IRC Logs for #crux-devel Tuesday, 2018-03-20

*** john_cephalopoda has quit IRC00:05
jaegeroops, just noticed my earlier paste didn't have any text update with it, so to clarify those links, a -test1 ISO is available for testing at
*** _________mavric6 has quit IRC03:37
*** _________mavric6 has joined #crux-devel03:38
jaegerOK, I think I've got all the initial features implemented that I wanted in the portdb, woot04:01
jaegerThe last couple things were administrative things, nothing visible04:01
j_vjaeger: I've spent a little time reviewing portdbc. I think it will require a complete rewrite probably, but in all honesty, it is probably due for some attention anyways04:02
j_vthe portdb looks and works great. the portdbc will just need to be rewritten since it does everything xml based04:05
*** maraku has quit IRC04:08
*** maraku has joined #crux-devel04:11
j_vI think what would help with implementing a portdbc to work with your new portdb would be have some id's, title's or such to be able to pull the desired elements04:12
jaegerHow does portdbc currently work? I've never looked at it so don't know much04:18
jaegeryou mentioned xml, does it pull xml from the current portdb somehow?04:19
j_vwell, it retrieves xml from the current portdbc, though I'm not clear about the backend, though I can look again at a source view of the current portdb to see if I can get a better understanding of it04:21
j_vI've already found that I can query your portdb using wget address like this:<name_to_search_for>04:22
j_vso it shouldn't really be necessary to have it in xml, it would just help to have an id attribute for the various table entries to glean the different portions from it04:23
jaegermakes sense04:24
jaegeras a side note I should remove the csrf token thing from the search form and view since it's not useful for GET requests04:26
jaegerOriginally I was thinking about doing that with POST but I don't really see a reason to04:28
j_vthis is kind of what I was thinking:
j_vnot sure I that is quite to html spec, but could play with it04:31
jaegerI think it would probably need to be class="collection" instead of id="collection" - I think class can refer to many but id is supposed to refer to one04:32
jaegerWith that said, I'm NOT an HTML expert04:32
j_vme either04:33
jaeger sorta explains it better04:33
j_vor if I remember better English, I should have written it as: me neither04:34
jaegertechnically, though in most cases people would know what you meant04:35
j_vI'd can get a fresh start on a portdbc rewrite based on the current state of your portdb, unless you think I'd be jumping the gun04:37
jaegerwell, I don't want to discourage you from doing that if you want to, but my portdb is just a hobby project, not like anyone asked me to replace the current one. :)04:38
j_vright, understood.04:38
jaegerWith that said, I don't mind adding those tags if they don't break anything, so I'll play with that tomorrow04:40
jaegerNow that I've added all the initial features I wanted to add, I'll let it bake for a day or two/three and then publicize it (though anyone in here and anyone who looks at the logs for the channel could find it)04:41
j_vWell, I'll start writing some toy code for submitting/retrieving the search and parsing the returned results. that much should be pretty basic, so shouldn't be too bad.04:42
j_vI won't get carried away with making it production ready for now, just layout some ideas04:43
j_vI have been thinking about the idea of being able to download an actual port directory with portdbc, and with those class labels and using execve to run rsync or httpup, it wouldn't be too hard.04:45
jaegeryeah, seems like it would be pretty simple04:46
j_vthat was were I was getting stuck when looking at the existing portdb, since I didn't see an immediate way to just pull a portion out since the table tag for the sync command doesn't have a class or such to parse it04:49
j_vOn side note, I've got chromium rebuilding right now, noticed it build private (third_party) copies of a lot of stuff we package... makes me wonder if there's a way to override that04:57
jaegerNo idea there, I don't use it or build it, myself04:58
jaegerAnyway, bedtime for me05:00
Worksterwhy not json the portdbc data from a query? so it's easy to parse05:34
j_vhmmm, sounds reasonable... I'll look into it05:42
j_vis there a json lib that takes html as input and parses it or converts it to json?05:44
j_vahhh, wait, maybe I'm being obtuse... I've never worked with json that wasn't already formatted. I'll need to investigate this more. Thanks for the suggestion.05:45
Worksteryou would output from the backed from a sql or what ever db to json then read the json at the program and do what ever with the data05:47
j_vok, well, right now I'm looking at implementing portdbc type features to work on jaeger's hobby portdb05:48
j_vand that is completely html right now05:49
j_vhmmm, looks like libxml can also handle html, so that's basically what I was looking for05:55
j_vThanks for the suggestion. I'm not discarding it. You've made me realize that I need to do a bit more research about implementation.06:05
*** BurnZeZ has quit IRC09:19
*** BurnZeZ has joined #crux-devel09:21
*** BurnZeZ has quit IRC11:02
*** BurnZeZ has joined #crux-devel11:04
*** g0relike-2 has quit IRC13:32
*** g0relike has joined #crux-devel13:33
*** john_cephalopoda has joined #crux-devel17:19
*** pitillo has joined #crux-devel17:59
*** onodera has joined #crux-devel18:03
*** dlcusa has quit IRC18:08
*** dlcusa has joined #crux-devel18:11
*** darfo has quit IRC22:16
*** darfo has joined #crux-devel22:20
*** onodera has quit IRC22:20
*** john_cephalopoda has quit IRC23:55

Generated by 2.14.0 by Marius Gedminas - find it at!