| <<<Back 1 day (to 2011/08/29) | 2011/08/30 |
henrys | kens, alexcher, all:if you see an email address on bugzilla with a potential support customer let me or somebody know so we can ping miles or scott, hcl can afford a support contract, I try to catch them all but sometimes I miss it. | 14:49.53 |
Robin_Watts | miles was talking about tying a meeting to a show next year; anyone remember offhand which show/meeting/dates that was ? | 14:58.04 |
| Was it the On Demand show ? | 15:00.52 |
henrys | yes jun 12 | 15:01.14 |
| new york | 15:01.28 |
Robin_Watts | Right, so we'll have a meeting in June. | 15:01.42 |
| There is a holiday Helen is looking at for next year 22 March to 5th April. Trying to figure out if that's going to conflict with a meeting. | 15:02.34 |
henrys | would the interpolation performance problems be addressed if we looked for some threshold scaling quantity for example say anything greater than 3x would be interpolated? | 15:17.54 |
Robin_Watts | henrys: I don't think so. | 15:18.22 |
| With this bug, we're scaling up lots. | 15:18.50 |
| Hence the heuristic you suggest wouldn't change this case. | 15:19.18 |
| (AIUI, IIRC, IANAL etc) | 15:19.33 |
henrys | right - my usual broken record question what does adobe acrobat do when rendering to a similar device and resolution? | 15:20.03 |
Robin_Watts | henrys: No idea. | 15:20.14 |
| Currently up to my neck in planar getbits/overprinting. | 15:21.04 |
kens | henrys I was planning to send the address to Scott, Alex beat me to it by closing the bug. I already knew the file worked OK with 9.04, but was waiting to see if htey responded. | 15:21.06 |
henrys | Robin_Watts:I'll check it. | 15:21.41 |
kens | Sorry for the delay in replyign, been to the dentist for a checkup | 15:22.01 |
henrys | kens:okay I already sent something but thanks. | 15:22.30 |
kens | Henrys, I replied to the 'shift the origin' support message, I haven't yet replied toe hte 'Embedded CID Font' message. It looks like we failed to reply to the customer. | 15:23.37 |
| I was going to open a bug for it and assign it to Alex for the initial investigation, unless you would rather do it ? | 15:23.58 |
henrys | marcosw_ said he was still working. | 15:24.13 |
kens | OK, I'll leave it for him then. | 15:24.30 |
henrys | I was wondering why you were answering did I miss something. | 15:24.35 |
| ? | 15:24.37 |
kens | No, but I knew the asnwer and wasn't sure about Marcos' status. | 15:24.58 |
| So I thought I would get them a reply. Also I was fed up banging my head against patterns. | 15:25.20 |
henrys | I'll ping him with an email and verify. | 15:25.27 |
kens | OK | 15:25.31 |
| We did already fail to reply to them once, so we should be careful to make sure they get a reply this time. | 15:25.50 |
henrys | okay sent | 15:28.13 |
kens | Great, thanks. | 15:28.19 |
| Hi Michael | 15:35.04 |
Robin_Watts | mvrhel2: Hi. | 15:35.29 |
| Other than some overprinting problems, the plank vs pamcmyk4 test seems pretty OK. | 15:36.09 |
mvrhel2 | good morning | 15:36.12 |
| great | 15:36.16 |
henrys | amazing a flock of about 50 swainson hawks just flew over.. | 15:37.26 |
Robin_Watts | henrys: Twitcher? | 15:38.22 |
henrys | not really these are obvious to anyone: a flock of hawks flying in concentric circles. | 15:39.56 |
kens | Hawks in flocks is pretty unusual, over ehre at least | 15:40.12 |
Robin_Watts | henrys: Coo. | 15:40.16 |
henrys | yea I don't know of any other flocking hawk. | 15:41.08 |
| we are pretty fortunate here bird wise I rarely go a year without seeing bald eagles, large owls ... lots of birds. | 15:42.37 |
kens | Lots bigger then theones we get round here | 15:42.54 |
ray_laptop | kens: I just replied to Grant's message on the CIDFont (sent your explanation). I guess we dropped the ball on that | 15:43.04 |
henrys | kens:you've had more time to kill them. | 15:43.12 |
kens | ray_laptop : yes I was just talking about it with henry | 15:43.17 |
| henrys, but you Yanks are more efficient (passenger pigeon ?) | 15:43.33 |
ray_laptop | kens: I also amplified your response to Tommy -- many PS files use initgraphics (or setpagedevice) so usually the 'Install' proc is better | 15:43.55 |
kens | Besides its the Mediterranean types who shoot anything with wings | 15:43.56 |
| ray_laptop : I did offer further help, but he solved it another way | 15:44.13 |
ray_laptop | kens: I hadn't gotten that message yet. | 15:44.47 |
Robin_Watts | Even the kind of people that shoot birds over here don't shoot raptors. | 15:44.58 |
ray_laptop | kens: (still haven't gotten anything but your response to Tommy) | 15:45.21 |
kens | D'oh, he sent it to me privatelty.... | 15:45.21 |
| I'll forward it later | 15:45.35 |
ray_laptop | kens: np | 15:45.39 |
| kens: I had also mentioned -dEPSCrop in case that helped | 15:46.00 |
kens | He used that in fact. | 15:46.08 |
ray_laptop | except I mistyped that | 15:46.14 |
kens | I guess he figured it out | 15:46.31 |
henrys | let's try to let marcos respond to this stuff he has asked to be the go between. | 15:47.59 |
ray_laptop | I guess I sort of blended -dUseCropBox and -dEPSCrop and came up with -dUseEPSCrop :-/ | 15:48.06 |
| henrys: OK, it's just that for Grant, we (marcos?) had dropped the ball. | 15:48.31 |
| henrys: and do you want marcos to be the go between on simple things like Tommy's issue ? | 15:49.16 |
| I wasn't sure, with Marcos traveling, that he would pick up on Grant's response (like, did he have ken's original reply) | 15:50.11 |
kens | To be honest, I'd forgotten I'd replied ! | 15:50.31 |
henrys | I'd like everything to go through marcos I think issues will get dropped with the everyone responding. but we can discuss that again with him. Obviously I just want good support, whatever gets us that is fine by me. | 15:50.45 |
ray_laptop | kens: quite promptly, too (about 30 minutes) | 15:51.04 |
kens | And then completely forgot about it, oh wel... | 15:51.21 |
| ray_laptop : just got your email to Tommy ;-) | 15:51.44 |
ray_laptop | henrys: and I assume that once one of us is working with a customer interactively (as Chris is with kanazawa-san) we don't have to go through Marcos, right ? | 15:52.40 |
henrys | of course. | 15:53.06 |
kens | ray_laptop : did you see my email to Craig Jones ? teh cidfmap and -I problem ? | 15:53.36 |
ray_laptop | kens: we still have a pending question with Craig. | 15:53.46 |
kens | Another one ? | 15:53.54 |
| I must have missed it, which is that one ? | 15:54.04 |
ray_laptop | kens: no, it's just still open after your response (you need info from him, right?) | 15:54.39 |
kens | Well, I'm hoping that this morning's mail will resolve the problem, but obviously he'll have to try it himself. | 15:55.04 |
| He had used hte escaped PDF filename /#62l... instead of a hex string and cvn | 15:55.34 |
| Obviously the escaped name doesn't work in PostScript ;-) | 15:55.46 |
ray_laptop | kens: OK, right. The hex string stuff | 15:56.25 |
kens | Yep. | 15:56.31 |
| It worked for me with a copy of his setup. | 15:56.43 |
ray_laptop | still, the ball is in their court, so we're waiting to hear back. | 15:57.06 |
kens | Yep. | 15:57.13 |
henrys | ray_laptop:has your customer gone quiet or has the email recipient list been trimmed? | 15:57.18 |
kens | But I'm reasonably confident | 15:57.19 |
ray_laptop | henrys: they've gone quiet (except for Kanazawa-san) | 15:57.41 |
mvrhel2 | meeting time? | 15:57.57 |
henrys | 3 minutes | 15:58.03 |
kens | not quite | 15:58.06 |
mvrhel2 | ah ok | 15:58.10 |
| need to adjust my clock | 15:58.14 |
chrisl | And Kanazawa-san has dropped support from the list - I didn't bother adding it back, as this stuff is so specific to their implementation | 15:58.25 |
ray_laptop | bbiab then | 15:58.26 |
| chrisl: since I was getting it, I assumed it was from support. Let's at least cc support on responses. | 15:59.16 |
henrys | chrisl:It's really good to add support back. You may get fed up with us someday then we'd be lost. | 15:59.33 |
ray_laptop | chrisl: also, Miles has asked to be cc'ed on all communications (I assume that still holds) | 15:59.51 |
henrys | okay meeting time. | 16:00.30 |
chrisl | henrys, ray_laptop: Ray is cc'ed, as is Miles. I'll add support next time | 16:00.32 |
henrys | is marcosw_ here? | 16:00.48 |
kens | think not | 16:00.55 |
henrys | well scratch the first item. | 16:01.08 |
| kens are you still working on the port monitor? | 16:01.23 |
kens | Nope, was told to leave it to Russell | 16:01.35 |
henrys | yea I think we (I) got a little mixed up here. | 16:01.57 |
| certainly up for discussion but I'd like to see an "Artifex" port monitor and gsprint. Thoughts? | 16:02.50 |
kens | can do, won't be soon | 16:03.32 |
| cos I stopped working on it | 16:03.38 |
henrys | okay maybe as a background project(s) | 16:04.07 |
kens | basically back to squaer one | 16:04.08 |
henrys | ? | 16:04.08 |
kens | I'll try, but its not that sort of project really | 16:04.25 |
henrys | didn't you make a git branch ;-)? | 16:04.28 |
kens | no. | 16:04.40 |
| What I have is still here though | 16:04.45 |
henrys | kens has gotten alexcher's bug list down I think ray's is the only other one that probably is a bit long. | 16:05.22 |
marcosw_ | henrys: I was not here earlier, but I am now. | 16:05.26 |
henrys | marcosw_:well we'd like to iron out support handling. | 16:05.51 |
marcosw_ | okay, did I miss something? | 16:06.14 |
henrys | you said you wanted to handle but there have been some late responses. | 16:06.26 |
| I sent you and kens some email about that. | 16:06.43 |
| I think if lots of folks respond we are likely to miss stuff. | 16:07.21 |
| marcosw_:what do you think? | 16:08.48 |
| mvrhel2, Robin_Watts:anything to say about planar for the meeting? | 16:09.50 |
Robin_Watts | From what I can see with the tests I'm doing on my machine here, most of the problems have gone away. | 16:10.27 |
| We get some 1 pixel shifts. | 16:10.38 |
marcosw_ | the messages that I tend to miss are the ones addressed to someone other than me or support. if a message starts "Hi Ken", I skip it. | 16:10.47 |
Robin_Watts | and we get some overprinting problems that I'm looking into now. | 16:10.47 |
mvrhel2 | so we can go ahead and push on with the halftone testing then | 16:10.51 |
| you did say something about overprint? | 16:11.02 |
henrys | kens:I am not sure the project should necessarily go to you, it seems like a good project(s) for alexcher also. | 16:11.02 |
Robin_Watts | but otherwise they broadly look the same. | 16:11.13 |
mvrhel2 | Robin_Watts? | 16:11.18 |
Robin_Watts | mvrhel2: Yes. I think I can see what's wrong. | 16:11.27 |
| I'll run the fix past you after the meeting. | 16:11.36 |
| When I get the fix in, it would be good for marcosw_ to rerun the plank vs pamcmyk4 comparison, so we can be sure. | 16:12.27 |
| then we can test the new halftone stuff. | 16:12.35 |
marcosw_ | Robin_Watts: are we ready to run another plank vs. pamcmyk4 test now? | 16:12.38 |
henrys | Robin_Watts:I do have the memory it should go out today. It would have gone out yesterday if my local fedex office had a collective IQ greater than 50. | 16:12.38 |
Robin_Watts | henrys: The trick is to try and figure out which one of the people working the counter is holding the braincell for today. | 16:13.15 |
| marcosw_: Not just yet. | 16:13.19 |
marcosw_ | yeah, I saw your comment, which came in just before my message. | 16:13.52 |
henrys | marcosw_:yeah it seems like there is no simple answer to the support problem. | 16:15.24 |
marcosw_ | the problem is that sometimes it makes more sense for customers to talk to engineers directly, like when the don't understand some internal bit of Ghostscript or are having trouble calling it and Ray knows exactly what the answer is. | 16:16.28 |
| what happens is that the next time they have a general support they email that person, and cc support. But I don't read the email and assume that it's a follow on to the previous issue and they don't need me to be involved. | 16:17.15 |
henrys | Also we don't do any tracking - we should be able to see all unanswered support questions. | 16:19.27 |
| companies usually do something more structured - every call is tagged with a number or something... | 16:20.06 |
ray_laptop | kens: If you want to play with some of my bugs, fine with me. Starting with bug 691652 (don't know why that one's mine anyway), then maybe bug 690246. | 16:20.12 |
kens | ray_laptop : I will if you want, but I was hoping to so some of mine now ;-) | 16:20.30 |
ray_laptop | oops. sorry for the repeat. | 16:20.32 |
| kens: no, that's OK. | 16:20.42 |
marcosw_ | henrys: I can look into a more formal support email tracking system. | 16:21.06 |
Robin_Watts | henrys: Yes, lots of companies use a ticketing system. | 16:21.06 |
marcosw_ | we can discuss it at the staff meeting next week. | 16:21.16 |
ray_laptop | anybody that wants to can have 691652 (ps to pxl via ijs and hplip) | 16:21.21 |
Robin_Watts | Every email gets a ticket number. People are told to leave the ticket number in the subject line. | 16:21.41 |
henrys | ray_laptop:assign it to me. | 16:21.45 |
ray_laptop | henrys: gladly | 16:22.32 |
henrys | marcosw_:okay if we had a 24 hour email with untended requests we could probably fix the entire issue. | 16:22.36 |
| so the other bug issue is:should we fix customer bugs and just update comments and then assign to marcosw_ for changing the status a presumably contacting the customer? | 16:23.43 |
Robin_Watts | marcosw_: a quick google finds osTicket.com | 16:24.04 |
marcosw_ | henrys: we actually have a policy about that already. | 16:24.10 |
kens | henrys I don't mind that. | 16:24.32 |
| Any way that works | 16:24.39 |
henrys | marcosw_:it didn't look like it really works though. I thought leaving them open would be more obvious. | 16:24.57 |
marcosw_ | the engineers are supposed to set them to resolved and I'm supposed to set them to closed. | 16:25.17 |
kens | Right, I set them to resolved and fixed on invalisd or whatever | 16:25.34 |
ray_laptop | does that mean we need to copy the customer's reply into the bug tracker (so the reply makes sense) ? Long bug tracker comment chains get confusing, IMHO. I'm thinking of issues like the ones from Craig Jones | 16:26.04 |
henrys | well then we probabl need a bug aging email section that says resolved and not closed. | 16:26.05 |
ray_laptop | henrys: that's easy to search for. | 16:26.28 |
kens | ray_laptop : I don't htink that ever got as far as Bugzilla :-) | 16:26.35 |
ray_laptop | kens: right, and neither dd Grant's | 16:27.02 |
henrys | marcosw_:so can you add that to bug aging? | 16:27.19 |
kens | ray_laptop : I guess that's because they weren't really bugs (or at least, not our bugs) | 16:27.35 |
henrys | that's all I got if anyone sees tor8 tell him to release mupdf ;-) | 16:28.00 |
marcosw_ | sure. customer bugs that are resolved but not closed will show up next time it's run. | 16:28.01 |
henrys | anybody have anything for the meeting:alexcher? | 16:28.23 |
ray_laptop | kens: right, but Grant's could have been managed as a bug (file runs in Acrobat, not in GS) | 16:28.25 |
Robin_Watts | Oh, urm, maybe. | 16:29.04 |
| I raised the issue of business cards as a topic for the meeting. | 16:29.20 |
henrys | yes it will be on the agenda. | 16:29.34 |
alexcher | ray_laptop: Please take a look at the bug 692409 | 16:29.47 |
marcosw_ | henrys: I presume no IRC meeting next week? I'll be on an airplane to Chicago :-) | 16:29.47 |
ray_laptop | should be time for Miles to print them to hand out at the meeting. | 16:29.49 |
Robin_Watts | Is it worth mentioning it to Miles in advance so he has time to produce said cards before the meeting. | 16:29.59 |
| Is it just me that's missing cards? | 16:30.10 |
kens | No me too. | 16:30.17 |
chrisl | I didn't get any | 16:30.24 |
henrys | oh he won't get them before the meeting. | 16:30.25 |
kens | Miles asked when I joined but I didn't anticip[ate meeting customers | 16:30.30 |
Robin_Watts | oh, ok. | 16:30.34 |
henrys | everyone will get them meeting after next. | 16:30.44 |
kens | Bad decision as it turns out | 16:30.47 |
marcosw_ | historically Miles has only printed business cards for engineers when they go to Japan. | 16:31.05 |
ray_laptop | henrys: why not? printers here turn them around in 3 days or less | 16:31.09 |
| if we are going to shows, many people still trade cards at shows | 16:31.35 |
henrys | it takes him a week to get everyones names right in japanese, but I'll call him if you want to try. | 16:31.49 |
Robin_Watts | ray_laptop: That was what made me want them. | 16:31.54 |
henrys | I had thought they were "standard issue", they ought to be. | 16:32.31 |
Robin_Watts | Anyone have anything they want brought from the UK for the meeting? | 16:32.53 |
| (unlikely I know) | 16:33.00 |
ray_laptop | alexcher: I agree that the patch in comment 4 should be applied. | 16:33.07 |
| henrys: I don't think we need to 2 sided Japanese cards | 16:33.43 |
marcosw_ | with the constant promotions and title changes Miles doesn't like printing business cards as standard issue. | 16:33.47 |
ray_laptop | ha ha | 16:33.58 |
marcosw_ | (Or am I the only one who get's promotions). | 16:33.59 |
kens | promotion ? What's one of those :-) | 16:34.25 |
chrisl | marcosw_: you're the only one who /needs/ them ;-) | 16:34.29 |
Robin_Watts | kens: That's where they sell you cut price. | 16:34.44 |
marcosw_ | hey, "mail room technician" is a perfectly good title. | 16:34.57 |
ray_laptop | miles is still probably seething about all of henry's old business cards that went to waste | 16:34.58 |
kens | Robin_Watts : been there, last placement | 16:35.13 |
| chrisl as well | 16:35.31 |
henrys | miles:always bring my business cards assuming I'll leave them. | 16:35.38 |
| at home. | 16:35.45 |
Robin_Watts | doesn't want *that* many cards :) | 16:35.46 |
chrisl | kens: that wasn't cut price - it was torture...... | 16:35.58 |
ray_laptop | some people that I give the cards with Japanese on them don't like them. They liike to make notes on the reverse (about what we talked about) | 16:37.37 |
henrys | marcosw_:yes no IRC needed next week. | 16:38.41 |
ray_laptop | OTOH, it's fun giving people a card with the Japanese on top to watch them double-take | 16:38.45 |
marcosw_ | I thought we all just bumped iPhones to trade contact information? | 16:38.49 |
chrisl | Oooh, company iPhones? | 16:39.18 |
Robin_Watts | marcosw_: If Miles wants to give me an iPhone instead of cards, I'll put up with that. :) | 16:39.19 |
ray_laptop | marcosw_: that was before the virus scare. Now you have to make sure your iPhone is running iRubber | 16:39.53 |
marcosw_ | he only buys iPhones for engineers that live in Sweden. | 16:40.18 |
henrys | Robin_Watts:I'm surprised cards are still around. | 16:40.33 |
Robin_Watts | I've seen 'card scanners'. The ultimate executive toy. | 16:41.11 |
| They ought to make 'card shredders', so you can show people what you really think of them. | 16:41.32 |
ray_laptop | Scott has one, but then he collects hundreds of cards from shows | 16:41.33 |
| scanner that is | 16:41.44 |
marcosw_ | I think we just print a bar code on our business cards. | 16:42.01 |
ray_laptop | my shredder works fine on business cards (or even credit cards) | 16:42.04 |
henrys | kens:I did simplify the pcl but something spooky in pdfwrite is going on there. If you eliminate the single character in the file the problem goes away. So somehow the pattern from the character is sticking around instead of the white pattern. | 16:42.33 |
kens | henrys the character is in a pattern. | 16:42.50 |
| The pattern causes the problem | 16:42.55 |
| hence bashing head of patterns all day | 16:43.14 |
henrys | the pattern is pure white. | 16:43.18 |
kens | Its still set as a pattern space though | 16:43.28 |
henrys | I can see lots of bashing potential with that problem. | 16:44.18 |
kens | Indeed, I'm working on it | 16:44.43 |
Robin_Watts | If the meeting is over, I have a question for henrys... | 16:44.52 |
| http://git.ghostscript.com/?p=ghostpdl.git;a=blobdiff;f=gs/src/gsbitops.c;h=7e04f919c1f3e334d73cae7a5f818ebe88b6af62;hp=9b634fb523b2dec34c29fc87e1ce0bc97f349764;hb=01c26a7315612467b74c26a71b8cfa5a67140c8d;hpb=7b1022e20e4ed2b60d48190016f7260be8de8bd0 | 16:44.59 |
henrys | meeting is over at the half hour | 16:45.11 |
Robin_Watts | That's where you introduced bits_fill_rectangle_masked | 16:45.20 |
| You say that src_mask is 1 for those bits of the original that are to be retained. | 16:45.39 |
| In the call I'm seeing here, I see src_mask= 0x88888888 pattern=0x88888888 | 16:46.04 |
| That's wrong, right ? | 16:46.08 |
henrys | hmm I'm pretty sure that's peter's code but I'll look | 16:46.40 |
Robin_Watts | It smells like Peters style. | 16:47.13 |
| gx_overprint_sep_fll_rectangle_1 was introduced in the same commit (in gxoprect.c). | 16:49.26 |
henrys | ah this was changed in the DeviceN project so not peter... | 16:51.33 |
Robin_Watts | mvrhel2, henrys: In the overprint device, we have opdev->retain_mask. If that's really the bits that should be retained, then it's odd to have that set to the same value as the pattern. | 16:52.34 |
| s/pattern/color/ | 16:52.48 |
ray_laptop | is Miles lurking here on IRC -- he just emailed saying he's ordering business cards (wanted to make sure I have enough) | 16:54.40 |
henrys | no I just forwarded your request. | 16:55.15 |
ray_laptop | recalls the days when a business card request required a purchase order request, signed by at least two people, then a business card triplicate form that had to be signed by HR _and_ IT. | 16:57.39 |
| there's still LOTS of organizations that do this kind of nonsense | 16:58.59 |
Robin_Watts | And then 3 weeks later you get cards for Roy Jonson. | 16:59.39 |
kens | Heading off now, goodnight all. | 17:00.37 |
ray_laptop | I kept one each of my CalComp business cards. In 23 years I had 19 titles and 4 pairs that had the same info but they changed to 'look' so reprinted. | 17:03.06 |
Robin_Watts | Who here best understands the overprint device? | 17:09.51 |
henrys | still studying it, mvrhel2 may be able to help also. | 17:11.42 |
mvrhel2 | I have fooled with the overprint device in the past, but I would not say I quite understand how it maintains the information about which inks are overprinted | 17:12.46 |
henrys | believe this is jan's code. | 17:12.54 |
| Robin_Watts:does forcing to sep_fill_rectangle_2 print properly? | 17:19.49 |
Robin_Watts | I'll try that in a mo. | 17:20.10 |
| This overprint code seems very odd. | 17:23.19 |
| gx_set_overprint_cmyk attempts to figure out what colorants are used in the current color, I think. | 17:23.44 |
| To do that it calculates 'drawn_comps' to be the CMYK colorants (and I get 0xF for that - sounds correct) | 17:24.30 |
| Then if the color is set, it tries to find the nonzero components to and with drawn_comps. | 17:24.57 |
| Again, that sounds right to me. | 17:25.02 |
henrys | it is also okay to create a planar overpring bug and move on. The target customer will not use overprint. | 17:25.21 |
| well the immediate target. | 17:25.33 |
Robin_Watts | But to find the nonzero components, it calls decode_color | 17:25.51 |
| which converts the cmyk color into rgb color components. | 17:26.05 |
| Then it searches those rgb color components as if there were 4 of them. | 17:26.17 |
| Maybe there is something screwed with planks decode_color routine. | 17:27.53 |
| Yes, pamcmyk4 has a decode_color that leaves it in cmyk. | 17:29.35 |
henrys | yes it should return cmyk | 17:29.40 |
Robin_Watts | So, OK, panic over, it's my cockup. Sorry to have wasted your time. | 17:29.50 |
henrys | there are some example procs in gdevdlt.c also. I did work on that piece a bit so can say something more intelligent. | 17:30.39 |
mvrhel2 | ah good that you were able to track this down | 17:33.04 |
Robin_Watts | Ok, that seems to solve it for this particular file. | 18:12.58 |
| Let me tidy it up and commit it. | 18:13.15 |
| The plang device is broken on this file (and possibly many others). But that can wait. | 18:17.52 |
| marcosw_: The plank vs pamcmyk4 you ran last night - the 19 differences are probably down to the fixes I made yesterday. | 18:35.51 |
| marcosw_: OK. Hopefully the plank vs pamcmyk4 tests should only have small differences now if you want to retest them. | 19:28.25 |
marcosw_ | Robin_Satts: okay,will retest | 19:39.38 |
Robin_Watts | bales for the night. | 19:41.26 |
mvrhel2 | good night Robin_Watts | 19:47.07 |
alexcher | marcosw_: do you know that tiger.ps does "clippath fill" i.e. has infinite size. | 21:35.00 |
mvrhel2 | bbiab | 22:00.51 |
marcosw_ | alexcher: I did not know that. What is the significance? | 22:51.18 |
Robin_Watts | marcosw_: -w10 is a LARGE window. | 22:57.44 |
marcosw_ | I'm not interested in the output, just playing with how parameters are passed to from clusterpush.pl to clustermaster.pl to build.pl | 22:58.19 |
Robin_Watts | Ah, ok. | 22:58.48 |
marcosw_ | wait!? What are you still doing up? | 23:00.32 |
| I'm used to seeing the UK team on IRC at 1:00AM, but that's when I'm in california. | 23:00.59 |
Robin_Watts | I often check mail/test results etc just before I head to bed. | 23:01.50 |
| I'm going to bed early tonight cos I'm still jetlagged. | 23:02.03 |
marcosw_ | Will we be seeing photos of Vietnam/Cambodia next week? Jill and I are envious. | 23:02.43 |
Robin_Watts | I have all the photos on the laptop, yes. | 23:03.19 |
| But I haven't sorted them yet. | 23:03.30 |
marcosw_ | I'm sure they will be more interesting to look at than the 400+ pictures I took of F1 cars, most of which are blurry or only capture half the car. Those cars are really fast! | 23:04.18 |
Robin_Watts | marcosw_: Oh yes. You need an SLR with a very fast shutter speed. | 23:04.49 |
| so given you can't get close to the car that means a wide aperture zoom, which means a monstrously huge thing. | 23:05.33 |
marcosw_ | I have an 80-200 F2.8, which combined with setting the ISO to 1600 and the fact that it was sunny-ish on Sunday let me take some at 1/8000 sec. | 23:06.43 |
| which is fast enough to freeze the tire rotation, so you can read the Pirelli logo. | 23:07.07 |
Robin_Watts | tor8: In the meeting today, henrys said: "that's all I got if anyone sees tor8 tell him to release mupdf ;-)" | 23:07.33 |
marcosw_ | otoh, they don't look good, since there is no motion apparent in the photo. i.e. the car looks like it's standing still. The better photos is | 23:07.36 |
tor8 | Robin_Watts: I saw. got held up by train chaos. | 23:08.13 |
marcosw_ | when I used a lower shutter speed and panned, that way the background is pleasantly streaked and the tires show motion. OTOH, it turns out I'm really bad at panning :-) | 23:08.17 |
Robin_Watts | That's a fairly big lens, right? | 23:09.52 |
| I've got a 70-300 (which actually gets multiplied by 1.6 cos it's a film lens on a digital canon). | 23:10.29 |
cryptopsy | you use a teleconvertor? | 23:12.12 |
| how crips do photos at 1/8000 compared to 1/1000 look? | 23:12.50 |
| crisp | 23:12.53 |
Robin_Watts | cryptopsy: no. Canon EF-S cameras can take EF lenses. | 23:13.01 |
| I had my zoom from film days (it's an EF lens) | 23:13.27 |
cryptopsy | Robin_Watts: do you use mutt mail client? | 23:13.54 |
Robin_Watts | marcosw_: My 'walking around lens' has a stabiliser in it, so in theory would be better for such panning shots. | 23:14.11 |
| I have crap wobbly hands; my photography is of the "shoot lots of pictures and see what's in focus later". | 23:14.39 |
| school. | 23:14.45 |
| cryptopsy: Nope. | 23:14.51 |
cryptopsy | i have a 36x optical zoom camera, the nikon p500, and the digital stabilisation is really good, even at full zoom | 23:15.18 |
marcosw_ | Robin_Watts: it is a big lens; it has a tripod mount (which I usually use with a monopod when I'm photographing my kids playing sports), but it can be handheld for a while without being to tiring. | 23:15.55 |
Robin_Watts | cryptopsy: I used to be vaguely snobbish about having a proper SLR not a compact cam. And then it dawned on me that my wife takes consistently better pics than me with a point and shoot compact thing. | 23:17.17 |
| I suspect I'm using my camera at some tiny percentage of it's capabilities. | 23:17.45 |
cryptopsy | the p500 is in a new class called 'superzoom', its like the midpoint between a point-and-shoot and a dslr | 23:18.36 |
| got it for 350$ | 23:18.42 |
Robin_Watts | cryptopsy: Yes indeed. Not really a new class anymore. | 23:18.52 |
marcosw_ | cryptopsy: at 1/8000 the tires appear not to be turning, at 1/1000 they look very blurred (maybe a quarter rotation). | 23:18.58 |
Robin_Watts | New class is Micro 4 thirds :) | 23:18.59 |
cryptopsy | they have good lenses for pocket cams now too | 23:19.21 |
| the lense is longer than the camera! | 23:19.29 |
Robin_Watts | Yeah, my wife had a Casio camera for a while that had the lens system from the Pentax Optio. When the camera was zoomed the lens ws longer than the camera was thick when it was powered off. | 23:21.15 |
marcosw_ | While this is an interesting conversation and I could probably talk about camera for hours it's 1:20 AM and I have an appointment at 10:00 that I have to be at least partially coherent for, so I'm going to go back to work so I can get to sleep within the next 2 hours :-) | 23:21.21 |
Robin_Watts | marcosw_: Night. | 23:21.38 |
marcosw_ | night to you as well (and kens and chris, who probably went to sleep 2 hours ago). | 23:22.08 |
cryptopsy | marcosw_: damn that's cool, i think i'm going to exchange it for the hs20exr whic has 12k iso | 23:24.39 |
| it has a thread in the front too so i can mount it to telescopes and telephotos | 23:26.23 |
henrys | marcosw_:alexcher comment is about a support message that came in today. | 23:27.57 |
marcosw_ | I just figured that out, with TIGER.PS. BTW, is Ray still the go to guy for bbox.dev? who_owns_what.txt seems to be a bit out of date... | 23:28.38 |
Robin_Watts | marcosw_: Just saw your mail about d0f9685b - are you retesting now with 2b898bd ? | 23:28.56 |
henrys | marcosw_:I was thinking about getting rid of who_owns_what.txt but if you use it I'll bring it up to date. | 23:29.35 |
marcosw_ | Robin_Watts: yes, I'm running a test with 2b898bdf | 23:29.52 |
| henrys: I do use it for more obscure issues, such as bbox.dev. The bit at the bottom is useful as well. | 23:30.40 |
henrys | okay I'll update it. | 23:30.52 |
| anyway that is the correct output for the bbox device for tiger | 23:31.08 |
marcosw_ | henrys: it is? seems a bit large to me. | 23:31.28 |
| gs8.54 produces 17 177 565 753 | 23:31.54 |
henrys | that's wrong tiger is supposed to scale with the output, I guess you could use the default page size but that is not how it is designed. | 23:33.12 |
| it being the bbox | 23:33.47 |
| but I guess we should investigate if 8.54 produced a different result | 23:34.07 |
| are you using the tiger distributed with 8.54 and has tiger changed? | 23:34.42 |
marcosw_ | Distiller says it's 7.57 x 7.88 inches, which I realize isn't the bounding box, but is clearly isn't the default page size either. | 23:34.57 |
henrys | I know it was changed to eps recently | 23:35.06 |
marcosw_ | I'm using the TIGER.PS the customer sent. | 23:35.09 |
henrys | I guess open a bug and assign it to ray | 23:35.42 |
marcosw_ | will do. | 23:35.56 |
henrys | I thought the bbox imposed a really large bbox then intersected that with what you sent. Since the job does a clippath the intersection is equal to the arbitrarily large bbox size. | 23:36.51 |
cryptopsy | i have the choice of switching from a 1500ISO camera with great image stabilisation t oa 12000ISO camera with poor image stabilization, slow menus, much larger | 23:39.40 |
henrys | marcosw_:FWIW I think the current code is more useful, it is supposed to show the maximum box for the job - for tiger that is infinite. | 23:41.10 |
marcosw_ | henrys: I've opened the bug, please feel free to add a comment if you feel strongly about it. I explained to the customer in the email that this may not be a bug. | 23:46.22 |
| Forward 1 day (to 2011/08/31)>>> | |