
vukini___ has joined #nanoc
vukini has quit [Read error: Connection reset by peer]
vukini___ is now known as vukini
LBo has quit [Quit: Oogjes dicht en snaveltjes toe...]
Evolution has joined #nanoc
<Evolution>
I'm running into a very simple Rules processing issue that I can't seem to work myself out of. Anyone got a second to help me parse something?
<Evolution>
I can't seem to find docs on this, mostly because it's simple enough I *should* be able to figure it out.
<Evolution>
I created a blank site with nanoc create-site foo. created a config.rb for compass, and then used compass to create a blueprint sass template.
<Evolution>
that works, but I'm having trouble with the subdirectory it creates for partials. I'm not smart enough to write a proper Rules route/compile setup for it.
<Evolution>
minus the default compile * and a couple other bits
<ddfreyne>
Hmm, yeah
<ddfreyne>
It kinda sucks that nanoc’s identifier matching is a bit odd (not like filesystem globs)
<ddfreyne>
bobthecow: nanoc 4.x should have glob-like syntax!
<darix>
ddfreyne: the check if 2 files are identical ... you really compile every file for that check right?
<darix>
is there an option to maybe limit that check to mtime comparison between source and target?
<Evolution>
the one other thing I've run into seems to be that with nanoc if you're using sass, the container directive seriously hates fluid/percentage based column styles.
<ddfreyne>
darix: It’s not compared if nanoc can infer with 100% certainty that it’s not necessary
<ddfreyne>
darix: mtimes are used
<darix>
ddfreyne: well it seems it redoes all my image processing all the time
<ddfreyne>
darix: All the time is weird… changes to config.yaml and lib/ will cause everythign to be recompiled though
<darix>
i called nanoc compile twice in a row
<darix>
nothing changed
<ddfreyne>
darix: Hmm. Does it still occur if you rm -rf tmp/ ?
<darix>
ok rm -rf tmp ;
<darix>
nanoc compile first run 53s
<darix>
2nd run 0.51
<darix>
so yeah tmp was to be blamed
<ddfreyne>
Hmm, shouldn’t have happened though
<darix>
well i like that it doesnt reprocess all images :)
<darix>
that test set is still rather small
<darix>
only 187mb
<darix>
i have 3gb images waiting for it once i am done ;)
<ddfreyne>
darix: adding new items will cause the site to be recompiled, too, though. The dependency tracking isn’t ideal :(
<darix>
hmm
<darix>
lets see
<ddfreyne>
3 GB is going to be pushing it… I’d say preprocess those up front
<ddfreyne>
Something much better is scheduled for nanoc 4.x though
<darix>
ddfreyne: well should my preprocessing then write to the out put paths as defined in the routing?
<ddfreyne>
darix: I mean create the thumbnail versions outside of nanoc… at least in nanoc 3.x
<ddfreyne>
(Maybe even with a simple makefile)
<darix>
yeah... though write the thumbnails to content our output?
<darix>
s/our/or/
<ddfreyne>
content, I’d say
<darix>
so how to teach nanoc then which items are related to each other?
<ddfreyne>
darix: if they have similar names, you could find their relationship that way
<darix>
and a proposal for --verbose output
<darix>
if you jump into "i need to reprocess to compare" state
<darix>
could you log something like
<darix>
"reprocessing item <identifier> to compare: <reason>"
<darix>
"unsure [<timing>] reprocessing item <identifier> to compare: <reason>"
<ddfreyne>
darix: nanoc show-data ← this gives a reason for outdatedness already
<darix>
ah
<darix>
next time i run into the issue i will check that
<ddfreyne>
There’s some srs changes coming up in nanoc 4.x there though
<stbuehler>
iow, everything is going to break :D
<ddfreyne>
stbuehler: Yeah, I have a “screw backwards compatibility” attitude there :)
<ddfreyne>
Which is good, because it allows me to do a serious spring cleaning.
<stbuehler>
hehe
<ddfreyne>
nanoc 3.x is 3 years old, which is quite a bit
<stbuehler>
so
<ddfreyne>
Actually, closer to 4 years
<stbuehler>
does nanoc4 have a good solution for generating assets from items?
<ddfreyne>
And when 4.x will old, it’ll be more than 4 years
<ddfreyne>
stbuehler: What do you mean by assets and what do you mean by good?
<stbuehler>
my old example still being embedded latex -> png
<ddfreyne>
stbuehler: Embedded LaTeX inside a page and letting it generate stuff on the fly? Hmm, not really. I don’t have a great way of tackling that either
<stbuehler>
well, my asset registry is working fine... but i think you want to add delayed (auto)prune, so this might be a good time to combine these ideas
<stbuehler>
i think that is what i'm currently using :)
bghost has joined #nanoc
<ddfreyne>
stbuehler: You should move that into a separate branch so you can easily rebase that on the latest release branch
<stbuehler>
ddfreyne: i don't think branch names will give me any rebase problems :)
<ddfreyne>
stbuehler: No, but you could put your changes onto 3.6.x easily then
<stbuehler>
ddfreyne: i don't see how the branch name would be involved with that. but release-3.4.x indicates that i tested it with 3.4.x, using a different name could be misleading
<ddfreyne>
stbuehler: If you let your feature-asset-registry branch start off from the original release-3.4.x branch, you can easily rebase it onto later release-3.4.x branches and even onto release-3.6.x, so you can keep up with nanoc development while retaining your branch
<stbuehler>
you mean if someone wanted to use my feature (branch), i could update them to a new release without them realizing it? :)
<ddfreyne>
stbuehler: You could keep your feature-asset-registry branch in sync and if other people want to use it (or yourself) then they can both use your asset registry stuff and the latest nanoc 3.x features and bug fixes
<ddfreyne>
Merging is better than rebasing actually… for public stuff at least
<stbuehler>
well, for now it is more a way to show you what could be useful upstream. i don't intend to keep the api stable as long as it isn't merged upstream somehow, so rebasing is actually fine (so people will realize it might break)
<stbuehler>
also i'm quite familiar with git and know that the public master branch should be fast-forward only :)
<stbuehler>
perhaps i should rename the branch "wip"
<ddfreyne>
stbuehler: I’ll take another look at your branch and see how it can be integrated with nanoc 4.x
<ddfreyne>
There won’t be any big changes in 3.x anymore (maybe a couple of useful features though)
<stbuehler>
no pressure :)
<stbuehler>
i just thought if you are making big changes anyway it might be a good time to mention it again
<stbuehler>
also it sounded like you need some registry for the delayed prune feature anyway
<ddfreyne>
It is indeed… I added it to the wishlist
<ddfreyne>
stbuehler: I haven’t really though about that yet
* ddfreyne
is off to sleep, night!
<stbuehler>
n8 :)
jugglinmike has quit [Quit: Leaving.]
jugglinmike has joined #nanoc
<guardian>
can I define helpers in Rules?
<guardian>
or just in the preprocess block maybe?
<darix>
guardian: uhm you can have functions in preprocess and call those yes.
jugglinmike has quit [Quit: Leaving.]
jugglinmike has joined #nanoc
jugglinmike has quit [Client Quit]
louquillio has quit [Remote host closed the connection]
jugglinmike has joined #nanoc
jugglinmike has quit [Client Quit]
<bobthecow>
ddfreyne: nanoc 4 should *only* support regex or globbing — via the built-in ruby globbing library — for matching identifiers.