| | <<<Back 1 day (to 2013/12/07) | 2013/12/08 |
| tor5 | morning | 17:11.54 |
| kens | morning tor | 17:18.56 |
| robin_watts_mac | morning | 17:24.39 |
| chrisl | Morning all | 17:27.18 |
| kens | morning robin | 17:27.20 |
| tor5 | breakfast at eight? | 17:49.50 |
| chrisl | Fine by me | 17:50.40 |
| kens | yep, that's good for me | 17:52.41 |
| tor5 | see you in a bit then | 17:52.51 |
| robin_watts_mac | kens, chrisl, tor5: When they seat you, they will ask if you've got any coupons or anything. Say you have starwood preferred guest cards back in your room. | 17:56.27 |
| | That gets a 50% discount. | 17:56.32 |
| sebras | http://ghostscript.com/pipermail/fitz-dev/2002-July/000008.html its quite interesting that 10 years on the GTK+ front-end is _still_ in progress... ;) | 18:35.50 |
| | tor5: ^^ | 19:03.22 |
| | tor5: good morning btw. | 19:03.29 |
| tor5 | sebras: morning at least, not sure about good yet :) | 19:07.04 |
| | sebras: oh, ancient history, eh? | 19:07.11 |
| chrisl | sebras obviously has too much time on his hands - we'll have to get him into Ghostscript..... ;-) | 19:09.04 |
| sebras | chrisl: I was looking at the mupdf on wikipedia and noticed that it linked to fitz-dev. | 19:11.04 |
| | and I happened to read about the GTK+ viewer. | 19:11.14 |
| chrisl | sebras: okay, I won't force gs onto you, then - or, at least, not yet. | 19:11.48 |
| tor5 | sebras: that must've been glenn's qt viewer mentioned in the email? | 19:18.29 |
| toothrot | i was looking at using mupdf in place of poppler for a text extraction tool, and have run into a difference in how mupdf treats grouping of text. see these screenshots for comparison: http://imgur.com/a/BW6xM#0 | 19:21.19 |
| | i'm relying on position in some instances, so the places where blocks of text are grouped together across a bunch of whitespace is throwing me off | 19:22.48 |
| sebras | tor5: yes, you are correct. | 19:24.50 |
| toothrot | so, i'm trying to figure out why this is happening on some lines but not others and is it intended? | 19:28.50 |
| tor5 | toothrot: the heuristics involved in assembling text back into lines are complicated (and sometimes fragile) | 19:29.40 |
| | robin_watts_mac may know the details better, but we're all travelling for a staff meeting so you may have more luck if you come back and ask again later during the week | 19:31.04 |
| toothrot | sure, i can check back during the week | 19:31.31 |
| | would the way mupdf is handling those blocks be considered a bug? | 19:31.50 |
| sebras | toothrot: looks like poppler gets some parts " | 19:33.22 |
| | "wrong" as well, see "397.54 FM" towards the end of the page | 19:33.44 |
| toothrot | yes, true | 19:34.14 |
| tor5 | toothrot: yes, I would consider it buggy, so go ahead and open a bug report and attach the PDF file | 19:35.41 |
| sebras | tor5: looks like mupdf would use whitespace distance to determine if two words that are lined up really belong together. but what value to choose? | 19:36.40 |
| tor5 | toothrot: mupdf can give you access to the raw text positioning data as well, so you can reassemble into paragraphs and lines separately yourself | 19:37.05 |
| toothrot | right, i was going to look into that next if you guys said it was tough luck | 19:37.47 |
| tor5 | sebras: there is some code in there to detect rows of a table, I believe that may be what's triggering here to join the lines across the columns | 19:37.48 |
| sebras | tor5: ah. true. damn heuristics. | 19:38.07 |
| tor5 | toothrot: if it's in any way time critical for you, I'd suggest you do that as well; it may take a while until we find a good solution | 19:38.45 |
| toothrot | sure, i understand, it's not critical | 19:39.40 |
| kens | tor5 chrisl checkout is 11 am, so I'm going to go see if we can get into the meeting room | 20:02.28 |
| tor5 | kens: alright, I'll see if I can find you down there in a bit then | 20:04.43 |
| kens | can't get in, door is locked and we are not lsited outside | 20:09.48 |
| | I hve to assume Miles cancelled teh 2nd day, oh well.. | 20:10.07 |
| tor5 | kens: rats. | 20:10.46 |
| kens | Yeah, the on's room Rayly way we can get aircon will be to us | 20:11.12 |
| kens | tries to write that again | 20:11.36 |
| tor5 | kens: your touchpad is not your friend today :) | 20:11.47 |
| kens | The only way we can get aircon will be to use Ray's room | 20:11.56 |
| tor5 | there's usually an option to disable the touchpad while typing | 20:12.01 |
| kens | tor5, no you are quite correct :-( | 20:12.07 |
| tor5 | kens: I reckon housekeeping will be done with Ray's room by now so we can migrate there | 20:12.46 |
| kens | Hmmm, OKI'll come knock on your door | 20:13.31 |
| | after I finish packing | 20:13.38 |
| chrisl | What room number was Ray's? | 20:13.52 |
| kens | 3225 | 20:14.05 |
| chrisl | ta | 20:14.11 |
| kens | test | 20:32.43 |
| | Forward 1 day (to 2013/12/09)>>> | |
IRC Logs