| *** tpb has joined #melange | 00:00 | |
| shikher | I assume that all of data is loaded onto list at once in this case, for it to take 2 minutes? | 00:01 |
|---|---|---|
| madrazr | shikher: correct, I would rephrase "continuous HTTP requests" to "multiple AJAX requests" | 00:01 |
| madrazr | rather | 00:01 |
| madrazr | "multiple XHRs" | 00:01 |
| madrazr | shikher: define once? | 00:01 |
| shikher | no paging | 00:01 |
| madrazr | shikher: that is correct then | 00:01 |
| shikher | y not do paging for tht list? | 00:02 |
| madrazr | shikher: there is a subtle distinction here | 00:02 |
| madrazr | shikher: when you say paging what are you thinking about? | 00:02 |
| madrazr | shikher: displaying the data in different pages | 00:02 |
| madrazr | shikher: or fetching the pages on demand? | 00:03 |
| madrazr | shikher: we already do the former | 00:03 |
| shikher | fetching a certain amount of data out of the total all query | 00:03 |
| madrazr | shikher: latter is quite tricky because we don't have server side filtering and multi-column sorting | 00:03 |
| shikher | oh by paging till now I was thinking u were fetching pages on demand | 00:04 |
| shikher | madrazr: here is my solution. please tell me the fault, if any | 00:05 |
| shikher | we separate filtering completely from lists | 00:06 |
| shikher | and do server side filtering | 00:06 |
| shikher | if lets say user opens a page with lists | 00:07 |
| shikher | we only fetch first watsoever lets say 150 entities and load it | 00:07 |
| shikher | however we keep making XHR's to get all the other data. | 00:08 |
| shikher | and we give priority to the page at which the user is on so that data is loaded first | 00:08 |
| shikher | we will not activate filtering until all data is not loaded in the background | 00:09 |
| shikher | once its loaded we activate the filtering option | 00:09 |
| shikher | this way user doesn't wait for entities to load and can see whatever he wants to see | 00:10 |
| shikher | only the filter will not be available until we have all the data, this way we don't need server side filtering either | 00:11 |
| shikher | As for sorting we can do it page by page when we fetch pages by demand | 00:11 |
| madrazr | shikher: I am not entirely sure if I understand everything you said | 00:16 |
| madrazr | but we can still proceed I guess | 00:16 |
| madrazr | on the way we will figure out | 00:16 |
| shikher | madrazr: yes please ask | 00:16 |
| madrazr | first of all, let's say you give me a page with 50 entries | 00:16 |
| madrazr | I wait until you give me 50 entries | 00:17 |
| madrazr | since you have blocked filtering now I have no way other than waiting | 00:17 |
| madrazr | now I am a sophisticated/power user like Carol | 00:17 |
| madrazr | and I know exactly what row of data I want | 00:17 |
| madrazr | I am willing to even wait for the first 50 entries to load | 00:17 |
| madrazr | saying "Ok whatever" | 00:18 |
| shikher | note: filtering would not be available to you until all the data is loaded anyway, in the current case right? | 00:18 |
| madrazr | then what problem is it solving? | 00:18 |
| madrazr | we do this already I told you | 00:18 |
| madrazr | :) | 00:18 |
| madrazr | just that you made the user experience even worse by not letting me type in the filter box until you fetch all the data | 00:19 |
| shikher | well u said u dont fetch pages on demand | 00:19 |
| madrazr | shikher: correct | 00:19 |
| madrazr | and your new solution is not fetching either? | 00:19 |
| shikher | that means u wait for all data to load before u render the list? | 00:19 |
| madrazr | we don't | 00:19 |
| madrazr | that is why we call it AJAX | 00:19 |
| madrazr | shikher: we render the batches as they arrive | 00:20 |
| madrazr | shikher: that's why you keep seeing the order changing until all the list loads | 00:20 |
| madrazr | *until the entire list loads | 00:21 |
| shikher | and wat if someone want to go to page 12, u keep showing loading until data arrives | 00:21 |
| madrazr | shikher: when we do paging, we fetch all the data | 00:21 |
| madrazr | but we don't render them all | 00:21 |
| madrazr | we render only the first page | 00:21 |
| madrazr | shikher: correct | 00:21 |
| madrazr | that's correct | 00:21 |
| madrazr | because there is no page 12 until we have all the data | 00:21 |
| madrazr | because we have not sorted it yet | 00:21 |
| madrazr | and we cannot sort until all the data arrives | 00:22 |
| madrazr | and until we sort we do not know what is page 12 | 00:22 |
| *** monkegjinni has joined #melange | 00:22 | |
| madrazr | it is a classical problem in Computer Science | 00:22 |
| madrazr | how do you sort or tell what a section of sorted data is before all the data arrives? | 00:22 |
| madrazr | shikher: if you solve that problem, Knuth is going to give you a Ph.D. likely :P | 00:23 |
| shikher | is it not possible to make a query that we tell the datastore to sort data in this way and send this part of that data only | 00:24 |
| madrazr | shikher: this is exactly what I have been telling you from an hour now? | 00:25 |
| madrazr | server side sorting, server side sorting and server side sorting! | 00:25 |
| shikher | madrazr: right u are. | 00:26 |
| madrazr | shikher: Oh btw, one detail I might have missed here which is essential for this discussion | 00:26 |
| madrazr | shikher: you can't run arbitrary sorts and filters in a datastore query | 00:27 |
| madrazr | shikher: you will need an index for every sort or filtering | 00:27 |
| shikher | index? | 00:27 |
| shikher | nvm | 00:28 |
| madrazr | shikher: and if the user sends an arbitrary sort request for which we don't have an index | 00:28 |
| madrazr | App Engine throws a DatastoreNeedIndexException on the query | 00:29 |
| madrazr | shikher: what about indexes? | 00:29 |
| madrazr | shikher: how much of database internals do you know? I can help you with indexes :) | 00:29 |
| madrazr | Database and Big Data is my research area :D | 00:29 |
| madrazr | don't shy away if you have database questions, I can give you all the details you may ever want | 00:30 |
| madrazr | and you may ever NOT want :P | 00:30 |
| shikher | madrazr: cool ! I was afraid to ask and was going through the docs | 00:30 |
| *** monkegjinni has quit IRC | 00:30 | |
| shikher | from previous encounters about how i should use my favourtie search engine. :D | 00:31 |
| madrazr | shikher: no need to be afraid to ask questions. We do expect students to do some home work | 00:31 |
| madrazr | at least basic web search | 00:31 |
| madrazr | reading docs | 00:31 |
| madrazr | etc | 00:31 |
| *** monkegjinni has joined #melange | 00:31 | |
| madrazr | but if you have questions for which you don't find the answer easily through search/docs | 00:31 |
| madrazr | you shouldn't be wasting time and just ask us | 00:31 |
| madrazr | worst comes to worst we will give you the link the right docs page | 00:32 |
| *** Guest55737 is now known as skullbocks | 00:32 | |
| shikher | well i havent looked at the docs for indexing and not sure what it is completely. so i would go through it and then come to u for any unanswered questions | 00:32 |
| *** skullbocks is now known as Guest17725 | 00:32 | |
| madrazr | *you shouldn't be wasting time and you should just ask us (clarifying :) ) | 00:32 |
| madrazr | shikher: sure | 00:32 |
| madrazr | but basically I can give you a simple overview | 00:33 |
| shikher | please tht would be great | 00:33 |
| madrazr | which may help you to understand indexes | 00:33 |
| madrazr | so all the data is stored as bits in the disk right | 00:33 |
| shikher | yes | 00:33 |
| madrazr | one very popular and most widely used software abstraction for handling this data is through files | 00:34 |
| madrazr | you logically organize these stored bits into iles | 00:34 |
| madrazr | *files | 00:34 |
| madrazr | so databases use this file abstraction to store the data | 00:34 |
| shikher | yes | 00:34 |
| madrazr | on disk | 00:34 |
| madrazr | for example MySQL in the background manages its own files to store all the data you insert into a MySQL database | 00:35 |
| *** monkegjinni has quit IRC | 00:35 | |
| *** edk_kdh has joined #melange | 00:35 | |
| madrazr | popularly accepted or well established convention in database world is to store each record (also called as row) in a line in the file | 00:36 |
| *** ankesh11 has quit IRC | 00:36 | |
| madrazr | shikher: makes sense so far? | 00:36 |
| madrazr | each record/row in a database table is stored as a line | 00:36 |
| shikher | yes I have taken OS course | 00:36 |
| madrazr | Ok cool | 00:37 |
| madrazr | shikher: now you know that files are sequential | 00:37 |
| madrazr | shikher: if you want to access a particular line of the file, you need the byte offset of that position in the file | 00:37 |
| shikher | yes | 00:37 |
| *** ankesh11_ has joined #melange | 00:37 | |
| madrazr | so if you want to access a particular row in the database you need the byte offset of that particular row in the file | 00:38 |
| madrazr | now let's say you have a database table with just these columns | 00:38 |
| madrazr | | Key | Firstname | Lastname | Age | | 00:38 |
| madrazr | now user gives a database query which says get me the record of a person whose last name is C.S. | 00:39 |
| madrazr | shikher: now how should a database know at what byte offset in the User table's file does the row corresponding to the person with last name C.S. exist? | 00:40 |
| madrazr | shikher: you are still with me? :P | 00:41 |
| shikher | it maintains an index? | 00:41 |
| madrazr | shikher: or did I make you sleep already :P | 00:41 |
| madrazr | shikher: alright you seem to be here :P | 00:41 |
| madrazr | shikher: exactly | 00:42 |
| madrazr | this is where index comes into picture | 00:42 |
| madrazr | shikher: the database maintains a separate file called an index file | 00:42 |
| madrazr | where it stores this exact information | 00:42 |
| shikher | alphabetical index? | 00:42 |
| madrazr | shikher: there are several ways of implementing an index | 00:42 |
| madrazr | alphabetical is a possibility yes | 00:42 |
| madrazr | shikher: people also implement sophisticated hashing techniques etc | 00:43 |
| shikher | ok | 00:43 |
| madrazr | shikher: and there are several datastructures to manage the index itself | 00:43 |
| madrazr | let's take the simplest one, the one you said | 00:43 |
| madrazr | the database creates an index where the first column is the last name | 00:44 |
| madrazr | and the first column is sorted alphabetically | 00:44 |
| madrazr | and the second column is the list of all the rows' byte offsets who have that same last name | 00:44 |
| madrazr | *byte offsets which have the same last name | 00:45 |
| madrazr | it is like a mapping | 00:45 |
| madrazr | this is exactly an index | 00:45 |
| shikher | ok | 00:45 |
| madrazr | and this particular index is called an inverted index | 00:45 |
| madrazr | shikher: as clear as mud? :) | 00:45 |
| shikher | for every column in a table there is an index, or does it get finer? | 00:45 |
| shikher | crystal, sir. | 00:46 |
| madrazr | shikher: it depends on the database technology used | 00:46 |
| shikher | :D | 00:46 |
| madrazr | there can be an index for every column | 00:46 |
| madrazr | there can be an index for combinations of columns | 00:46 |
| madrazr | etc | 00:46 |
| madrazr | also I intentionally chose MySQL as an example | 00:46 |
| madrazr | sadly at Google our problem is not that simple | 00:46 |
| madrazr | how awesome would it be if we could store all the data in one single machine? :D :D | 00:47 |
| madrazr | as the joke goes | 00:47 |
| madrazr | "At Google, you get paged at 2AM in the night waking up everybody at your home. And you look at your pager to see what the page is and it keeps showing high alert that your cluster has only 2 petabytes of storage remaining and it is running out of space fast. You need to take action now, right now!" | 00:48 |
| madrazr | 2 petabytes is a joke at Google | 00:48 |
| madrazr | :P | 00:48 |
| shikher | :D | 00:49 |
| madrazr | so | 00:49 |
| madrazr | App Engine datastore itself is implemented on Bigtable | 00:49 |
| madrazr | (well now Megastore) | 00:49 |
| madrazr | and Bigtable indexes are more complicated than what I explained | 00:49 |
| *** cosenal has quit IRC | 00:49 | |
| *** downey has quit IRC | 00:49 | |
| shikher | ok | 00:49 |
| madrazr | because it has to be distributed across the cluster | 00:49 |
| shikher | exactly, | 00:50 |
| shikher | i got tht frm the joke | 00:50 |
| madrazr | shikher: you can read the published paper on Bigtable if I have got you interested enough on how Bigtable implements indexes etc | 00:50 |
| madrazr | shikher: anyway since App Engine is implemented on Bigtable | 00:50 |
| madrazr | it needs indexes | 00:50 |
| madrazr | and App Engine does not build all the N! combination of indexes possible by default | 00:51 |
| shikher | madrazr: sure. if i get the time | 00:51 |
| *** downey has joined #melange | 00:51 | |
| shikher | which is pretty hard in my college | 00:51 |
| madrazr | shikher: so for most queries we, the application (Melange in our case) need to define the required index for the model | 00:51 |
| madrazr | model queries' I mean :) | 00:52 |
| madrazr | shikher: I keep saying this. | 00:52 |
| madrazr | There is no such thing as "I have enough of" of two things | 00:53 |
| *** mwilkes|away is now known as MatthewWilkes | 00:53 | |
| madrazr | 1. Time | 00:53 |
| madrazr | 2. Money | 00:53 |
| madrazr | :) | 00:53 |
| *** cosenal has joined #melange | 00:54 | |
| shikher | true. but I am not generalizing here. this is a statement which I am making after doing comparisons. | 00:54 |
| MatthewWilkes | madrazr: 1: people in pain. 2: Millionaires. What's next to dismiss? | 00:54 |
| madrazr | shikher: fair enough :) | 00:54 |
| *** suranga_ has quit IRC | 00:55 | |
| madrazr | MatthewWilkes: 1. they say they have enough time? 2. If millionaires or even billionaires had ever said we have enough money this world would have been such a great place to live! | 00:55 |
| shikher | MatthewWilkes: people in pain. wat? | 00:56 |
| madrazr | shikher: MatthewWilkes means people in pain say they have enough time :) | 00:57 |
| MatthewWilkes | It's quite common for the terminally ill to not want any more time. And it's common for billionaires to think they have enough money, that doesn't stop them wanting more. | 00:57 |
| madrazr | in the sense that every second in pain is like a century or whatever | 00:57 |
| MatthewWilkes | Bill Gates, for example, has more than he needs | 00:57 |
| MatthewWilkes | but, yes, that's OT | 00:57 |
| MatthewWilkes | I know! | 00:58 |
| * olly crushes MatthewWilkes with a big pile of money | 00:58 | |
| madrazr | MatthewWilkes: I was about to give Bill Gates as example and claim that he might be doing all the charity work but he might not have said "I have enough money" | 00:58 |
| madrazr | but either way, there is no way to prove | 00:58 |
| madrazr | MatthewWilkes: olly: shikher: Oh btw! | 00:59 |
| madrazr | there might have been someone who said "enough money" | 00:59 |
| madrazr | http://www.youtube.com/watch?v=3OyrX11cMkE | 00:59 |
| tpb | Title: The Dark Knight - The Joker - Everything Burns - HQ2 - Cut - YouTube (at www.youtube.com) | 00:59 |
| madrazr | for your own enjoyment! | 00:59 |
| *** RaulT has quit IRC | 00:59 | |
| madrazr | this is actual better http://www.youtube.com/watch?v=2TJ3Yv_vgUc | 01:00 |
| tpb | Title: The Jokers Money Burning - YouTube (at www.youtube.com) | 01:00 |
| madrazr | *actually | 01:00 |
| madrazr | s/enjoyment/joy/ | 01:00 |
| madrazr | :P | 01:01 |
| shikher | madrazr: I can never get too much of joker! | 01:01 |
| madrazr | shikher: :) | 01:02 |
| shikher | the right emoticon would be :'( | 01:02 |
| madrazr | shikher: why? | 01:03 |
| shikher | madrazr: really! please think on it. while I think on improving lists. If I have a solution u will see it in my proposal, otherwise I guess u will have to wait for ur ninja. :D | 01:04 |
| madrazr | shikher: an important note here | 01:05 |
| madrazr | shikher: we don't expect our students to give perfect solutions in the proposal | 01:05 |
| madrazr | this is not a school exam come on :) | 01:05 |
| madrazr | shikher: as long as you are going in the right direction we are willing to explore that territory | 01:06 |
| madrazr | in fact most of the work exploration for both the student and the mentor right? | 01:06 |
| madrazr | if we knew the solution, we would have implemented it already | 01:06 |
| madrazr | :) | 01:06 |
| madrazr | shikher: so you should feel free to try | 01:07 |
| madrazr | shikher: also there is a reason why mentors exist | 01:07 |
| shikher | madrazr: will keep tht in mind. going to sleep. have exams starting today. | 01:07 |
| madrazr | to point you in the right directions or at least tell what may not work | 01:07 |
| madrazr | shikher: alright | 01:07 |
| madrazr | shikher: good luck | 01:08 |
| shikher | thanks. well u hve certainly done ur job in pointing me in the right direction. bye now. | 01:08 |
| madrazr | shikher: bye | 01:08 |
| *** shikher has quit IRC | 01:08 | |
| *** Gentlecat has joined #melange | 01:11 | |
| *** monkegjinni has joined #melange | 01:20 | |
| *** monkegjinni has quit IRC | 01:29 | |
| *** monkegjinni has joined #melange | 01:30 | |
| *** Guest17725 is now known as skullbocks | 01:33 | |
| *** skullbocks is now known as Guest21243 | 01:33 | |
| *** monkegjinni has quit IRC | 01:34 | |
| *** suranga_ has joined #melange | 01:35 | |
| *** joshuatj has joined #melange | 01:58 | |
| joshuatj | hi there! Is it normal that I don't see an option to upload enrollment forms on "My Profile"? | 01:58 |
| *** Gentlecat has quit IRC | 01:59 | |
| *** joshuatj has quit IRC | 02:05 | |
| *** joshuatj has joined #melange | 02:06 | |
| *** joshuatj has joined #melange | 02:07 | |
| *** joshuatj has quit IRC | 02:11 | |
| *** monkegjinni has joined #melange | 02:11 | |
| *** edsiper has left #melange | 02:12 | |
| *** joshuatj has joined #melange | 02:12 | |
| *** joshuatj|2 has joined #melange | 02:14 | |
| *** joshuatj has quit IRC | 02:14 | |
| *** joshuatj|2 has quit IRC | 02:14 | |
| *** joshuatj has joined #melange | 02:14 | |
| *** abhinavsv3 has joined #melange | 02:19 | |
| *** rvraghav93_ has joined #melange | 02:21 | |
| *** rvraghav93 has quit IRC | 02:22 | |
| abhinavsv3 | Hi , I want to apply for GSoC 2014 . I am preparing my proposal . Where should I submit my proposal for review and suggestions from the community ? | 02:22 |
| abhinavsv3 | Also I havnt made a patch yet . I am working on it . Can I submit the patch link later ? | 02:23 |
| madrazr | joshuatj: it is *NOT* normal | 02:26 |
| madrazr | :) | 02:26 |
| *** monkegjinni has quit IRC | 02:26 | |
| abhinavsv3 | Hi , I am very much interested in doing melange project . I want to contribute to the project . I also want to participate in melange at GSoC 2014 . Should I definitely submit a patch for my Proposal to be selected ? | 02:32 |
| abhinavsv3 | or can i spend more time in carving of the proposal about my Part of work in the proposal | 02:33 |
| *** Guest21243 is now known as skullbocks | 02:34 | |
| *** skullbocks is now known as Guest14860 | 02:34 | |
| madrazr | abhinavsv3: it depends on which GSoC organization you are participating for. Each organization has its own set of requirements | 02:35 |
| madrazr | abhinavsv3: since you don't mention the organization name and asking the question here on #melange I am assuming this question is for Melange | 02:36 |
| madrazr | for Melange yes, we want you to submit patch(es) | 02:36 |
| madrazr | it is a requirement for us | 02:36 |
| madrazr | we want to know how you write code | 02:36 |
| olly | "I am very much interested in doing melange project" | 02:36 |
| madrazr | how you use version control system | 02:36 |
| madrazr | etc | 02:37 |
| madrazr | olly: oops | 02:37 |
| madrazr | thanks! | 02:37 |
| madrazr | my fail! | 02:37 |
| madrazr | abhinavsv3: sorry about that | 02:38 |
| abhinavsv3 | olly : Hey , Any mistake in that statement , why is mine statement quoted again ?? | 02:40 |
| olly | abhinavsv3: nothing wrong with what you said, just madrazr missed where you said "melange" | 02:42 |
| *** rubbersheep has joined #melange | 02:42 | |
| * olly was just pointing out you had actually said | 02:42 | |
| madrazr | abhinavsv3: nothing wrong, olly just showed me how stupid I am :P | 02:43 |
| olly | s/stupid/overloaded/ | 02:43 |
| madrazr | olly: he he true :) | 02:43 |
| madrazr | olly: on April 22nd I am going to party | 02:44 |
| madrazr | I have promised myself | 02:44 |
| madrazr | :D | 02:44 |
| *** MatthewWilkes is now known as mwilkes|away | 02:54 | |
| abhinavsv3 | Oh ok . Thank you . Working on the patch . Can I submit my patch after submitting my proposal . i.e, can I edit my proposal later ?? | 02:57 |
| *** robbyoconnor has quit IRC | 02:57 | |
| abhinavsv3 | OR , can I update my proposal later with the patch ? | 03:01 |
| olly | you can edit your proposal up to the deadline | 03:03 |
| abhinavsv3 | ok that means I have get a bug fix before the deadlin(21st March) right ? | 03:04 |
| * olly can't speak for the melange mentors | 03:04 | |
| olly | but probably | 03:04 |
| abhinavsv3 | ok . | 03:05 |
| abhinavsv3 | thank you | 03:05 |
| *** abhinavsv3 has quit IRC | 03:05 | |
| *** Gentlecat has joined #melange | 03:15 | |
| *** monkegjinni has joined #melange | 03:21 | |
| *** monkegjinni has quit IRC | 03:30 | |
| *** monkegjinni has joined #melange | 03:31 | |
| *** Guest14860 is now known as skullbocks | 03:35 | |
| *** monkegjinni has quit IRC | 03:35 | |
| *** skullbocks is now known as Guest24607 | 03:35 | |
| *** monkegjinni has joined #melange | 03:43 | |
| *** monkegjinni has quit IRC | 03:57 | |
| *** Gentlecat_ has joined #melange | 03:58 | |
| *** Gentlecat has quit IRC | 04:00 | |
| *** iSwapnil has joined #melange | 04:13 | |
| madrazr | olly: you were right even for Melange :) | 04:20 |
| *** vivekjain has joined #melange | 04:33 | |
| *** madrazr has quit IRC | 04:34 | |
| *** Guest24607 is now known as skullbocks | 04:36 | |
| *** skullbocks is now known as Guest28867 | 04:36 | |
| *** vivekjain has quit IRC | 04:42 | |
| *** edk_kdh has quit IRC | 04:58 | |
| *** edk_kdh has joined #melange | 04:59 | |
| *** denysbutenko has joined #melange | 05:13 | |
| *** Guest28867 is now known as skullbocks | 05:37 | |
| *** skullbocks is now known as Guest456 | 05:37 | |
| *** raghunayyar has joined #melange | 05:39 | |
| *** denysbutenko has quit IRC | 05:40 | |
| *** iSwapnil2 has joined #melange | 05:55 | |
| *** iSwapnil has quit IRC | 05:56 | |
| *** rubbersheep has quit IRC | 05:58 | |
| *** denysbutenko has joined #melange | 06:07 | |
| *** denysbutenko has quit IRC | 06:11 | |
| *** denysbutenko has joined #melange | 06:11 | |
| *** jasvir has joined #melange | 06:12 | |
| *** denysbutenko has quit IRC | 06:17 | |
| *** edk_kdh has quit IRC | 06:18 | |
| *** jasvir has quit IRC | 06:19 | |
| *** edk_kdh has joined #melange | 06:22 | |
| *** aps-sids has joined #melange | 06:25 | |
| aps-sids | Hello everyone. I'm writing a proposal in GSoC website but I'm not able to figure out how to do formatting (bold text, bullets, etc.) in the Content box. Can someone please help me about the same? | 06:26 |
| *** denysbutenko has joined #melange | 06:27 | |
| *** saapw has joined #melange | 06:28 | |
| *** denysbutenko has quit IRC | 06:32 | |
| aps-sids | nevermind, formatting toolbar had not loaded properly. | 06:37 |
| *** denysbutenko has joined #melange | 06:37 | |
| *** Guest456 is now known as skullbocks | 06:38 | |
| *** skullbocks is now known as Guest38596 | 06:38 | |
| *** denysbutenko has quit IRC | 06:40 | |
| *** denysbut_ has joined #melange | 06:40 | |
| *** monkegjinni has joined #melange | 06:42 | |
| *** denysbut_ has quit IRC | 06:42 | |
| *** denysbutenko has joined #melange | 06:42 | |
| *** denysbutenko has quit IRC | 06:46 | |
| *** denysbutenko has joined #melange | 06:46 | |
| *** edk_kdh has quit IRC | 06:49 | |
| *** edk_kdh has joined #melange | 06:54 | |
| *** edk_kdh has quit IRC | 06:58 | |
| *** madrazr has joined #melange | 06:59 | |
| *** ChanServ sets mode: +o madrazr | 06:59 | |
| *** s0b0lev has quit IRC | 07:01 | |
| *** edk_kdh has joined #melange | 07:02 | |
| *** monkegjinni has quit IRC | 07:03 | |
| *** monkegjinni has joined #melange | 07:03 | |
| *** shrihari has joined #melange | 07:04 | |
| *** monkegjinni has quit IRC | 07:08 | |
| *** edk_kdh has quit IRC | 07:08 | |
| *** edk_kdh has joined #melange | 07:13 | |
| *** s0b0lev has joined #melange | 07:13 | |
| *** edk_kdh has quit IRC | 07:20 | |
| *** jasvir has joined #melange | 07:21 | |
| *** edk_kdh has joined #melange | 07:22 | |
| *** suranga_ has quit IRC | 07:25 | |
| *** denysbutenko has quit IRC | 07:31 | |
| *** jasvir has quit IRC | 07:32 | |
| *** jasvir has joined #melange | 07:38 | |
| jasvir | madrazr: hello | 07:38 |
| madrazr | jasvir: hello | 07:38 |
| *** Guest38596 is now known as skullbocks | 07:39 | |
| *** skullbocks is now known as Guest38255 | 07:39 | |
| jasvir | madrazr: I wish to work on student and mentor flow. What are expectations from your side? | 07:46 |
| *** edk_kdh has quit IRC | 07:46 | |
| madrazr | jasvir: what do you mean by expectations from our side? | 07:46 |
| madrazr | jasvir: basic basic requirement is | 07:47 |
| madrazr | to read the ideas page description | 07:47 |
| madrazr | and also all the associated issues listed along with ideas description | 07:48 |
| madrazr | and to make a coherent proposal out of it | 07:48 |
| *** edk_kdh has joined #melange | 07:51 | |
| jasvir | madrazr: I meant to say that issues listed there do have any priorities? You want all the issues to be addressed or only some of them should get removed on priority basis. | 07:53 |
| madrazr | jasvir: ideally all | 07:54 |
| madrazr | jasvir: your proposal should maximize the number of issues it resolves | 07:55 |
| jasvir | madrazr: Ok. | 07:57 |
| shrihari | madrazr: How many slots are expected for Melange? | 07:58 |
| madrazr | shrihari: we don't know | 07:59 |
| madrazr | and that information is not public until Google announces the accepted students list publicly | 07:59 |
| *** jasvir has left #melange | 08:00 | |
| shrihari | Yeah, alright. | 08:00 |
| *** edk_kdh has quit IRC | 08:02 | |
| *** edk_kdh has joined #melange | 08:06 | |
| *** edk_kdh has quit IRC | 08:10 | |
| *** edk_kdh has joined #melange | 08:13 | |
| *** shrihari has quit IRC | 08:16 | |
| *** aps-sids has left #melange | 08:17 | |
| *** edk_kdh has quit IRC | 08:17 | |
| *** rvraghav93_ has quit IRC | 08:31 | |
| *** rvraghav93 has joined #melange | 08:32 | |
| *** Guest38255 is now known as skullbocks | 08:40 | |
| *** skullbocks is now known as Guest56393 | 08:40 | |
| *** shrihari has joined #melange | 08:49 | |
| *** denysbutenko has joined #melange | 08:52 | |
| *** denysbutenko has quit IRC | 08:56 | |
| *** madrazr has quit IRC | 09:06 | |
| *** s0b0lev has quit IRC | 09:16 | |
| *** Guest56393 is now known as skullbocks | 09:41 | |
| *** skullbocks is now known as Guest15177 | 09:41 | |
| *** rubbersheep has joined #melange | 09:46 | |
| *** rubbersheep_ has joined #melange | 09:47 | |
| *** rubbersheep has quit IRC | 09:50 | |
| *** Gentlecat_ has quit IRC | 09:53 | |
| *** Gentlecat has joined #melange | 09:53 | |
| *** rubbersheep_ has quit IRC | 10:09 | |
| *** mwilkes|away is now known as MatthewWilkes | 10:12 | |
| *** Guest15177 is now known as skullbocks | 10:42 | |
| *** skullbocks is now known as Guest7185 | 10:42 | |
| *** hiddenpearls has joined #melange | 10:52 | |
| *** hiddenpearls has quit IRC | 11:04 | |
| *** s0b0lev has joined #melange | 11:05 | |
| *** Guest7185 is now known as skullbocks | 11:43 | |
| *** skullbocks is now known as Guest85751 | 11:43 | |
| *** larryxiao1 has joined #melange | 11:48 | |
| *** shrihari has quit IRC | 12:04 | |
| *** s0b0lev has quit IRC | 12:17 | |
| *** s0b0lev has joined #melange | 12:29 | |
| *** monkegjinni has joined #melange | 12:37 | |
| *** Guest85751 is now known as skullbocks | 12:44 | |
| *** skullbocks is now known as Guest69490 | 12:44 | |
| *** Niharika has joined #melange | 12:45 | |
| *** monkegjinni has quit IRC | 12:48 | |
| *** monkegjinni has joined #melange | 12:53 | |
| *** shrihari has joined #melange | 13:01 | |
| *** denysbutenko has joined #melange | 13:04 | |
| *** s0b0lev has quit IRC | 13:09 | |
| *** Akz- has joined #melange | 13:28 | |
| *** joshuatj|2 has joined #melange | 13:32 | |
| *** joshuatj has quit IRC | 13:35 | |
| *** shrihari has quit IRC | 13:39 | |
| *** Guest69490 is now known as skullbocks | 13:45 | |
| *** skullbocks is now known as Guest54029 | 13:45 | |
| *** shrihari has joined #melange | 13:46 | |
| *** Gentlecat has quit IRC | 13:55 | |
| *** joshuatj|2 has quit IRC | 13:55 | |
| *** joshuatj|2 has joined #melange | 13:57 | |
| *** joshuatj has joined #melange | 14:01 | |
| *** ayushpix has joined #melange | 14:02 | |
| *** suranga_ has joined #melange | 14:02 | |
| *** joshuatj|2 has quit IRC | 14:03 | |
| *** fgnievinski has joined #melange | 14:13 | |
| fgnievinski | GSoC is emailing me broken links; is this the right place to report, please? | 14:14 |
| *** trout has quit IRC | 14:19 | |
| *** variable has joined #melange | 14:22 | |
| *** rvraghav93 has quit IRC | 14:23 | |
| *** joshuatj|2 has joined #melange | 14:23 | |
| *** joshuatj has quit IRC | 14:24 | |
| gevaerts | fgnievinski: it probably is, but see the email address in the topic if you don't get a useful response here | 14:24 |
| *** rvraghav93 has joined #melange | 14:25 | |
| fgnievinski | it says: "Someone has submitted a new proposal to GNU Octave named Project under GNU Octave at /gsoc/proposal/review/org/google/gsoc2014/cantbeat007/5629499534213120:" | 14:27 |
| fgnievinski | the http://domain.com/ seems to be missing | 14:28 |
| tpb | Title: Domain Name Registration and Web Hosting | Domain.com (at domain.com) | 14:28 |
| fgnievinski | maybe it's better to email [email protected] ? | 14:33 |
| *** saapw has quit IRC | 14:34 | |
| *** raghunayyar has quit IRC | 14:37 | |
| *** rvraghav93 has quit IRC | 14:38 | |
| *** fgnievinski has quit IRC | 14:39 | |
| *** rvraghav93 has joined #melange | 14:39 | |
| *** rihbyne has joined #melange | 14:39 | |
| *** iSwapnil2 has quit IRC | 14:45 | |
| *** Guest54029 is now known as skullbocks | 14:46 | |
| *** skullbocks is now known as Guest60221 | 14:46 | |
| *** rbyne has joined #melange | 14:50 | |
| *** rihbyne has quit IRC | 14:51 | |
| *** rbyne is now known as rihbyne | 14:51 | |
| *** rihbyne has quit IRC | 15:00 | |
| *** variable has quit IRC | 15:03 | |
| *** iSwapnil has joined #melange | 15:04 | |
| *** shrihari has quit IRC | 15:05 | |
| *** variable has joined #melange | 15:06 | |
| *** rihbyne has joined #melange | 15:07 | |
| *** shrihari has joined #melange | 15:07 | |
| *** rvraghav93 has quit IRC | 15:16 | |
| *** rvraghav93 has joined #melange | 15:16 | |
| *** monkegjinni has quit IRC | 15:18 | |
| *** rbyne has joined #melange | 15:29 | |
| *** rihbyne has quit IRC | 15:31 | |
| *** Guest60221 is now known as skullbocks | 15:47 | |
| *** skullbocks is now known as Guest30320 | 15:47 | |
| *** rishabh_ has joined #melange | 15:51 | |
| *** iSwapnil has quit IRC | 15:52 | |
| *** bitgeeky has joined #melange | 15:58 | |
| *** rbyne is now known as rihbyne | 16:02 | |
| *** raghunayyar has joined #melange | 16:07 | |
| *** rihbyne has quit IRC | 16:07 | |
| *** iSwapnil has joined #melange | 16:12 | |
| *** carols has joined #melange | 16:12 | |
| *** ayushpix has quit IRC | 16:40 | |
| *** Guest30320 is now known as skullbocks | 16:48 | |
| *** skullbocks is now known as Guest98804 | 16:48 | |
| *** bitgeeky has quit IRC | 16:56 | |
| *** rihbyne has joined #melange | 16:58 | |
| *** sekharz2 has joined #melange | 17:05 | |
| *** flegmash has joined #melange | 17:06 | |
| *** rbyne has joined #melange | 17:06 | |
| *** jasvir has joined #melange | 17:07 | |
| *** rihbyne has quit IRC | 17:07 | |
| *** Arc has joined #melange | 17:08 | |
| Arc | hey guys | 17:08 |
| Arc | are you aware the link to the project sent to mentors is missing the protocol and host? | 17:09 |
| Arc | Marsh has submitted a new proposal to Copyleft Games named Dungeon Master at /gsoc/proposal/review/org/google/gsoc2014/marsh/5629499534213120 -- and the url it points to is the same | 17:09 |
| *** Niharika has quit IRC | 17:09 | |
| *** saapw has joined #melange | 17:12 | |
| *** rihbyne has joined #melange | 17:13 | |
| *** rbyne has quit IRC | 17:13 | |
| *** rbyne has joined #melange | 17:18 | |
| *** rihbyne has quit IRC | 17:18 | |
| *** rihbyne has joined #melange | 17:24 | |
| *** rbyne has quit IRC | 17:24 | |
| *** saapw has left #melange | 17:25 | |
| *** iSwapnil has quit IRC | 17:30 | |
| *** iSwapnil has joined #melange | 17:32 | |
| *** rbyne has joined #melange | 17:33 | |
| *** rihbyne has quit IRC | 17:33 | |
| *** rihbyne has joined #melange | 17:39 | |
| *** MatthewWilkes is now known as mwilkes|away | 17:39 | |
| *** rbyne has quit IRC | 17:39 | |
| *** flegmash has quit IRC | 17:40 | |
| *** RaulT has joined #melange | 17:44 | |
| *** jasvir_ has joined #melange | 17:45 | |
| *** denysbut_ has joined #melange | 17:47 | |
| *** jasvir_ has quit IRC | 17:47 | |
| *** Guest98804 is now known as skullbocks | 17:49 | |
| *** skullbocks is now known as Guest42500 | 17:49 | |
| *** denysbutenko has quit IRC | 17:50 | |
| *** edu159 has joined #melange | 17:52 | |
| *** edu159 has left #melange | 17:52 | |
| *** jasvir has quit IRC | 17:53 | |
| *** shrihari has quit IRC | 17:53 | |
| *** rihbyne has quit IRC | 17:54 | |
| *** rihbyne has joined #melange | 17:57 | |
| *** asmeurer__ has joined #melange | 18:00 | |
| *** shrihari has joined #melange | 18:02 | |
| *** edu159 has joined #melange | 18:02 | |
| *** kuntal has joined #melange | 18:10 | |
| *** shrihari has quit IRC | 18:18 | |
| *** asmeurer__ has quit IRC | 18:25 | |
| *** hiddenpearls has joined #melange | 18:26 | |
| *** pavnik has joined #melange | 18:30 | |
| *** pavnik has quit IRC | 18:34 | |
| *** meflin has joined #melange | 18:46 | |
| *** jahewson has joined #melange | 18:47 | |
| jahewson | Hi, I'm a GSoC mentor and I've spotted a couple of issues with the melange automated e-mails... | 18:49 |
| Arc | ditto | 18:49 |
| Arc | liken the link url? :-) | 18:49 |
| jahewson | Yep, the proposal e-mails start with a partial link | 18:50 |
| *** Guest42500 is now known as skullbocks | 18:50 | |
| *** skullbocks is now known as Guest18948 | 18:50 | |
| jahewson | But also the footer is missing the unsubscribe url: | 18:50 |
| jahewson | You are receiving this message because you are participating in Google Summer of Code 2014. | 18:50 |
| jahewson | To stop receiving these messages, go to: . | 18:50 |
| *** madrazr has joined #melange | 18:50 | |
| *** ChanServ sets mode: +o madrazr | 18:50 | |
| carols | madrazr: we've had some email issues reported from the mentors | 18:53 |
| RaulT | madrazr: When you have some free time to discuss some gsoc ideas please ping me | 18:53 |
| *** iSwapnil has quit IRC | 18:53 | |
| carols | jahewson Arc: if madrazr doesn't respond you can also forward the email to daniel and he can take a look | 18:54 |
| madrazr | carols: email issues? Are the details here in the logs? Or emails? | 18:55 |
| carols | madrazr: jahewson and Arc can tell you about it | 18:55 |
| madrazr | RaulT: can't say I am "free" right now, but we can have this conversation. I will be around but a bit slow in responding | 18:56 |
| madrazr | RaulT: how can I help you? | 18:56 |
| madrazr | jahewson: Arc: is there some background reading I can do about this issue? | 18:56 |
| * madrazr is reading the logs | 18:57 | |
| RaulT | madrazr: first about the ranking system. My idea is implementing a ranking system for the proposals | 18:57 |
| *** mwilkes|away is now known as MatthewWilkes | 18:57 | |
| *** hiddenpearls has quit IRC | 18:57 | |
| RaulT | madrazr: Similar to what mentors have | 18:57 |
| carols | madrazr: "Hi, I'm getting automated e-mails from melange with the following footer: | 18:58 |
| carols | [11:42am] jahewson: You are receiving this message because you are participating in Google Summer of Code 2014. | 18:58 |
| jahewson | madrazr: I'm a GSoC mentor and noticed a few minor issues in the proposal notification e-mails. The unsubscribe link at the bottom is missing, it's just a ".", also the e-mail starts with a broken link to the proposal page on melange. | 18:58 |
| carols | [11:42am] jahewson: To stop receiving these messages, go to: ." | 18:58 |
| RaulT | madrazr: And in doing so the deduplication process can be fully automated : The mentors decide if they should accept the proposal and give it a grade | 18:58 |
| RaulT | madrazr: And the students rank the proposal as well. | 18:58 |
| *** robbyoconnor has joined #melange | 18:59 | |
| RaulT | madrazr: With this information you can deduce a way to distribute accepted students to orgs so that everyone is happy | 18:59 |
| *** sekharz2 has quit IRC | 18:59 | |
| RaulT | madrazr: And after slot realocation the deduplication happens again ? | 18:59 |
| RaulT | madrazr: Is it clear so far ? | 18:59 |
| * gevaerts thinks that RaulT is an optimist | 19:01 | |
| *** kuntal has quit IRC | 19:01 | |
| gevaerts | "everyone is happy" and gsoc do not go together! | 19:01 |
| RaulT | gevaerts: What i mean is mentors get the best of the applications they are willing to accept and the students to the orgs they whant to go if they get accepted :P | 19:02 |
| RaulT | gevaerts: for everything else there's mastercard | 19:03 |
| *** rihbyne has quit IRC | 19:03 | |
| *** hiddenpearls has joined #melange | 19:04 | |
| gevaerts | RaulT: I'm just an interested observer here, so feel free to tell me to shut up, but how do you handle things like one of the orgs having a possible replacement project and the other not? | 19:04 |
| gevaerts | Deduplication is about a lot more than how good the proposal is on its own | 19:05 |
| madrazr | jahewson: carols: I see | 19:05 |
| madrazr | taking a look (also in a talk) | 19:05 |
| RaulT | gevaerts: what do you mean by replacing project ? | 19:06 |
| gevaerts | RaulT: there might be another student with a similar proposal that's also acceptable | 19:06 |
| gevaerts | Often, if that's the case for one organisation but not the other, the organisation without such a replacement will get the student | 19:07 |
| RaulT | gevaerts: That's where the mentor ranking comes into play. If the org gets a list of proposals that they are willing to accept and order them. | 19:08 |
| olly | jahewson: there's an open issue for that | 19:08 |
| olly | for the lack of the unsubscribe link at least - I think the links are in the HTML version | 19:08 |
| gevaerts | RaulT: there is of course the other interesting issue of different orgs using very different scoring and ranking systems | 19:08 |
| RaulT | gevaerts: but within an org the scoring system should be consistent | 19:09 |
| gevaerts | If it's there at all, yes :) | 19:09 |
| meflin | umbrella's :) | 19:09 |
| jahewson | olly: ok, thanks | 19:09 |
| olly | jahewson: http://code.google.com/p/soc/issues/detail?id=1519 | 19:09 |
| tpb | <http://ln-s.net/BIar> (at code.google.com) | 19:09 |
| *** jahewson has quit IRC | 19:09 | |
| gevaerts | Ah yes. It's not raining here, so I forgot about umbrellas :) | 19:10 |
| meflin | that could lead to disaster of suborgs with more students then mentors :D | 19:11 |
| olly | RaulT: we might have some mentors who are only up for mentoring a limited set of proposals though | 19:11 |
| gevaerts | So no, the scoring system within an org isn't guaranteed to exist at all, and if it exists it's not guaranteed to be consistent overall | 19:11 |
| olly | so if we don't get student X, we might not want another student | 19:11 |
| olly | but if we don't get student Y, we would | 19:11 |
| olly | a ranked list isn't enough to capture that | 19:11 |
| gevaerts | Hmmm | 19:12 |
| gevaerts | Maybe the org admin should provide a set of rules that handle all of these things :) | 19:12 |
| olly | submit a patch to melange to implement your preferences? | 19:12 |
| olly | i feel the current iterative solution we have is practical at least | 19:13 |
| madrazr | RaulT: gevaerts: yes! The Stable Marriage problem :) | 19:13 |
| madrazr | http://en.wikipedia.org/wiki/Stable_marriage_problem | 19:13 |
| tpb | <http://ln-s.net/BIb3> (at en.wikipedia.org) | 19:13 |
| madrazr | RaulT: also be aware of the fact that the current Melange system does not do anything to encourage ranking | 19:14 |
| olly | it used to be we set a ranked order and the top N were taken | 19:15 |
| madrazr | all Melange cares about is before we run the "accept students" jobs we have proposals marked as "acceptable" by organization | 19:15 |
| olly | but we got to fiddle between deduplication rounds | 19:15 |
| olly | personally, i find the current system of marking students explicitly better | 19:16 |
| olly | less chance of a surprise acceptance | 19:16 |
| RaulT | madrazr: So the students ranking proposal is done to help with the deduplication process | 19:16 |
| RaulT | ? | 19:16 |
| meflin | its for internal org use | 19:16 |
| olly | RaulT: no, ranking is an optional tool for the org | 19:16 |
| madrazr | olly: 1519 needs to be bumped somehow | 19:16 |
| olly | madrazr: i was surprised it didn't get fixed for so long - it seems it shouldn't be very hard as the HTML has them | 19:17 |
| RaulT | olly: What purpose the this issue intend to serve ? https://code.google.com/p/soc/issues/detail?id=1530 | 19:18 |
| tpb | <http://ln-s.net/BIbI> (at code.google.com) | 19:18 |
| *** hiddenpearls has quit IRC | 19:18 | |
| RaulT | olly: Just to help in some particular cases of deduplication ? | 19:18 |
| madrazr | olly: yeah, the problem is we rely a lot on App Engine to do email handling for us | 19:18 |
| madrazr | olly: we should generate the email content ourselves for text/plain and text/html I guess | 19:19 |
| olly | RaulT: seems unnecessary to me - if students have a strong preference, they can just tell the org they prefer | 19:19 |
| madrazr | olly: but someone should figure out how to do it | 19:19 |
| olly | but they should realise that their preference won't necessarily be respected, and only apply for projects they'd be happy to do | 19:20 |
| * olly is kind of amazed 1530 is medium but 1519 is low | 19:20 | |
| gevaerts | RaulT: the original text there has some value I think, but the followup comments make a lot of invalid assumptions | 19:21 |
| *** ayushpix has joined #melange | 19:21 | |
| olly | madrazr: ah, so the text is autogenerated from the html? | 19:21 |
| *** bitgeeky has joined #melange | 19:22 | |
| madrazr | olly: yeah, App Engine does that HTML to text sanitization for us | 19:22 |
| olly | perhaps just sending text/plain with the URLs in would be better? | 19:22 |
| madrazr | and obviously App Engine does not know what to do with the anchor texts | 19:22 |
| *** robbyoconnor has quit IRC | 19:22 | |
| gevaerts | Just drop html email. It's been a horrible mistake all along :) | 19:22 |
| olly | gui clients will make the emails clickable anyway | 19:22 |
| madrazr | olly: may be, but then some people with fancy clients will be upset | 19:23 |
| RaulT | gevaerts: So according to the original text. If a student gets accepted to two organizations the student rank is revealed to the mentors as a preview of what the student wants to chose ? | 19:23 |
| olly | gevaerts: i feel the whole wheel thing was where we went wrong | 19:23 |
| olly | madrazr: well, anyone who reads the text/plain is upset now | 19:23 |
| RaulT | gevaerts: Or alternatively the student ranking is used to decide to wich organization the student goes to | 19:23 |
| *** robbyoconnor has joined #melange | 19:23 | |
| gevaerts | RaulT: the student ranking is *optional* | 19:24 |
| gevaerts | Oh, wait | 19:24 |
| gevaerts | *that* ranking :) | 19:24 |
| gevaerts | No | 19:24 |
| gevaerts | Organisations are allowed to take the student's preference into account, but they explicitly don't have to | 19:24 |
| olly | and if I have a stack of good students and the other org has very few, I'm likely to let them have a student we both tried to accept | 19:25 |
| gevaerts | The reason that that suggestion might be useful is that right now, if they want to take the student's preference into account, they have to ask the student, which reveals a lot about acceptance, which isn't actually allowed | 19:25 |
| olly | that's not a process that's easy to automate | 19:25 |
| RaulT | gevaerts: So the information is revealed and organizations take that into account | 19:25 |
| RaulT | gevaerts: By whatever process it consideres fit | 19:25 |
| olly | gevaerts: but the simple instruction of telling students to mention it to the org they prefer solves that | 19:26 |
| olly | no need for elaborate machinery in melange for a corner case | 19:26 |
| gevaerts | olly: yes, indeed, but some students have the idea that mentioning "competing" proposals hurt their chances :) | 19:26 |
| olly | well, let's fix that | 19:27 |
| gevaerts | And the thing is that I don't trust org admin and mentor common sense enough to be totally convinced that that's always untrue... | 19:27 |
| olly | the orgs aren't naive, we know they can apply to up to 4 other orgs | 19:27 |
| gevaerts | Well... | 19:27 |
| olly | perhaps knowing there's competition for a good student makes them more desirable? | 19:28 |
| * gevaerts isn't *entirely* convinced :) | 19:28 | |
| olly | melange also gives clues that the student may have applied elsewhere | 19:28 |
| olly | unless that's been changed | 19:29 |
| gevaerts | I seem to vaguely remember a mentor or org admin saying something on the mailing list or irc (I told you it was vague!) about picking a student that didn't have a risk of being a duplicate | 19:29 |
| gevaerts | Personally I think that's just stupid, but then I'm not in charge anywhere | 19:29 |
| RaulT | olly: So for this simple purpose a 5 star ranking system similar to what mentors get should be enough ? | 19:30 |
| RaulT | olly: And how the mentors take intor consideration that information is entirely up to them | 19:31 |
| RaulT | into* | 19:31 |
| gevaerts | Of course there are also lots of orgs that will just ask the student directly | 19:31 |
| gevaerts | And I'd say lying to them in such a case is positively dangerous | 19:31 |
| olly | gevaerts: we do | 19:31 |
| gevaerts | olly: yes, we always did too | 19:32 |
| olly | and so far, haven't seen any indication that students lie to us about it | 19:32 |
| * gevaerts nods | 19:32 | |
| * olly has also noticed that the best proposals come from students who are much less likely to have applied elsewhere | 19:32 | |
| *** robbyoconnor has quit IRC | 19:33 | |
| RaulT | olly: I't better to focus on one goal then spread your attention on many subjects | 19:34 |
| RaulT | than* | 19:34 |
| olly | indeed | 19:35 |
| olly | at least the limit's now 5, not 20 | 19:35 |
| RaulT | olly: So to onclude the debate on this. This feature is purely optional and serves as some additional information mentors could consider when deciding upon a duplication. So for this purpose a simple ranking system (eg 5 star ranking) whould suffice | 19:36 |
| RaulT | would* | 19:37 |
| gevaerts | I'd say so, yes | 19:37 |
| olly | RaulT: my view is that if anything is added, it should just be a field which only gets shown to the orgs when deduplication starts | 19:37 |
| RaulT | olly: Got it! Thanks ! | 19:38 |
| RaulT | The next thing I want to ask about is this issue : https://code.google.com/p/soc/issues/detail?id=515 | 19:39 |
| tpb | <http://ln-s.net/BIcu> (at code.google.com) | 19:39 |
| olly | actually, I guess it should show the ratings against the other orgs shown as duplicates as well as the one for yours | 19:39 |
| *** edu159 has left #melange | 19:40 | |
| olly | RaulT: ooh, i reported that one | 19:41 |
| gevaerts | That one *will* make people buy you drinks | 19:41 |
| RaulT | with some major parsing it can be done | 19:42 |
| RaulT | considering the output of something like this http://docs.python.org/2/library/difflib.html | 19:42 |
| tpb | Title: 7.4. difflib — Helpers for computing deltas Python v2.7.6 documentation (at docs.python.org) | 19:42 |
| RaulT | you can get a summary of the changes between two consecutive proposals | 19:42 |
| olly | yes | 19:42 |
| RaulT | and If you have the current state of the proposal and the list of changes it should be easy to reverse engineer the changes | 19:43 |
| olly | conceptually it's pretty straightforward, but I can believe that actually implementing it isn't so easy | 19:44 |
| RaulT | I consider saving the most recent state because in most cases you do not want to compute all history | 19:44 |
| olly | yes, you want the current version efficiently | 19:44 |
| olly | TBH, you could probably just save each version | 19:44 |
| RaulT | olly: And do the diff between whatever versions you want ? | 19:46 |
| olly | yes | 19:46 |
| olly | something like http://en.wikipedia.org/w/index.php?title=Google_Summer_of_Code&action=history maybe | 19:47 |
| tpb | <http://ln-s.net/BIdL> (at en.wikipedia.org) | 19:47 |
| RaulT | olly: Isn't a problem if we save a different version for every litle change the user makes. I think it's not ok to save 500 words for a typ-o | 19:47 |
| gevaerts | Why not? | 19:47 |
| RaulT | It just seems very inefficient | 19:48 |
| gevaerts | Then use a VCS backend :) | 19:49 |
| olly | just chant "disk is cheap" as you do it | 19:49 |
| olly | that's what everyone else seems to do these days | 19:49 |
| *** Guest18948 is now known as skullbocks | 19:51 | |
| *** meflin has left #melange | 19:51 | |
| * olly doesn't have the stats, but I'd see if you can find info on how many proposal revisions there are and how big a proposal typically is before worrying a lot about the size | 19:51 | |
| *** skullbocks is now known as Guest31223 | 19:51 | |
| RaulT | that seems like a reasonable alternative :) | 19:51 |
| gevaerts | I'd still go for a VCS backend :) | 19:52 |
| RaulT | gevaerts: But that provides a full list of features that are not required | 19:53 |
| olly | gevaerts: appengine might thwart that, but it's worth considering | 19:53 |
| olly | RaulT: don't use them all then | 19:53 |
| gevaerts | RaulT: in that case, you also can't use appengine, or html, or javascript :) | 19:53 |
| RaulT | olly: Don't use what ? | 19:53 |
| olly | makes more sense to use something existing than reinvent it | 19:53 |
| olly | don't use the features of a VCS backend you don't require | 19:53 |
| olly | as gevaerts says, melange doesn't reject HTML because it doesn't need a <blink> tag | 19:54 |
| *** ayushpix has quit IRC | 19:54 | |
| gevaerts | What you want is a way to efficiently store revisions of a text file, and a way to get specific revisions out | 19:54 |
| gevaerts | If you don't want much more than that, go for rcs :) | 19:55 |
| *** denysbut_ has quit IRC | 19:56 | |
| RaulT | gevaerts: Or just save them all as olly suggested | 19:56 |
| RaulT | gevaerts: but get a rcs that does just that is hard to find | 19:59 |
| gevaerts | RaulT: you're misunderstanding me | 19:59 |
| olly | you don't need an RCS which does *JUST* that | 19:59 |
| olly | it just needs to do *AT LEAST* that | 20:00 |
| gevaerts | I'm not saying to use *an* RCS. I'm suggesting to use rcs :) | 20:00 |
| olly | my car has a reverse gear, but I can just ignore it if I only need to drive forwards | 20:00 |
| olly | gevaerts: ooh, retro | 20:00 |
| gevaerts | But yes, I'd go for one of the modern ones. rcs is getting a bit old now :) | 20:00 |
| olly | it's true that this is pretty much exactly what rcs does though | 20:01 |
| RaulT | olly: That's what i suggested. Implementing a RCS for this particular purpose. | 20:03 |
| RaulT | olly: because i don't thing that adding something like git or mercurial might help | 20:04 |
| * gevaerts wants people to stop confusing people by calling the *concept* RCS :) | 20:04 | |
| gevaerts | RCS is an actual implementation! | 20:04 |
| olly | RaulT: it's mad to build your own versioning system | 20:04 |
| * gevaerts nods | 20:05 | |
| olly | you could do this with an existing system | 20:05 |
| olly | git allows you to diff versions of a file, and list the history | 20:05 |
| olly | I imagine mercurial does | 20:05 |
| olly | sure they support stuff like branches which you don't need | 20:06 |
| olly | but you can ignore all that | 20:06 |
| gevaerts | The bit you need will still be more efficient and better tested than whatever you can come up with on short notice | 20:07 |
| olly | if you write your own, you'll end up having more bugs, and it'll likely be slower, because these popular tools have been optimised a lot | 20:07 |
| RaulT | olly: So my job here is to find a way to integrate an versioning system with the proposal system | 20:07 |
| RaulT | a* | 20:07 |
| olly | well, gevaerts and I are just heckling from the peanut gallery really | 20:07 |
| olly | we aren't melange devs | 20:07 |
| gevaerts | Yes. We all do what we're good at :) | 20:07 |
| olly | but I'd say that's the best approach | 20:07 |
| RaulT | cool | 20:09 |
| RaulT | so that's done. Thank you for your help | 20:12 |
| RaulT | next one : https://code.google.com/p/soc/issues/detail?id=1816 This should be solved with a "Save" and a "Submit for review" button | 20:13 |
| tpb | <http://ln-s.net/BIfG> (at code.google.com) | 20:13 |
| RaulT | ? | 20:13 |
| *** robbyoconnor has joined #melange | 20:14 | |
| RaulT | does this seem reasonable ? | 20:20 |
| *** variable is now known as trout | 20:29 | |
| *** denysbutenko has joined #melange | 20:29 | |
| *** denysbutenko has quit IRC | 20:33 | |
| *** suranga_ has quit IRC | 20:35 | |
| *** sekharz has joined #melange | 20:38 | |
| RaulT | where has everyone gone ? | 20:42 |
| *** Guest31223 is now known as skullbocks | 20:52 | |
| *** skullbocks is now known as Guest27869 | 20:52 | |
| *** vivekjain has joined #melange | 21:03 | |
| cosenal | https://code.google.com/p/soc/issues/detail?id=2093 | 21:03 |
| tpb | Title: Issue 2093 - soc - past GSoC proposal link - SoC (Spice of Creation) - Google Project Hosting (at code.google.com) | 21:03 |
| cosenal | ^ it says fixed, but I still don't see old proposals | 21:03 |
| *** monkegjinni has joined #melange | 21:06 | |
| *** suranga_ has joined #melange | 21:08 | |
| *** monkegjinni has quit IRC | 21:21 | |
| *** monkegjinni has joined #melange | 21:22 | |
| *** monkegjinni has quit IRC | 21:26 | |
| madrazr | cosenal: what do you mean by "don't see"? | 21:27 |
| madrazr | cosenal: do you see an error like server error? Or some sort of a message? | 21:27 |
| madrazr | cosenal: also can you please give us the URL of the page where you see this behavior? | 21:28 |
| cosenal | madrazr: everything as in the issue I reported | 21:28 |
| *** denysbutenko has joined #melange | 21:30 | |
| *** dblia_ has joined #melange | 21:30 | |
| *** Dogedrop has joined #melange | 21:30 | |
| Dogedrop | there seems to be a problem with enrolment form uploading | 21:31 |
| Dogedrop | says service unavailable after loading for a long time | 21:31 |
| carols | Dogedrop: madrazr can help you. | 21:31 |
| madrazr | Dogedrop: which browser do you use | 21:31 |
| madrazr | ? | 21:31 |
| Dogedrop | tried both safari and firefox | 21:31 |
| dblia_ | i have the same with firefox 27.0.1 | 21:31 |
| madrazr | Dogedrop: can you try the same in private browsing mode? | 21:32 |
| madrazr | dblia_: to you too ^ | 21:32 |
| madrazr | cosenal: the URL has changed | 21:33 |
| madrazr | cosenal: in the sense that we made some changes to our data layer that might have given your proposal a different ID | 21:33 |
| madrazr | cosenal: the last part in your URL | 21:33 |
| cosenal | madrazr: how do I figure out the new ID? | 21:34 |
| madrazr | cosenal: you should be to go to your GSoC 2012 dashboard and access your proposal from there though | 21:34 |
| madrazr | cosenal: you don't have to figure out the ID :) | 21:34 |
| *** denysbutenko has quit IRC | 21:34 | |
| madrazr | cosenal: you should just be able to go to the dashboard and find the link to your proposal from there | 21:34 |
| Dogedrop | tried both safari and firefox private mode, same thing | 21:34 |
| dblia_ | madrazr: the same | 21:34 |
| madrazr | dblia_: Dogedrop: would you mind sending a screenshot of the error so that I can use it to analyze the problem? | 21:35 |
| madrazr | you can PM the link | 21:35 |
| madrazr | or email me the screenshot | 21:35 |
| Dogedrop | sure | 21:35 |
| madrazr | thanks | 21:35 |
| *** dblia has joined #melange | 21:36 | |
| *** dblia_ has quit IRC | 21:36 | |
| *** dblia has left #melange | 21:36 | |
| *** dblia has joined #melange | 21:36 | |
| madrazr | I don't see errors anything in our logs though | 21:36 |
| madrazr | Dogedrop: dblia ^ | 21:36 |
| cosenal | madrazr: ok, cool. I see it from the dashboard. A small issue: when I click on the public link, it says "no page name. report the issue to soc bug tracker…" | 21:36 |
| madrazr | *I don't see any errors in our logs though | 21:36 |
| dblia | madrazr: it happend after editing my profile info | 21:37 |
| madrazr | cosenal: yeah make some noise here please https://code.google.com/p/soc/issues/detail?id=2096 :) | 21:37 |
| tpb | <http://ln-s.net/BIjU> (at code.google.com) | 21:37 |
| dblia | madrazr: the previously uploaded form was missing and tried to re-upload it | 21:38 |
| cosenal | madrazr: all right, thanks! | 21:38 |
| madrazr | dblia: Dogedrop: don't upload anything now | 21:39 |
| madrazr | It is time for me to be terribly upset with someone | 21:39 |
| madrazr | let me go do something about that :D | 21:40 |
| Dogedrop | alright | 21:40 |
| dblia | madrazr: ok, thank you | 21:40 |
| * gevaerts didn't do it! | 21:40 | |
| madrazr | gevaerts: :P | 21:40 |
| madrazr | dblia: Dogedrop: I wish I could give you a link to the health status of the service we use to store files | 21:41 |
| madrazr | like the health monitoring dashboard we have for large services like GMail, Drive etc | 21:41 |
| madrazr | so that you will know when you can try again | 21:41 |
| madrazr | but looks like there is no such thing | 21:42 |
| madrazr | so may be you will have to check for yourselves again after sometime | 21:42 |
| madrazr | or ping me | 21:42 |
| dblia | madrazr: i will retry tommorow | 21:42 |
| Dogedrop | what is the problem? | 21:42 |
| madrazr | dblia: thanks! | 21:42 |
| madrazr | Dogedrop: looks like a blobstore outage | 21:42 |
| madrazr | but I can't confirm it | 21:43 |
| madrazr | because there is no health monitoring dashboard | 21:43 |
| Dogedrop | alright I will try later then | 21:44 |
| Dogedrop | thanks | 21:44 |
| *** Dogedrop has quit IRC | 21:45 | |
| RaulT | madrazr: Can we talk now ? | 21:48 |
| madrazr | RaulT: yeah> | 21:49 |
| madrazr | ? | 21:49 |
| RaulT | madrazr: What is your take about integrating an existing revision sistem with melange to provide the "track differences " feature for proposals ? | 21:50 |
| madrazr | we do not want a version control system | 21:51 |
| madrazr | all we want is to be able to view diffs between any two versions of a proposal | 21:52 |
| madrazr | any version control system today is much more complicated than that | 21:52 |
| RaulT | madrazr: So what is needed is an implementation of _just_ this feature | 21:52 |
| RaulT | ? | 21:53 |
| *** Guest27869 is now known as skullbocks | 21:53 | |
| RaulT | madrazr: Without the added overhead a revision system comes with. | 21:53 |
| *** skullbocks is now known as Guest89283 | 21:53 | |
| madrazr | storing versions | 21:53 |
| madrazr | showing diffs | 21:53 |
| RaulT | madrazr: For storing should I save the full proposal and just add a diff between the versions I want | 21:54 |
| RaulT | madrazr: Or try to store just the diffs and compute the result based on the requirement | 21:55 |
| *** vivekjain has quit IRC | 21:55 | |
| RaulT | requirements* ? | 21:55 |
| *** rishabh_ has quit IRC | 21:59 | |
| *** bitgeeky has quit IRC | 22:01 | |
| *** denysbutenko has joined #melange | 22:16 | |
| RaulT | madrazr: Just a kindly reminder in case you got distracted, I am still waiting a response. Plese respond when you find the time. | 22:16 |
| *** Darrel has quit IRC | 22:25 | |
| *** Darrel has joined #melange | 22:26 | |
| *** raghunayyar has quit IRC | 22:27 | |
| *** monkegjinni has joined #melange | 22:36 | |
| *** sekharz has quit IRC | 22:48 | |
| *** robbyoconnor has quit IRC | 22:52 | |
| *** Guest89283 is now known as skullbocks | 22:54 | |
| *** skullbocks is now known as Guest11864 | 22:54 | |
| *** monkegji_ has joined #melange | 23:03 | |
| *** monkegjinni has quit IRC | 23:03 | |
| *** carols has quit IRC | 23:06 | |
| *** downey has quit IRC | 23:11 | |
| *** downey has joined #melange | 23:15 | |
| *** monkegji_ has quit IRC | 23:18 | |
| *** monkegjinni has joined #melange | 23:27 | |
| *** denysbutenko has quit IRC | 23:47 | |
| *** monkegjinni has quit IRC | 23:53 | |
| *** monkegjinni has joined #melange | 23:54 | |
| *** Guest11864 is now known as skullbocks | 23:55 | |
| *** skullbocks is now known as Guest47698 | 23:55 | |
| *** monkegjinni has quit IRC | 23:58 | |
Generated by irclog2html.py 2.12.1 by Marius Gedminas - find it at mg.pov.lt!