| --- Log opened Tue Mar 02 00:00:09 2010 |
| 00:07 | | You're now known as TheWatcher[zZzZ] |
| 00:30 | | Serah [Z@26ECB6.A4B64C.298B52.D80DA0] has quit [Ping timeout: 121 seconds] |
| 00:38 | | Attilla [Attilla@FBC920.65CFFF.ADEEA3.DBDA33] has quit [[NS] Quit: ] |
| 00:46 | | * ToxicFrog ambushes Reiver with pointers |
| 01:10 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has joined #code |
| 03:21 | | gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has quit [[NS] Quit: Z?] |
| 03:36 | | * Orthia screams and dies, for he is embarrased by his mastery of Java syntax |
| 03:39 | < ToxicFrog> | o.O |
| 03:40 | < Orthia> | I am stlil trying to implement the ackermann function |
| 03:40 | < Orthia> | The bits that keep tripping me up are goddamn System.Out.println() etc. Having to look them up constantly is rather horrific. >.< |
| 03:43 | < ToxicFrog> | ...you shouldn't really need to look up anything for ackermann apart from System.out.println and Integer.parseInt |
| 03:43 | < Orthia> | Well, I was trying to use it as a testbed for accepting commandline args as well. |
| 03:47 | < ToxicFrog> | Right, but those are passed straight in |
| 03:47 | | Orth [orthianz@Nightstar-94f97f6e.xnet.co.nz] has joined #code |
| 03:48 | < ToxicFrog> | <Orthia> Well, I was trying to use it as a testbed for accepting commandline args as well. |
| 03:48 | < ToxicFrog> | <ToxicFrog> Right, but those are passed straight in |
| 03:49 | | Orthia [orthianz@Nightstar-168d9349.xnet.co.nz] has quit [Ping timeout: 121 seconds] |
| 03:50 | < ToxicFrog> | Anyways, I always found the JAva docs to be quite complete; the problem lies in the fact that while they're great at telling you what a given class does, they're terrible at telling you which class you want. |
| 03:54 | < Orth> | That is pretty much precisely my problem, yes. |
| 04:00 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has joined #code |
| 04:07 | < celticminstrel> | Hehe. A lot of documentation seems to be like that... |
| 04:07 | < ToxicFrog> | celticminstrel: the JAva docs are worse than most because they are huge. |
| 04:07 | < ToxicFrog> | Something like Pilot or Lua you can hold the entire library in your head, so it's not that hard to figure out what you need. |
| 04:07 | < celticminstrel> | Maybe. I haven't looked at them recently, so I don't remember how huge they are. |
| 04:12 | < ToxicFrog> | Orth: anyways. For Ackermann specifically, you should be able to get away with: |
| 04:12 | < ToxicFrog> | - basic program structure |
| 04:13 | < ToxicFrog> | - System.out.println (string output) |
| 04:13 | < ToxicFrog> | - Integer.parseInt (string -> int conversion) |
| 04:19 | < Vornicus> | Python has a really big library but usually there's only one thing that you actually want and it's vaguely properly named. |
| 04:21 | | celticminstrel [celticminstre@Nightstar-f8b608eb.cable.rogers.com] has quit [[NS] Quit: *hums* Can't stay now!] |
| 04:29 | < Orth> | Remind me how you write a function and I think I will be fine. |
| 04:30 | < Vornicus> | <options> <return type> <function name>(<type> <parameter name>, <type> <parameter name> ...) |
| 04:30 | < Vornicus> | then a { and the function body |
| 04:30 | < Vornicus> | oh, and after the ) you need throws if it can throw anything. |
| 04:32 | < Orth> | Ahem. |
| 04:32 | < Orth> | Hn. |
| 04:32 | < ToxicFrog> | frex: public static int parseInt(String str) throws NumberFormatException { /* code goes here */ } |
| 04:32 | < Orth> | Ach! Danke, TF |
| 04:32 | < Orth> | I was missing the int. |
| 04:33 | < ToxicFrog> | ("public" being "other classes can see and call this method" and "static" being "this method is not associated with any object, but is called through the class") |
| 04:36 | | Orthia [orthianz@Nightstar-d75b0f99.xnet.co.nz] has joined #code |
| 04:38 | | Orth [orthianz@Nightstar-94f97f6e.xnet.co.nz] has quit [Ping timeout: 121 seconds] |
| 04:39 | | * ToxicFrog gets out the staples |
| 04:43 | < Orthia> | Frakking internets ;_; |
| 04:43 | < Orthia> | So, hn. |
| 04:45 | < Orthia> | TF: Also, I did not seem to find a plugin to let it give me a cheat-sheet of valid functions as I type |
| 04:45 | < Orthia> | This may be because I did not install the plugin correctly; I'm not really sure how you activate them once installed. |
| 04:46 | < ToxicFrog> | Plugins->Plugin Manager->Manage tab |
| 04:46 | < ToxicFrog> | They should default on once installed, though |
| 04:46 | < ToxicFrog> | You might try ctrl-B (Complete Word) too |
| 04:47 | < ToxicFrog> | I'm afraid I'm not going to be of all that much help here; I use jedit mostly for bash, lua, and DSLs |
| 04:49 | < Orthia> | np, was just wondering given you know it better than I |
| 04:55 | < ToxicFrog> | The editor as a whole? Yes. Intellisense and Java-specific features? Nyet. |
| 05:05 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has quit [Ping timeout: 121 seconds] |
| 05:27 | | * Orthia nod. Thank you, TF |
| 06:08 | < Rhamphoryncus> | Okay, maybe someone here will appreciate this: |
| 06:08 | < Rhamphoryncus> | <b><br> |
| 06:08 | < Rhamphoryncus> | </b> |
| 06:09 | | * ToxicFrog eyebrows |
| 06:09 | | AnnoDomini [annodomini@Nightstar-f09bebe3.adsl.tpnet.pl] has joined #code |
| 06:09 | | mode/#code [+o AnnoDomini] by Reiver |
| 06:09 | < Rhamphoryncus> | Courtesy of google docs |
| 06:09 | < jerith> | A *bold* linebreak! |
| 06:10 | < Rhamphoryncus> | The editor behavior is just as sane |
| 06:10 | < Rhamphoryncus> | Deleting at the end of the line consistently remove 1 space to the left when it joins them, so I have to add an extra first, and sometimes it first adds a blank line instead of deleting |
| 06:11 | < jerith> | There are emacs modes that do something similar. |
| 06:11 | < jerith> | I can never remember whic ones until I get bitten. |
| 06:11 | < Rhamphoryncus> | If you're at the start of a link, adding a space sometimes makes it part of the link, sometimes not. Thus having to add a space before removing the linefeed |
| 06:12 | < Rhamphoryncus> | Every now and then delete will instead select your whole line |
| 06:12 | < jerith> | Deleting a line removes the first space from the next line. |
| 06:12 | < jerith> | It's a mode without autoindent, so I can't just hit tab to make things happy again. |
| 06:13 | < Rhamphoryncus> | Hitting undo will move your cursor to the top of the document, but doesn't scroll you there. Using the keyboard before clicking the mouse will cause you to scroll |
| 06:13 | < jerith> | Web-based editing in general really sucks for me. |
| 06:13 | < Rhamphoryncus> | I agree. However, I want my document versioned and shared :( |
| 06:13 | < jerith> | Probably a combination of MacOS and whatever. |
| 06:14 | < ToxicFrog> | github ?? |
| 06:14 | < Rhamphoryncus> | and web accessible |
| 06:14 | < Rhamphoryncus> | and uhh, NO FUCKING GIT |
| 06:14 | < jerith> | The usual methods for navigating more than one character-dimension in any given direction don't move the cursor, only the display. |
| 06:14 | < ToxicFrog> | ...why not? |
| 06:14 | < Rhamphoryncus> | I've used git. It hates me |
| 06:15 | < ToxicFrog> | Oh? |
| 06:15 | < Rhamphoryncus> | It's constantly getting hosed in weird ways |
| 06:15 | < jerith> | And the emacsian navigation chords tend to do Bad Things to the page. |
| 06:15 | < jerith> | C-e in particular navigates away and obliterates my text. |
| 06:15 | < ToxicFrog> | That's a neat trick; examples? |
| 06:16 | < Rhamphoryncus> | I have no examples right now. I haven't used it in a while |
| 06:16 | < jerith> | I find git too complex for my needs. |
| 06:17 | < jerith> | It's really aimed at people who do a lot of patch management. |
| 06:17 | < Rhamphoryncus> | And I don't think IRC anecdotes would effectively explain what's wrong with it |
| 06:17 | < jerith> | Not surprising, when one considers its origins. |
| 06:17 | < Rhamphoryncus> | What it needs is a camera, recording the screen, the keyboard, and the user's face |
| 06:18 | < Rhamphoryncus> | hrm didn't I rant in here a while back about my checkout with 3, maybe 4 branches of the same thing, 2 of them with the identical name, and it still refusing to operate on my branch? |
| 06:20 | < jerith> | As I said, aimed at patch management rather than development. |
| 06:20 | < jerith> | Having more than one branch in the same working dir just doesn't do it for me. |
| 06:20 | < ToxicFrog> | I note that I have been using it for all of my development for quite some time now very happily. |
| 06:21 | < ToxicFrog> | Er? |
| 06:21 | < ToxicFrog> | You mean, having the working dir change to reflect one branch at a time, rather than containing all the branches at once a la SVN? |
| 06:21 | < Rhamphoryncus> | I actually found the single working dir reasonable, once someone explained on IRC that git used it. The docs I came across didn't explain it to me |
| 06:23 | < jerith> | Gah! My autoident on chatspike isn't working. And now their server isn't letting me change back to my nick. :-( |
| 06:31 | < ToxicFrog> | Rhamphoryncus: whatever git docs I read when I was learning it explained it pretty well, but yeah, the command man pages mention it but don't go into detail and the gittutorial page is just bad. |
| 06:32 | < Rhamphoryncus> | *nods* |
| 06:32 | < Rhamphoryncus> | I'm sure I mostly read the man pages |
| 06:32 | | * jerith prefers bzr. |
| 06:32 | | * ToxicFrog read Git for Computer Scientists to understand how branches/commits work, and some other tutorial I don't remember to get a handle on the commands and day to day usage |
| 06:33 | < jerith> | Although that's mostly because the the Big Win integration at launchpad. |
| 06:33 | < Rhamphoryncus> | A well designed too shouldn't require a significant effort to learn properly. It should be intuitive |
| 06:33 | < Rhamphoryncus> | I'm using bzr, but Python's moving to hg |
| 06:35 | < jerith> | It's actually starting to become feasible to use your own dvcs of choice and translate on the border to whatever upstream uses. |
| 06:36 | < ToxicFrog> | bzr has been on my list of things to check out for a while, but I can't see myself switching away from git unless it has some killer feature that I didn't know I craved |
| 06:36 | < ToxicFrog> | Also, github makes me happy inside |
| 06:36 | < Rhamphoryncus> | I still find it ironic that they all claim to be very fast, yet the first thing a user experiences is the initial checkout of a repository, which they all do *vastly* slower than a tarball |
| 06:38 | < Rhamphoryncus> | You could make a script that makes a local checkout, tarballs it, downloads it, and untarballs it, and probably be 10x faster |
| 06:38 | < Rhamphoryncus> | And that's not an exaggeration |
| 06:39 | < ToxicFrog> | (one thing I do like about bzr, at least from looking at it, is that it plays much nicer with non-unices than git does - which is a nonissue for me, but not for some of the people using my software) |
| 06:40 | < jerith> | ToxicFrog: bzr's killer feature is launchpad. |
| 06:40 | < jerith> | Project hosting with decent supporting tools. |
| 06:43 | < ToxicFrog> | ...on the other hand, it doesn't have a staging area :/ |
| 06:47 | < ToxicFrog> | And now, slep |
| 06:47 | < jerith> | 'Night TF. |
| 06:48 | < Kazriko> | Heh, we've been using SVN just because it plays nice with windows and that's what most of the projects at work are... Some of the companies we work with use Perforce though... :( |
| 06:49 | < Rhamphoryncus> | svn's okay in a closed shop where everybody can and does create branches as desired |
| 06:50 | < Kazriko> | yeah... |
| 06:50 | < Rhamphoryncus> | If the policy discourages creating branches and a patch sits in your working dir for days, getting tweaked.. well you're not using version control are you? |
| 06:53 | < Kazriko> | We tend to have multiple branches, though we usually commit in bite size, unit tested chunks. |
| 06:53 | < Kazriko> | My trending app is one of the few that is big enough to need its own branch. heh |
| 06:54 | < Kazriko> | we tend to move code around between systems through svn too, so we'll commit it at night before going home, then update at home, etc. |
| 06:54 | < Kazriko> | as soon as something's greenlined it can go up into the trunk. |
| 06:55 | < jerith> | We generally don't push to trunk unless something's ready for integration testing. |
| 06:56 | < Kazriko> | Ahh, different philosophy. usually that sort of testing happens in a frozen branch here. |
| 06:56 | < jerith> | That's because we have people on two continents working on code which depends on the same stuff. |
| 06:56 | < Kazriko> | Nod. |
| 06:56 | < jerith> | All the integration systems pull in trunk. |
| 06:56 | < Kazriko> | all of us are in the same office, so that probably makes it easier for us. |
| 06:57 | < jerith> | They build and deploy on every commit, which I dislike intensely. |
| 06:57 | < Kazriko> | and systems are deployed on branches not trunk. |
| 06:57 | < jerith> | Because it means someone else can hose my test system completely. |
| 06:57 | < Kazriko> | trunk is usually versions like 0.9r6, where we'd branch when we had 0.9s ready for final testing... |
| 06:57 | < Kazriko> | the last number means its a beta. |
| 06:58 | < jerith> | We have a lot of nontrivial interdependent services. |
| 06:58 | < Kazriko> | sounds like a pain. |
| 06:59 | < jerith> | So having a single "this is either in production or intended to be there soon" staging area is useful. |
| 06:59 | < Kazriko> | our stuff is mostly firmware for hardware devices, so we can control when something gets installed on that hardware more. |
| 06:59 | < jerith> | It is pain, but dramatically less pain than having all that crap in one process. |
| 07:00 | < Kazriko> | We have customers with their own stable branches, for example... |
| 07:00 | < Kazriko> | because they want to control when they merge in new features from our trunk... |
| 07:01 | < jerith> | The autobuild is a consequence of there being very few people who will update their test systems regularly. |
| 07:01 | < jerith> | So if it wasn't there, stuff would break horribly as soon as it hit trunk and then releases would be delayed. |
| 07:01 | < Kazriko> | heh. I always do an update and a run of the unittestall.out file before a commit... |
| 07:02 | | * jerith nods. |
| 07:02 | < Kazriko> | it's harder for us to do automated testing though. have to keep a hardware unit plugged into a server somewhere all the time to do that. |
| 07:02 | < Kazriko> | some of the unit tests require 2 hardware units, one to act as a serial slave to the other... |
| 07:03 | < jerith> | If Bob is working on the payments stuff and there's a change in the domains stuff that he depends on, he needs to make sure his thing works with the new version. |
| 07:03 | < Kazriko> | yep. |
| 07:03 | < Kazriko> | A bigger code base would be more difficult than what we're doing. |
| 07:03 | < jerith> | To force this, his test system will update itself as soon as a new commit hits trunk. |
| 07:03 | < Kazriko> | we only have 5 developers, mostly working on different projects. |
| 07:04 | < jerith> | This means that he's forced to test it at his system's convenience, not his. |
| 07:04 | < Kazriko> | Datamux and padmanager are the only shared projects. |
| 07:04 | < Kazriko> | ahh. |
| 07:04 | < jerith> | And it also means that broken commits to unrelated projects can screw up *my* testing for two days. |
| 07:04 | < Kazriko> | ours doesn't do that, that sounds like a recipe for mysterious bugs. |
| 07:05 | < Kazriko> | We test it with what we have checked out, greenline it, update then check it again... |
| 07:05 | < jerith> | Yeah. That's what I'd prefer. |
| 07:05 | < Kazriko> | that way we know if it's our code breaking or someone else's code breaking us... |
| 07:05 | < Kazriko> | I guess that would get harder when you have a bigger team. |
| 07:06 | | * jerith returns with power for the laptop. |
| 07:07 | | * Kazriko works on the 4th part of the django tutorial... |
| 07:07 | < jerith> | If we make people manage their own updates, it doesn't happen. |
| 07:07 | < Kazriko> | (Zope 2 is getting too stale, and bluebream/zope3 seems harder than django to learn. :( ) |
| 07:07 | < Kazriko> | I see. :( |
| 07:07 | < Kazriko> | SVN kind of forces it a bit. |
| 07:07 | < jerith> | Django's the least painful web framework I've used. |
| 07:08 | < Kazriko> | if you commit and have an old version, it makes you update. |
| 07:08 | < jerith> | Ah, I'm talking about a different thing. |
| 07:08 | < jerith> | We use svn. |
| 07:08 | < Kazriko> | but you have more layers of libraries... |
| 07:08 | < jerith> | And we do all nontrivial development in branches. |
| 07:08 | < Kazriko> | nod.. |
| 07:08 | < jerith> | Not libraries, multiple apps. |
| 07:09 | < Kazriko> | ahh. |
| 07:09 | < Kazriko> | interlinked though. |
| 07:09 | < jerith> | So our test environments have their own CI server, which builds and deploys everything from trunk. |
| 07:10 | < Kazriko> | One example of our stuff, I didn't branch for my Trending thing even though I probably should have, the code just sits in the folder with the production stuff but in a different file (since it's a whole new module.) |
| 07:10 | < jerith> | To test my stuff, I point my integ box's build at my branch for my project. |
| 07:10 | < Kazriko> | I commit everything except for the glue code in the main library that links it in. |
| 07:10 | < Kazriko> | Nod. |
| 07:10 | < jerith> | Then I have a system that's trunk versions of everything except my code. |
| 07:10 | < Kazriko> | I see, makes sense. |
| 07:11 | < Kazriko> | what if you have multiple apps worth of work you're doing that are related? |
| 07:11 | < Kazriko> | can you point multiple chunks at different branches? |
| 07:11 | < jerith> | Yes. |
| 07:11 | < jerith> | They're all different repos. |
| 07:11 | < jerith> | One repo per app. |
| 07:11 | < Kazriko> | cool... if we ever get our stuff more interrelated, then that would be nice. |
| 07:11 | < Kazriko> | that goes a bit against svn though. |
| 07:11 | < Kazriko> | svn says that you should be able to have the whole state of the tree... *shrug* |
| 07:11 | < Kazriko> | we have everything now in one repository... |
| 07:12 | < Kazriko> | we used to have it split, but... |
| 07:12 | < jerith> | You can do that too. |
| 07:12 | < jerith> | We used to do it at Amazon. |
| 07:12 | < Kazriko> | we merged back in the active projects. |
| 07:12 | < Kazriko> | We have about 10 repositories, but only 1 has any active development... |
| 07:12 | | You're now known as TheWatcher |
| 07:12 | < jerith> | You just check out smaller per-project subtrees. |
| 07:12 | < Kazriko> | Yeah. |
| 07:12 | < Kazriko> | Trac doesn't play well with multiple svn repos, so... |
| 07:13 | < jerith> | It doesn't? |
| 07:13 | < Kazriko> | it has to have a separate wiki, separate userbase, etc. |
| 07:13 | < jerith> | Works fine for what I've used it for, but that hasn't been very much. |
| 07:13 | < jerith> | Ah, right. |
| 07:13 | < Kazriko> | We wanted everything to be in one ticket tracking system. |
| 07:13 | < jerith> | Yeah, I wanted all those to be separate. |
| 07:13 | < jerith> | We use Jira for all that. |
| 07:13 | < Kazriko> | Too much overhead to have people checking 3-4 separate trackers. |
| 07:14 | < jerith> | A bit "enterprisey" for my tastes, but it does the job. |
| 07:14 | < Kazriko> | We were looking at Redmine. |
| 07:14 | < Kazriko> | haven't really gotten into it much yet. |
| 07:15 | < Kazriko> | Django seems like a much cleaner framework for sure. |
| 07:15 | < Kazriko> | Zope2 is quicker for freeform python mangling, but it certainly isn't very clean to do it that way. |
| 07:16 | < Kazriko> | though, one of my first challenges will be to wrangle django's relational databasing to simulate some of my zope hierarchal data storage. |
| 07:16 | < jerith> | I prefer the Twisted stuff myself, but you have to roll your own a lot more there. |
| 07:16 | < Kazriko> | I've fiddled with twisted, i like it for non-web stuff a bit. |
| 07:16 | < Kazriko> | I think wsgi is better as far as the roll your own web stuff goes. |
| 07:16 | < jerith> | You can replace the ORM entirely, but you have to do a bunch of stuff by hand. |
| 07:17 | < Kazriko> | I had fun making the wsgi interface for ennesbot, but i need to make it more stable... |
| 07:18 | < jerith> | I did that when I was rewriting stuff on a deadline. Kept the old SQLAlchemy-based data layer and ignored Django's stuff completely. |
| 07:18 | < Kazriko> | I like django's database stuff though. It seems very convenient compared to my old zsql stuff. |
| 07:19 | < Kazriko> | I did look at sqlalchemy once, haven't gotten into it fully though. |
| 07:19 | < Kazriko> | I would love to shoehorn zpt into django though. I'm more used to zpt templates. |
| 07:19 | < jerith> | All ORMs suck. |
| 07:19 | < jerith> | Replacing the templating is easier than replacing the db. |
| 07:20 | < jerith> | Or rather, more stuff assumes the db is there than the templating engine. |
| 07:20 | < Kazriko> | true, though for my homepage is pretty simple, right now it uses a hierarchy of flat files with a tiny bit of metadata indicating priority, and a hack of two-layered grouping... |
| 07:20 | < jerith> | If you're not wanting to use the standard auth plugins and whatnot, it's a bit of a non-issue. |
| 07:21 | < Kazriko> | Templating seems like you can just drop another one in, at least thus far into the tutorial. |
| 07:21 | < Kazriko> | just replace the render_to_response call with one that uses zpt... |
| 07:21 | | * jerith nods. |
| 07:21 | < jerith> | Or, in my case, with a thing that builds JSON responses for an API. |
| 07:22 | < jerith> | (Although trying to use Django for a webservice rather than a webapp is harder than I had anticipated.) |
| 07:22 | < Kazriko> | i guess that's what repoze sort of tried to do. decompose zope into layers and let you use the bits you want. |
| 07:23 | < Kazriko> | i wonder if repoze and django play nice together in the wsgi stack. |
| 07:23 | < jerith> | Django doesn't make you drink /all/ the koolaid, but it's generally easier if you do. |
| 07:23 | < jerith> | All bets are off when you're porting an existing app to the new framework, though. |
| 07:23 | < Kazriko> | I think i'm going to approach django at first from the koolaid drinking side. :) |
| 07:23 | < Kazriko> | try to rebuild my app into a pure django sort of thing. |
| 07:24 | < Kazriko> | since I'm pretty unhappy with many of the hacks I made in python to make it work initially. |
| 07:24 | < Kazriko> | It was originally DTML, so switching templating languages is fairly trivial. |
| 07:24 | | * jerith nods. |
| 07:25 | < jerith> | You don't have a year's worth of production data, though. Do you? |
| 07:25 | < Kazriko> | well, 13 years worth of web snippets, but they're not hard to shift over. |
| 07:25 | < Kazriko> | not really mission critical stuff. |
| 07:26 | < jerith> | (I managed switching over to the rewrite seamlessly. No downtime, no data migrations, no point anywhere at which customers couldn't give us their money.) |
| 07:26 | < Kazriko> | I'm going to try django with my work mssql database though and let it try to reverse-engineer the database into django structures. |
| 07:26 | < jerith> | There's a tool to do that? |
| 07:26 | < Kazriko> | and that database has 3-4 years of production data. |
| 07:26 | < Kazriko> | I think there was one. |
| 07:26 | < jerith> | I've always done it by hand. |
| 07:27 | < jerith> | (Where "always" is "the one time I did it".) |
| 07:27 | < Kazriko> | python manage.py inspectdb > models.py |
| 07:27 | < jerith> | Oh, shiny. |
| 07:27 | < Kazriko> | it's apparently not foolproof, but gets you a starting point. |
| 07:28 | < Kazriko> | http://docs.djangoproject.com/en/dev/howto/legacy-databases/ |
| 07:28 | < Kazriko> | I'd definitely try it on a copy of the production database, rather than the actual production database first though. :D |
| 07:29 | < jerith> | "But we have backups, right? What's the worst that can happen?" |
| 07:29 | < jerith> | I once heard those words uttered without any irony whatsoever. |
| 07:29 | < jerith> | I needed a *long* walk to cool down. |
| 07:30 | < jerith> | Then I packed up and went home (it was nearly knocking-off time anyway) and helped pick up the pieces the next day. |
| 07:30 | < Tarinaky> | Oh dear. |
| 07:30 | < Kazriko> | I mean, I do editing by hand on the production database frequently, but I would never let a program do that... heh |
| 07:31 | < jerith> | I've been trying very hard to build tools that will remove the necessity to edit the production databases by hand. |
| 07:32 | < jerith> | The problems are political, not technical. |
| 07:32 | < Kazriko> | Yeah. I had those back on another project, but this one... oy. |
| 07:32 | < jerith> | Along the lines of prioritisation and stuff. |
| 07:32 | < Kazriko> | one problem is that a customer with one of the database has tightly tied many reports to the database structure at least once before.. |
| 07:33 | < Kazriko> | so we have 3 different versions of this database with slightly different names for tables and columns... :( |
| 07:33 | < jerith> | Six months ago I started building a single integrated tool that would let us do everything the myriad hacked scripts did. |
| 07:33 | < jerith> | Last week, after nearly six months of screaming by the Support team, I was finally given official time to do it. |
| 07:34 | < Kazriko> | nice... |
| 07:34 | < Kazriko> | I'm trying to get them to give me time to work on a backend tool to manage all of the database stuff, but it's always the lowest priority thing on the list... |
| 07:34 | < jerith> | Yeah. |
| 07:34 | < Kazriko> | adding a new device to the system takes a couple of hours, a new pad takes all day to add. |
| 07:35 | < Kazriko> | and you have to touch about 5 different tools and 4 database tables to do it. |
| 07:35 | < jerith> | I had the start of a working tool six months ago, but it was ugly (I suck at UI) and we didn't have the APIs to modify stuff properly. |
| 07:36 | < Kazriko> | yeah. at least 2 of the applications we don't have a good way to modify their data directly. :( |
| 07:36 | < Kazriko> | so the plan is to generate files that you can quickly import into them instead. |
| 07:37 | < Kazriko> | the rest are in databases, but there's no good interface for forcing them to reload from the database if it's altered... |
| 07:38 | < Kazriko> | I hate automation software... |
| 07:39 | < jerith> | Heh. |
| 07:40 | < jerith> | I hate web software. |
| 07:40 | < Kazriko> | Iconics... OPC... argh. |
| 07:40 | | * Kazriko is glad they hired a new developer to work on the opc and iconics stuff, hopes they learn it well enough they don't need to ask him anything in the future. |
| 07:41 | < Kazriko> | I'd be perfectly happy to maintain the well manager in django and the Datamuxer in python and leave all that annoying opc stuff to someone else... |
| 07:45 | < Kazriko> | http://kazriko.arkaic.com/ptgrid/source.html << that's the sloppy mess I wanted to replace. It's a lot cleaner in ZPT than it was in dtml though. |
| 07:45 | < Kazriko> | have to viewsource to see the zpt code though. |
| 07:46 | | * jerith is on expensive bandwidth from out in the bundu at present. |
| 07:47 | < Kazriko> | ok... |
| 07:47 | | * Kazriko should probably chat at you less then. heh |
| 07:48 | < jerith> | Oh, bundu. It roughly translates to "rural middle of nowhere". |
| 07:48 | < Kazriko> | nod. |
| 07:48 | < Kazriko> | Kind of like where I grew up. :) |
| 07:48 | < jerith> | Chatting's fine, but I'm avoiding web stuff mostly. |
| 07:48 | < Kazriko> | (though, they now have high speed internet there, about double what I pay however.) |
| 07:49 | < jerith> | Staying in a chalet on a nature reserve near a town with one restaurant. |
| 07:49 | < Kazriko> | Sounds close to where I grew up, though I think the restaurant closed now. |
| 07:49 | < jerith> | A good one, though. And rather cheaper than the equivalent back in the city. |
| 07:50 | < jerith> | The meteorologists don't know that it exists, but the 3g providers do. |
| 07:50 | < Kazriko> | no hotels there now. there's an apothecary shoppe, and a grocery store that's more like a convenience store elsewhere. |
| 07:50 | < Kazriko> | heh. |
| 07:50 | < Kazriko> | the grocery store was closed for a couple years because the owners went bankrupt... |
| 07:50 | < Kazriko> | (luckily the next town is only 5 miles away and has another convenience-store sized grocery store.) |
| 07:51 | < Kazriko> | no 3g there though last time I was there, only analog. |
| 07:51 | < jerith> | If I had more battery, I'd be out on the balcony enjoying the view. |
| 07:52 | < jerith> | As soon as the Young Lady wakes up, we need to head into town to find a place selling bread and milk. |
| 07:52 | < Kazriko> | nod. |
| 07:52 | < jerith> | (I'm doing the antisocial internet thing while she sleeps.) |
| 07:52 | < Kazriko> | nod. |
| 07:53 | < Kazriko> | It could be worse I suppose. My mother grew up in a town that only had a convenience store, and was 30 miles from the nearest grocery store... |
| 07:53 | < jerith> | Well, it's antisocial to the people around one in meatspace. :-) |
| 07:53 | < Kazriko> | yeh |
| 07:54 | < Kazriko> | Ever watched Thelma and Louise? |
| 07:54 | < jerith> | Not that I recall. |
| 07:55 | < Kazriko> | (If you do, there's a wooden looking convenience store, all gray wood, very ancient looking... that's the only store in my mother's home town. technically, the town next door to her town, Bedrock, CO) |
| 07:56 | < Kazriko> | I would guess that colorado has more very remote rural places than most of the US though. |
| 07:56 | < Kazriko> | aside from wyoming... |
| 08:02 | | Attilla [Attilla@FBC920.65CFFF.ADEEA3.DBDA33] has joined #code |
| 08:03 | | mode/#code [+o Attilla] by Reiver |
| 08:15 | | Vornicus is now known as Vornicus-Latens |
| 08:19 | | You're now known as TheWatcher[afk] |
| 08:25 | < Tarinaky> | http://www.guapacho.net/wp-content/uploads/2008/05/jesus_backup.jpg |
| 10:11 | | You're now known as TheWatcher |
| 11:28 | | Orthia [orthianz@Nightstar-d75b0f99.xnet.co.nz] has quit [Ping timeout: 121 seconds] |
| 11:43 | | You're now known as TheWatcher[afk] |
| 12:03 | | celticminstrel [celticminstre@Nightstar-f8b608eb.cable.rogers.com] has joined #code |
| 12:30 | | You're now known as TheWatcher |
| 12:34 | | celticminstrel [celticminstre@Nightstar-f8b608eb.cable.rogers.com] has quit [[NS] Quit: *hums* Can't stay now!] |
| 13:13 | | Orthia [orthianz@Nightstar-bcfceb93.xnet.co.nz] has joined #code |
| 13:43 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has joined #code |
| 13:45 | | Rhamphoryncus [rhamph@Nightstar-8931f88f.abhsia.telus.net] has quit [Client exited] |
| 13:46 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has quit [[NS] Quit: The world is so funny sometimes; I often forget to laugh.] |
| 13:49 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has joined #code |
| 14:08 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has quit [[NS] Quit: The world is so funny sometimes; I often forget to laugh.] |
| 14:11 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has joined #code |
| 14:39 | | Orth [orthianz@Nightstar-9a0f83c9.xnet.co.nz] has joined #code |
| 14:41 | | Orthia [orthianz@Nightstar-bcfceb93.xnet.co.nz] has quit [Ping timeout: 121 seconds] |
| 15:05 | | gnolam [lenin@Nightstar-38637aa0.priv.bahnhof.se] has joined #code |
| 15:19 | | You're now known as TheWatcher[afk] |
| 15:25 | | * gnolam ponders shader error handling. |
| 15:27 | < gnolam> | Specifically, what to do with non-existent uniforms. |
| 15:28 | < PinkFreud> | The emperor has no clothes? |
| 15:31 | <@AnnoDomini> | Are you playing Maid the RPG? |
| 15:31 | < gnolam> | They're an irregular army of shaders. |
| 15:32 | < gnolam> | (In GLSL-speak, a uniform is a per-program shader variable.) |
| 15:33 | < gnolam> | Right now, I'm just silently ignoring errors. |
| 15:34 | < gnolam> | (Trying to set a value of a non-existent uniform makes your program die a horrible, hard-to-trace death, so just passing it on regardless is not an option) |
| 15:35 | < gnolam> | Problem is, I can't really think of a good way of handling this that doesn't lead to a world of pain. |
| 15:46 | < jerith> | WORLD OF PAIN! |
| 15:57 | < gnolam> | And now for something completely different: http://threewordchant.com/2010/02/24/why-the-internet-will-fail-from-1995/ |
| 16:08 | < Namegduf> | Actually, they were pretty right. |
| 16:08 | < Namegduf> | I'd call taking 15 years "no time soon" for ebooks. |
| 16:08 | < Namegduf> | We're just living in the future now. |
| 16:09 | < Namegduf> | Cyberbusiness didn't take close to that long, though. |
| 16:09 | < Tarinaky> | Most of them folded though. |
| 16:10 | < Namegduf> | Well, I mean in their sense. |
| 16:10 | < Namegduf> | We pretty much do buy airline tickets, anything over the Internet now. |
| 16:10 | < Namegduf> | Order pizzas. |
| 16:10 | < Tarinaky> | Yeah, but not till 5 years after that piece was published. |
| 16:10 | < Namegduf> | Yeah. |
| 16:10 | < Namegduf> | That's a little closer than they seem to have thought it'd be. |
| 16:11 | < Tarinaky> | I think the piece was mostly about the nature of investment in the internet. |
| 16:11 | < Tarinaky> | .com crash and all that. |
| 16:12 | < Tarinaky> | And, in truth, I don't think the internet has really changed the way governments work. |
| 16:12 | < Tarinaky> | Anyone who thinks that is being a little bit big-headed. |
| 16:13 | < Tarinaky> | At least, not macroscopically. |
| 16:13 | < Tarinaky> | It's no more democratic today than it was yesterday. |
| 16:19 | < Serah> | The internet has removed alot of paper pushing and waiting for responses. |
| 16:20 | < Serah> | At least the danish government has benefitted greatly from converting to an online platform. |
| 16:31 | | Serah [Z@3A600C.A966FF.5BF32D.8E7ABA] has quit [Ping timeout: 121 seconds] |
| 16:36 | | AbuDhabi [annodomini@Nightstar-604de11f.adsl.tpnet.pl] has joined #code |
| 16:38 | | AnnoDomini [annodomini@Nightstar-f09bebe3.adsl.tpnet.pl] has quit [Ping timeout: 121 seconds] |
| 16:55 | | Vornicus-Latens is now known as Vornicus |
| 17:33 | | Attilla [Attilla@FBC920.65CFFF.ADEEA3.DBDA33] has quit [Ping timeout: 121 seconds] |
| 17:36 | | Attilla [Attilla@FBC920.0A905B.55C6A3.AD1094] has joined #code |
| 17:36 | | mode/#code [+o Attilla] by Reiver |
| 18:00 | | Serah [Z@26ECB6.A4B64C.298B52.D80DA0] has joined #code |
| 18:09 | | AbuDhabi [annodomini@Nightstar-604de11f.adsl.tpnet.pl] has quit [Ping timeout: 121 seconds] |
| 18:11 | | celticminstrel [celticminstre@Nightstar-f8b608eb.cable.rogers.com] has joined #code |
| 18:14 | | AnnoDomini [annodomini@Nightstar-67564f12.adsl.tpnet.pl] has joined #code |
| 18:14 | | mode/#code [+o AnnoDomini] by Reiver |
| 18:20 | | AnnoDomini [annodomini@Nightstar-67564f12.adsl.tpnet.pl] has quit [Ping timeout: 121 seconds] |
| 18:25 | | AnnoDomini [annodomini@Nightstar-3c34be84.adsl.tpnet.pl] has joined #code |
| 18:25 | | mode/#code [+o AnnoDomini] by Reiver |
| 18:53 | | Orth [orthianz@Nightstar-9a0f83c9.xnet.co.nz] has quit [Ping timeout: 121 seconds] |
| 19:15 | | Thaqui [Thaqui@27B34E.D54D49.F53FA1.6A113C] has quit [Connection closed] |
| 22:47 | | AnnoDomini [annodomini@Nightstar-3c34be84.adsl.tpnet.pl] has quit [[NS] Quit: s/ee/p] |
| 23:15 | | * gnolam stabs SDL's documentation. |
| 23:16 | | * celticminstrel stabs documentation in general. |
| 23:18 | < gnolam> | Yeah. But this is pretty much the definition of a mature project, and its documentation is still awful in places. |
| 23:23 | < celticminstrel> | Perhaps projects should get people who aren't involved in development to write the documentation, or at least to flesh it out. |
| 23:24 | < celticminstrel> | Or just get documentation beta-testers to read it and ask a zillion questions that the documentation doesn't cover (or doesn't cover clearly). |
| 23:27 | < celticminstrel> | ...what's the normal menu mnemonic for Cut/Copy/Paste on Windows? |
| 23:27 | < celticminstrel> | The one for which the menu has to be opened first. |
| 23:27 | <@McMartin> | Like, which letter gets underlined? |
| 23:28 | <@McMartin> | A quick check says "none", with the accelerators being ^X/^C/^V respectively |
| 23:28 | <@McMartin> | Typically, Shift-Del/Ctrl-Del/Shift-Ins are the other accelerators expected to do the same thing. |
| 23:28 | < celticminstrel> | Really? No underline? This surprises me. |
| 23:28 | | * McMartin just checked Notepad and Visual Studio; neither has any. |
| 23:29 | < celticminstrel> | Well, okay then. |
| 23:29 | < celticminstrel> | Thanks. |
| 23:31 | < celticminstrel> | Do you have any idea why the "minimal" example in wxWidgets assigns "Alt-X" as the accelerator for Exit? This is odd because usually Alt isn't used for accelerators... |
| 23:31 | | You're now known as TheWatcher |
| 23:33 | < gnolam> | Cut: t, Copy: c, Paste: p |
| 23:34 | < gnolam> | I too just checked in Visual Studio and Notepad. |
| 23:34 | < gnolam> | But you have to access the menu itself by accelerator or they don't show up, at least in Win7. Cannae remember what the previous Windowses did. |
| 23:34 | < gnolam> | But the keys are the same. |
| 23:35 | < celticminstrel> | ...you only see the underlines if you press Alt-E to get the menu, then? Odd... |
| 23:38 | < gnolam> | Google logic. :P |
| 23:40 | | SmithKurosaki [Smith@4FC299.B7448D.9CC136.3DD58B] has quit [Connection closed] |
| 23:45 | | SmithKurosaki [Smith@Nightstar-1a7d4505.dsl.teksavvy.com] has joined #code |
| 23:45 | <@McMartin> | So it would seem. |
| 23:59 | | * gnolam stabs SDL's documentation again. |
| 23:59 | <@McMartin> | Are you using raw SDL or trying to integrate it with OpenGL/ |
| 23:59 | <@McMartin> | ? |
| --- Log closed Wed Mar 03 00:00:00 2010 |