| 00:04:28 | gradha | I'm going to make a visible link to both rss feeds in the main table, is it ok if I copy http://en.wikipedia.org/wiki/File:Feed-icon.svg ? |
| 00:07:50 | dom96 | where do you plan on putting it? |
| 00:07:59 | gradha | topbar, right aligned |
| 00:08:49 | gradha | any preference? |
| 00:08:52 | dom96 | Maybe footer would be better? |
| 00:09:12 | gradha | ok, it's still named topbar |
| 00:11:05 | gradha | to be really really cool we could make the thread table header have the rss icon to the right of the word "topics", which leads to the thread rss, and the word "posts" could lead to posts rss, or maybe that would be confusing as hell |
| 00:11:29 | gradha | meh, I'll default on crappy style and let others decide |
| 00:14:06 | gradha | does .svg go into public/images ? |
| 00:14:29 | dom96 | sure |
| 00:20:50 | gradha | I don't understand these em/ex relative size units, trying to use them for the icon but they all render as 1px? |
| 00:28:57 | * | gradha quit (Quit: bbl, have youtube videos to watch) |
| 02:07:20 | * | q66 quit (Remote host closed the connection) |
| 05:23:02 | * | fowl joined #nimrod |
| 08:53:55 | * | fowl quit (Read error: Connection reset by peer) |
| 08:57:32 | * | fowl joined #nimrod |
| 10:39:06 | * | Trix[a]r_za is now known as Trixar_za |
| 10:52:47 | * | q66 joined #nimrod |
| 11:32:22 | * | gradha joined #nimrod |
| 12:42:44 | Araq | gradha: iirc the compiler tries hard to not mess with the line information in .tmpl files ... |
| 12:45:05 | Araq | do you have a simple example where it's off? |
| 12:47:11 | gradha | so far I haven't had a different line number, only cols, obviously due to translation |
| 12:47:25 | gradha | what I think I've seen is weird line continuation issues |
| 12:47:49 | gradha | maybe writing "const a = 2" in different lines as to begin a block of consts, which is why all templates prefix with const |
| 12:48:04 | gradha | there the indentation seems to be generated incorrectly in the output from --verbose:2 |
| 12:51:21 | gradha | example http://pastebin.com/kxmbi0Xi, want a ticket? |
| 12:51:48 | Araq | nah |
| 12:52:31 | Araq | btw --verbosity:2 indeed only lists the output |
| 12:52:49 | Araq | so there is no real need for separate switch |
| 12:53:33 | gradha | it lists more than the output http://pastebin.com/CA57y8BM includes hints and errors, where as a tool you would expect it to generate a file, or split erros to stderr |
| 12:54:17 | gradha | the main function of this would be to compare the compilation error line/column with the translated source, if its offset by hints and other stuff it's not very handy |
| 12:54:18 | Araq | yeah ok, but come on ... :P |
| 12:54:52 | Araq | well the problem is that I had the parser be indentation aware |
| 12:55:17 | gradha | don't worry about this, put it as a feature scheduled for nimrod 3K |
| 12:55:33 | Araq | but this indendentation + "not really and require 'end'" thing was hard to understand |
| 12:56:04 | Araq | so now it know that some keywords require indentation |
| 12:56:44 | Araq | #! stdtmpl |
| 12:56:46 | Araq | #proc genMain(): string = |
| 12:56:47 | Araq | #result = "" |
| 12:56:50 | Araq | #end proc |
| 12:56:51 | Araq | --> works too |
| 12:57:29 | gradha | I guess you would need an "end const" to mark the end of the block |
| 12:58:57 | Araq | only wIf, wWhen, wTry, wWhile, wFor, wBlock, wCase, wProc, wIterator, |
| 12:59:07 | Araq | wConverter, wMacro, wTemplate, wMethod are recognized |
| 12:59:33 | Araq | if we would allow const/var/let these *require* and 'end' |
| 13:00:50 | gradha | maybe you could squish https://github.com/Araq/Nimrod/issues/375 and https://github.com/Araq/Nimrod/issues/301 together by making the compiler output a few lines of the context, already translated |
| 13:01:08 | Araq | so the inline version wouldn't work anymore ... :-/ |
| 13:01:08 | Araq | which I think is much worse :P |
| 13:01:08 | Araq | but lets see |
| 13:01:45 | Araq | these lines have a ':' or '=' if it's an inline let/var/const, so that should be easy to do |
| 13:02:42 | gradha | so if you detect that, changing from a const block to var block could be trouble, the var block being indented by the const one |
| 13:03:09 | gradha | unless the var automatically ended the previous block |
| 13:08:00 | Araq | what's issue 301? github is failing |
| 13:08:55 | gradha | that one is to display a few lines of context along with the line causing an error, zahary said he's meanint to do that |
| 13:12:23 | Araq | I don't really like that though |
| 13:19:12 | gradha | I still think you can patch that by opening and reading the file again without having to change the compiler internals, speed is not an issue there bacause compilation already failed |
| 13:20:14 | gradha | in fact, let's do this, where do I hack this? |
| 13:24:55 | dom96 | hello |
| 13:25:04 | gradha | hallo |
| 13:25:21 | zahary | what I did is to implement inside the compiler a mechanism that remembers the source lines so they can be used for purposes like this |
| 13:25:51 | zahary | I currently use it in a special compilation option that outputs the original source in the generated code |
| 13:27:11 | gradha | ok, won't hack anything then, will patiently wait for your commits |
| 14:01:18 | gradha | oh, nimforum's session cookie is just the password, nasty |
| 14:07:36 | dom96 | is that really a problem? |
| 14:07:47 | dom96 | the passwords are hashed |
| 14:08:21 | dom96 | and the hash is salted :P |
| 14:09:13 | gradha | sure, but if somebody grabs the cookie you can impersonate somebody forever, webs tend to generate cookies based on password + random number and store in a login cookie, then they can be purged |
| 14:09:58 | gradha | the hashed password is effectively the password to impersonate somebody on the forum |
| 14:10:13 | dom96 | You can implement that if you wish. Also the cookies should live for longer. |
| 14:10:55 | gradha | the usual is something like http://stackoverflow.com/questions/12990557/are-passwords-stored-in-cookies/12990591#12990591 |
| 14:13:04 | dom96 | oh, I just remembered. There is another table which stores some additional info in regards to logins. |
| 14:13:10 | dom96 | I think it's the IP address. |
| 14:13:24 | * | fowl quit (Quit: Leaving) |
| 14:13:38 | dom96 | If your IP address doesn't match, you will not be logged in. |
| 14:14:04 | dom96 | I don't think that solution is the best though. |
| 14:14:42 | dom96 | Well, I dunno. I haven't looked at the code in a while. |
| 14:15:23 | gradha | did you know sqlite doesn't enforce varchar limits? I've just posted 300Kb of text in a post because no other checks are done at the app level |
| 14:15:26 | dom96 | Yeah, there is a session table. |
| 14:19:04 | gradha | next time you are logged in to the host try "select id from post where length(content) > 1000;" to see if somebody already posted above the limit |
| 14:29:25 | * | zahary quit (Read error: Connection reset by peer) |
| 14:41:39 | * | gradha quit (Quit: bbl, have youtube videos to watch) |
| 15:41:32 | NimBot | Araq/Nimrod d739d7d Araq [+0 ±1 -0]: better treatment of var/let/const/type sections in the templating system |
| 15:41:32 | NimBot | Araq/Nimrod b3729eb Araq [+0 ±1 -0]: bugfix: mark&sweep GC |
| 15:41:32 | NimBot | Araq/Nimrod 3b642ee Araq [+0 ±1 -0]: fixes memory leaks concerning closures |
| 18:34:38 | * | XAMPP joined #nimrod |
| 18:50:46 | * | Trixar_za is now known as Trix[a]r_za |
| 19:47:19 | * | reactormonk quit (Ping timeout: 264 seconds) |
| 20:09:36 | * | q66 quit (Remote host closed the connection) |
| 20:13:25 | * | q66 joined #nimrod |
| 20:15:54 | * | q66 quit (Remote host closed the connection) |
| 20:17:09 | * | q66 joined #nimrod |