The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2015-05-28

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:23 Ptolemarch joined #mojo
00:27 jberger yapc soon!
00:27 jberger http://imgur.com/5HiEXRi
00:27 jberger taken the last time I was there
00:33 Grinnz joined #mojo
00:52 asarch joined #mojo
01:18 fhelmber_ joined #mojo
01:45 klapperl_ joined #mojo
02:19 zivester joined #mojo
02:25 davido_ joined #mojo
02:29 zivester joined #mojo
02:45 absolut_todd joined #mojo
02:46 noganex_ joined #mojo
03:10 basic6_ joined #mojo
03:23 howitdo joined #mojo
03:34 mattastrophe joined #mojo
04:06 sri wow, that new hacker show Mr Robot is actually good :o
04:07 sri that name made me expect the worst
04:11 diginet_ joined #mojo
04:54 mattastrophe joined #mojo
05:01 bpmedley http://pastie.org/10211300 <— It’s late, so no one is there.. however, would people use a CM API that looked somewhat like that?   This is an example that bootstraps Mojo with Postgres.  Perhaps this is offtopic; however, I’m wondering about deployment.
05:04 lluad How you do depency management and how you make sure things are idempotent are two of the more important things about CM for me, and they tend to affect the overall structure and the DSL pretty heavily.
05:05 preaction I think Rex does a good job, honestly
05:05 bpmedley That makes sense.  I’m pretty new to all that stuff; however, I don’t want a DSL.  Just plain perl.
05:05 preaction Rex is basically just plain perl
05:06 bpmedley Hrmm
05:06 lluad Well... you're writing a DSL, even if it's implemented as just perl.
05:06 preaction the docs leave a bit to be desired though
05:06 bpmedley I saw a bunch of tasks and whatnot.  I guess those are simply subs or somesuch?
05:06 preaction basically yep
05:07 preaction this guy is making a series of blog posts that might serve as a better intro than the regular site: http://big-elephants.com/2015-05/rex-in-practice-infrastructure-as-code/
05:07 preaction i've put my hat into the ring to help with their docs, but i dunno if i'm approved yet. i might just spend a bit of the weekend trying to clean it up a bit
05:08 preaction the last straw was earlier this week when a coworker found Rex::CMDB, which i hadn't seen despite using Rex for months now, and which solved our prevailing problem of getting rid of our old homebrewed config/deployment engine
05:08 preaction it says DSL, but it's DSL the way Moo(se) is a DSL
05:10 preaction i will say i'm not fully satisfied with some of the design decisions that were made, but since i'm not yet crazy/stupid/annoyed enough to fork it...
05:10 bpmedley Rex does look awesome; would it be possible to modify to something like minion for deployment?
05:11 vmbrasseur joined #mojo
05:11 preaction i don't think that'd really get you much
05:11 bpmedley Do you think minion with postgres would scale better?
05:14 preaction for config management and deployment?
05:15 bpmedley My understanding is that Rex uses SSH only.  Is there a practical upper-bound for deploying a script to a set of boxes?   Also, please keep in mind that I’ve never really used a CM tool.. ;)
05:18 preaction it can do it in parallel or serially. that post says you can even get gearman involved, though i only manage about 20 machines, so i don't even need to parallelize it at all
05:18 bpmedley Understood.  I assume Rex works well for you?
05:18 inokenty joined #mojo
05:18 preaction and yes, it uses ssh to run tasks, though other things (like rsync) might not require an ssh connections
05:19 preaction i use it for my releases, and we're starting to use it more for our config management stuff
05:19 bpmedley Cool stuff.
05:19 preaction and yeah, it's been working well, even though i mostly only use the run() command
05:20 preaction a guy in the chicago.pm is introducing it to where he works, duckduckgo, i believe for provisioning dev boxes
05:21 bpmedley I’ll have to ponder if it makes sense to continue, or just use Rex.  As an aside, may I ask what you think of the pastie?
05:23 preaction it looks extremely similar. i might forgo the package stuff for simpler constructors: cmd(), file(), service(), install()
05:23 bpmedley That’s a thought!
05:23 preaction i'm a fan of CFEngine, but i hate its config syntax
05:23 bpmedley Is that the YAML one?
05:23 preaction but the idea is: Describe the desired end state. for most things, it should be very obvious what one needs to do to achieve it
05:23 preaction no, that's Ansible i think
05:24 preaction thinking like "If this machine is a <web node> it must <have a firewall line unblocking port 80>, it must not <have a firewall line unblocking port 22>" that kind of thing
05:25 polettix joined #mojo
05:25 preaction huh. now that i think about it, i've never seen a system that can do "must not"
05:27 bpmedley Thx for looking
05:28 preaction ... sigh... and now i'm thinking about my own... and wondering which one it'll end up looking like...
05:32 bpmedley Your own project or your own CM tool?
05:32 preaction own CM tool
05:32 bpmedley How would you organize it?
05:33 preaction it'd be very role-oriented
05:33 vmbrasseur joined #mojo
05:33 bpmedley What would you use for deploying the recipes/scripts?  Pure perl or  DSL, or ?.
05:33 preaction servers consume roles. environments contain servers.
05:33 preaction servers can be part of multiple environments, i guess. i have that situation unfortunately
05:34 preaction roles describe: tasks, files, packages, etc...
05:34 preaction it'd probably be very close to Rex in that regard: A perl-based DSL
05:38 sri there is something in rex that actually uses minion
05:38 preaction oh, right, their web app
05:38 preaction they've got a web app to run rex tasks
05:38 sri https://metacpan.org/requires/distribution/Minion?sort=[[2,1]]
05:38 bpmedley Wow, totally missed that.. gaw..
05:38 sri not many reverse dependencies yet :o
05:39 preaction as mentioned, their docs leave quite a bit to be desired
05:40 inokenty-w joined #mojo
06:09 schetchik joined #mojo
06:13 dod joined #mojo
06:15 kaare joined #mojo
06:19 dod joined #mojo
06:26 gatitskiy joined #mojo
06:54 amon joined #mojo
06:59 AndrewIsh joined #mojo
07:11 eseyman joined #mojo
07:13 dod joined #mojo
07:28 trone joined #mojo
07:32 inokenty joined #mojo
07:38 dotandimet joined #mojo
07:39 gatitskiy joined #mojo
08:00 schetchik joined #mojo
08:11 Vandal31341 joined #mojo
08:13 schetchik joined #mojo
08:13 gatitski_ joined #mojo
08:36 zivester joined #mojo
08:44 geira joined #mojo
08:46 kwa joined #mojo
09:02 fhelmber_ joined #mojo
09:12 ladnaV joined #mojo
09:12 geira joined #mojo
09:24 geira joined #mojo
09:27 polettix joined #mojo
09:34 geira joined #mojo
09:53 cpan_mojo Mojolicious-Plugin-Restify-0.01 by KWAKWA https://metacpan.org/release/KWAKWA/Mojolicious-Plugin-Restify-0.01
10:13 cpan_mojo Mojolicious-Plugin-Restify-0.02 by KWAKWA https://metacpan.org/release/KWAKWA/Mojolicious-Plugin-Restify-0.02
10:28 Snelius spam release
10:28 Snelius )
10:32 kwa I fixed it in the 2nd one. :)
10:32 kwa First upload to CPAN, so didn't really know what I was doing...
10:33 oetiker joined #mojo
10:38 kwa Snelius: unless "spam release" has other connotations I don't know about?
10:42 bpmedley kwa: First upload?
10:43 pink_mist kwa++ #congrats on your first upload!
10:43 kwa bpmedley: Yeah. Always been a bit scared of doing it.
10:43 bpmedley kwa++  … let’s get drunk.. ;)
10:43 kwa Cheers pink_mist!
10:44 kwa Haha. I do feel like celebrating. I'm sat in the office, so all I can have is a coffee. I'll make it really strong I think.
10:44 bpmedley lol
10:46 kwa The trepidation I've always had is what if something goes wrong. So when I uploaded it and it wasn't right, that helped me get over that...
10:46 kwa Shit always happens I guess. You just fix it if it does. :)
10:48 pink_mist this is why I like Dist::Zilla ... it helps one avoid the pitfalls involved in making something cpannable; all you really need to worry about is the code itself
10:48 polettix joined #mojo
10:50 kwa Yeah, that does sound good.
10:50 kwa I read a lot of stuff to try and get my head around the process, and learned just enough to get it on cpan. Now it is, it'll give me context for learning more about it.
10:53 Snelius kwa: no no, good job! :)
10:56 kwa Good good.
10:56 purl Great Googly Moogly!
10:57 pink_mist botsnack
10:57 purl :)
11:16 neilhwatson joined #mojo
11:22 inokenty joined #mojo
11:40 dustinm joined #mojo
12:26 eseyman joined #mojo
12:31 Mikey joined #mojo
12:31 marmez joined #mojo
12:42 gatitskiy joined #mojo
12:46 al joined #mojo
12:59 eseyman joined #mojo
13:04 ZoffixWork joined #mojo
13:33 zivester joined #mojo
13:33 jberger kwa: welcome to the club
13:34 jberger warning, it can get addictive
13:36 Grinnz kwa, Dist::Milla is a nice easy way to get into Dist::Zilla, if you're using a github repo and its other assumptions apply
13:47 kwa jberger: cheers. It's actually easier than I though it would be. Already thinking about another one.
13:47 kwa Thanks Grinnz. Will give it a read.
13:50 Grinnz Dist::Milla actually does most of the same things that my DZ setup does, but i have it setup to do the release commit and version bump commit separately, thats really the only reason i can't directly switch to Dist::Milla lol
13:53 ZoffixWork kwa, the beauty is that you can make your own plugin bundle and then making a new dist involves just modifying a few values in a 5-line file: https://metacpan.org/source/ZOFFIX/JSON-Meth-1.001004/dist.ini
13:53 ZoffixWork That dist.ini does all the stuff specified by my bundle, which includes quite a bit: https://metacpan.org/pod/Dist::Zilla::PluginBundle::Author::ZOFFIX
13:55 Grinnz Dist::Milla is just another bundle, with a wrapper script, really
13:56 ZoffixWork Yeah, but Dist::Zilla evokes memories Godzilla, which is really cool :P
13:56 jabberwok thoroughly modern Milla ?
13:57 jabberwok Q: would it be appropriate and OK to use the Perl Raptor in a logo for a perl mongers group on Meetup, with the attribution & license in the "About Us" section −?
13:58 jberger plugin bundles are the devil, I thought we had already established this
13:58 jberger :p (kinda)
13:58 ZoffixWork jabberwok, yes, it's licensed under Creative Commons: https://github.com/kraih/perl-raptor/blob/master/README.md
13:59 jabberwok =bow=  ZoffixWork
14:18 absolut_todd joined #mojo
14:43 dotan joined #mojo
14:45 asarch joined #mojo
14:47 dotandimet joined #mojo
14:50 dvinciguerra joined #mojo
14:57 zackiv31 joined #mojo
15:00 robinsmidsrod joined #mojo
15:07 sri \o/ https://github.com/rurban/Cpanel-JSON-XS/issues/37#issuecomment-106102700
15:09 sh4 joined #mojo
15:10 jberger very cool
15:28 jberger wow, check this out: http://www.theregister.co.uk/2015/05/28/sourceforge_accused_of_shackling_gimp_in_kinky_adware/
15:28 absolut_todd joined #mojo
15:31 jberger http://arstechnica.com/information-technology/2015/05/sourceforge-grabs-gimp-for-windows-account-wraps-installer-in-bundle-pushing-adware/
15:31 Grinnz_ oh jeez
15:32 Grinnz_ “Mirrored projects,” SourceForge explains, “are sometimes used to deliver easy-to-decline third-party offers
15:32 Grinnz_ in other news, everyone still hates flash and java
15:32 marcusr (how low can you go)
15:34 jasanj joined #mojo
15:36 jasanj hi, all. Is it possible to implement a syntax like this in ep template ?
15:36 jasanj http://pastie.org/10212107
15:37 jasanj Or Why I didn't need this
15:37 marcusr jasanj: You probably want to make helpers instead.
15:37 jberger jasanj: what is imports and what is macros
15:37 jberger jasanj: what is imports and what is **macro**
15:38 jasanj marcusr: yes I have ask the question several days before , and got the answer is using helper
15:38 jberger so macro is itself a helper?
15:39 jberger what is imports?
15:39 purl imports is a lot saner
15:39 jberger shut it purl
15:39 jasanj jberger: imports and macro is a helper
15:39 jberger jasanj: the include helper already exists btw
15:39 jberger and macro is just a helper that, what, builds helpers?
15:40 jberger I think you are over thinking this
15:41 jberger jasanj: can you please start at the beginning and tell use what you are trying to achieve
15:41 jberger (oh gods I've become mst)
15:41 jasanj jberger: include just render the template to string I think, I want use a helper 'imports' to load the macros in other templates
15:41 marcusr purl: step back?
15:41 * purl backs into a monster!
15:41 jberger but helpers are already available in templates, you don't need to import them
15:43 jberger jasanj: if you had stayed around before, you might have seen my reply
15:43 jberger http://irclog.perlgeek.de/mojo/2015-05-24#i_10649206
15:44 jasanj I'm feels worried for using helper to define a global subrutine, that makes every template file can using that helper, is it not safe ?
15:44 jberger in which I link to a very simple implementation of a helper like the ones you link to in the python templating: http://pastie.org/10205063
15:45 jberger jasanj: yes helpers are global
15:45 jberger though they can be namespaced
15:46 ToApolytoXaos joined #mojo
15:47 jberger and also note that lots of the helpers you seem to want (from looking at the python doc you link to) are already available
15:47 jberger http://mojolicio.us/perldoc/Mojolicious/Plugin/TagHelpers
15:49 jberger if you are really worried about global helpers, you can instead make Controller methods which will then only be accessible in templates rendered by that controller
15:49 jberger but I rarely go that far myself
15:50 jberger (though Roles like that, which could be consumed by Controllers might be a fun idea)
15:51 jasanj controller method means I need write this $c->somefunc() in template ?
15:51 jberger yes, but so would namespaced helpers
15:52 jberger remember though, ep templates are just perl, you could in theory "use" inside them
15:52 jberger (I can't remember how the sandboxing works though, templates might reuse sandboxes)
15:53 * jberger looks
15:55 jberger yes sandboxes are reused, so I wouldn't export too many functions into them
15:56 jberger (that is how helpers work too btw)
15:58 initialo joined #mojo
15:59 punter joined #mojo
16:03 Onigiri joined #mojo
16:09 jasanj jberger: thanks, and global helper maybe is enough, another question about python's macro syntax, for using macro, I can define several helper function in one template file, How could I do this in ep template ? using render_to_string will only define one help in one template
16:11 jasanj so I need seperate a template file into serveral parts, and write helper for them, is there a better way to do this ?
16:11 jberger jasanj: you can bundle your helpers into a plugin, that plugin can itself have templates possibly even in its data section
16:12 jberger so yes, I think what I'm saying is that you probably want to have a template file per helper
16:12 jberger but again, investigate templates in data sections
16:14 jasanj jberger: I see, thank you for this
16:14 jberger http://mojolicio.us/perldoc/Mojolicious/Guides/Rendering#Bundling-assets-with-plugins
16:26 dod joined #mojo
16:32 sri we once used a new namespace for every template, the memory usage was quite substantial
16:34 jberger right, and helpers weren't free then either
16:35 kmx joined #mojo
16:35 jberger that's what I thought I remembered (I was part of that change)
16:35 jberger (but it was a while ago)
16:45 sri hehe, someone on the list doesn't believe Mojo::Pg is fast because you have to write the query twice
16:46 jberger twice?
16:46 purl somebody said twice was a rogue pfy or happenstance
16:47 jberger I told you to shut it purl!
16:47 sri you pass the sql to ->query() with every call
16:57 marcus joined #mojo
16:58 marcus hmm
17:06 ZoffixWork joined #mojo
17:30 inokenty joined #mojo
17:36 trone joined #mojo
17:52 damaya joined #mojo
17:53 gatitskiy joined #mojo
17:55 damaya I'm making the choice between Mojo::Pg and DBIx::Class. What are your thoughts on this choice?
17:55 damaya I.e., what would be the things you'd consider, and what would make you choose one over the other?
17:56 Grinnz DBIx::Class is an ORM
17:56 Grinnz Mojo::Pg is not
17:56 damaya Well, Mojo::Pg can also do non-blocking requests, I guess.
17:56 Grinnz Mojo::Pg is what you should use instead of DBI, DBIx::Class is a completely different decision
17:57 damaya I'm curious exactly what kind of difference that would make in the app. Say I have 500,000 users, would the non-blocking aspect make it faster, better?
17:57 Grinnz non-blocking queries certainly helps it scale, yes
17:57 damaya ok, then my decision is made
17:58 damaya I don't necessarily need an ORM considering I'm comfortable writing SQL
17:58 Grinnz as long as you do it right, but delays make that easy
17:58 damaya I do love DBIx::Class though
17:58 Grinnz i feel the same, i used DBIx::Class once and it was great, but i havent had a need for it
17:58 lluad But scaling is overrated. Realistically, your app is not going to need to scale. Mojo::Pg, though, is simple to use, so you can write apps quickly. And it doesn't hide the database, which even the best ORMs do.
17:58 Grinnz "your app is not going to need to scale" what?
17:59 damaya That's my other consideration. Sometimes ORMs abstract out the SQL so much that it's incredibly difficult to do something that's supported by the database but not the ORM.
17:59 Grinnz my app at work already has to scale to thousands of clients, and one of my side projects if used would have to scale to much more
17:59 lluad Hobby and startup apps, and version 1 of MVP apps ... almost never need to scale.
18:00 lluad People whose apps need to scale almost never ask abut that on IRC, they just engineer to suit.
18:00 Grinnz i don't agree
18:00 mst damaya: DBIx::Class is more about metaprogramming SQL than hiding it
18:01 mst basically the question is "do I want to write a complete custom model, or do I want to re-use existing code"
18:01 Grinnz you don't get the best knowledge for making a Mojo app scale on google, you get it from this channel and the mailing list
18:01 mst both are valid choices
18:02 mst lluad: I don't remember seeing you in #dbix-class so I find it moderately unlikely you have opinions about DBIx::Class rather than about ORMs-in-general
18:02 lluad I like, and use, DBIx::Class.
18:03 mst the 'hiding the database' thing sounded like the usual david fetter ranting :P
18:03 damaya mst: My comment was not in relation to DBIx::Class, because I can do anything I want with DBIx::Class
18:03 Grinnz s/DBIx::Class/zombocom/g
18:03 mst there are plenty of things wrong with DBIx::Class, most of which are my fault ... but 'hiding the database' isn't really one of them
18:03 damaya I've got ltree stuff working with DBIx::Class, and as we discussed the other day I have a helper that returns nextval for a sequence.
18:04 mst yeah
18:04 damaya I have worked with ORMs that do hide things though, and make it incredibly difficult to run SQL code.
18:04 mst plus, if I need to do something really complicated, I stick it in a view and let DBIC talk to that
18:05 mst since DBIC basically gives no fucks what it's talking to, you can lie to it if you need to ;)
18:05 mst right. I hate such ORMs
18:05 damaya I think the only thing that would sway me towards Mojo::Pg is the non-blocking aspect, but that's mostly because I don't fully understand the pros/cons of that, and it sounds good. haha
18:05 mst I've often described DBIx::Class as "one of the few ORMs developed by people who actually like databases"
18:05 damaya haha
18:05 Grinnz damaya, well it's pretty simple: a worker can handle other requests while some query is running
18:06 lluad Mojo::Pg is great if you want to throw SQL at the database. It doesn't give you much of a framework beyond that. And while it can do async calls, it can't do async transactions, so depending on what you need to do that might be useful, or not.
18:06 Grinnz yeah, the other benefits of async don't really apply since you can still only do one per connection at a time
18:06 mst in general, I find the database server is the limiting factor and so a lot of time asyncness doesn't help a lot
18:07 Grinnz right, but for slow queries, the non-blocking part helps
18:07 mst not really
18:07 mst tying up a web process is irrelevant when the limiting factor is the database server
18:07 Grinnz not helps it return faster
18:07 Grinnz helps it not block the worker
18:08 mst which is irrelevant
18:08 mst spin up an extra web node and move on
18:08 Grinnz that doesn't seem simpler
18:09 mst web nodes are cattle
18:09 lluad The webworker that can be reused is probably a lot cheaper than the database backend, so even if you can do it perfectly it's not going to save a huge amount of resources.
18:09 bpmedley You’ll save RAM if you use the async methodology.
18:09 mst wat
18:09 bpmedley You can do more with fewer web workers.
18:09 Grinnz the database backend doesn't matter, if you need to scale that there are a number of solutions that don't affect the web application
18:10 sri lluad: async transactions?
18:10 Grinnz but a thousand requests that each have to wait 1s for a response, and a thousand requests that can wait 1s for a response at the same time, is a big difference
18:10 mst fundamentally, I'd rather spin up an extra web node and spend the time I've saved not having to think about async code doing something useful like optimising the database
18:10 mst 99% of the time this is a better trade-off
18:10 sri lluad: if you mean async queries during a transaction, that's totally possible
18:11 sri whole reason i stole scope guard transactions from DBIx::Class
18:11 sri they work great with async
18:11 lluad AIUI, you can't use the samy database connection during multiple steps of a delay?
18:11 Grinnz thats mostly because DBI sucks
18:12 sri lluad: you mean sequential?
18:12 lluad Yeah.
18:12 sri that works
18:13 sri only concurrent queries on the same connection are not possible
18:13 sri which is not actually a DBI limitation
18:13 Grinnz wasnt there an issue with an async query not fully returning?
18:13 sri postgresql couldn't handle it even if DBI did
18:14 lluad Yeah, that's not the issue I'm thinking of.
18:14 sri postgresql is kinda using the apache prefork model of connection management
18:14 sri one worker process per connection
18:14 lluad You can't do ->delay(sub { db stuff A }, sub {db stuff B}) and have syuff A and stuff B use the same connection.
18:14 sri and no concurrent queries per worker
18:15 sri lluad: you can
18:15 lluad I was trying to do that, hit a bunch of weird bugs, asked her and was told it was a known issue with DBI.
18:15 lluad Maybe I'll try that again and see if I can pull out a test case
18:17 sri Grinnz: or did i forget something?
18:20 mattastrophe joined #mojo
18:22 Grinnz sri, https://github.com/kraih/mojo-pg/commit/5aa941743183842f3a7109e7795bcd951a3e18d7
18:23 sri oh right
18:24 sri i was just about to add a test case, but can confirm that it doesn't work
18:25 sri for the record https://gist.github.com/anonymous/a83538e7a2e0e6d4d19f
18:26 sri no error message, DBI just fails
18:26 Grinnz the related issue: https://github.com/kraih/mojo-pg/issues/3
18:26 sri Grinnz++ # better memory
18:27 sri Grinnz: no that was a different bug
18:28 sri argh
18:28 sri it was two issues
18:28 sri just don't listen to me today i guess :)
18:32 sri it's only bugs in DBI/DBD::Pg though, theoretically it should be possible
18:33 lluad Yeah, it's just going to take some DBD::Pg spelunking.
18:34 schetchik joined #mojo
18:34 lluad I've not found it much of a limitation in the code I'm writing right now, though.
18:37 sri damn, it's only been 2 months and i had already completely forgotten about it :O
18:40 sri guess that information has been overwitten with knowledge about hearthstone cards :p
18:41 lluad Much more important. :)
18:46 berov joined #mojo
18:48 vmbrasseur joined #mojo
18:49 Grinnz_ lol
18:52 disputin joined #mojo
18:55 schetchik joined #mojo
19:04 amon joined #mojo
19:10 bpmedley_ joined #mojo
19:40 _dave_ ever played spacechem? you'd likely be very good at it. actually...don't go looking for that game...it will timesink you really badly. :)
19:42 schetchik joined #mojo
20:01 firnsy joined #mojo
20:05 sri nope, i could use a real game again though now that i'm done with diablo 3
20:07 sri too bad there's no new bioshock in the works
20:09 howitdo joined #mojo
20:09 basic6 joined #mojo
20:11 Grinnz_ bioshock infinite was pretty awesome
20:27 _dave_ spacechem is a puzzle game to be sure...not sure what your definition of "real" is
20:35 Adurah Puzzle games huh, Huniepop is a puzzle game.
20:39 jberger _dave_: while I see the appeal, that looks a lot like designing algorithms
20:39 jberger which I'm not sure I can do on the time that I'm not writing them for work or for open source
20:40 _dave_ well some people need a simple FPS to balance these things
20:44 jberger I play widelands sometimes
20:44 jberger though I guess it gets old quickly, but it helps sometimes
20:44 jberger things are always moving (RTS) but you have enough time to handle them, unlike StarCraft
20:49 VinsWorldcom joined #mojo
20:49 _dave_ never seen widelands, thanks for the pointer to it...I'll check it out
20:49 VinsWorldcom left #mojo
20:50 _dave_ I had been playing a lot of space engineers, but I could use a genre change
20:50 jberger it was a clone of settlers 2
20:50 jberger but I think it has now expanded beyond slightly
20:51 _dave_ given my current frustration level with javascript toolkits, I might go play some tribes again
20:52 _dave_ it always helps to shoot people :/
20:52 * Grinnz_ is in the middle of his second Dragon Age Inquisition playthrough, and is planning the third
20:52 _dave_ I like that game a lot it's just...
20:52 _dave_ well when I one shot the dragons with my rogue
20:53 _dave_ it gets boring
20:53 Grinnz_ difficulty setting too low :P
20:53 _dave_ nightmare =P
20:53 Grinnz_ then you didn't fight them early enough in the game!
20:53 _dave_ lol!
20:53 _dave_ rogues are OP
20:53 Grinnz_ obviously you need to start a new game :P
20:54 jberger I might not be playing widelands on a hard enough setting either, to be fair
20:54 Grinnz_ i like FPS but it has to be a certain type of FPS... like bioshock, or borderlands
20:54 _dave_ I suppose I could play on nightmare with 4 tanks ...
20:54 Grinnz_ an FPS designed for a single player campaign, i guess :P
20:55 Grinnz_ _dave_: i'm contemplating playing nightmare with no tanks
20:55 _dave_ well Tribes is the only FPS I'll actually play against others...too many cretins out there
20:55 _dave_ and even tribes has it's issues
20:55 Grinnz_ but as a squishy mage thats probably a bad idea
20:55 Grinnz_ i miss the Arcane Warrior from DAO
20:55 Grinnz_ i tanked the shit outta the late game bosses
20:56 _dave_ hm it's been a while since my last DA:I incursion but you'll have fun in the early game
20:56 _dave_ with 4 mages
20:56 Grinnz_ well i was thinking 2 mages, iron bull and sera/varric
20:56 _dave_ iron bull can likely tank :)
20:56 Grinnz_ that was my other thought
20:56 _dave_ even though you aren't supposed to build him that way
20:56 Grinnz_ the two-handed tree does have a whole tanking side pretty much
20:57 _dave_ yeah but...reaver...so tempting ;)
20:57 Grinnz_ yeah...
20:57 Grinnz_ my first playthrough i played a reaver... fun class
20:57 _dave_ I am always a mage on new dragon age games
20:57 Grinnz_ dragon rage is OP as shit
20:58 Grinnz_ i would end every dragon battle doing thousands of damage a hit and ending up with like 1 hp
20:58 _dave_ heh what's the rogue skill? 1000 strikes or something?
20:58 Grinnz_ i don't have that tree... but in artificer there's one where every archery ability is doubled for 40 secodns
20:59 _dave_ I had a build for sera where the hardest dragon in the game didn't even wake up before it died
20:59 Grinnz_ considering some of the archery abilities are pretty damaging already when used properly...
20:59 Grinnz_ lol
20:59 Grinnz_ it's what i always liked about dragon age though, you can build up your characters to do insane stuff
21:00 _dave_ indeed
21:00 _dave_ I bet they've patched a few issues since I've played last
21:01 Grinnz_ actually what confuses me the most is that tanking ability in tempest
21:01 Grinnz_ like, wtf is an archer going to be suddenly tanking for
21:02 Grinnz_ its not permanent or that would be a viable tank
21:03 _dave_ if memory serves, you use that after you've gained agro away from the tank ;) and you -will- do that
21:03 Grinnz_ heh i suppose... but the upgrade for it makes it TAUNT things
21:03 _dave_ tank only had two taunts as I remember
21:03 _dave_ one area and one single target
21:03 Grinnz_ i think champion gets another
21:04 _dave_ you don't always have to upgrade skills...I think they put a few bogus skill upgrades there just to make you wonder
21:04 Grinnz_ heh
21:04 Grinnz_ yeah some of them are wtf
21:04 _dave_ now if you run 3 mages and a rogue...maybe that's useful
21:05 _dave_ but an ice mage can "tank" sort of
21:05 _dave_ for a small bit lol
21:05 Grinnz_ rogues have plenty of ways to get out of the way if they pull aggro, but mages
21:05 _dave_ are you familiar with the term "agro mez"?
21:05 Grinnz_ not really
21:05 _dave_ it's when you get agro but you are faster or have some distance between the mob
21:06 _dave_ so you run in circles
21:06 _dave_ and it's an effective "mez" (the mob is not doing damage)
21:06 Grinnz_ ah
21:06 Grinnz_ well on my rogue as an artificer i just drop the spike trap when i pull something, and then gtfo
21:07 _dave_ right but you could also taunt the mob as a rogue and use that jump back skill...agro mez
21:07 Grinnz_ yea
21:07 Grinnz_ you can actually get two jump back skills, one in archery one in stealth
21:07 Grinnz_ but its a waste of slots :/
21:08 Grinnz_ thats the only thing that really pisses me off about DAI, only 8 ability slots and no way to use the rest
21:08 _dave_ I seem to remember a way to get infinite respecs
21:08 Grinnz_ in DA2 you could use any ability from the pause menu
21:08 Grinnz_ im just talking about mapping to the action menu
21:08 _dave_ ah yeah
21:09 _dave_ that bit
21:09 _dave_ it's too bad bioware is no longer bioware but bioware/ea
21:09 _dave_ the very first dragon age was awesome
21:09 Grinnz_ i'm in the minority but i really liked DA2
21:10 _dave_ that was good too, I seem to prefer the first one over all of them
21:10 _dave_ not saying all are bad either, just my preference
21:10 Grinnz_ dont give a crap about the environments, the combat was better, the AI was better, the story was just as good
21:10 Grinnz_ imo of course
21:10 _dave_ of course :D
21:10 Grinnz_ and then in DAI they stripped the AI :/
21:10 Grinnz_ that was half the fun
21:11 _dave_ there's AI in DAI?
21:11 Grinnz_ just whether they'll "use" or "prefer" each ability, and some basic behaviors
21:11 _dave_ right
21:11 _dave_ I hated that too
21:12 _dave_ I manual all the characters in DAI
21:12 _dave_ everything is off
21:12 _dave_ *pause* plot plot plot *unpause* watch *pause* ...
21:12 Grinnz_ yeah i'm going to try to do that for my nightmare run, but on lower diffs i prefer to focus on one.. that was easier when you could set up the AI properly
21:12 _dave_ the thing is, even when you plot actions sometimes they do something different
21:14 Grinnz_ i didnt really realize how great the DA2 settings were until they're gone... you could set a rule to use some ability when a certain class of character got low on health, or do something else when there was a grouped up group of 5+ enemies... etc
21:15 _dave_ actually, while they are good, I rarely used the AI in any DA game. Usually it was only for battles I knew were trivial.
21:16 _dave_ any new mob I fought was completely manual...so I could react properly or ... in some cases... discover how to react properly
21:16 Grinnz_ well the tactical view was only in DAO PC version and not in DA2 at all, so i never had that
21:17 _dave_ you don't play on PC?
21:18 Grinnz_ for DAI i do, not back then
21:36 Grinnz_ i'm debating holding off my third playthrough so i can try witcher 3 though :P
21:38 marty joined #mojo
21:41 disputin joined #mojo
22:09 vmbrasseur joined #mojo
22:17 bjakubski joined #mojo
22:45 mattastrophe joined #mojo
23:03 sri still not many PG_ASYNC users :S http://grep.cpan.me/?q=PG_ASYNC
23:11 * sri is suspicious of AnyEvent::DBD::Pg, because it has nested async queries https://api.metacpan.org/source/MONS/AnyEvent-DBD-Pg-0.03/ex/sample.pl
23:12 sri no actual tests though
23:33 disputin joined #mojo
23:34 zackiv31 joined #mojo
23:40 dvinciguerra joined #mojo

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary