<DocScrutinizer05>
zrafa: anyway the move to kicad results in the project getting way more open and FOSS-spirit compliant, in the end of the day might turn out to be the best thing ever "happened" to the project, and WOULD MAKE FOR A REALLY GREAT NEWSLETTER
sn0wmonster has joined #neo900
<DocScrutinizer05>
akas we lost our only talented and willing PR officer dos1
<DocScrutinizer05>
alas*
xes has joined #neo900
<zrafa>
DocScrutinizer05: (WOULD MAKE FOR A REALLY GREAT NEWSLETTERWOULD MAKE FOR A REALLY GREAT NEWSLETTER) yeah! I agree completely
<DocScrutinizer05>
the problem isn't there are no news, the problem is we got too many news and no writer who could convert those into a nice format suitable to publish
<hellekin>
DocScrutinizer05: what about accumulating news for a month and send them in bulk, with regularity, whatever the quality of the writing?
<hellekin>
editing a newsletter is quite demanding, but accumulating small bits and pieces over time is rather painless
<hellekin>
especially if you use git </troll>
<DocScrutinizer05>
if you say that. I don't see how to do that
<hellekin>
well, whenever you get some news, add a sentence to a file. At the end of the month, send that file. That's a conversation starter at least.
<DocScrutinizer05>
I get news every second
<hellekin>
if you or somebody else has a bit of time during the month, they can add more precision and reference urls...
<DocScrutinizer05>
I got no time to review and summarize them
<hellekin>
the interest of *news* is that you don't have to go through 200 pages to find 1 page of relevant stuff.
<DocScrutinizer05>
that's why I answered zrafa's questions
<DocScrutinizer05>
what you're asking for is a "please provide the newsletter text and I will fix the spelling"
<hellekin>
80% of this channel's content at the very least is nothing like news
<hellekin>
even reducing to only what you and wpwrak write here, we're very far from it.
<DocScrutinizer05>
yeah, too bad that it's the author who needs to make the story. Not the world creating the story for the author to pick it up
<hellekin>
the move to kicad is certainly news.
<hellekin>
but the meaning of it is not obvious to me beyond "we're moving away from a proprietary system to an open source system". That requires a bit of understanding of the software to get the strategic relevance. E.g., how does that change the relationship with the manufacturers, and so on.
<DocScrutinizer05>
I seem to recall we had quite a few very interesting widely known evens recently, where we also had something to contribute. Wasn't it something about Mr Snowden etc?
<hellekin>
yes, that happened because we were available at that time to "play it" properly. Still very few reach and response.
<DocScrutinizer05>
I don't see how *I* could help to convert that into a newsletter then
<hellekin>
well actually 10 followers on twitter
<DocScrutinizer05>
((how does that change the relationship with the manufacturers)) irrelevant
<hellekin>
see, you know what's relevant news.
<DocScrutinizer05>
yes, the move to kicad is relevant since it's a FOSS tool available to everybody
<DocScrutinizer05>
and we now have the project files in git
<hellekin>
zero response from bunnie or frank rieger.
<DocScrutinizer05>
so it's hard to figure how we could get any more community-freindly and open
<DocScrutinizer05>
well, it's Sunday!
<hellekin>
same silence I hit when I talked to cryptophone users really close from him.
<DocScrutinizer05>
see? even you have news
<hellekin>
how is that news?
<DocScrutinizer05>
how's it not?
<hellekin>
that was a private conversation in January.
<hellekin>
reading a reddit thread to skim the content is pretty much a day job, and now it's the middle of the night :)
<DocScrutinizer05>
hmm, I didn't either
<DocScrutinizer05>
however the software flaw they detected is highly relevant for Neo900 since even when our modem wuld suffer from it, we simply don't care since our design is centered from the beginning around coping with a compromised rogue modem
lkcl has joined #neo900
<hellekin>
then that's press release material.
<DocScrutinizer05>
?
<hellekin>
when the press is telling the world how bad the situation is and oh, we don't know how to fix it, but actually you know, it's good to tell them
<hellekin>
not sure what you're talking about. Anyway it's time for me to sleep. If you want to expand please do it by email, at least it sticks to my mailbox.
<hellekin>
ttyl good night!
<DocScrutinizer05>
wouldn't know how to do a dialog via email
<DocScrutinizer05>
what I'm talking about? >>[2016-07-25 Mon 02:41:34] <DocScrutinizer05> however the software flaw they detected is highly relevant for Neo900 since even when our modem wuld suffer from it, we simply don't care since our design is centered from the beginning around coping with a compromised rogue modem <<
SylvieLorxu has quit [Read error: Connection reset by peer]
sn0wmonster has quit [Ping timeout: 258 seconds]
<DocScrutinizer05>
now when that is no headline for a story "Snowden and Bunnie try to design something for iPhone that looks like a knockoff or poorly done clone of 50% of the security concepts of Neo900. Also a recently found vuln in probably all Qualcom modem chipsets (among many other infrastructure systems like Base Stations) would render all those phones incl the Snowden secure phone shell (if iPhone was hit by the vuln) compromised and no fix is
<DocScrutinizer05>
likely to ever show up, while Neo900 still stays secure, no matter if its modem has the vulnerability or not<<
<DocScrutinizer05>
...I dunno what to say more, except "ask me for details"
sn0wmonster has joined #neo900
<DocScrutinizer05>
hellekin: highlight so you can't say you can't filter the relevant stuff out of the 'noise' in this channel
<wpwrak>
(lack of anelok news) been rather busy with neo900 in the last weeks. also, my current work item on anelok a primitive GUI for managing accounts on a PC, which is messy work (i'm not too good with GUIs) and not really newsworthy before i have something running that at least connects the dots. so, a bit more patience is needed :)
<wpwrak>
how about a pdf ? i'm not so good at picturing schematics from reading text :)
Pangolin has quit [Ping timeout: 260 seconds]
<DocScrutinizer05>
a pdf won't show any of the relevant aspects
<wpwrak>
btw, what you're currently doing is what people use branches for. make a variant of the design, see if it pans out. if yes, merge.
<DocScrutinizer05>
this is not fit for merging, I did lots of mess in there
<wpwrak>
ah, nice. a bit of new activity on the "Greift zu kurz" thread
sn0wmonster has joined #neo900
<DocScrutinizer05>
it's just a PoC to evaluate KiCAD functionality with those particular changes
<DocScrutinizer05>
and a branch doesn't help me since I'm not used to the workflow
announ has joined #neo900
<wpwrak>
i know, i know. it's hard ...
lkcl has quit [Ping timeout: 244 seconds]
zrafa has joined #neo900
Defiant has quit [Read error: Connection reset by peer]
Defiant has joined #neo900
sn0wmonster has quit [Ping timeout: 260 seconds]
lkcl has joined #neo900
zrafa has quit [Ping timeout: 258 seconds]
<DocScrutinizer05>
wpwrak: maybe a infobit you eventually would find useful: it seems the sequence of $sheet ... $endsheet sections in root sheet defines the sequence of subsheets and thus the sheetnumber
<DocScrutinizer05>
and I also found no hint about a way to move the root sheet out of its position as sheet#1
<DocScrutinizer05>
and is confirmed at least as far as removing/deleting $sheet_2 sequence renumbers all other sheets
<wpwrak>
we basically have no meaningful control over sheet numbers. which sucks. that's why i suggested to use manual numbering, at least for now.
<wpwrak>
(since there are major changes to eeschema's file format on their way, this may also be an opportunity to see improvements for sheet numbers. but i'm not sure when the file format change will really happen.)
<DocScrutinizer05>
the sheet number should only be a default fallback
<DocScrutinizer05>
I think along the line of adding a field to the sheet properties
<DocScrutinizer05>
but that's KiCAD R&D
<wpwrak>
yeah. what i take from all this is that you either 1) keep your sheet structure very stable (which is my usual approach when well into a project), 2) simply don't care (which is my approach in the early stages of a project), or 3) do it manually.
<wpwrak>
in our case, we have the added complication that the numbers come from eagle, where things work differently. so option 1) isn't even available.
<wpwrak>
at least the eagle-side numbers seem to be pretty consistent. could be worse :)
Pangolin has joined #neo900
<DocScrutinizer05>
I think I'll do a "manual" fixing of numbering for all sheets now at start of KiCAD
<DocScrutinizer05>
search&replace in KiCAD is braindead too :-/
<DocScrutinizer05>
I can't even figure how the allegedly supported wildcards for the searchstring work
lkcl has quit [Ping timeout: 240 seconds]
fling has quit [Ping timeout: 260 seconds]
<DocScrutinizer05>
and "close" on the search&replace dialog nevertheless does a search and centers the sheet on the hit
<DocScrutinizer05>
>.-(
<wpwrak>
"find and replace" sounds scary to me :) i always do such things manually
<DocScrutinizer05>
it's more useless than scary
<wpwrak>
<-- luddite :)
<DocScrutinizer05>
nah, it's just not versatile enough, but has a "find" and "replace" for the cautious, in addition to the "replace all"
<DocScrutinizer05>
so you could just semi-automate what you otherwise would do manually
<DocScrutinizer05>
I think sed or awk is it for the really heavy lifting
<DocScrutinizer05>
and yes, that's a quite demanding task to do this right and make sure no false positives creep in
<DocScrutinizer05>
well, only slightly demanding since the file format it clear enough i'd think
* wpwrak
nods
<wpwrak>
i usually run a grep first to see if my regexp gets the right things, and if anything looks suspicious
<DocScrutinizer05>
yep
<wpwrak>
then i sacrifice a black goat to murphy and change grep to sed -i ;-)
<DocScrutinizer05>
hehe, then you do a diff and eat the rest of the goat, on barbeque
<wpwrak>
sounds like a plan :)
<DocScrutinizer05>
also counting how many actions been taken makes for a good first indicator what happened
<DocScrutinizer05>
usually it's easier to find out about the needed number of actions than to proofread every single one of them
<DocScrutinizer05>
then you check the first and last replace been done and make sure those are the ones you'd expect
<DocScrutinizer05>
odd are no flase negative and false positive in between both at the same time
<DocScrutinizer05>
so e.g. when you raplce R401 to R417 by R501 to R517, you know there should be 17 actions and you could see from logs if both 401 and 417 been done right (assuming the find sequence is ascending order)
<DocScrutinizer05>
replace*
<DocScrutinizer05>
ooh, it's clear we never start at N00 for a component, like R500, right? ;-)
<DocScrutinizer05>
we shouldn't
<DocScrutinizer05>
first component X${N}01
zrafa has joined #neo900
<wpwrak>
yeah, no ~00 in sight
Pangolin has quit [Ping timeout: 260 seconds]
Pangolin has joined #neo900
<DocScrutinizer05>
for the "E"dit I'd expect to have access to *all* properties of whatever object. For the find&replace I'd expect it to allow all existing properties incl object type semantic filters and true perl regex or better for the basic search and replace action
<DocScrutinizer05>
but meh, kicad not even has wire color or wire line width in schematics as user adjustable properties of simple wires
<DocScrutinizer05>
so it's your design that wins here :-)
<wpwrak>
i wouldn't use a bus just for visual effect. if you don't connect it properly, you'll probably get complaints or worse.
<DocScrutinizer05>
complaints for sure
<DocScrutinizer05>
and the visual effect is poor anyway
<wpwrak>
(connect properly) with "Place ... entry"
<DocScrutinizer05>
:nod:
<DocScrutinizer05>
those entries look not like matching this usecase anyway
<DocScrutinizer05>
I tested all that. Actually I hoped for bus being a multi-line parallels thing
<DocScrutinizer05>
didn't pan out
<DocScrutinizer05>
did you check the sheet symbols in my PoC? do you think the fontsizes used are prone to pending trouble? (while at it, did you check the negative fontsize? ;-D )
<wpwrak>
negative font size sounds dangerous :)
<DocScrutinizer05>
s/the negative/ a negative/
<DocScrutinizer05>
I didn't use those
<DocScrutinizer05>
but they are funny
<DocScrutinizer05>
I rather had expected mirrored fonts
<DocScrutinizer05>
;-)
<DocScrutinizer05>
or inverted
<DocScrutinizer05>
anyway worst thing I used was fontsize=0
<DocScrutinizer05>
which already gives me a frowning
<DocScrutinizer05>
but seems the only way I can find to make the text vanish
<DocScrutinizer05>
and since there's also no way to move the text *inside* the box, I did that
<DocScrutinizer05>
if the box size was a numerically editable property, I maybe had used negative Y size there ;-)
pagurus has quit [Ping timeout: 276 seconds]
<DocScrutinizer05>
but those sheet symbols are very strange objects
pagurus has joined #neo900
<DocScrutinizer05>
obviously a unique primary object type
<DocScrutinizer05>
weird
<DocScrutinizer05>
I rather had expected they reuse the frame object
<wpwrak>
which text are you trying to hie ? sheet name or file ?
<DocScrutinizer05>
both, well for file I donated a 0.06mm font size ;-)
<DocScrutinizer05>
for no particular reason
<wpwrak>
why hide them at all ? the information is useful, and it's a little easier to find if it's already shown
<DocScrutinizer05>
well it looks weird when sheet name gets listed twice, so hide it. filname I considered irrelevant for the index and you can see in in nice bold size in the sheet's frame infobox anyway, but I kept a 0.06mm as a reminescense that somebody might consider it useful there
<DocScrutinizer05>
I'd rather hide it completely to get a clean pretty index of sheet names
<DocScrutinizer05>
regarding the procedure I actually first placed and resized the window symbol, then copied and edited the text, then edited the sheet symbol properties and set the font sizes
<DocScrutinizer05>
so the filename and sheetname were visible when I edited the text
<wpwrak>
the sheet name kicad displays also tells you what kicad things is going on. so if you have an inconsistency, you'll spot it quickly. doesn't really take anything anyway from what you're trying to do.
<wpwrak>
well, they're always visible in some way :)
<DocScrutinizer05>
I don't see the use of showing sheetname twice. Gets puzling like hell which is active for clicking and which isn't
<DocScrutinizer05>
I fail to see why anybody would need or want it in index on sheet1
<DocScrutinizer05>
but feel free to tinker with it, as long as we can keep our sheet numbers
<DocScrutinizer05>
as already mentioned we could even do a "beautify step" prior to printing, where we change the index to any type you like, from simple list of text with small frames around each row, to complete functional block diagram and moving the schematics from sheet_1 to a newly inserted sheet_2, since then we don't want to edit the result anymore and so I don't mind kicad features like support to annotation getting broken by this
<DocScrutinizer05>
for daily editing I don't care about the index, I got my fast pageflipping anyway, where I can browse through all 38 sheets in <3s. And then I also could add the already mentioned tiny sheetlabels "#01" to "#38" and jump directly to a page by ctrl+F,#24;enter
<DocScrutinizer05>
same also works for tagwords, like "wwan, modem, gemalto, cinterion, siemens, umts, lte[,whatever]..." in fontsize 0.001mm ideally in center of each sheet
<DocScrutinizer05>
so ... ctrl+F,'cinter',enter
<DocScrutinizer05>
nice: you could have same tag on multiple (2, or 3) sheets and switch between them swiftly with fafter this
<DocScrutinizer05>
F5
<DocScrutinizer05>
sorry, I hit f "F5" and typing closed the popuüp requester immediately
<DocScrutinizer05>
so 'ctr+F s t o r Enter' would kick you to the first page with e.g. eMMC (storage), and F5 gets you to the NAND (storage), and then uSD (storage)
<DocScrutinizer05>
I think this is a useful concept
<wpwrak>
suggest it on #kicad ?
<DocScrutinizer05>
why not, though what to suggest? everybody can do that and I don't think it's particularly innovative
<wpwrak>
you're talking about a feature you want added in kicad, no ?
<DocScrutinizer05>
no
<DocScrutinizer05>
I'm talking about stuff I already use in kicad
<wpwrak>
oh, i see
<wpwrak>
but that's stuff that depends on kde, right ?
<DocScrutinizer05>
no
<wpwrak>
so how does that page flipping work then ?
<DocScrutinizer05>
my ctrl+shift+pagedwon depens on kde
<wpwrak>
i.e., how to enable it
<wpwrak>
what is ctrl+shift+pagedown ?
<DocScrutinizer05>
you add a text "umts" on three pages. for convenience to make it size 0.001mm and place it in sheet center
<DocScrutinizer05>
then you press ctrl+F
<DocScrutinizer05>
then umts
<DocScrutinizer05>
then enter
<DocScrutinizer05>
or carriage return
<DocScrutinizer05>
will bring you to first sheet with umts tag on it
<DocScrutinizer05>
F5 repeats search and so kicks you to next sheet
<wpwrak>
well, you can do that. but i'd really stay away from "hiding" things. put it at a location where it doesn't get in the way, but make it big enough that one can see that it's there.
<DocScrutinizer05>
up to you
<DocScrutinizer05>
the concept is so that everybody can customize it to their own needs, with minimum effort
<DocScrutinizer05>
but you give me ideas, the "hidden" text should actually start with a "magic": something like "##tags: ", so you can search for *all* those hidden things
<DocScrutinizer05>
so far I had my #01, #02 etc tags in 0.6mm in upper left corner
<DocScrutinizer05>
downside: "Find" centers the sheet to there
<DocScrutinizer05>
also >:-( is that kicad can't find any of the sheetname, filename, date, whatever strings in the frame infobox
<DocScrutinizer05>
while, strange enough, it *does* find the sheetname and filename as components of the sheetsymbols
<DocScrutinizer05>
whatever style you use to add your tags, once you searched for one of then with ctrl+F, you can close the requester and still repeat the find via F5, when you checkmarked "wrap arounf end of searchlist" in the requester, this will work to cycle through multiple sheets with same tag
<DocScrutinizer05>
and _if_ "Find" actually would support proper regex instead of obscure or defunct wildcards, you could even list two alternative sheet number tags or arbitrary other unique tags to create that searchlist to cycle through
<DocScrutinizer05>
((something like "##tags: ", so you can search for *all* those hidden things)) s/##tags/##hidden/
<DocScrutinizer05>
at least I understand now where from your concerns regarding hiding filename in index are :-)
<DocScrutinizer05>
I don't think hiding passive graphical elements (i.E. text) could cause a lot of trouble
<DocScrutinizer05>
worst case you'd move around or even copy it to another sheet, without noticing
<DocScrutinizer05>
you'll notice when you actually use the features those hidden things are introduced for. Then simply press Del and it's gone
<wpwrak>
i'm a bit surprised that you need so much stuff just to be able to navigate. i understand that you'd want bigger labels on the subsheets, but all the rest sound like the sort of stuff you won't need by the time you're done implementing it, because you'll then simply remember that chip X is on sheet Y, and won't need tags and such.
<DocScrutinizer05>
err aha
lkcl has joined #neo900
<DocScrutinizer05>
"done implementing"... let's see
<wpwrak>
just saying. this is actually a common phenomenon - and the source of lots of semi-finished documents and such. people get started, realize they'd really want a manual, start documenting their journey, but at some point they know the way, and the manual gets abandoned soon thereafter.
<DocScrutinizer05>
done adding tag "modem" to all sheets related to modem and monitor
<DocScrutinizer05>
what took me the most time was actually finding the pages to tag
<DocScrutinizer05>
anyway n8
<DocScrutinizer05>
ctrl+F mod Enter Esc F5 F5 F5 F5 F5 F5......
<DocScrutinizer05>
damn, the commit doesn't fit, it says ctrl+alt+pgdown
<DocScrutinizer05>
anyway it's just a macro hotkey for the key sequence alt+V, H, down (or up), enter
<DocScrutinizer05>
this works also manually, of course not >10 times per second
<wpwrak>
funny. the conversion got the arc of BMG160 and BNO055 wrong. however my converter "resurrects" the original style. sometimes, doing it wrong twice does make it right ;-)
arcean has joined #neo900
<DocScrutinizer05>
start and end swapped
<DocScrutinizer05>
hellekin: another 20 N900 on their way, 20 more ordered
<DocScrutinizer05>
Wuerzburg, Muenchen, Ansbach... mad suckers getting closer pretty fast
arcean has quit [Read error: Connection reset by peer]
arcean has joined #neo900
louisdk has joined #neo900
lobito has quit [Ping timeout: 258 seconds]
arcean has quit [Read error: Connection reset by peer]
<wpwrak>
DocScrutinizer05: do you have any changes that need committing ? i have a bunch of small fixes (all visual stuff) i'd like to make (will affect several sheets and also neo900.lib)
<wpwrak>
DocScrutinizer05: at least the last one had the ethnical background compatible with mob expectations. some improvement :)
Pali has joined #neo900
<wpwrak>
(all visual) add one bug fix. a wire i accidently eliminated in conversion.
herpderphurr has quit [Ping timeout: 240 seconds]
<enyc>
wpwrak: i'm just lpeased to see ''things happening'' =))
<wpwrak>
yeah, shooting the lame bird (eagle) helped to unblock a lot of things. finally we get to roam freely :)
<enyc>
wpwrak: at what point will ke know feasibility of project completion // ability to start showing this to potentional investors etc.?
<wpwrak>
i guess when v2 is done (designed, built, doing something), it'll be something worth showing.
<wpwrak>
as a smaller intermediate step, once the schematics are done for v2, that will also be a major milestone, though of a less visible nature
<enyc>
worth briefly documentting // showing on eo900.org imho
<enyc>
wpwrak: iirc you said preparing v2 is now not 'new ground' just using everything and sorting out all those tedious fiddly details ?
<wpwrak>
basically yes. there are some things that have diverged from the white papers (or vice versa), that need updating. some of them are relatively major (sim switch). also, one bit that's largely unexplored is the bb-xm interface.
<wpwrak>
and there's tiny stuff, basically visual cleanup. there are many things that aren't drawn very nicely, which can cause confusion and mistakes.
<wpwrak>
my favourites are texts that overlap each other. i've already disentangled many of these things, but there's still more.
<enyc>
'there (was) some discussion over usb bus connections... being able to use LTE-modem and usb simulatnously, etc. iirc
<enyc>
good luck // keep going =)
<wpwrak>
i also expect that we'll now work more schematics-centric, i.e., with the schematics being the principal reference for things, and make less use of white papers. they still have their place, but if we can apply a change directly in the schematics, we don't need to document what should be in the schematics if there's not much else to say about it.
<wpwrak>
(usb) err, which usb ? we have three separate usb busses: the micro connector, modem, and hackerbus.
<wpwrak>
(luck) thanks ! :)
<enyc>
wpwrak: i can't remember ;p there was some time it wosn't possible to use certian combinations of (something) at once whic i thouht was a problem ;p
<enyc>
wpwrak: maybe it was seiral or i2c busses i can't remember!
<wpwrak>
hmm, we have uart3 shared between IR and hackerbus
galiven has joined #neo900
SylvieLorxu has quit [Remote host closed the connection]
SylvieLorxu has joined #neo900
galiven has quit [Quit: Konversation terminated!]
galiven has joined #neo900
<enyc>
good luck =)
xman has joined #neo900
tsuggs has quit [Ping timeout: 258 seconds]
<DocScrutinizer05>
wpwrak: (committing) yes, I'd probably like to commit sheet neo900_SS_5
<DocScrutinizer05>
not finished yet but...
<wpwrak>
no problem. i'll just wait then
<DocScrutinizer05>
ooh, I already saved it, so just commit now
<DocScrutinizer05>
[master 4668d7b] neo900_SS_5.sch batt gauge major cleanup, 75% finished
<DocScrutinizer05>
wpwrak: I tested the index idea of mine further and noticed you can only doubleclick the subsheet rectangle in an area where no plain Text overlay, so at least we need to make those rectangles wider than the text
<wpwrak>
btw, given that i now have a nearly complete reader and renderer for schematics, i'm tempted to extend it to interactive navigation (still read-only). that should be relatively simple. this would create room for lots of comfort features.
<wpwrak>
don't you want to get rid of the "Sheet #". save some space (which may be useful later)
<DocScrutinizer05>
sure, that's basically just placeholders
xman has quit [Read error: Connection reset by peer]
<wpwrak>
and please use something close to the default size (60 mil) for the text fields on the subsheets.
<DocScrutinizer05>
funny how that's _no_ flase friend though it terriobly sounds like one
<DocScrutinizer05>
hmmm
<DocScrutinizer05>
60mil are 0.06" ?
<wpwrak>
hehe. sch2fig can now process all the sheets. time to set up a regression test so that the remaining few fixes don't break all the stuff that's working.
<wpwrak>
yes
<wpwrak>
50 mil is also readable. beyond that, it gets painful.
<wpwrak>
ah, almost all sheets. i don't support subsheets yet. well, that can wait.
<wpwrak>
designed for engineers, no walled apple garden ;-)
<DocScrutinizer05>
engineers need those strings? what for?
announ has joined #neo900
<DocScrutinizer05>
particularly since they are in prettyprint on the sheet itself, in infobox
<wpwrak>
to see that the manual label matches what kicad thinks is going on. also, please consider that we'll want reviewable PDF, where you can't just "click to look it up on the sheet"
<DocScrutinizer05>
how would filename or true sheetname be any relevant for evaluation in a pdf that has no idea of either of both? and still has them in the A3frame infobox
<DocScrutinizer05>
au contraire I consider the unentanglement between the big blue arbitrary text and the sheetname a feature
<wpwrak>
if true name and manual label disagree, we have an issue. and you may want to look up the file name.
<DocScrutinizer05>
but meh, whatever you like, however I still would like to understand the rationale which I can't follow so far
<DocScrutinizer05>
you *can* look up the sheet name and filename any time, **on the sheet** - just click on the sheetsymbol and look to the source
<wpwrak>
mostly such things are instinct :) it's the sort of thing that doens't "feel right". alas, i didn't store a detailed explanation along with it. so it's hard to justify this in a court of inquisition.
<DocScrutinizer05>
I don't think I need to check if kicad has an internal lobotomy to use another sheet/file-name on index page than on the page itself
<wpwrak>
now let's see which pdf merger works best ... so many to choose from. pdfunite, pdfjoin, ...
<DocScrutinizer05>
for pdf review it's massively irrelevant
<wpwrak>
still need to stabilize the output size (right now, it just uses a badly computed bounding box), and there are a few known bugs in the rendering
<wpwrak>
please don't mix the subsheets with the rest of the circuit. that just gets incredibly messy. also, if we want to arrange them in a nicer way at some point, the circuit stuff will get in the way.
<wpwrak>
the sheet numbers aren't to be trusted anyway. i..e., if we need to add a sheet somewhere in the middle, or merge sheets, all the sheet numbers change anyway.
<wpwrak>
i rarely use such "hotkeys". do they actually work ? :)
<DocScrutinizer05>
wpwrak: >>please don't mix the subsheets with the rest of the circuit.<< this was the whole purpose of this whole exercise, and it will stay to keep sheetnumbers in sync
<DocScrutinizer05>
we cpould movve the switches schematics to sheet 2
<wpwrak>
so we can't add or merge sheets because that would change the sheet numbers ? that's really the tail wagging the dog
<DocScrutinizer05>
no, adding a sheet will cause work to fix numbering. I won't accept sloppy messed up pseudo "part reference numbers ((maybe, or maybe not)) are in sync with sheet names"
SylvieLorxu has quit [Remote host closed the connection]
<DocScrutinizer05>
when we follow a policy of part reference number = sheet number *100 +nn then we must stick to that
SylvieLorxu has joined #neo900
<wpwrak>
you could solve the "problem" simply by letting it reset the numbers, and then renumbering
<wpwrak>
then they're "in sync" again
<wpwrak>
and you can repeat this as often as needed
<DocScrutinizer05>
alternative would be part reference numbers by functional subblocks, but that would mean major re-annotation, all manually done
<DocScrutinizer05>
yes, I do exactly that
<DocScrutinizer05>
once the core issue of that stupid index sheet got solved
<wpwrak>
what issue is left there then, if you renumber anyway ?
<DocScrutinizer05>
a complete automatic renumbering been on the todo since 3 days already
<wpwrak>
i mean, besides adding larger text
<DocScrutinizer05>
LARGER text??
<wpwrak>
heck, you could even enlarge the name kicad already puts there
<wpwrak>
the larger names you've been putting there
* DocScrutinizer05
is out
<wpwrak>
i get that change. but what's the point of trying to force the buttons onto the index sheet, if you renumber anyway ?
<DocScrutinizer05>
what?
<DocScrutinizer05>
I'm not forcing *anything*
<DocScrutinizer05>
those crappy useless "buttons" been there all the time, as gigantic empty rectangles with illegible microscopic text
<DocScrutinizer05>
butt ugly and useless like hell
<wpwrak>
i mean the buttons in the circuit, lock switch, volume, etc.
<DocScrutinizer05>
what I did is resizing them and aligning them to form a semi-decent index structure with already pretty large arbitrary text user could actually *read*
<wpwrak>
the othes are hierarchical sheet symbols
<wpwrak>
yes, that part is fine
<wpwrak>
what worries me is that you're also merged the next sheet (with the buttons) into this
<DocScrutinizer05>
I moved the scheatics tehre to be able to delete one sheet so the numbering was ok to start with, again. This can get fixed later on when we do a renumbering
<wpwrak>
if you want to have "circuit" on the front sheet, that should be connections between subsheets. like i've done for example with anelok
<DocScrutinizer05>
I do not WANT that, it was the most obvious alternative
<wpwrak>
and even without "circuit", it's a good place for global information
<wpwrak>
well, just get it over with and renumber then ?
<DocScrutinizer05>
yeah, please let me finish stuff then criticise it
<wpwrak>
renumbering has its own set of issues (in general, not kicad specific), but we'll just have to live with that
<DocScrutinizer05>
I provided full PoC for you to review and comment
<wpwrak>
very well :)
<DocScrutinizer05>
:-/
<DocScrutinizer05>
it's pretty frustrating to design something, provide for review, ask for comments, get "OK looks good". then start work to turn it into something complete and THEN suddenly receive criticism
<wpwrak>
well, what you showed first looked good. what you did then, not so much. so ... :)
<DocScrutinizer05>
please! what I showed and what I provided never really changed, except I unhide text as you requested
<DocScrutinizer05>
and don't ask me to prove that now. I could
<DocScrutinizer05>
but that would cost me another ten minutes going through chanlog
<DocScrutinizer05>
you asked for a pdf which doesn't make any sense since the issue of e.g. sheet rectangles with text overlay not working for doubleclick is not visible in pdf
<DocScrutinizer05>
I thought I achived to communicate that
<DocScrutinizer05>
anyway renumbering is on the ToDo list, but only later on, when the basics are established
<DocScrutinizer05>
[2016-07-25 Mon 06:50:43] <DocScrutinizer05> as already mentioned we could even do a "beautify step" prior to printing, where we change the index to any type you like, from simple list of text with small frames around each row, to complete functional block diagram and moving the schematics from sheet_1 to a newly inserted sheet_2, since then we don't want to edit the result anymore and so I don't mind kicad features like support to
<DocScrutinizer05>
annotation getting broken by this for daily editing I don't care about the index, I got my fast pageflipping anyway, where I can browse through all 38 sheets in <3s. And then I also could add the already mentioned tiny sheetlabels "#01" to "#38" and jump directly to a page by ctrl+F,#24;enter
<DocScrutinizer05>
NB >>moving the schematics from sheet_1 to a newly inserted sheet_2<< !
<DocScrutinizer05>
you commented my efforts to sanitize the useless big empty rectangles with cynicism and denial from very beginning. I wonder why
<wpwrak>
i never said anything against adding larger text
<DocScrutinizer05>
and I never rejected having a break when due
<DocScrutinizer05>
I think this is the right moment now
<DocScrutinizer05>
bbl
<wpwrak>
yeah, dinner time here, too :)
<DocScrutinizer05>
[2016-07-24 Sun 20:28:05] <DocScrutinizer05> if you *really* want a dedicated page filled only with scale 1:8 sheet symbols, we can move to that style at end of schematics editing
<DocScrutinizer05>
[2016-07-24 Sun 20:32:26] <DocScrutinizer05> btw the approach I suggested also en passant 'fixes' the page number issue as found in pdf
<DocScrutinizer05>
it's puzzling like hell when you want to got to sheet3 in pdf, enter page number: 3 and are on sheet2. You said the approach of origin:0 doesn't work in all pdf viewers
<DocScrutinizer05>
starting with "sheet_SS_2" as first subsheet is fixing this conveniently
<DocScrutinizer05>
this - among others - is a reason why sheet_SS_1 had to go
<DocScrutinizer05>
and I had to place the "buttons" *somewhere*
<DocScrutinizer05>
since they are numbered X1nn I placed them on real sheet_1 which is the index page
<wpwrak>
a18:157.7
<DocScrutinizer05>
we're free to move them anywhere we like, as long as we re-annotate all parts
<DocScrutinizer05>
bownout again?
<DocScrutinizer05>
I pushed neo900.sch
paulk-aldrin has quit [Quit: Leaving]
Pali has quit [Ping timeout: 240 seconds]
louisdk has quit [Remote host closed the connection]