Ghostscript IRC logs

Log of #ghostscript at irc.freenode.net.

	<<<Back 1 day (to 2014/03/24)	2014/03/25
Robin_Watts	OK, new sot.git online.	00:19.36
spanners	Jogux: Bug fixes to the fonts in resources should be self-contained, I think.	00:21.23
kens	paulgardiner (for the logs)< canyou look at this please ?	08:13.07
	http://stackoverflow.com/questions/22614720/adding-views-to-mupdfreaderview	08:13.07
	Hmm, the Virgin flight says '3 seats left' presumably at the price which is 461.75	08:18.37
	Oh th	08:19.24
	that's one way...	08:19.30
	Wow, total cost is 911.45	08:19.46
chrisl	Maybe fly out on the Friday?	08:23.19
kens	Not a lot different	08:23.29
chrisl	More seats!	08:23.39
kens	I'm guessing we're hitting the holiday season	08:23.44
	More seats is good foc ourse	08:23.58
	Have to get Niles to spring an extra night at his place	08:24.13
	Of course, going out on a Saturday we hit all the holidaymakers	08:24.55
chrisl	Yeh, I was thinking the same thing	08:25.09
kens	If that '3 seats left' is true we can't all fit on that flight anyway, there are 4 of us	08:25.34
chrisl	Well, one or more could book PE	08:25.53
*kens*	would not fancy the last remaining seat assignments either	08:25.54
	PE is 1800 flyign out O.O	08:26.34
chrisl	Yeh, so maybe an extra night in a motel is the better option for Miles!	08:26.57
kens	'only' 730 flying on Friday	08:26.58
	We could try a different airline (reluctantly)	08:27.29
	I presume it would be helpful for tor to meet up with us. I am hoping Robin doesn't mind driving again....	08:28.52
	BA is slightly more expensive	08:29.15
chrisl	IIRC, neither Robin nor Tor will want to fly BA	08:29.41
kens	OK well lets see what elsse we cna find	08:29.59
chrisl	Also, IIRC, Tor has already visited Miles's place, so he might want to skip that, keep the trip as short as possible	08:31.04
kens	I think Tor's last visit was the same time as me, before the cabin was there, he'd expressed an interest in another vist, but we cna check with him of course	08:31.39
	United are the cheapest	08:36.06
	Looks like choices are United, Lufthansa, BA or Virgin. I don't fancy FinnAir or Iberia and American starts over £1000	08:37.42
	A lot of the airlines are showing the Saturday fligths as extremely full.	08:38.36
chrisl	That's not really surprising	08:38.51
kens	Most are shoing 2/3/4 seats left	08:39.01
	The early morning United flight has a reasonable number of seats	08:39.49
	THe later (and more expensive) flight has fewer seats	08:41.00
chrisl	I'd rather not suffer a packed flight if I can avoid it	08:42.03
kens	THe BA flight looks ok going by the seat map	08:42.23
	But most of the aisle seats are marked' disability seat', not sure what that means exactly	08:43.04
	The Virgin flight won't give me a seat map	08:44.02
	I guess I'll wait until Robin/Paul/Tor are online	08:44.58
*kens*	disappears off to talk to tree surgeons, back in a minute	08:45.35
pedro_	hi folks	09:27.50
kens	Morning	09:27.57
paulgardiner	hi	09:28.07
ghostbot	bonjour, paulgardiner	09:28.07
chrisl	Howdie	09:28.17
*pedro_*	forgot just how much fun the ooxml/epage debugging was	09:33.40
chrisl	pedro_: do you also enjoy whipping yourself with willow branches and other forms of self flagellation??	09:35.22
pedro_	chrisl: its muh the same ;)	09:35.38
	much	09:35.50
*pedro_*	really shoudl go back to a wired keyboard...	09:36.01
chrisl	As opposed to a "weird keyboard"?	09:36.23
pedro_	:)	09:36.44
chrisl	I'm wondering if it's the wireless nature of my mouse that's causing my problems with that.....	09:37.31
kens	paulgardiner : did you see Miles' email ?	09:38.26
	chrisl the Lufthansa flight turns out to be the United flight in disguise. So the airlines reduce to United, BA or Virgin, and on the whole I'd really rather not fly with United again	09:39.04
paulgardiner	kens: what's the problem with United?	09:40.41
kens	THe in flight 'entertainment'	09:40.51
	and they soak you for extra leg room seats	09:41.16
paulgardiner	Coming back, I had the best screen of any flight I've been on... ah, but you were an exit seat. Was that the problem?	09:41.42
kens	No, its the content....	09:41.51
	And the flight out was dreadful	09:42.04
*kens*	accepts the flight back had nice screens	09:42.16
	The flickering electrics didn't fill me with confidence either	09:42.43
paulgardiner	Out seemed below average and back above to me	09:43.30
	Oh yeah, there was no content on flight out	09:43.49
kens	I didn't liek the selection on the way back much, and the outbound flight was execrable	09:43.51
paulgardiner	I remember now	09:43.54
	I thought there was loads of choice coming back. Took ages to browse through it.	09:44.54
kens	There was lots of choice, but I wasn't keen on any of it very much, and it had that terrible touch screen 'flip'# interface, where nothing happens until you gesture firmly, and then 20 films fly past	09:45.40
	But its the fligh out I really didn't like	09:46.23
chrisl	As we'll probably be relying on Robin to drive, there's probably not much point discussing until he appears..... I don't fancy a 5+ hour journey on public transport instead!	09:47.04
kens	chrisl yes, just wanted to make sure Pual had seen the mail :-)	09:47.33
	And note that the LH flight is actually a UA flight	09:47.46
paulgardiner	We should pick the least popular flights to maximie chance of a spare adjoining seat	09:54.27
	Whatever maximie means!	09:54.51
kens	I can't tell how popular the Virgin flight is	09:54.57
	Won't give me a seat map without a booking	09:55.09
Robin_Watts	chrisl: eh?	10:15.12
chrisl	Robin_Watts: Um?	10:15.39
kens	Robin_Watts : you saw Miles' email ?	10:16.23
Robin_Watts	just seen it.	10:17.27
	5 hour drive? crumbs.	10:17.31
	oh, on public transport it would be 5+ hours, gotcha.	10:18.01
kens	I don't remember it being 5 hours	10:18.02
chrisl	Robin_Watts: yeh, bus + train apparently takes ~5 hours	10:18.25
kens	It is a fair way	10:18.37
chrisl	Of course, legally speaking, I could hire a car and drive, but I don't fancy my first experience of driving on the colonial side of the road being immediately after an 11 hour flight	10:20.02
Robin_Watts	chrisl: I'll drive it, no problem.	10:20.41
	We'll put at least one other person on the car rental just in case I get struck by a migrane or something.	10:20.59
	United flights are MUCH cheaper.	10:23.11
kens	well, not that much	10:23.22
	150 overall more or less	10:23.29
Robin_Watts	has anyone seatguru'd to see what they are like?	10:24.26
kens	didn't get a defintieive answer, 3 possible configurations	10:24.45
	I believe the united flight is cheaper because its early, and therefore less popular	10:26.05
	The later flight is not so much cheaper	10:26.11
*kens*	suspects United may also be less popular generally.....	10:26.28
Robin_Watts	United has power everywhere.	10:27.02
kens	Are you certain ?	10:27.15
Robin_Watts	That trumps decent seat back video IMHO as I can bring my own and watch on the laptop.	10:27.20
kens	THe flight I looked up did not	10:27.21
Robin_Watts	On United 900, Sat 31 May: http://www.seatguru.com/airlines/United_Airlines/United_Airlines_Boeing_777-200_2_New.php	10:27.52
kens	Well, assuming that's the plane they use	10:28.35
	config 3 has no power	10:28.59
Robin_Watts	kens: That's the only config option searguru gave me.	10:29.24
kens	Robin_Watts : in the overview at the left I see 3 versions for the 77-200	10:29.44
Robin_Watts	and likewise on the way back same layout.	10:29.49
	Yes, but when you search on a flight it only gives you the one layout.	10:30.10
kens	Yes, but that assumes the info is correct, and they don't switch planes	10:30.25
Robin_Watts	It does assume that.	10:30.32
kens	Hmm expedia not showing the early United departure now	10:30.57
chrisl	It's probably not a normal AC outlet, is it?	10:30.58
Robin_Watts	Config 2 is a 3x3x3 layout.	10:31.15
kens	chrisl it was a normal US one before	10:31.17
Robin_Watts	It is a US outlet, supposedly.	10:31.28
kens	Oh, there's the early united flight now	10:31.35
Robin_Watts	Though I would take adapters in case :)	10:31.36
	config 3 is 2x5x2 with no power.	10:31.57
kens	Yeah that's what I mean	10:32.08
Robin_Watts	COnfig 1 has power too.	10:32.14
kens	Yes, only v3 hasn't	10:32.21
*pedro_*	thought the last couple of united international flights I had were combi us/euro/uk sockets	10:32.27
	domestic were just US	10:32.39
Robin_Watts	but I suspect that they'll be very reluctant to change from v2 to v3 as that would bugger seat assignments up.	10:32.44
kens	Robin_Watts : assuming it is that config	10:33.00
Robin_Watts	kens: Indeed.	10:33.08
kens	The onlky way to be certain is to try a booking I htink	10:33.12
	Give me a minute	10:33.34
	19:35 return trip on Tuesday	10:34.19
	Huh, won't give me a seat map without a purchase	10:34.45
	expedia does show it as 3,3,3	10:35.08
Robin_Watts	Well, Virgin will sell me 5 adult seats on those dates.	10:35.19
	Personally, I'd be tempted to go for the united one, but I'll go with the flow.	10:35.39
kens	I'd prefer Virgin, it also departs later, 10:25 is possible, but kind of early	10:35.59
	OTOH Virgin gets in at 4pm which is sort of late	10:36.22
Robin_Watts	Frankly, I'd rather go early in this case as it means we get to drive to guernville before the rush hour hits.	10:36.29
kens	Its a couple of hours by car as I recall, so early seems better I guess :-(	10:36.55
Robin_Watts	We either need to drive through SF itself, or we need to drive around the bay etc. Which might be nice, but will be longer.	10:37.15
kens	Hmm, I htought we would just go due North over the bridge, but I didn't look at a map last time	10:37.47
Robin_Watts	kens: Yes, if we want to drive through SF, we go straight north.	10:38.05
kens	(Miles was driving and he knew the way)	10:38.09
Robin_Watts	but if the traffic is bad in the city (like at rush hour) and we want to avoid that, it's a longer, more circuitour route.	10:38.36
chrisl	As I said, we could fly out on the Friday - the flights should be quieter	10:38.44
kens	I'd think if we get in early we could be clear of SF before 5pm	10:38.56
Robin_Watts	kens: yes.	10:39.09
kens	chrisl that's OK too, but needs a night in a hotel extra	10:39.10
	I guess we'd have to talk to Miles. WOUld also be good to know what Tor wants to do	10:39.32
	An early flight might be hard for him	10:39.41
chrisl	I'd thought if we went on Friday, we'd stay near the airport Friday night, and head up Saturday morning.	10:40.22
kens	Seems fair	10:40.34
chrisl	But it entirely depends on other people's plans, and the costs	10:40.57
Robin_Watts	Personally, I'd ignore the "only 3 seats left" stuff, as it's almost always a sales pitch to make you bok now.	10:41.18
kens	I'm fairly sure that's true	10:43.38
Robin_Watts	We need to speak to tor8 really.	10:44.50
kens	WSell, I'd prefer Virgin because I like the entertainment better, and they won't rip me off so much ofr extra legroom, but flying in early would trump that, if Tor is OK with that.	10:44.56
Robin_Watts	I was just going to say that I think tor8 has refused to ever fly virgin again :)	10:45.16
chrisl	what did virgin do to tor, then?	10:45.40
Robin_Watts	kens: I've had too many experiences with virgin where the movies don't work at all.	10:45.44
	"Sorry, we'll need to reboot the system. Again."	10:45.56
	chrisl: He missed a connecting flight, and virgin couldn't get him out on any other airline. Just told him to come back tomorrow.	10:46.42
chrisl	That's pretty crap :;-(	10:47.53
	Robin_Watts: btw, you can pull in the new LCMS to master whenever you are ready	10:52.20
Robin_Watts	chrisl: Ta.	10:52.54
chrisl	I'll be looking at updating the other third party libs next week, too	10:53.22
	kens: Bug 695108: you'll note he says the fix was in Sept, but he's trying 9.10, which was released in 2nd Sept....... and only had commits up to about August 25th	10:59.28
kens	chrisl if he actually uploads the PDF I'll take a look	10:59.44
chrisl	I just thought it was worth pointing out he's not testing what he thinks he ios	11:00.28
kens	:-)	11:00.35
Robin_Watts	Jogux: Is the NAS missing on points for you?	11:01.14
Guest10482	Hi, I use pcl6.exe to convert pcl to pdf , I have an issue on a new server, I think a library is missing but I don't know which...	11:04.36
	here is the error :+ ..\pcl\pcfsel.c:443: pcl_reselect_font(): Fatal \| ..\pcl\pctext.c:946: pcl_text(): Fatal Warning interpreter exited with error code -100 Flushing to end of job	11:04.49
	if anyone have an idea :)	11:05.06
kens	Yes, report a bug, this is not a missing library	11:05.21
Guest10482	same version with same pcl file on another server work fine ?	11:06.07
kens	Then its probably a font difference, or a configuration problem.	11:06.28
chrisl	Guest10482: are you build the exe yourself?	11:06.58
Guest10482	no I download binary I try v9.10 and v9.06	11:07.30
kens	What OS ?	11:07.46
tor8	Robin_Watts: holy crap, tuesday is a bad day to fly!	11:07.51
Robin_Watts	tor8: Morning.	11:08.07
	We're trying to decide between virgin or united.	11:08.16
tor8	mon, wed, thu, fri are 400gbp, tuesday, 2500.	11:08.25
Guest10482	Windows 2012 R2	11:08.27
Robin_Watts	united gets in earlier and has power at every seat.	11:08.37
kens	tor8 why so much ?	11:08.52
tor8	kens: probably sold out :(	11:09.08
kens	oh	11:09.14
tor8	all other days are more reasonable. or it may be scheduling, there are no tuesday departures from CPH to SFO.	11:09.39
kens	tor8 was assuming you would meet us in Heathrow again,to share drive	11:09.57
pedro_	Robin_Watts: can't even list /mnt from here	11:10.23
tor8	kens: yeah, it looks like that may be a better option if we're having the meeting on mon/tue given the lack of (non-extortionist) direct flights on tuesday	11:10.36
Robin_Watts	tor8: Plan is to fly out on saturday, drive up to guernsville, check in there.	11:10.50
tor8	I'd rather connect than stay yet another night to fly back on wednesday	11:11.05
	and sharing drive would be convenient	11:11.15
kens	OK tor8 the problem then is that the saturday flight with united is early	11:11.22
	departs at 10:25	11:11.30
Robin_Watts	Who are direct flights from CPH to SFO with ?	11:11.44
tor8	Robin_Watts: SAS	11:11.50
Robin_Watts	Are they in a group with united?	11:12.08
tor8	one of the few direct flights to the US out of CPH, and it's a fairly new line	11:12.15
	they're star alliance	11:12.32
	not sure what united is	11:12.39
Guest10482	what font is use for a default document? I try to convert a simple text file with only "Test" and it is not working	11:12.46
Robin_Watts	tor8: United and SAS are both Star Alliance.	11:12.47
kens	onestar I think	11:12.48
Robin_Watts	So you could fly out to SFO direct, then fly back via london.	11:13.05
*kens*	wonders what time tor's SAS flight get sin	11:13.22
tor8	Robin_Watts: yeah, I might be able to do that	11:13.30
	or fly back with BA or lufthansa	11:13.44
Robin_Watts	What time is that flight?	11:13.53
	If it's within an hour or so of us, that'd be fine.	11:14.18
kens	united gets in at 13;30	11:15.05
tor8	Robin_Watts: SK935 arrives at 14:45	11:16.06
kens	close enough I'd say	11:16.21
Robin_Watts	assuming it's not late, yeah.	11:16.29
	cos we'd clear customs etc, then go get car.	11:16.39
kens	well we can never control late flights, if tor not on our flight its always a risk	11:16.58
	Hopefully we'd still be on the road by 4	11:17.30
Robin_Watts	And if he's late, we can get a milkshake from prestons!	11:17.44
kens	ROFL	11:17.50
tor8	kens: yeah, united would be cutting it very tight if it leaves at 10:25	11:18.52
kens	That was my concern	11:19.02
	but if you go direct, np	11:19.14
tor8	yeah, problem is finding a reasonable return combination, which looks like it could be tricky	11:19.36
	expedia errors out :(	11:19.40
	I thought I'd found one, but it leaves SFO at 7 in the morning on tuesday. doubt that'll be popular :)	11:22.28
kens	No, I guess not ;-)	11:22.49
tor8	I can find one set of connecting flights on virgin (VS19 and VS20) but that's priced ridicously	11:23.44
	Robin_Watts: I take it you all still refuse flying BA?	11:25.54
kens	really ? THe virgin combo for me was £900 LHR-SFO-LHR	11:26.01
tor8	1700gbp for me	11:26.19
Robin_Watts	tor8: Have no strong objection to BA.	11:26.26
	Have no luggage for them to lose :)	11:26.40
kens	Hmm I saw 1700 if I didn't fly saturday	11:26.43
tor8	because I can always find cheap decent BA flights that connect CPH-LHR-SFO	11:26.55
Robin_Watts	tor8: You don't need to tie up with us on the way back.	11:27.31
tor8	my problem is finding a return trip that has the direct flight on the way out...	11:28.20
	if I can't do that, I might as well come in with your flights	11:28.42
kens	Hmm, yes, awkward	11:28.45
tor8	oh, I might have found one!	11:33.03
	it leaves at 13.19 on the tuesday though, which could end up a tad early I fear	11:33.39
kens	Yes, maybe....	11:34.05
tor8	but would be okay for 9/10 meetings	11:34.13
	kens: ah, now I see why the virgin was so expensive. it only has premium economy seats left.	11:36.38
kens	wow!	11:36.45
	I guess hte '3 seats left' was real then	11:37.01
tor8	so, the absolute earliest I can arrive at heathrow on saturday is 8:30	11:39.36
	and that's at T5	11:39.47
kens	Hmm, I htink United is T1 ?	11:39.56
Robin_Watts	tor8: Come a day early and stay with ken?	11:40.06
kens	that's fine of ocurse too	11:40.15
Robin_Watts	kens: There is no T1 at the moment. It's a big pile of rubble, isn't it ?	11:40.23
kens	Umm, not sure hwta the status is, good point though	11:40.35
chrisl	No T2 is the big pile of rubble	11:41.20
*kens*	thinks whole airport could be described that way	11:41.51
tor8	Robin_Watts: kens: that would work too, if it makes it easier to find a decent flight for all (given that virgin has no more seats)	11:42.06
Robin_Watts	It would appear to be T1.	11:42.11
kens	tor8 we'd be glad to have you stay the night	11:42.19
tor8	or I can take the direct flight that arrives an hour after you and leave the meeting before lunch on tuesday	11:42.41
Robin_Watts	sorry to volunteer ken like that. Tor would be welcome here, but it's a long way to come :)	11:42.44
kens	Its a smart idea robin, glad you mentioned it	11:43.02
Robin_Watts	I suspect we should not assume that tuesdays meeting will be short.	11:43.12
tor8	Robin_Watts: that it is! maybe I should come two days earlier then ;)	11:43.19
Robin_Watts	It probably will be, but if we assume it then it'll run for hours.	11:43.39
kens	You're welcome any time tor	11:43.45
	hop[efully opur plumber should be round before then to fix the guest shower too :-)	11:43.59
tor8	kens: Thanks.	11:44.07
	I'm all out of luck finding return flights other than at 7 or 13 with SAS...	11:44.40
	and I'd rather stay an extra day with ken than at the hyatt :)	11:44.59
*kens*	consider that a coimpliment :-)	11:45.19
	Just let me know	11:45.29
Robin_Watts	https://www.google.co.uk/maps/search/Guerneville,+CA,+USA+Hotels/@38.5116964,-122.9856879,15z/data=!3m1!4b1	11:45.34
*Robin_Watts*	wonders where we're staying in Guernville.	11:45.54
kens	Miles said over the road	11:46.56
	Not sure I recall the location well enough	11:47.06
	maybe creekside inn ?	11:47.28
tor8	kens: don't ask me why, but if I go via united.com it routes me on the 10.25 flight but wants 3000gbp for economy fare!	11:58.49
kens	O.O	11:58.56
	Outrageous!	11:59.04
Robin_Watts	tor8: eh!?	11:59.05
kens	came in at 780 odd here	11:59.19
tor8	yeah. ridiculous.	11:59.39
Robin_Watts	I was about to say "should we all book flights then?", but...	11:59.50
kens	Hey robin I didn't notice 'Highlands Resort' in Guerneville before :-)	12:00.00
Robin_Watts	kens: mmm.	12:00.12
	Those people who feel that clothing is optional are almost by definition those people who should not.	12:00.36
	brb.	12:01.01
Guest10482	does anyone know which is the default font use by ghostpdl when no font is specfied?	12:01.30
kens	THere's also the 'New Dynamic Inn' we should fit right in there	12:01.34
	Guest10482 : Helvetica usually	12:01.43
*kens*	wonders if 'homemade muffins' at the Highlands resort is a euphemism	12:02.17
Guest10482	thanks kens	12:02.23
kens	OMG Google maps has a clippy :-O	12:03.06
tor8	Robin_Watts: and that was just one way, so double that fare for the total...	12:03.44
kens	Hmm tor8 must be first class surely	12:03.57
tor8	flight UA930, Fare Class: United Economy (Y).	12:04.50
kens	Good grief....	12:04.57
Guest10482	mmm I don't have Helvetica on my computer( not on the server where conversion failed), if not exist another font is used?	12:05.09
kens	Wait UA930,l I htought we were talking UA900	12:05.10
tor8	kens: on the return trip	12:05.18
kens	Guest10482 : severl problems, firstly what OS are you using	12:05.31
tor8	UA900 on the way out	12:05.33
kens	tor8 ah right	12:05.38
Guest10482	windows 2012R2	12:06.05
tor8	so yeah, can't do united if I book it, might work out if we do what we did for maui and you make the reservation and I come in to LHR on a separate booking	12:06.16
kens	Hmm this is PCL not GS.....	12:06.28
Guest10482	yes	12:06.38
kens	chrisl do we build the fonts into the ROMFS for GhostPCL ?	12:06.44
	tor8 I can do that sure	12:06.55
tor8	kens: and checking lufthansa for other flights that arrive at the same time, most of the tuesday flights are sold out	12:09.46
kens	OK....	12:10.00
tor8	the cheapest economy options are all gone	12:10.41
kens	Oh wow, let me check here	12:10.52
	LOL now I'm getting prices in $US :-)	12:11.33
tor8	if I take the trip via frankfunt, I can be at SFO at 12:45 which should be an hour before you lot	12:11.38
	otherwise I'd suggest BA if you can all stomach it	12:11.56
kens	THat's fine for us of course	12:12.00
	BA is OK	12:12.04
tor8	that way we can all get on the same flights and I don't have to do scheduling contortions	12:12.22
*kens*	thinks robin doesn't do BA ?	12:12.43
chrisl	kens: yes we do build the fonts in	12:13.02
kens	OK so in that case Guest10482 you can't fail to find Helvetica, its built into the executable	12:13.25
chrisl	Erm, I don't know what the PCL default font is.....	12:14.04
kens	chrisl whatever it is, I would think its one fo the base 35, and therefore included	12:14.23
chrisl	PCL has 80+ fonts, but yes, the default ought to be there	12:15.08
tor8	BA has two flights a day both ways, 14.35 and 17.10 out and 16.40 or 19.20 back	12:16.54
Guest10482	ok so it is not a font issue, if there are builtin...	12:16.57
kens	Guest10482 : It may be referencing a font on disk which is corrup on one system (for example)	12:17.18
Robin_Watts	kens: I can do BA.	12:17.43
kens	tor8 problem with that is that's a late arrival, we'd be in the SF rush hour	12:17.44
chrisl	I don't think PCL can reference fonts on disk.....	12:17.52
kens	OK then scratch that	12:17.59
Robin_Watts	What's the BA flight?	12:18.10
kens	looking now	12:18.15
tor8	the 17.10 flight arrives at 18.30 so that'd be a bit late	12:18.23
Guest10482	I try to convert a simple text file with only "test" inside so I think it use the default font, and conversion failed	12:18.26
chrisl	Except for the default font set if you don't include them in the exe (which we do)	12:18.32
kens	Firefox grumbles about the BA site....	12:18.57
chrisl	Guest10482: what's the Windows version on the system(s) that work?	12:19.10
tor8	Robin_Watts: BA284 through BA287	12:19.22
Robin_Watts	BA0285 ?	12:19.31
tor8	285 is the early flight out, yes	12:19.46
kens	BA0285 looks good, 11:30 departure	12:19.50
Robin_Watts	11:30 -> 14:35 ?	12:19.51
kens	yeah that one	12:19.54
Robin_Watts	That would be best I reckon.	12:20.00
kens	14:35 arrival, so shuld be OK for onward travel	12:20.21
Robin_Watts	0286 back?	12:20.25
kens	I would say so yes	12:20.35
	16:40 is a better time than usual	12:20.41
tor8	284 is earlier, but 286 is my usual flight :)	12:20.46
kens	as long as we don't go into the afternoon on day 2	12:20.49
Guest10482	I try on my compter windows 7 and on another server 2012 R2 , it is ok	12:20.57
Robin_Watts	kens: no, you're looking at 284	12:21.02
kens	AM I ?	12:21.06
Guest10482	I download pcl6.exe from here : http://downloads.ghostscript.com/public/binaries/ghostpcl-9.10-win32.zip	12:21.08
kens	Oh yes, eyes slipped	12:21.14
Robin_Watts	286 is 19:20	12:21.16
kens	Well I guess 19:20 maybe better given the likely meeting discussions	12:21.33
Robin_Watts	ok, 19:20 then.	12:21.42
kens	Tor8 how do these work for you ?	12:21.55
tor8	BA285 out, BA286 back?	12:22.24
kens	Yes exactly	12:22.34
tor8	that works great for me	12:22.40
chrisl	Guest10482: TBH, unless you can identify the difference(s) between the two systems that trigger the problem, it's hard for us to help :-(	12:22.52
kens	SHould we check with Paul ? :-)	12:22.57
Robin_Watts	What was wrong with the united flights again?	12:23.04
*kens*	has lost track now	12:23.15
	Its an hour earlier, tricky connection for tor maybe ?	12:23.42
tor8	Robin_Watts: 6400gbp fare for me	12:24.08
kens	oh yes, cost.....	12:24.17
tor8	or having to make separate bookings and a tricky connection	12:24.24
	Robin_Watts: http://techcrunch.com/2014/03/24/gbatteries-batterybox/	12:25.45
Robin_Watts	I'm getting 926 quid on BA.	12:25.56
Guest10482	chrisl yes I understand, the two windows 2012 are VM and come from the same base, so I don't understand... maybe an update...	12:25.58
Robin_Watts	for a flight with no power.	12:26.00
kens	Robin I saw slightly more cost, but yes	12:26.15
Guest10482	is pcl6 use ms vc++ ?	12:26.16
Robin_Watts	as opposed to 748 on united for a flight with nicer seating and power at every seat.	12:26.28
kens	Guest10482 : WHy would that matter ? Its s compiled application	12:26.39
tor8	Robin_Watts: wow. I'm only getting 822gbp for the BA flights (and that is with connections)	12:26.49
kens	: I'm seein £996 for BA	12:27.10
	adding the two components	12:27.17
Robin_Watts	I am tempted to fly united with paul. We'll land at 13:30, can get car etc and then meet up with tor.	12:27.26
kens	558 + 438	12:27.27
*kens*	doesn't see BA as better than United	12:27.49
paulgardiner	That suits me	12:27.50
kens	Except from Tor's POV of course	12:28.03
Robin_Watts	Tor will still be flying on the flights that are convenient for him, right?	12:28.33
tor8	if the united seats are anything like delta or AA, I'd rather not... those seats were really uncomfortable.	12:28.35
kens	Huh p[rice went down when I selcted an outbound flight :-)	12:29.01
tor8	I had a stiff neck for days after last meeting :(	12:29.03
Robin_Watts	tor8: You come on the BA flight. If It's an hour different, right? That's about right for us getting the car.	12:29.17
kens	I get £926	12:29.20
Robin_Watts	926 is what I got.	12:29.27
tor8	Robin_Watts: yeah. BA or LH look to be my best options, and they arrive +/- one hour of the United flight	12:29.46
Guest10482	yes but maybe it use external dll? some application need vcredist	12:29.55
kens	For me LH was the United flight, maybe they have another	12:30.04
tor8	a real shame about tuesday being full, or I could get a direct flight!	12:30.06
Robin_Watts	tor8: Then, it sounds like I won't be inconveniencing anyone by opting for united.	12:30.10
kens	Guest10482 : No	12:30.13
chrisl	Guest10482: then the exe wouldn't start at all - we only used the core C lib, anyway	12:30.26
*kens*	is not a fan of BA any more than United	12:30.51
Robin_Watts	ok, I'm booking united.	12:31.19
tor8	Robin_Watts: then there's the fact that it's still impossible to do online checkins with connecting flights on any US based airline...	12:31.24
Guest10482	ok, sorry for the questions, I just try to understand what can be change on the 2 servers :)	12:31.24
kens	OTOH at least the inflight entertainment won't be US TV	12:31.28
chrisl	Ugh? United is suddenly quoting me in USD..... stoopid site!	12:31.53
Robin_Watts	tor8: Fly whatever is easiest for you. The timing should be fine with the car hire delay, and if it's not, I'll go get a milkshake.	12:32.01
kens	Yeah dfid that to me too	12:32.04
chrisl	But it quoted UKP when I looked a couple of hours ago	12:32.25
*kens*	thinks BA doesn't offer extra legroom, looking like United is winning	12:32.28
	chrisl, yes me too....	12:32.36
	I guess it htinks we're in the US now	12:32.43
chrisl	Let's hope the pilots know which is which!	12:33.06
*kens*	wonders if Malaysia airways hastha problem....	12:33.48
chrisl	Guest10482: the Ghostscript/GhostPDL builds on Windows are very self contained - straight C (not C++) and link to a very minimal set of libs.	12:34.58
kens	OK I' going to go and get some lunch, maybe I'll feel more enthusiastic about flying when I've eaten	12:36.37
chrisl	I wonder how you get United to quote a price of Economy Plus.....	12:37.38
Robin_Watts	$105 for economy plus.	12:37.54
chrisl	Do you add that after you book?	12:38.20
Robin_Watts	When you select seats.	12:39.12
chrisl	Ah, I could be tempted....	12:39.30
tor8	kens: heh, the LH flight is cheaper, but an hour longer, but hey, A380! I've never flown on one of those before :)	12:44.51
	but yeah, some food and this might not be quite as depressing anymore...	12:45.26
Guest10482	ok thanks you chrisl and kent, I will continue to investigate...	12:46.24
chrisl	Guest10482: there are some debugging flags, but I doubt any of the useful ones will work on a non-debug exe	12:49.08
Guest10482	with a -d ?	12:49.42
chrisl	No, the pcl6.exe debug options are -Z... options, but as I say, you'd need to rebuild the exe from source for that	12:50.36
Guest10482	ok	12:52.30
aleray	hi, I'm looking for the herschey fonts	12:57.35
	running archlinux, I can't find the files	12:57.43
	are they somewhere available?	12:58.12
	thanks	12:58.13
chrisl	Gawd, I don't think we've used those fonts in years	13:00.26
aleray	chrisl, nice fonts for plotting :)	13:01.19
	I found the files through ubuntu package gsfonts-others	13:03.03
	I found the files through ubuntu package gsfonts-other	13:03.05
	oups	13:03.07
Robin_Watts	minivan booked.	13:09.22
kens	chrisl what's economy plus on United ?	13:09.24
Robin_Watts	kens: In what sense?	13:09.46
kens	Robin_Watts : what's the differe3nce to economy ?	13:10.03
Robin_Watts	You book for economy, then at seat picking time, can upgrade to extra legroom for $105.	13:10.16
	I think it's just extra legroom.	13:10.20
kens	aha extra legroom OK	13:10.23
	Even more expensive than Delta :-(	13:10.39
Jogux	Robin_Watts : being out the office this morning; I'l check nas/points now	13:14.46
chrisl	kens: I thought there was more elbow room, too	13:14.53
	aleray: is there an upstream source listed in the Ubuntu package?	13:15.28
kens	chrisl possibly. On balance I'll fly United, so we're all together. BA wants to charge me £50 for an exit row	13:15.30
chrisl	Oh, great, and now United has lost my search.....	13:16.22
kens	SoUA900 out, and UA 930 back ?	13:16.27
akerbeltz_	hi quick question, is the localization project for ghostscript on Launchpad or is there an upstream l10n project/platform where we ought to work?	13:16.55
kens	WHat localisation project ?	13:17.05
	Robin_Watts : you and Paul booked already ?	13:17.42
akerbeltz_	I'm kind of assuming ghostscript can be localized into other languages?	13:17.48
Jogux	robin_watts: points is decidely unhappy. it seems the nas is down, or at least no longer reachable :(	13:17.50
kens	akerbeltz_ : No	13:17.56
	There isn't really anything to localise	13:18.07
akerbeltz_	oh	13:18.24
	ok	13:18.25
	thanks	13:18.39
kens	NP	13:18.43
chrisl	Hmm, UA930 isn't listed for me	13:19.13
kens	Oh, I wonder if I chose a stopping fligh, oops	13:19.29
aleray	chrisl, don't know... http://packages.ubuntu.com/saucy/gsfonts-other	13:19.53
kens	Well I certinaly chose the wrong lfight	13:19.56
	chrisl UA 930 departs 7:35pm arrives 2:05 pm (Wenesday)	13:20.42
chrisl	Redoing the search does bring up 930 - this doesn't inspire confidence!	13:20.48
Robin_Watts	kens: yes, paul and I have booked already.	13:21.51
	(Well, I have, paul was doing so as we spoke)	13:22.01
kens	Working on mine now	13:22.12
chrisl	And you got UA900/930 ?	13:22.14
Robin_Watts	yes.	13:22.25
chrisl	Okay, will go find my credit card.....	13:22.39
kens	United web site is convinced I'm American....	13:23.10
chrisl	Top right, you can change your "language"	13:24.03
kens	Ah I see economy plus is 'from' $105 and is actually $124	13:27.18
	Oh no, I take it back a middle row seat is only $118	13:27.43
chrisl	Yeh, and contrary to what I thought, there isn't extra elbow room - not bothering	13:27.52
kens	exit rows are $145	13:28.03
	Even more expensive than delta......	13:28.17
chrisl	"Would I like to upgrade? $12,348 per person" - nah, not just now, thanks.......	13:31.14
kens	LOL	13:31.19
	Oooh Premier access fo $68.....If they could give me faster immigration I might be tempted....	13:32.06
Robin_Watts	I guarantee it won't.	13:32.23
kens	Yeah I was being facetious	13:32.31
Robin_Watts	It'll be 'faster at gates', but not security or immigration.	13:32.47
kens	Indeed	13:32.55
	ie pointless	13:32.59
Robin_Watts	So, I've got twiki.ghostscript.com set up.	13:33.15
	I've added webs for Ghostscript, MuPDF and GhostDocs.	13:33.27
*kens*	will look after I finish the booking challenge	13:33.30
Robin_Watts	and the GhostDocs stuff is currently only visible to users that I put in a specific group.	13:33.46
	How much use it is all depends on how much people use it of course.	13:34.05
chrisl	I thought we weren't doing a wiki for Ghostscript/MuPDF??	13:34.21
Robin_Watts	chrisl: Having set up the wiki, it cost me 2 minutes to add a web for those.	13:34.54
	so it seemed silly not to.	13:35.02
Jogux	robin_watts : awesome. are we going to import the existing picsel stuff into ghostdocs? (or perhaps better, as another module - it might be good to keep ghostdocs focussed on what /we/ care about)	13:35.26
Robin_Watts	We can use it as a dumping ground for all sorts of stuff. Like cluster setup instructions, useful PS scripts etc.	13:35.32
	Jogux: I was wondering that.	13:35.38
	We could maybe make the GhostDocs web public and have a separate GhostDocsPrivate web with the old stuff in.	13:36.02
	But I don't know how hard it is to import stuff.	13:36.18
kens	Oh great, NOW It realises I'm in the UK and insists I start back at square one and deal in £ instead of $	13:36.35
Jogux	actually maybe there's three distinct areas: the old picsel stuff; stuff we need for smartoffice/fvdk/etc customers; stuff that actually relates to the final ghostdoc product.	13:36.54
	robin_watts: I'm not sure either. :-S	13:37.09
Robin_Watts	Jogux: I was still working on the assumption that we'd try and get the VM up and running.	13:37.20
chrisl	kens: BTW, it looks like -dHaveTrueTypes=false isn't working as it should......	13:37.25
Robin_Watts	And then ssh tunnel so we can access that.	13:37.42
kens	chrisl I can't even remember what that's supposed to do	13:37.43
Jogux	robin_watts : ahhh.	13:37.44
	that makes sense I think	13:37.53
Robin_Watts	The old stuff all heavily assumes it's on intranet.picsel.com	13:38.06
	and rewriting all the links seems hard work.	13:38.16
chrisl	kens: in the ps2ps2 docs it says it converts TTFs to bitmaps fonts, but I just get inline images for the glyphs, no fonts at all	13:38.29
kens	chrisl close enough :-)	13:38.50
chrisl	kens: I have a CUPs buggy printer that I hoped it might be a workaround for, but I suspect it won't be usable like that :-(	13:39.35
kens	Its probably somethgni to do with all the enormous amount of work that was involved with type 3 fonts and PCL/PXL including stick fonts and such	13:40.16
chrisl	Well, I don't have the printer, it's a Ubuntu bug	13:40.51
Jogux	robin_watts : interesting; didn't know that. I agree with the approach, anyway. assuming we can get the NAS back to life :-S	13:41.04
kens	chrisl I can look at it, but let me finish booking first	13:41.14
	Or I'll never get there	13:41.23
	Stupid form just reset on me	13:41.42
chrisl	kens: I just found out, haven't looked any further..... no rush, you have until September, at least!	13:41.48
Robin_Watts	Jogux: It may simply be that there has been a power cut at Miles' and the NAS powered down.	13:41.52
*kens*	may need that long at this ratge	13:42.01
Jogux	robin_watts : possiblye; though points didn't lose power.	13:42.38
Robin_Watts	UPS on the servers, but not the NAS maybe?	13:42.51
chrisl	Robin_Watts: thing is, if we wanted a wiki, I'd have preferred to convert the entire website to a content managed setup.......	13:43.01
*Jogux*	nods. possibly but you'd perhaps hope not :-)	13:43.05
Robin_Watts	chrisl: The Picsel stuff has a large twiki, with lots of vital info in.	13:43.48
	so I started to setup twiki with a view to importing that.	13:44.01
	BUT it looks like it'll be slightly non trivial to get that all across into a new setup as it assumes the hostname it's running on.	13:44.31
	An easier approach is to run the old VM (and more secure too)	13:44.50
	but by the time I'd realised that I'd pretty much got the twiki set up, and tor (for one) seemed keen to have it anyway.	13:45.20
	I think a TWiki would be a useful thing to have. For all those things like "how do I set up a cluster user?" (or maybe, "how do I do a release?")	13:46.07
	and "how do I build MuPDF on android?" etc	13:46.21
chrisl	We have docs for most of that stuff	13:46.36
Robin_Watts	for some, certainly.	13:46.57
henrys	Robin_Watts: we tried a FAQ for that and nobody contributed â¦ do you think we'd be more likely to contribute to a WIKI.	13:47.57
chrisl	Things like how to builds shouldn't go on a wiki	13:47.57
henrys	?	13:48.02
Robin_Watts	henrys: Yes.	13:48.35
*chrisl*	doubts it	13:48.47
Robin_Watts	chrisl: Why not? I would say that's exactly the kind of place where stuff should go.	13:48.53
chrisl	Robin_Watts: how to build should be in the release archives	13:49.18
Robin_Watts	If it's unused, we can remove it.	13:49.19
kens	Thanks for putting me in on the email robin, saves me sending one to Miles.	13:49.38
Jogux	henrys : I'd certainly contribute to any ghostdoc wiki	13:50.11
kens	chrisl can you send me the file and command line ?	13:50.17
Robin_Watts	chrisl: Right, but if I'm a user, and I hit problems with a build, and I solve them, I am more likely to edit a wiki to say	13:50.21
	"I had problems here, and this is what I did to solve them"	13:50.31
Jogux	and I was just about to say what Robin just said.	13:50.34
chrisl	We're opening this to users? Oh god :-(	13:50.39
Jogux	it needs a bit of curation for that kind of thing.	13:50.44
chrisl	kens: I just chucked together a PS test, I'll you one after I have lunch	13:51.19
kens	OK thanks	13:51.29
*kens*	tries to remember what I was doing this morning before all this started	13:51.46
spanners	wikis are cool, but can end up a mess of obsolete crud without careful curation	13:54.01
henrys	spanners: careful curation does not seem to be one of our strong suits	13:55.07
spanners	the drawer you have in your house with bits of wire, old string, papers, blu-tac, keys, mystery batteries	13:55.19
	that's what a wiki is to a company	13:55.30
	IMO :-)	13:55.56
*kens*	thinks spanners just described Ghostscript	13:58.52
henrys	Jogux, pedro_ :artifex normally has a staff meeting on IRC at 7:30 Pacific time, in 1/2 hour, if you guys want to attend that's great. If not can you give status to paulgardiner so he can discuss what's developed?	14:00.28
kens	ermm, what's the time now ?	14:00.43
*kens*	is still confused by time zones....	14:00.54
henrys	7 pacific	14:01.06
kens	OK 30 minutes then, thanks	14:01.15
pedro_	henrys: sure - we'll be here anyway	14:02.07
henrys	Jogux: sot does not find my printers either, however if I write a very small client and run it with their library on linux - it works. Of course that is just using snmp, not bonjour	14:03.55
	Jogux: doesn't find them on the ipad	14:04.41
	Robin_Watts: I do think a wiki makes a lot of sense now, with folks getting up to speed they are very helpful. I don't know what sort of wisdom we would gain sharing it with the user base.	14:07.38
tor8	meeting in 25?	14:07.53
kens	23, 22...	14:08.00
henrys	tor8: yes	14:08.01
Jogux	marcosw_ : morning. the NAS doesn't seem to be reachable from points anymore - could you perhaps prod something?	14:11.17
marcosw_	Jogux: I'll give a look	14:19.04
mvrhel_laptop	morning. I may be about 5 minutes late for our meeting... brb	14:19.50
	morning. I may be about 5 minutes late for our meeting... brb	14:19.56
kens	chrisl 8.81 emits the text as linework :-)	14:22.00
	8.71*	14:22.07
	9.00 emits a series of inline images	14:23.10
chrisl	kens: in http://svn.ghostscript.com/ghostscript/tags/ghostscript-9.01/doc/Ps2pdf.htm#Options	14:23.33
	"If not, TrueType fonts are converted into raster fonts with resolution specified in HWResolution"	14:23.52
kens	chrisl yeah but I think that may not have been true for quite some time.....	14:24.07
chrisl	Quite willing to believe that's wrong	14:24.13
kens	I'm curious that 8.71 emits linework	14:24.36
	But neither emits a 'font' as such	14:24.45
	Possibly I could make it do so, but tis in a hairy complicated piece of code	14:24.58
chrisl	Ho hum. It looks like some Toshiba printers have an issue with how ps2write emits TTFs :-(	14:26.09
kens	Ah well, that's not entirely surprising	14:26.26
	Given the way we emit TrueType fonts	14:26.35
chrisl	I wondered it it might be down to the way some of the tables aren't regenerated for a subset, so I tried disallowing subset fonts......	14:27.17
	the opdfread code fails with the whole font	14:27.34
kens	Oh....	14:27.40
chrisl	I assume it's expecting only <=255 glyphs	14:27.56
kens	I would think that's likely yes	14:28.13
	I had been thinking of looking at the TT font emission again, and fixing it so its right	14:28.41
chrisl	Yeh, trouble is, I have no easy way to confirm if that's the problem	14:29.08
kens	I haven't a clue, but there are other enhancements/bugs waiting on me to do that, so I need to sometime anyway. I take it the images work at least ?	14:29.50
chrisl	I haven't sent the "image" file for him to try yet. I need to regenerate his test, and add all the printer specific crap to it	14:30.40
kens	Hmm OK	14:30.47
chrisl	I'll do it after the meeting	14:31.25
kens	I'll see what I can do about a t3 font, maybe nothing	14:31.48
henrys	chrisl, marcosw : so we're are we release wise? - I was sort of worried about those late pxl device changes by hin-tak and I'm not sure when pxl is tested (overnight). That's an important printer device for open source folks	14:32.17
chrisl	henrys: those changes didn't do into the RC, so no problem there	14:32.47
	s/do/go	14:32.54
marcosw_	marcosw: testing showed no significant differences, so we are good to go.	14:33.38
henrys	chrisl: okay good for us but I think hin-tak will be disappointed, he was trying to get them in.	14:33.57
	tor8: are you ready?	14:34.08
chrisl	henrys: tough, frankly	14:34.08
*marcosw_*	seems to be talking to himself	14:34.10
tor8	henrys: yes, I'm back now.	14:34.32
henrys	tor8: are you ready to release?	14:34.48
tor8	that's a trickier question	14:35.00
	we just had some big merges, and I need to verify that nothing has broken with the various builds	14:35.23
	but other than that, yes, I think we should be in good shape. paul might want to weigh in on the status of the ios port, and mvrhel on the winrt	14:35.57
mvrhel_laptop	ok back	14:37.05
henrys	tor8:but it sounds like you have a few more days and chrisl is ready to go with gs. chrisl can decide if he wants to hold off gs until you're ready.	14:37.14
chrisl	As I'm half way through the release, I do not want to wai!	14:37.37
	wait!	14:37.40
tor8	don't wait! mupdf should be ready for release within a week.	14:38.00
	just make scott wait a few days with the newsletter	14:38.23
henrys	mvrhel_laptop, paulgardiner anything stopping a mupdf release?	14:39.24
mvrhel_laptop	no. everything is fine on the winrt side of things	14:39.39
Robin_Watts	We need to fix the cluster. I'll try to do that later today.	14:39.52
	and I'll rewind it so it can rerun those tests that have failed.	14:40.16
paulgardiner	henrys: not that I'm aware of... well for iOS we were hoping to have Airprint. I could have a look at that.	14:40.22
marcosw_	Robin_Watts: is this the build target issue for nojs? I can fix that if you'd like	14:40.30
tor8	paulgardiner: if we can get airprint done this week, that would be very good to have in the release	14:40.41
henrys	all:norbert will be using the cluster, if he comes around IRC (obviously) help him out	14:40.47
Robin_Watts	marcosw_: That, plus whatever is stopping the mujstests working.	14:40.55
tor8	if not, we can make an app store release with airprint after the release (it's not like anybody builds the ios app from source...)	14:41.04
	Robin_Watts: mujstest-v8 is renamed to just mujstest	14:41.30
Robin_Watts	tor8: right.	14:41.37
tor8	marcosw: and there is no "all-nojs" target, just an "all" target	14:41.55
henrys	tor8: so our domain isn't set yet?	14:42.22
tor8	ray_laptop needs to update the DNS to point mujs.com to casper	14:42.39
	(or someone else who knows the DNS configuration)	14:42.49
henrys	marcosw: I feel like there was a miscomm with guillaume, if there is some other way I should pass stuff to you let me know.	14:43.03
	mvrhel_laptop: so how much more to do with gsview?	14:43.43
marcosw_	henrys: I think it was entirely my fault, I've been distracted the last week or so.	14:43.47
ray_laptop	tor8: so there is ahttp://www.mujs.com page on casper ?	14:43.53
Robin_Watts	ray_laptop: There is.	14:44.00
tor8	ray_laptop: yes, the web server should be all set up	14:44.13
henrys	marcosw_: yeah that's no big deal I just wanted to know if there was something procedural that would make things easier for you.	14:44.23
ray_laptop	Robin_Watts: OK. Right after we finish here, I'll dig into the DNS updates	14:44.25
	tor8: thanks	14:44.39
kens	chrisl as far as I can tell that documentation for 'HaveTrueTypes' is misleading, we never made a type 3 bitmap font.	14:45.10
chrisl	kens: no surprise there	14:45.28
kens	I'm not seeing an easy way to add it either....	14:45.58
mvrhel_laptop	henrys: I have to clean up some stuff on the hyperlinks today. Then I need to add just a couple more features (copy page to clipboard, select text, and saving out as with a few different mupdf types). This should not take much more than the rest of this week. Then there will be all the final little clean ups (adding a bit of help information, about information, and figuring out how we...	14:46.00
	...get it in the store)	14:46.02
marcosw_	Jogux: the NAS is back. There was a problem with the ip address.	14:46.03
mvrhel_laptop	we have all the pdf flavor saving working (PDF-X, PDF-A etc). Also we can select regions of a page and save that region as a PS, PDF, EPS output	14:46.59
chrisl	kens: I do find it infuriating that it should be so hard to use an already existing fallback case :-(	14:47.01
kens	chrisl I'm not entirely certain that is an existing fallback, at least not now	14:47.23
marcosw_	henrys: I appreciate the offer but I think it's up to me to pay more attention.	14:47.33
chrisl	kens: I thought any unrecognised font type would fall back to bitmap type 3	14:48.04
henrys	mvrhel_laptop: wow cool	14:48.22
kens	to bitmap yes, but I'm unconvinced it falls back to a type 3 bitmap	14:48.23
	We only seem to capture charprocs for type 3 fonts	14:48.39
chrisl	kens: even for things like the stick font?	14:48.52
kens	everything else seems to wind up as inline images	14:48.56
	chrisl the stick font is specially handled, as a type 3	14:49.05
chrisl	Oh, oh well :-(	14:49.29
kens	As are Microtype and HPGL fotns	14:49.29
mvrhel_laptop	henrys: one weird thing that I am a little concerned about is that our XPS output creation is huge. not sure why. but we are using that for printing, so it may be worthwhile for one of us to spend a little time on this at some point later.	14:49.37
henrys	paulgardiner: so I saved the best for the last 10 minutes, how is SOT ? ;-)	14:50.16
mvrhel_laptop	oh is there a new git repository for SOT?	14:50.33
henrys	mvrhel_laptop: yup that is my department, I'll try to get something on that soon before next week.	14:50.34
mvrhel_laptop	I saw some discussion on the logs	14:50.37
marcosw_	Robin_Watts: i'll let you deal with the mupdf build issue on the cluster, tuesday is my day to go to uni, so I wouldn't be able to spend time on it until tonight.	14:50.47
Robin_Watts	ok.	14:50.55
Jogux	marcosw_ : awesome, many thanks	14:50.58
paulgardiner	henrys: in headless chicken mode.	14:51.06
mvrhel_laptop	:)	14:51.15
Robin_Watts	mvrhel_laptop: There is, but hold off on it for now.	14:51.15
marcosw_	henrys: thanks for walking scott through the sot on iPad demo.	14:51.15
mvrhel_laptop	robin_watts ok thanks	14:51.23
Jogux	henrys : mace should be starting on ats setup tonight.	14:51.33
henrys	marcosw_: oh how did that go?	14:51.34
marcosw_	henrys: not sure, I received an email from scott saying that I wouldn't be needed that you had sorted him out.	14:52.05
henrys	paulgardiner: when mvrhel_laptop is done with gsview he wants to try his hand with SOT, it will be interesting to see how quickly someone completely new to the code can get up to speed. Though mvrhel_laptop is a quick study.	14:53.16
paulgardiner	henrys: we got the builds out requested by a customer. I believe we have the iOS and Android builds for the apps working, so we could potentially do a release once we have regression testing up	14:53.53
mvrhel_laptop	yes. I will be the sacrifice.	14:54.05
pedro_	mvrhel_laptop: good luck :)	14:54.27
mvrhel_laptop	:)	14:54.34
Jogux	hehe	14:54.39
henrys	paulgardiner: we won't schedule releases this time. It goes when it's ready for now.	14:55.26
Jogux	henrys : not sure if you saw when I said the other day, but it turns out the ATS VMs were not imaged from the actual working systems; there hence is some doubt about how complete they're going to be :-(	14:55.29
paulgardiner	I've been meaning to take a look to see if the FBS already includes attempts towards Windows phone	14:55.52
henrys	but I would like to get SOT on the regular march september schedule as the other products.	14:56.03
Jogux	paulgardiner : I believe it does; but I also believe it's not finished.	14:56.11
	henrys : in terms of releases to the appstore?	14:56.35
Robin_Watts	Miles has been talking to an existing Picsel OEM. Interestingly someone that we'd tried to sell MuPDF too. Apparently they have licensed the PDF viewer from Picsel.	14:56.37
	Rather than being a SmartOfficeLibrary customer like the one Paul was building for, they are an FVDK customer.	14:57.07
	They get win32 and RVCT builds.	14:57.15
Jogux	robin_watts : by my reckoning, all non-editting customers will be FVDK (pretty much)	14:57.32
Robin_Watts	I fear we don't have an RVCT build machine anywhere. I was really hoping that would be part of the ATS VMs.	14:57.46
	We need (possibly) RVCT 3.1 and 4.1	14:57.59
Jogux	robin_watts : I highly doubt it. However; I believe the licsenses are there	14:58.07
Robin_Watts	ok.	14:58.17
Jogux	(I forget where, but I'm sure I commented that I found them the other day)	14:58.21
	ah, yes, I think there's a VM that I reckon is the ADS license server.	14:58.28
henrys	Jogux: not necessarily app store every time. But a tagged release we can go back to would probably be useful. SOT is very different than ghostscript and mupdf and may not be served well by schedule releases.	14:58.35
pedro_	henrys: in terms of bugx/layout isue, we've been through debugging a number of issues in the documents and there are quite a number of individual issues there, broadly in 3 categories: bugs in bounding box calculations (leading to missing content), unsupported properties in XML graphics and a number of unimplemented parsing issues with layout attributes	14:59.02
Jogux	henrys : nod. It may take a while for us to start to understand the requirements. I have no idea how many existing customers there are, what there requirements are, etc...	14:59.19
paulgardiner	robin_watts have you mentioned your ideas of recreating picsel's intranet for us to access privately?	14:59.35
pedro_	we have a couple of issues fixed locally but would prefer to commit them when we have the ability to regression test through ats if possible (lots of scope to break things here)	14:59.40
Jogux	henrys : A lot of the picsel customers historically tended to insist on fixes being made to ancient versions :-S	14:59.41
Robin_Watts	paulgardiner: I've mentioned it on here, I can go over it again.	14:59.54
henrys	pedro_: any bug fixes layout wise? Miles wants updates on that constantly.	14:59.56
Robin_Watts	Picsel used to have an internal network, of which the 'hub' machine was intranet.picsel.com	15:00.27
Jogux	I have absolutely no idea when ATS will be up btw; probably a reasonable chance by Monday next week. I'm likening it to trying to tame a feral kitten (something I'm also trying to do just now)	15:00.45
paulgardiner	robin_watts I just thinking that it might give us easy access to various info that is difficult at the moment, even perhaps the vm setups	15:00.55
Jogux	oh, that reminds me, I need to sync the old intranet VM onto the new ATS boxes.	15:01.11
henrys	Jogux: I think we are likely to push back on that unless it is a big player.	15:01.18
Robin_Watts	I think a good way to proceed would be to run the intranet.picsel.com VM (and others) within marcosw's garage.	15:01.25
	Possibly on a separate subnet.	15:01.34
	With no access from the internet.	15:01.41
Jogux	henrys : That was exactly picsel's policy. though their definition of "big player" was fairly random :-)	15:01.55
henrys	Jogux: we've only talked to a few customers and they seem to be happy we're just speaking to them ;-( so I'm not sure how high the bar is.	15:02.13
paulgardiner	pedro_: if you push your changes to your casper repo, I can try to rebase them onto the new epage+resources repo, unless your happier doing it yourself	15:02.17
Robin_Watts	Then we can run an ssh tunnel through one of the machines at marcosw's house, and access the intranet securely from where we are.	15:02.27
Jogux	henrys : :-)	15:02.33
paulgardiner	I'm trying to move everything esle across between those repos at the moment	15:02.45
Jogux	henrys : tbf the worst were the japanese mobile phone OEMs; I think there are a lot fewer of them these days ;-)	15:02.54
kens	chrisl can you think of a test file with a CIDFont in it offhand ?	15:03.03
chrisl	kens: type11withvmtx2.pdf	15:04.01
pedro_	henrys: some of the missing images (although not all) and a missing header/footer style parsing change (although I'm still testing them on different docs). There are inconsistencies depending howthe document was created and/or saved (eg saving in compatible mode xml is less well supported than saving in 'new' xml)	15:04.02
kens	CHRISL OK THANKS	15:04.07
ray_laptop	kens: Or tests_private/comparefiles/HeiseiMinStd.pdf ?	15:04.47
chrisl	kens: or japan-.pdf is even simpler	15:04.52
pedro_	paulgardiner: I'm happy to push the changes; I'd just like to be able to regression test for a getter level of confidence	15:04.53
*kens*	ONLY NEEDS ONE :-)	15:04.55
paulgardiner	pedro_: yeah, we can hold off pushing them to the main repo until then	15:05.30
henrys	the meeting always ends, officially, after 1/2 hour. Usually folks stick around after though, mainly for Jogux and pedro_	15:05.57
ray_laptop	kens: has capslock stuck :-)	15:06.04
kens	Yeah, happens all the time	15:06.15
*pedro_*	nods - we're not going anywhere ;)	15:06.17
Jogux	pedro_ : tbf I think the first step in the ATS bring up will be testing on the tree as we got it from Picsel, so we will then later explicitly test it with any changes that have been made. we may need to do that in stages as robin already changed the layout of everything :-)	15:06.19
ray_laptop	or it's noisy in his house and he has to shout to be heard	15:06.31
pedro_	jogux: sure	15:06.34
Robin_Watts	Jogux: none of my changes are in the new repo yet.	15:08.42
	If ATS is going to be up by next monday, then we can test what we were given, then re-apply changes one commit at a time.	15:09.25
kens	ray_laptop : no the noise is finished now, the chainsaws are gone	15:09.32
paulgardiner	Jogux: at this stage we probably don't care if, for the sake of testing, we have to reset the main repo back to as we received it, provided we keep our updates on a repo somewhere	15:09.38
ray_laptop	kens: so did the patient survive the surgery ?	15:10.04
paulgardiner	Oh I suppose the new sot.git is as we received it	15:10.22
kens	ray_laptop : this week, yes, but the tree will be saying goodbye to its friends soon	15:10.26
henrys	pedro_: so for these bug fixes does the bugzilla workflow work okay for you?	15:11.14
Jogux	robin_watts , paulgardiner : nah, I'd just go ahead and get the repository sane; we can use tags/shas/wahtever to give ATS exactly what we want it to test.	15:11.44
*Jogux*	is sacrificing points's upstream to the ATS gods again; feel free to kill it if it's affecting anyone.	15:12.37
	it'll probably take beyond Monday to get ATS + git playing together properly, but I really hope we can at least run tests by then.	15:13.33
pedro_	henrys: yeah, its fine although I'm effectively working across a number of the bugs simultaneously just now as each docuemnt contains a number of distinct issues. I'm trying to find common issues, but I could raise specific bugs based on the root cause if thats more useful?	15:13.35
paulgardiner	pedro_: that sounds sensible to me.	15:15.44
henrys	pedro_: we like to have a bug fix commit associated with a bug number in bugzilla - the number referenced in the log. So if that demands a new bug, that's fine, then you can close other bugs as being fixed by the other fix.	15:15.52
	pedro_: the bug number should be somewhere in the commit log	15:16.22
pedro_	henrys: cool - as it stands I think we'll have a number of fixes for each of the existing documents, so I'll make sure they all reference the appropriate fixes.	15:18.16
henrys	pedro_: bugzilla's resolved as a duplicate functionality seems like it might be fitting for many of your cases.	15:19.32
*pedro_*	nods	15:19.57
henrys	how does one pronounce the name Iain banks - ee ah in?	15:23.38
chrisl	Just like Ian	15:23.54
henrys	chrisl: thanks, as soon as shift is done I'm diving in.	15:24.27
paulgardiner	ee un I reckon	15:24.45
henrys	early death, very sad.	15:26.12
pedro_	iirc its the Gaelic equivalent of John	15:26.16
	henrys: indeed	15:26.22
henrys	paulgardiner: in the states it would be Ian, that medial 'i' had me wondering	15:29.26
kens	Its a Gaelic spelling	15:29.50
Robin_Watts	We have both Ian's and Iain's here.	15:30.05
ray_laptop	almost as confused as french spellings	15:30.10
paulgardiner	I think it was the first time I'd seen that spelling too	15:30.52
pedro_	paulgardiner: its the norm up here :)	15:31.14
chrisl	I never met anyone called Norm.....	15:31.31
pedro_	oddly enough neitherhave I	15:31.47
ray_laptop	OK. For mujs.com which of you transferred it to us, and who manages the domain ? godaddy or network solutions ?	15:31.48
paulgardiner	Nothing is normal up there :-)	15:31.50
Robin_Watts	ray_laptop: Neither of us transferred it to us.	15:32.12
pedro_	paulgardiner: you're not wrong ;)	15:32.16
Robin_Watts	ray_laptop: Miles did the deal for the domain. We've not heard anything more.	15:32.38
ray_laptop	I wasn't in the loop on the purchase of it (Miles took care of it)	15:32.50
	last I heard from Miles was "It's done. I'll forward the info to Ray when I get the confirmation."	15:33.26
	seems like I need to contact Miles	15:33.45
chrisl	pedro_: The father of a school friend of mine would correct your spelling of his name to whichever of Ian/Iain you hadn't used - drove us nuts :-)	15:33.54
Robin_Watts	ray_laptop: Right, that's why we assumed you'd know. Sorry.	15:33.59
pedro_	chrisl: excellent idea. Just need to do the same with Ian/iain/Eoin/John...	15:34.47
Robin_Watts	Henceforth I shall be Robyn	15:35.16
chrisl	I think I'll stick to Chris.....	15:36.06
tor8	pedro_: don't forget Jon and Jan	15:38.27
pedro_	tor8: ah - forgot. I think there are enough variants for us all to adopt one ;)	15:39.10
tor8	and Johan	15:39.13
pedro_	and Ewan and Owen	15:39.46
chrisl	Owain.....	15:39.59
ray_laptop	and Jean	15:53.11
Robin_Watts	tor8: Did you change the NOX11 stuff in the mupdf makefiles ?	15:54.48
tor8	Robin_Watts: yeah. HAVE_X11=no is the now NOX11	15:55.01
	s/now/new/	15:55.08
Robin_Watts	ok.	15:55.31
kens	chrisl I'm momentarily at a loss, the CIDFOnt code does fall back to a type 3 font, I can't for the life of me see why the TT code doesn't.	16:02.18
	OIf course, I can't actually see where the Type 3 gets built in the CIDFont case either....	16:02.38
Robin_Watts	mvrhel_laptop: I've just pushed the lcms 2.6 import to golden.	16:03.13
	If you want to play with making lcms2 use it's threading stuff, feel free.	16:03.36
kens	chrisl norbert is quick off the mark there....	16:03.36
mvrhel_laptop	robin_watts: ok great. thanks	16:03.44
chrisl	kens: the CIDFont always falls back to type 3	16:04.18
kens	chrisl yes, in ps2write	16:04.26
	But the path is the same as the TT font with no HaveTrueTypes, yet the CIDFont 'works' (creates t3) and the TT case doesn't. Currently bafled by that	16:05.01
chrisl	Strange - I'd assumed it would be the usual totally separate logic for two almost identical functions.....	16:05.39
kens	chrisl, no because I wrote (or heavily modified) most of this	16:05.59
	THere is something different, clearly, but I can't see it at the moment	16:06.18
tkamppeter	chrisl, should we not mention on http://www.ghostscript.com/doc/current/News.htm that we have now a new "pwgraster" output device for PWG Raster output and support for PPD-less printing? My commits of 2013-11-27 and 2013-12-19.	16:07.11
chrisl	tkamppeter: too late, the release it done	16:07.34
	tkamppeter: unfortunately, those news pages are included in the release archives, so it would mean repeating the release process - sorry	16:09.25
tkamppeter	chrisl, thought that this is only web site stuff, but never mind, it is enough that I know it personally as I am the only direct user of it. End users use it only through cups-filters.	16:09.50
chrisl	tkamppeter: I'll make a note to put a retrospective announcement in the next release - I do think it worthy of note	16:10.38
Robin_Watts	Jogux, pedro, paulgardiner: Any idea what IAPT is ?	16:11.40
paulgardiner	none sorry	16:11.58
Jogux	I think it's something like internal alien porting team	16:12.44
	what exactly it does to the build I can't remember	16:13.01
	in which context are you seeing/worrying about it? :)	16:13.08
tkamppeter	chrisl, OK.	16:13.27
Robin_Watts	Jogux: I've found an FVDK build line, so I'm having a go at it. and it mentioned -iapt	16:14.24
	and it saying "Using IAPT clean room" which sounded very exciting :)	16:14.41
Jogux	oo, neat.	16:14.50
paulgardiner	print "Using IAPT clean room" os.system("rm *")	16:17.05
Jogux	I /think/ it means it's using a different, cleaner, version of the alien app. maybe.	16:17.41
pedro_	the IAPT was basically just an external reference application model (reference apps we could supply to customers along with the picselwin lib)	16:18.42
Robin_Watts	OK. We'll see what it spits out at the end.	16:19.03
pedro_	apps are in epage/alien-reference-applications	16:19.32
Robin_Watts	They have listed 2 specific problems with the PDF agent, so I guess I ought to look at those.	16:19.35
Jogux	ahha. I believe the iapt stuff essentially means "use the alien in epage/alien-reference-applications" (ie. not the platform/tgv/apps/windows-alien etc)	16:21.08
	seems to cover linuxfb/linuxqtk/linuxgtk/win32.	16:21.47
marcosw	chrisl: is there a reason we are skipping 9.13 as a release?	16:21.54
Robin_Watts	13!	16:22.02
kens	voodoo numbers	16:22.17
Robin_Watts	We'll be skipping 9.370 too.	16:22.18
chrisl	marcosw: I was just making the point about 13 (after skipping 9.11)	16:22.20
pedro_	Robin_Watts> think they only build with vs2005 at present	16:22.26
Jogux	oh, and activex. Paul_Gardiner will be ecstatic :)	16:22.37
Robin_Watts	vs2005 is my preferred vs.	16:22.40
marcosw	robin_watts 9.370?	16:22.40
	what about 9.666?	16:22.50
Robin_Watts	marcosw: It's the malaysian version.	16:23.08
*kens*	will be retired by the time we get to 666	16:23.18
paulgardiner	Jogux: that must have died surely	16:23.32
ray_laptop	actually, gs only supports 2 digit minor numbers	16:24.13
Jogux	paulgardiner> last change in 2012. so maybe :)	16:24.51
marcosw	if we plan on selling to china we can't have a 3 in the release number at all (or is it 4?, I can't keep track). I do know that 8 is a good thing, which is why United flight Bejing -> SFO is UA888.	16:24.57
Robin_Watts	4 is bad.	16:26.07
kens	Robin_Watts : are you aware the cluster is still failing ?	16:27.24
Robin_Watts	kens: I am :(	16:27.36
kens	Ah, OK then	16:27.46
Robin_Watts	tor8: The cluster is having trouble building cmapdump.	16:31.24
	If I'm reading the makefiles correctly, we only build it if CROSSCOMPILE == yes now ?	16:31.58
	or actually, maybe never build it at all.	16:33.12
	I think the CROSSCOMPILE logic is screwed. I'd be in favour of a force-push to fix this, otherwise we are going to have a long series of broken tests in the cluster, some of which we could really do with having had tested.	16:39.58
tor8	Robin_Watts: oh yeah. there's a fix for the CROSSCOMPILE error later.	16:40.34
	feel free to rebase and squash it into the right place so we can test the lot	16:40.50
	Robin_Watts: aaab0da Fix typo in Makefile.	16:41.24
Robin_Watts	ok.	16:42.02
chrisl	bbl	16:51.24
Robin_Watts	Ok, let's hope that these tests now run through.	16:52.19
*kens*	finally finds the place where the type 3 character gets built for CIDFonts	16:55.19
	OK in the TrueType case our bitmap has no ID< so we don't know its a character, that's why its not coming through in a type 3 font.	16:56.21
	Aha!	16:57.48
	Its being rendered uncached by FAPI	16:57.57
	THat's the casue chrisl	16:58.03
	we genuinely don't make fonts for uncached glyphs	16:58.45
Guest10482	chrisl, for information, I build from source a new pcl6.exe and now it is working...	17:12.15
Robin_Watts	Hi zeniko.	17:13.11
	I'm afraid there has been some force pushing on the mupdf repo because the import of mujs broke our regression tests.	17:13.52
	The new tests are hopefully running through now.	17:14.06
zeniko	Robin_Watts: no worries, I just haven't figured out how to convince git to rebaes my patches on top of the new HEAD	17:15.39
Robin_Watts	Also, we (well, I) changed the implementation of streams in order to allow us to do processing at the operator level.	17:15.52
zeniko	Robin_Watts: in case you're not quite done with the release yet, there's two more simple patches on zeniko/mupdf and one openjpeg one on zeniko/ghostpdl (in case you're not going to use openjpeg 2.1)	17:16.19
Robin_Watts	zeniko: We haven't started the release yet. Getting the tests working was kind of a prerequesite for that :)	17:16.54
	henrys: Can you look at the openjpeg fix on zeniko/ghostpdl?	17:17.19
	The two fixes on zeniko/master look fine. I will pull those in, thanks.	17:18.58
kens	chrisl seems that we reduce the resolution to 72 dpi when creating the cached bitmap for CIDFonts (!!) and this is why the caching works for that, but not for TT fonts, because those get rendered at 720 dpi and are too big for the cache	17:25.32
	All I need to do now is find out where the 72 dpi gets set. Alternatively, try running wth -r72	17:25.55
	For me that produces a font	17:26.53
	Should be good enough for testing prposes	17:27.03
	I'll look into the resolution thing tomorrow, in the meantime I've had enough for one day, goodnight folks	17:28.16
ray_laptop	bye, kens	17:28.55
Robin_Watts	tor8: ping	17:28.56
	tor8: The first of the cluster tests just passed. Hopefully we'll run through quite quickly now.	17:29.39
	But 1 file is looking different in the mupdf test run of the mujs commit.	17:30.12
	tests_private/pdf/sumatra/1503_-_support_form_text_fields.pdf at all resolutions/bandings.	17:30.29
mvrhel_laptop	I am having trouble rebasing to golden	17:39.10
	something funky with the mujs stuff	17:39.23
	robin_watts; is there something different that I need to do?	17:39.35
Robin_Watts	mvrhel_laptop: Yes, we had to force push, unfortunately.	17:39.53
mvrhel_laptop	oh	17:39.59
Robin_Watts	How many commits do you have?	17:40.14
mvrhel_laptop	all the commits that I have to rebase are in my branch, there are a lot of them	17:40.48
	but they are contained in the winRT folder	17:41.01
Robin_Watts	mvrhel_laptop: Can you push your branch onto your private repo please?	17:41.26
mvrhel_laptop	right now my master is in a conflicted state	17:41.44
Robin_Watts	Then I'll pull it down, do the necessary, and then push it back up for you to grab.	17:41.47
	git rebase --abort	17:41.54
mvrhel_laptop	ok. let me push what I have then	17:42.15
	robin_watts: ok. so the win_desktop branch has all the commits	17:43.59
Robin_Watts	OK, give me 2 mins to finish this email	17:44.14
mvrhel_laptop	my master seems to have veered off from golden according to what I can see, hopefully you can fix that too	17:45.21
Robin_Watts	OK, to fix master... git checkout master && git reset --hard HEAD~100 && git pull --rebase golden master	17:49.16
	That changes you onto master, throws away the top 100 commits, then pulls in the new stuff.	17:49.33
mvrhel_laptop	ok	17:52.38
Robin_Watts	OK, so robin/win_desktop should have your changes on.	17:54.59
mvrhel_laptop	robin_watts: ok thank you	17:55.11
Robin_Watts	to pull that in, you probably want to: git checkout win_desktop && git pull --rebase robin win_desktop	17:55.41
mvrhel_laptop	ok I have my master here and on my repos back in sync with golden	17:58.06
Robin_Watts	fab.	17:59.03
	sorry about that.	17:59.10
mvrhel_laptop	robin_watts: so you did not push your changes into my repos?	18:02.43
	just so I understand	18:02.50
Robin_Watts	I can't push into your repo.	18:02.56
mvrhel_laptop	ok	18:03.00
Robin_Watts	All I can do is make it available so that you can pull it into yours.	18:03.26
	The main repo is set up as "shared", so it's slightly different. We can all push to that.	18:03.59
mvrhel_laptop	ugh. this is a mess	18:05.55
Robin_Watts	?	18:07.36
	Are you having trouble pulling my changed version in?	18:08.09
	If the git pull --rebase isn't working, then we can use a hammer and do it in 2 commands.	18:08.31
mvrhel_laptop	yes. I may have already used the hammer but I am not sure I did it correctly	18:08.46
	hold on	18:09.02
Robin_Watts	Well, first, let's back out of the rebase.	18:09.10
	git rebase --abort	18:09.13
mvrhel_laptop	hold on	18:09.17
Jogux	he's gone to get the chainsaw ;)	18:09.48
mvrhel_laptop	its odd there are a couple commits from Simon that you stuck in there that I dont see in the head	18:10.04
	or the master that is	18:10.11
	perhaps I had not pulled those in yet	18:10.46
	may have just missed them	18:10.50
Robin_Watts	Yeah, they went in very recently.	18:11.24
mvrhel_laptop	ok. I think I have it all fixed now	18:12.15
Robin_Watts	fab.	18:12.56
mvrhel_laptop	yes. ok everything looks good	18:13.33
	thanks robin_watts	18:13.36
Robin_Watts	no worries	18:13.59
mvrhel_laptop	robin_watts one question for you	18:14.02
	for some reason the jpeg folder in the submodule keeps thinking it is modified	18:14.40
	or openjpeg that is	18:14.47
	is there any special command I can do to zap it back to its senses	18:15.07
Robin_Watts	git submodule update --init ?	18:15.28
	Whenever you update you should do that, in case someone has changed the thirdparty submodules.	18:16.08
mvrhel_laptop	right. I have everything updated there	18:16.20
	its weird. next time we are together I will show you	18:16.45
	it is not a major issue	18:16.51
Robin_Watts	henrys: What's our luratech jpx decoder license?	18:23.37
	Can I pull it into epage?	18:23.44
	or would that require license fees?	18:23.58
paulgardiner	At last! ~paulg/private-repos/sot.git has the new combined epage/resources repo with our own commits rebased across (at least everything I could find in the epage repos on casper)	18:50.33
Robin_Watts	paulgardiner: Excellent!	18:51.08
Jogux	paul: woot. cloning now :)	18:51.31
	paul: btw, did you use some clever method to copy between the two trees?	18:51.49
paulgardiner	git rebase -s subtree	18:52.08
Robin_Watts	Probably we should review all the commits before just pulling them across willy nilly.	18:52.39
paulgardiner	But it fails on commits that only create files, so also some fiddling	18:52.55
Robin_Watts	I'm sure there are some of mine in there that would benefit from another look.	18:53.01
	paulgardiner: An alternative would have been to: git format-patch HEAD~10 (or however many), then edit the paths in the patches and reapply.	18:53.36
paulgardiner	robin_watts that's why I stuck them just on my repo, so people have access to them but without them goint to the main repo	18:53.45
Jogux	paul : that seems clever :-)	18:53.59
paulgardiner	robin_watts that might have been easier, to be honest	18:54.06
Robin_Watts	so people should treat: /home/git-private/sot.git as the golden repo.	18:54.18
Jogux	does format-patch cope with binary files?	18:54.32
paulgardiner	Although most of the hassle with both methods is navigating the huge tree	18:54.34
Robin_Watts	Jogux: IIRC, yes.	18:54.40
Jogux	neat	18:54.44
paulgardiner	robin_watts that's worrying. I'm fetching from the main repo again with the addition of --tags and receiving a lot of new objects	19:01.34
Jogux	which 'main' repo? :)	19:02.02
paulgardiner	:-) /home/git-private/sot.git	19:02.46
	golden, as we call it	19:02.57
Robin_Watts	Jogux: paulgardiner did a conversion step earlier (to remove CVSROOT) and didn't use "--tag-name-filter cat"	19:03.27
	We looked, and it didn't seem like there was a problem with the created repo.	19:03.48
	All the tags looked to be in the correct places.	19:03.55
Jogux	ah, neat / weird.	19:04.12
Robin_Watts	But if we were redoing the process, I'd add that back in.	19:04.27
paulgardiner	It's probably fine. Perhaps there a few tags not on branches	19:04.34
Jogux	uh, I just cloned ghostscript.com:/home/paulg/private-repos/sot.git - what's the /home/git-private/sot.git one?	19:04.45
paulgardiner	... if you see what I mean	19:04.48
Robin_Watts	Jogux: The /home/git-private/sot.git one is the golden repo. It's the one that pauls one is based on.	19:05.11
paulgardiner	Jogux: call the first paul and the second golden	19:05.14
Robin_Watts	paul = golden + our commits.	19:05.26
Jogux	right. thanks :)	19:05.32
paulgardiner	Also in the future, paul will be where I push things for safe keeping and for people to look at for review	19:05.58
Robin_Watts	Any commit to golden should probably be subject to our informal review system from now on.	19:06.02
paulgardiner	Weird. fetching tags has pulled in a load of commits that start "Summary:". Even if my CVSROOT removal was wrong, it shouldn't have made those reappear.	19:10.17
Jogux	paul : have I forgotten the name of the tag for the android build you did last week, or has it vanished?	19:10.41
paulgardiner	Not sure. I don't seem to have your tag either	19:11.43
Jogux	I suspect anything you did with rebase -s has lost the tags :-S	19:12.07
Robin_Watts	I don't think we were expecting tags to be copied across from epage.git to sot.git	19:12.17
Jogux	robin_watts : yes :-)	19:12.25
Robin_Watts	but all the tags that were in sot.git already should still be there.	19:12.37
paulgardiner	Ah yes, of course. I realised that earlier and forgot.	19:13.13
Jogux	:-)	19:13.18
paulgardiner	But still there is something strange going on with tags	19:13.26
Jogux	so I guess we just manually put those tags back on?	19:13.26
paulgardiner	Jogux: yep	19:13.37
	Ah no sorry, ignore me. The repo is fine	19:16.09
	Too tired	19:16.18
	I swapped back to the combined repo by accident	19:16.37
Jogux	yeah, too many repos just now :-)	19:16.53
paulgardiner	Yeah. All looks fine, other than the need to recreate our recent tags, as you said	19:18.16
tor8	Robin_Watts: 6801564 fix warnings in fitz/printf.c breaks linux builds	19:18.45
Robin_Watts	tor8: oh, ass.	19:19.38
	Do you have a fix ?	19:19.43
tor8	-#if _MSC_VER < 1800	19:19.56
	+#if defined(_MSC_VER) && _MSC_VER < 1800	19:19.57
Robin_Watts	Want to push that to tor/master, and I'll review it, and we can push it out?	19:20.21
	We can live without force pushing that and causing michael more hassle as it's the last one we pushed.	19:20.50
tor8	Robin_Watts: we have the same basic ifdef but just _MSC_VER in fitz/buffer.c	19:22.15
	so maybe we should just fix the underlying problem with some macros in fitz.h	19:22.29
Robin_Watts	tor8: That might be nicer.	19:23.03
tor8	I'll bash on that and give you a commit to review in 5 minutes	19:23.15
Robin_Watts	ok.	19:24.05
tor8	Robin_Watts: okay, va_copy ifdef fix on tor/master	19:30.38
	please double check that it builds on win32	19:30.51
	have you bmpcmped mujs difference that you mentioned?	19:31.36
Robin_Watts	tor8: Why do we use va_copy_end rather than just va_end ?	19:32.00
tor8	Robin_Watts: because in c89, va_end belongs to va_start, and there is no va_copy	19:32.23
	in c99, if you do va_copy, you must also va_end	19:32.30
Robin_Watts	Right, so why do we define a va_copy_end?	19:32.51
	Can't we just call va_end, and if it doesn't exist, define it.	19:33.05
tor8	so that on c89 we can do ap=old_ap and not call va_end	19:33.06
	and in c99 that becomes va_copy and va_copy_end	19:33.14
	doing ap=old_ap and then calling va_end would be wrong	19:33.34
Robin_Watts	ok.	19:33.40
tor8	and we still have places where we va_start and need to call va_end	19:33.44
Robin_Watts	looks good to me, but I haven't tested it yet.	19:33.59
	Just a mo.	19:34.00
tor8	now we could probably get away with doing ap=old_ap and never call va_copy	19:34.05
Robin_Watts	tor8: have you done a windows build of mujs before?	19:39.25
	warning city arizona.	19:39.31
	and you appear to be calling naked realloc and not checking the results :(	19:40.06
	but that fix commit works. Shall I push it?	19:41.11
	pushed	19:41.49
	There are 5 instances of naked realloc in jscompile.c. I suspect they should be js_realloc ?	19:44.03
	tor8: Let me try to fix the warnings as I'm on windows.	19:54.33
henrys	yikes bugzilla is almost unusable, Jogux much more to do?	20:14.04
Robin_Watts	tor8: Various reviews for you then.	20:15.19
Jogux	henrys: urrr. bugzilla is hosted at Miles's office?	20:18.38
henrys	visual studio'er â¦ any idea why the intellisense data base won't write to a shared directory, it says no permission - but it converts the makefile and writes to the directory. maybe no permission to create a directory.	20:19.19
Robin_Watts	Bugzilla was on casper, but it was too much of a load for the instance. So we moved it.	20:19.24
henrys	Jogux: yes	20:19.25
Jogux	also, my rsync finished 4 hours ago :-)	20:19.27
	henrys : okay. didn't know that; I shall try and find an appropriate bwlimit next time. but I don't think the current issue you see is my fault :-)	20:20.04
Robin_Watts	henrys: Usually that kind of thing is because of having 2 copies of vs open on the same project at once.	20:20.37
	otherwise I'd suggest rebooting and trying again?	20:20.47
	Unless miles has hit the fair usage limit or something and is now being throttled? (dunno if there is one, but...)	20:21.20
	Or maybe marcosw is copying stuff?	20:21.34
henrys	bugzilla is okay now don't know what happened.	20:22.41
marcosw_	wasn't me, I wasn't using an appreciable portion of miles' bandwidth. and as far as I know being a business account there is no limit.	20:35.56
	there is a twice a day mysqldump that locks the bugzilla database for a couple of minutes, but that happens at 6:00am and 6:00pm PDT.	20:37.31
tor8	Robin_Watts: oops, re the naked reallocs	21:09.21
Robin_Watts	tor8: easy mistake, simple fix. see my repos.	21:09.40
	Am I right in thinking that you're using a fixed size exception stack?	21:10.29
	Isn't that a bit risky?	21:10.38
	Could someone deliberately arrange for something heavily recursive to blow it?	21:10.56
tor8	Robin_Watts: yes. if we run out of exception stack space, we throw a stack overflow error	21:11.07
Robin_Watts	Is that the ->panic() thing ?	21:11.40
tor8	the panic thing what happens if we throw an exception without a try around it	21:12.06
	normally you'll call js_dostring or js_pcall which are 'protected' calls that have a try	21:12.26
Robin_Watts	right.	21:12.35
tor8	but if you get an error while setting up your native bindings or something like that, you could end up with an exception being thrown without the try protection	21:13.05
Robin_Watts	I see that fz_malloc etc throw in the failure case.	21:13.06
tor8	that's when panic gets called	21:13.08
Robin_Watts	ah, i see.	21:13.22
	cunning.	21:13.27
	You've not got any fz_var equivalents in there.	21:13.42
tor8	shouldn't be necessary, the way I've structured the code	21:14.04
	fz_var equivalents, i.e.	21:14.10
Robin_Watts	Is that because nothing is in local vars? It's always members being written to?	21:14.18
tor8	anything risky always lives on the stack, and any local vars are read-only after the try	21:14.30
Robin_Watts	makes sense.	21:14.40
	There are 4 warnings that I didn't fix.	21:14.56
	to do with reallocing things that are const char **'s.	21:15.06
tor8	I do: a=malloc(); if (try()) { free(a); } ... stuff that uses a...; free(a); endtry();	21:15.12
	oh yeah, void* and const behaves oddly with warnings on MSVC :(	21:15.29
Robin_Watts	no, the problem is that you can't realloc a const char ** on any platform :)	21:15.57
	or at least, you shouldn't be able to.	21:16.08
	being called for food, I gotta go, sorry!	21:16.20
tor8	I hope you mean "char * const *" not being reallocable	21:16.30
	Robin_Watts: clang and gcc are annoying with -Wunreachable-code	21:18.06
	clang warns about the abort sections you added, and without them gcc warns instead	21:18.31
	but gcc has a broken -Wunreachable-code anyway :(	21:18.43
Jogux	I thought gcc made -Wunreachable-code a no-op in the latest versions?	21:21.23
	( http://gcc.gnu.org/ml/gcc-help/2011-05/msg00360.html )	21:22.30
mvrhel_laptop	ok. some nice page copy and paste stuff added to gsview	21:32.38
aleray	Is it possible to convert gsf font files to something else like type 1?	21:39.19
henrys	Jogux:are you in contact with mace? You said something earlier about test setup. Just got off the phone to Miles and he would like the web site changes to be the high priority thing.	21:48.32
Jogux	henrys : I shall pass that on	21:48.57
henrys	Jogux: thanks - also miles says he still hasn't gotten email response from him, - he's worried mace isn't getting his mail.	21:50.06
Jogux	henrys : okay, odd, will pass that on too. I know he has been a combination of ridiculously busy and/or away from the office over the past 4 days or so.	21:51.08
ray_laptop	aleray: gsf isn't really a standard font type. Gerber fonts are sometimes labelled .gsf -- other older .gsf fonts were actually conforming Type 1, but weren't from a particular font provider	21:58.19
	google searches turn up that Gerber fonts don't readily convert to outline fonts.	22:01.40
aleray	ray_laptop, thanks. I tried using fontforge,which seems to support gsf, but I got an error: "didn't understand /makeHersheyFontGS993 where " in blended font defn"	22:02.03
ray_laptop	Some Type 1 fonts were shipped with old Ghostscript were labelled .gsf, but they really were Typ 1	22:02.25
aleray	ray_laptop, any idea how I could do then. My aim is to produce an otf, by expending the stroke to an outline	22:03.04
	fontforge can outline, but it doesn't seems to like the kind of custom commands in some gsf font in gfonts-other	22:04.14
ray_laptop	The hershey fonts were early fonts (outline, stroke fonts) shipped with gs	22:04.20
	aleray: you have now surpassed my understanding of these old fonts :-(	22:05.25
aleray	ray_laptop, yes I'm actually trying to make an outline version of the simplex hershey font	22:05.57
	here is what it looks like: http://paulbourke.net/dataformats/hershey/	22:06.05
Jogux	henrys : mace reports website stuff is essentially done, he just needs to put it online.	23:59.13
	Forward 1 day (to 2014/03/26)>>>

IRC Logs

Log of #ghostscript at irc.freenode.net.